44644 LESSONS FROM REFORMS IN LOW­ AND MIDDLE­INCOME COUNTRIES GOOD PRACTICES IN HEALTH FINANCING LOW-INCOME INC MIDDLE-INCOME MIDDLE INC HIGH-INCOME HIGH INC COUNTRIES COUNTRIES COUNTRIES PRIVATE OUT OF POCKET PRIVATE OUT OF POCKET PRIVATE POOLED PRIVATE OUT OF POCKET PRIVATE POOLED PRIVATE POOLED GOVERNMENT GOVERNMENT GOVERNMENT Good Practices in Health Financing Good Practices in Health Financing Lessons from Reforms in Low- and Middle-Income Countries Pablo Gottret,George J.Schieber,and Hugh R.Waters Editors The World Bank © 2008 The International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org E-mail: feedback@worldbank.org All rights reserved 1 2 3 4 5 11 10 09 08 This volume is a product of the staff of the International Bank for Reconstruction and Development / The World Bank. The findings, interpretations, and conclusions expressed in this volume do not neces- sarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The bound- aries, colors, denominations, and other information shown on any map in this work do not imply any judgement on the part of The World Bank concerning the legal status of any territory or the endorse- ment or acceptance of such boundaries. Rights and Permissions The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruc- tion and Development / The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; telephone: 978-750-8400; fax: 978-750-4470; Internet: www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202- 522-2422; e-mail: pubrights@worldbank.org. ISBN-13: 978-0-8213-7511-2 eISBN-13: 978-0-8213-7512-1 DOI: /10.1596/978-0-8213-7511-2 Library of Congress Cataloging-in-Publication Data Good practice in health financing : lessons from reforms in low and middle-income countries / Pablo Gottret, George J. Schieber, and Hugh R. Waters, coeditors. p. ; cm. Includes bibliographical references and index. ISBN-13: 978-0-8213-7511-2 (alk. paper) ISBN-10: 0-8213-7511-3 (alk. paper) 1. Medical care--Developing countries--Finance. 2. Health care reform--Economic aspects-- Developing countries 3. Medical policy--Economic aspects--Developing countries. 4. Medical care, Cost of--Developing countries. 5. Health status indicators--Developing countries. I. Gottret, Pablo E. (Pablo Enrique), 1959- II. Schieber, George. III. Waters, Hugh. IV. World Bank. [DNLM: 1. Health Care Reform--economics--Statistics. 2. Health Care Reform--standards-- Statistics. 3. Delivery of Health Care--economics--Statistics. 4. Developing Countries--Statistics. 5. Health Care Costs--Statistics. 6. Health Expenditures--Statistics. WA 530.1 G646 2008] RA395.D44G66 2008 362.1'04252091724--dc22 2008008389 Cover design by Rock Creek Creative. Contents Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xv Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xvii Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xix Acronyms and Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xxiii Part 1 Assessing Good Practice in Health Financing Reform 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 2 Health Financing Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9 3 Criteria for Defining "Good Practice" and Choosing Country Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15 4 Summaries of Country Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .27 5 Enabling Factors for Expanding Coverage . . . . . . . . . . . . . . . . . . . . . . . . . .57 References for Part I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76 Part 2 Nine Case Studies of Good Practice in Health Financing Reform 6 Chile: Good Practice in Expanding Health Care Coverage Lessons from Reforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89 Ricardo D. Bitrán and Gonzalo C. Urcullo 7 Colombia: Good Practices in Expanding Health Care Coverage . . . . . . .137 Diana Masis Pinto 8 Costa Rica:"Good Practice" in Expanding Health Care Coverage Lessons from Reforms in Low- and Middle-Income Countries . . . . . . . .183 James Cercone and José Pacheco Jiménez 9 Estonia:"Good Practice" in Expanding Health Care Coverage . . . . . . . .227 Triin Habicht and Jarno Habicht 10 The Kyrgyz Republic: Good Practices in Expanding Health Care Coverage, 1991­2006 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .269 Melitta Jakab and Elina Manjieva v vi Contents 11 Sri Lanka:"Good Practice" in Expanding Health Care Coverage . . . . . .311 Ravi P. Rannan-Eliya and Lankani Sikurajapathy 12 Thailand: Good Practice in Expanding Health Coverage Lessons from the Thai Health Care Reforms . . . . . . . . . . . . . . . . . . . . . . .355 Suwit Wibulpolprasert and Suriwan Thaiprayoon 13 Tunisia:"Good Practice" in Expanding Health Care Coverage Lessons from Reforms in a Country in Transition . . . . . . . . . . . . . . . . . .385 Chokri Arfa and Hédi Achouri 14 Vietnam:"Good Practice" in Expanding Health Care Coverage Lessons from Reform in Low- and Middle-Income Countries . . . . . . . .439 Björn Ekman and Sarah Bales Appendix A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .479 Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .487 Boxes 2A Measures of Financial Protection in Tunisia . . . . . . . . . . . . . . . . . . . . . . 12 6.1 Chile: Key Political Milestones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 8.1 Costa Rica: Cooperatives as Health Care Providers . . . . . . . . . . . . . . . .217 10.1 The Kyrgyz Republic: The Manas National Health Care Reform Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .298 12.1 Thailand: Minimum CUP Requirements . . . . . . . . . . . . . . . . . . . . . . . . .370 Figures 1.1 Determinants of Health, Nutrition, and Population Outcomes . . . . . . . .5 2A.1 Payments as Share of Total and Nonfood Expenditure in Tunisia, 2003 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13 3.1 Population Health Indicators Relative to Income and Spending . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20 3.2 Health Service Delivery Indicators Relative to Income and Spending . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21 3.3 Total Health Spending Relative to Income . . . . . . . . . . . . . . . . . . . . . . . . .21 3.4 Health Spending as Share of GDP and per Capita vs. Income . . . . . . . . .22 3.5 Revenue to GDP Ratio vs. Income . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23 3.6 Government Share of Health vs. Income . . . . . . . . . . . . . . . . . . . . . . . . . .23 3.7 Out-of-Pocket Spending Relative to Income . . . . . . . . . . . . . . . . . . . . . . .24 3.8 Hospital Bed and Physician Capacity vs. Income . . . . . . . . . . . . . . . . . . .25 3.9 Literacy vs. Income . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25 5.1 Real GDP Trends per Capita, 1960­2005 . . . . . . . . . . . . . . . . . . . . . . . . . .60 5.2 Political Freedom Trends in Case Countries, 1900­2004 . . . . . . . . . . . . .64 Contents vii 6.1 Chile: Economic Growth, 1810­2005 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .90 6.2 Chile: Growth of Real GDP, 1997­2005 . . . . . . . . . . . . . . . . . . . . . . . . . . .91 6.3 Chile: GDP per Capita, 2004 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .91 6.4 Chile: External Debt, 1996­2005 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .92 6.5 Chile: Composition of External Debt, 2004 . . . . . . . . . . . . . . . . . . . . . . . .93 6.6 Chile: Population Structure, 1990, 2005, and 2020 . . . . . . . . . . . . . . . . . .95 6.7 Chile: Infant Mortality, 1960­2002 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97 6.8 Chile: Life Expectancy, by Historical Period and Gender, 1950­2025 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97 6.9 Chile: Poverty Compared with Other Latin American Countries, 1999 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .98 6.10 Chile: Infant Mortality and Life Expectancy Compared with Other Latin American Countries, 2004 . . . . . . . . . . . . . . . . . . . . . . . . . . .99 6.11 Chile: The Mandatory Health Insurance System, 2006 . . . . . . . . . . . . . .101 6.12 Chile: Structure of Health Spending, by Source, 1998­2004 . . . . . . . . .103 6.13 Chile: Coverage of Social Security System, 1984­2005 . . . . . . . . . . . . . .107 6.14 Chile: Coverage of Open ISAPREs, 2006 . . . . . . . . . . . . . . . . . . . . . . . . .107 6.15 Chile: Structure of Financing for Public Health Spending . . . . . . . . . . .110 6.16 Chile: Social Security System Beneficiaries, by Income Decile and Insurance Type, 2000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .111 6.17 Chile: FONASA and ISAPRE Beneficiaries, by Age, 1990 and 2005 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 6.18 Chile: Available Beds in the SNSS System, 1990­2002 . . . . . . . . . . . . . .115 6.19 Chile: Chronology of Health Reforms, 1917­2006 . . . . . . . . . . . . . . . . .119 6.20 Chile: Cumulative Cost of the 56 GES-Covered Health Conditions, estimates for 2007 . . . . . . . . . . . . . . . . . . . . . . . . . . .126 7.1 Colombia: Age-Specific Fertility Rates . . . . . . . . . . . . . . . . . . . . . . . . . . .140 7.2 Colombia: Population Pyramid, 2005 . . . . . . . . . . . . . . . . . . . . . . . . . . .140 7.3 Colombia: Flow of Funds in the Solidarity and Guarantees Fund . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .144 7.4 Colombia: Funding for the Subsidized Regime, by Source, 2005 . . . . .145 7.5 Colombia: Main Reasons for Not Using Health Care Services, by Income Quintile, 1992 and 1997 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .154 7.6 Colombia: Reported Sources of Payment for Consultations and Hospitalizations, by Insurance Status, 2000 . . . . . . . . . . . . . . . . . . .154 7.7 Colombia: Respondents with a Health Problem and Seen by a Doctor, by Insurance Status and Location, 2000 and 2003 . . . . . . . . . . .156 7.8 Colombia: Population Reporting Use of Preventive Services, by Insurance Status, 1997 and 2003 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .156 7.9 Colombia: Population Reporting Hospitalization in Last Year, by Insurance Status, 1997 and 2003 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .157 7.10 Colombia: Total Enrollees, by Type of Health Plan and Regime, 2005 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .158 7.11 Colombia: Expansion of CR and SR Insurance Coverage, 1992­2006 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .164 viii Contents 7.12 Colombia: Total CR enrollment, Contributors, and Beneficiaries, 1993­2005 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .165 7.13 Colombia: Employment of SR Contributors, 1993­2006 . . . . . . . . . . . .165 7.14 Colombia: Insurance Coverage, by Income Quintile, 1992­2003 . . . . .166 7.15 Colombia: Growth of Insurance Coverage, by Regime and by Residence, 1993­2003 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .167 7.16 Colombia: National Health Spending Distribution and Trends, 1993­2003 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .168 7.17 Colombia: Balance of FOSYGA Compensation Fund, 1996­2005 . . . .169 7.18 Colombia: Unemployment, GDP Real Growth, and Percentage of Population in Poverty, 1991­2005 . . . . . . . . . . . . . . . . . . . . . . . . . . . .170 7.19 Colombia: SR Projected Resources versus Actual Expenditures, 1994­2000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .171 7.20 Colombia: SR Projected Resources versus Actual Expenditures of Solidarity Funding Sources, 2000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .172 7.21 Colombia: Projections for Universal Health Insurance, 2005­2010 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .176 8.1 Costa Rica: Population Pyramids, 1990, 2005, and 2020 . . . . . . . . . . . .185 8.2 Costa Rica: Basic Demographic Indicators . . . . . . . . . . . . . . . . . . . . . . .187 8.3 Costa Rica: National Distribution of Public Spending on Health, by Income Category, 2001 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .190 8.4 Costa Rica: Outpatient Consultations per Inhabitant, by Income Decile, 1998 and 2001 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .195 8.5 Costa Rica: Income and Social Expense Distribution by Function, 2000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .198 8.6 Costa Rica: Distribution of Public Spending on Health Nationally and Regionally, by Income Group, 2001 . . . . . . . . . .199 8.7 Costa Rica: Health Expenditures and Contributions to CCSS, by Income Decile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .199 8.8 Costa Rica: Health Insurance Coverage, of Economically Active Population 1990­2004 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .212 8.9 Costa Rica: Primary Health Care Program Coverage, 1990­2003 . . . . .215 9.1 Estonia: Population Pyramids, 2004 and 2025 . . . . . . . . . . . . . . . . . . . .230 9.2 Estonia: Burden of Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .231 9.3 Estonia: Average Life Expectancy at Birth Compared with EU Countries, 2003 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .232 9.4 Estonia: Distribution of Funding Sources, 1999 and 2004 . . . . . . . . . . .234 9.5 Estonia: Out-of-Pocket Payments Compared with European Union, 2004 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .236 9.6 Estonia: Overview of the Health Financing System, 2004 . . . . . . . . . . .237 9.7 Estonia: The EHIF Contracting Process . . . . . . . . . . . . . . . . . . . . . . . . . .238 9.8 Estonia: Payment Methods for Inpatient and Outpatient Specialist Care, 2005 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .240 9.9 Estonia: Payment Methods for Family Physicians, 2005 . . . . . . . . . . . . .241 Contents ix 9.10 Estonia: Number of Family Physicians, 1993­2004 . . . . . . . . . . . . . . . .245 9.11 Estonia: Number of Hospitals and Acute Care Admissions, 1985­2003 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .248 9.12 Estonia: Access to Medical Care, by Residence . . . . . . . . . . . . . . . . . . . .249 9.13 Estonia: Number of Doctors and Nurses per 100,000 Inhabitants, 1998­2004 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .250 9.14 Estonia: Cumulative Increase in EHIF Pharmaceutical and Health Care Services Expenditures, 1993­2006 . . . . . . . . . . . . . . . . . . . .251 9.15 Estonia: Use of Cardiovascular Medicines, by Groups, 1994­2005 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .252 9.16 Estonia: Milestones in Health Sector Reform, 1992­2003 . . . . . . . . . . .253 10.1 The Kyrgyz Republic: Key Economic Indicators, 1990­2004 . . . . . . . . .271 10.2 The Kyrgyz Republic: Population Pyramids, 1990 and 2005 . . . . . . . . .273 10.3 The Kyrgyz Republic: Leading Causes of Infant Mortality, 2004 . . . . . .274 10.4 The Kyrgyz Republic: Health Expenditures as Share of the State Budget, 1995­2003 . . . . . . . . . . . . . . . . . . . . . . . . . . . .278 10.5 The Kyrgyz Republic: Access to Health Services, by Income Level, 2000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .287 10.6 The Kyrgyz Republic: Comparative Indicators of Hospital Efficiency, 1995 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .288 10.7 The Kyrgyz Republic: Reallocation of Public Expenditures in the Single-Payer System from Fixed Costs to Variable Costs . . . . . . .293 10.8 The Kyrgyz Republic: Trends in Out-of-Pocket Payments, 2000­03 (KGS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .299 10.9 The Kyrgyz Republic: Access to Outpatient Care and Hospital Care, 2000 and 2003 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .301 10.10 The Kyrgyz Republic: Mean Payment by a Public Hospital Patient, 2000 and 2003 (KGS) . . . . . . . . . . . . . . . . . . . . . . . . . .302 11.1 Sri Lanka: Population Pyramids 1991, 2006, 2026, and 2051 . . . . . . . . .316 11.2 Sri Lanka: Government Recurrent Health Spending, 1927­2005 . . . . .324 11.3 Sri Lanka: Differentials in Infant Mortality Rate, by Asset Quintile, 1987­2000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .327 11.4 Sri Lanka: Differentials in Medical Attendance at Childbirth, by Asset Quintile, 1987­2000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .327 11.5 Sri Lanka: Differentials in Use of Modern Methods of Contraception by Currently Married Women, 1987­2000 . . . . . . . . . . .328 11.6 Sri Lanka: Trends in Infant Mortality Rates, Country and Nuwara Eliya District, 1920­2003 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .330 11.7 Sri Lanka: Government Hospital Provision, 1920­2000 . . . . . . . . . . . . .335 12.1 Thailand: Population Pyramids, 2005, 2010, and 2020 . . . . . . . . . . . . . .357 12.2 Thailand: Shift in Budget Allocations, 1982­89 . . . . . . . . . . . . . . . . . . .360 12.3 Thailand: Health Service Delivery Indicators Relative to Income and Spending, 1977­2003 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .361 x Contents 12.4 Thailand: The Health Insurance Model, 2007 . . . . . . . . . . . . . . . . . . . . .365 12.5 Thailand: Health Care Providers' Satisfaction with the UCS, 2003­07 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .371 12.6 Thailand: Number of Doctor Resignations from MOPH Rural Facilities, 1995­2006 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .372 12.7 Thailand: Proposed and Approved Capitation, 2002­08 . . . . . . . . . . . .373 12.8 Thailand: Number of Financially Constrained Hospitals Receiving Contingency Fund Support, 2002­04 . . . . . . . . . . . . . . . . . . .374 12.9 Thailand: The "Triangle That Moves the Mountain" Strategy . . . . . . . .378 13.1 Tunisia: Population Pyramids, 1990, 2005, and 2020 . . . . . . . . . . . . . . .389 13.2 Tunisia: Trends in Life Expectancy, Men and Women, 1966­2004 . . . .390 13.3 Tunisia: Coverage Rates by Health Insurance Scheme . . . . . . . . . . . . . .395 13.4 Tunisia: Trends in Health Care Spending, by Financing Agent, 1980­2004 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .400 13.5 Tunisia: Shifts in Sources of MSP Financing, 1990­2003 . . . . . . . . . . . .403 13.6 Tunisia: Annual Health Care Expenditure, per Capita, 1975­2005 . . . .408 13.7 Tunisia: Average Growth in the Number of Physicians, Dentists, Pharmacists, and Paramedical Personnel, Selected Periods . . . . . . . . . .412 14.1 Vietnam: Population Pyramids, 1990, 2005, and 2020 . . . . . . . . . . . . . .444 Tables 2A.1 Catastrophic Impact of Out-of-Pocket Payments in Threshold Expenditure Shares in Tunisia, 2003 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .14 3.1 Income and Health Spending, 2004 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18 3.2 Health Outcome and Delivery Indicators, 2005 . . . . . . . . . . . . . . . . . . . .19 3.3 Correlations between Population Health Outcomes and Income, Health Spending, and Literacy Levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26 6.1 Chile: Net Tax Revenue Structure, 1996­2004 . . . . . . . . . . . . . . . . . . . . . .92 6.2 Chile: Macroeconomic Performance, 2000­05 . . . . . . . . . . . . . . . . . . . . .93 6.3 Chile: Disease Burden, 1993 (DALYs lost) . . . . . . . . . . . . . . . . . . . . . . . . .96 6.4 Chile: Key Health Spending Indicators, 1998­2004 . . . . . . . . . . . . . . . .102 6.5 Chile: Health Spending as a Percentage of GDP, 2004 . . . . . . . . . . . . . .103 6.6 Chile: Composition of Spending on Social Welfare Programs, 2000­05 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .104 6.7 Chile: Health Problems Covered under the GES, 2005­07 . . . . . . . . . . .105 6.8 Chile: Health Coverage Provided by Social Security System, 2005 . . . .106 6.9 Chile: Services Provided by the FONASA and Average Annual Spending per Beneficiary, 2004 . . . . . . . . . . . . . . . . . . . . . . . . . .108 6.10 Chile: Average Number of Health Services Provided by ISAPRES per Beneficiary, 1996­2004 . . . . . . . . . . . . . . . . . . . . . . . . . . . .108 6.11 Chile: Number of Hospital Admissions, 2001­03 . . . . . . . . . . . . . . . . . .109 Contents xi 6.12 Chile: Contributions, Benefits, and Subsidies within the FONASA, 1995 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .112 6.13 Chile: Global Health Indicators: 1960 and 1995 . . . . . . . . . . . . . . . . . . .124 7.1 Colombia: Selected Economic Indicators . . . . . . . . . . . . . . . . . . . . . . . .138 7.2 Colombia: Demographic Profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .139 7.3 Colombia: Main Causes of Mortality, by Age Group and Gender, 2002 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .141 7.4 Colombia: Top Five Health Conditions, by Burden of Disease and Cause, 2002 (DALYs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .141 7.5 Colombia: Burden of Disease, by Disease Category and Cause, 2002 (DALYs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .141 7.6 Colombia: Total Population, by SISBEN category, 2003 . . . . . . . . . . . . .145 7.7 Colombia: UPC Premium Value, 2006 . . . . . . . . . . . . . . . . . . . . . . . . . . .146 7.8 Colombia: Summary of NSHI Schemes, December 2005 . . . . . . . . . . .148 7.9 Colombia: Classification of Medical Care Types, by Complexity . . . . .149 7.10 Colombia: Comparison of Breadth of CR, SR, and PS Benefits Packages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .150 7.11 Colombia: CR and SR Enrollment Rules, Copayments, and Choice, 2006 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .151 7.12 Colombia: Coverage of Services to Supplement the Benefits Packages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .152 7.13 Colombia: Coverage of Minimum Interventions to Address 2002 Burden of Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .153 7.14 Colombia: Selected Health Expenditure Indicators, 2003 . . . . . . . . . . .153 7.15 Colombia: Households Experiencing Catastrophic Payments or Impoverishment due to Ambulatory or Hospital Health Shock, by Insurance Status, 2003 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .155 7.16 Colombia: Services Offered by Health Care Facilities, by Type, 2004 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .160 7.17 Colombia: Milestones toward Achievement of Universal Health Insurance Coverage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .161 7.18 Colombia: Selected Determinants of FOSYGA Balance, 1998­2002 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .171 7.19 Colombia: SR Funding Sources, 1995­2005 . . . . . . . . . . . . . . . . . . . . . .172 8.1 Costa Rica: Economic Indicators, 1990­2004 . . . . . . . . . . . . . . . . . . . . .184 8.2 Costa Rica: Population Growth Rates and Projections, 1990­2020 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .186 8.3 Costa Rica: Basic Health Information . . . . . . . . . . . . . . . . . . . . . . . . . . .187 8.4 Costa Rica: Leading Causes of Morbidity and Mortality, 2002 . . . . . . .187 8.5 Costa Rica: Health Expenditure Indicators 1998, 2000, and 2003 . . . . .189 8.6 Costa Rica: Payroll Fees, by Insurance Scheme . . . . . . . . . . . . . . . . . . . .191 8.7 Costa Rica: Health Indicators, 1990­2004 . . . . . . . . . . . . . . . . . . . . . . . .193 8.8 Costa Rica: Health Insurance Coverage, by Income Group, 2005 . . . . .195 8.9 Costa Rica: Child and Infant Mortality, by Health Region . . . . . . . . . . .196 xii Contents 8.10 Costa Rica: Health Spending, by Level of Care . . . . . . . . . . . . . . . . . . . .197 8.11 Costa Rica: Sources of Health Financing . . . . . . . . . . . . . . . . . . . . . . . . .197 8.12 Costa Rica: Estimated Health Insurance Financial Balance in 2050 under Different Scenarios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .201 8.13 Costa Rica: Health System Inputs, 1995­2002 . . . . . . . . . . . . . . . . . . . . .205 8.14 Costa Rica: Major Strengths and Weaknesses of the Health Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .208 9.1 Estonia: Total Health Expenditure, by Source, 2004 . . . . . . . . . . . . . . . .233 10.1 The Kyrgyz Republic: Leading Causes of Mortality and Disability, 2002 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .273 10.2 The Kyrgyz Republic: IMR from Respiratory Infections, 1997­2004 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .275 10.3 The Kyrgyz Republic: Key Health Financing Indicators, 2000­03 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .277 10.4 The Kyrgyz Republic: Coverage in Phase-1 (1997­2001): Population Groups, Sources of Financing, and Benefits . . . . . . . . . . . . .280 10.5 The Kyrgyz Republic: Copayment Rates for Surgery, 2004 (KGS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .282 10.6 The Kyrgyz Republic: Copayments for Treatment without Surgery, Diagnosis, Minor Surgery, 2004 (KGS) . . . . . . . . . . . .282 11.1 Sri Lanka: Economic Indicators 1930­2005 . . . . . . . . . . . . . . . . . . . . . . .313 11.2 Sri Lanka: Social Indicators 1930­2005 . . . . . . . . . . . . . . . . . . . . . . . . . .313 11.3 Sri Lanka :Demographic and Health Indicators, 1930­2003 . . . . . . . . .315 11.4 Sri Lanka: Leading Causes of Mortality, 2001 . . . . . . . . . . . . . . . . . . . . .317 11.5 Sri Lanka: Patient Morbidity, Inpatients and Outpatients . . . . . . . . . . .317 11.6 Sri Lanka: Trends in Health Care Spending, 1953­2005 . . . . . . . . . . . . .319 11.7 Sri Lanka: Incidence of Public Health Expenditures, 1979­2004 . . . . .320 11.8 Sri Lanka: Progressivity in Health Financing Compared with Selected Asian Countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .321 11.9 Sri Lanka: Proportion of Population Pushed below the PPP$1.08 Poverty Line by Household Health Spending, Compared with Selected Asian Countries . . . . . . . . . . . . . . . . . . . . . . . .322 11.10 Sri Lanka: Infant Mortality Rates in Different Social Groups, 1920­22 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .325 11.11 Sri Lanka: Infant Mortality Rate, Selected Districts, 1921­2000 . . . . . .326 11.12 Sri Lanka: Health Services Use and Spending, Compared with Selected Comparable Countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .331 11.13 Sri Lanka: Technical Efficiency in Public Hospitals, Compared with Selected Countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .332 11.14 Sri Lanka: Proportion of MOH Expenditures Devoted to Hospitals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .333 11.15 Sri Lanka: Provision of Health Service Inputs and Activities, 2003 . . . .334 11.16 Sri Lanka: Trends in Treatment Sources Used by Sick Persons, 1978­2004 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .334 Contents xiii 11.17 Sri Lanka: The Chronology of Scaling-Up Health Reforms . . . . . . . . . .338 11.18 Sri Lanka: Expansion of Health Service Coverage, 1931­51 . . . . . . . . .345 12.1 Thailand: Top 10 Causes of Disease Burden, by Gender, 2004 . . . . . . . .358 12.2 Thailand: Evolution of Health Insurance Coverage . . . . . . . . . . . . . . . .362 12.3 Thailand: Health Expenditure, by Source, 1995­2005 . . . . . . . . . . . . . .363 12.4 Thailand: Characteristics of Health Insurance Schemes, 2007 . . . . . . . .364 12.5 Thailand: UCS Inclusion and Exclusion List of Expensive Health Care Interventions, 2001 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .367 12.6 Thailand: The 13 Elements for UCS Budget Allocation, 2002­08 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .369 12.7 Thailand: Utilization by UCS Beneficiaries, 2002­07 . . . . . . . . . . . . . . .371 12.8 Thailand: Progress in Health Insurance Coverage . . . . . . . . . . . . . . . . .375 12.9 Thailand: Catastrophic Health Care Expenditure by Households, 2000­06 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .375 12.10 Thailand: Consumer Satisfaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .376 13.1 Tunisia: Economic Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .386 13.2 Tunisia: Government-Sponsored Benefits . . . . . . . . . . . . . . . . . . . . . . . .387 13.3 Tunisia: Poverty Trends, by Area, 1975­2000 . . . . . . . . . . . . . . . . . . . . . .388 13.4 Tunisia: Selected Demographic Indicators . . . . . . . . . . . . . . . . . . . . . . . .391 13.5 Tunisia: Global Burden of Disease, 2002 Estimates . . . . . . . . . . . . . . . . .392 13.6 Tunisia: Top Five Burden of Disease Causes, 2002 Estimates . . . . . . . . .393 13.7 Tunisia: CSS Insurance Coverage Rates . . . . . . . . . . . . . . . . . . . . . . . . . .396 13.8 Tunisia: Health Expenditure, as a Share of GDP and Per Capita, 1980­2004 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .398 13.9 Tunisia: Health Expenditure Indicators, 1995­2004 . . . . . . . . . . . . . . . .399 13.10 Tunisia: Health Care Coverage and Financing . . . . . . . . . . . . . . . . . . . . .402 13.11 Tunisia: Health Care Provider Payment Methods, by Type of Coverage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .404 13.12 Tunisia: Provision for NFSPS Affiliates . . . . . . . . . . . . . . . . . . . . . . . . . .406 13.13 Tunisia: Provision for NFSS Affiliates . . . . . . . . . . . . . . . . . . . . . . . . . . . .407 13.14 Tunisia: Household Consumption, 1975­2000 . . . . . . . . . . . . . . . . . . . .407 13.15 Tunisia: Household Expenditures, by Service Category, 2000 . . . . . . . .408 13.16 Tunisia: Health Care Provision, by Facility Type, 2004 . . . . . . . . . . . . . .409 13.17 Tunisia: Number of Beds in MSP Facilities . . . . . . . . . . . . . . . . . . . . . . .410 13.18 Tunisia: Private Health Facilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .411 13.19 Tunisia: Number of Public and Private Hospital Beds, 1985­2004 . . . .411 13.20 Tunisia: The Physician Workforce, 1981­2004 . . . . . . . . . . . . . . . . . . . .411 13.21 Tunisia: Medical Personnel, 1990­2004 . . . . . . . . . . . . . . . . . . . . . . . . . .412 13.22 Tunisia: Large-Scale Medical Equipment, 1995­2004 . . . . . . . . . . . . . . .414 13.23 Tunisia: Life Expectancy and Infant Mortality Compared with MENA Region, 1960­2004 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .417 13.24 Tunisia: Coverage with Primary Health Care Services, 2005 . . . . . . . . .418 13.25 Tunisia: Health Outcomes Compared with Similar Countries . . . . . . .419 13.26 Tunisia: Inequity of Financial Contribution Indicators Compared with Similar Countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .420 xiv Contents 13.27 Tunisia: Health Inequalities Compared with Similar Countries . . . . . .421 13.28 Tunisia: Stakeholder Analysis for Health Insurance Reform . . . . . . . . .428 13.29 Tunisia: Contribution Rates for Health Insurance Scheme . . . . . . . . . .431 14.1 Vietnam: Economic Indicators, 1990­2004 . . . . . . . . . . . . . . . . . . . . . . .441 14.2 Vietnam: Selected Population and Social Indicators, 1990­2004 . . . . .443 14.3 Vietnam: Estimated Burden of Disease, 2002 . . . . . . . . . . . . . . . . . . . . .445 14.4 Vietnam: Per Capita Total Expenditure on Health, 1999­2003 . . . . . . .448 14.5 Vietnam: Health Spending, by Type of Service . . . . . . . . . . . . . . . . . . . .449 14.6 Vietnam: Sources of Funding, by Type of Service, 2003 . . . . . . . . . . . .449 14.7 Vietnam: Health Insurance Coverage, by Program Type, 1998­2004 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .450 14.8 Vietnam: Human Resources for Health, Selected Indicators . . . . . . . . .459 14.9 Vietnam: Selected Health Care Coverage and Health Outcome Indicators, 1990­2004. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .460 14.10 Vietnam: Chronology of Health Reforms, 1989­2006 . . . . . . . . . . . . . .464 Foreword With health at the center of global development policy, the international commu- nity and developing countries are focused on scaling up health systems, in line with the Millennium Development Goals (MDGs). As a result, both global aid and individual country health reform plans are geared to improving health out- comes, securing financial protection against impoverishment, and ensuring long- term, sustainable financing to support these gains. However, with the scaling up of aid, both donors and countries have come to realize that more money alone can- not buy health improvements or prevent impoverishment due to catastrophic medical bills. Well-targeted health spending is critical for success--including international donor aid to help in achieving sustainable results in the long term. These con- cerns have triggered a major debate over the need to define "results" and assemble appropriate data for decision making, monitoring, and evaluation. Central to this endeavor is the need for an analytically sound evidence base on national health financing policies that have "performed well" in terms of improving health out- comes and financial protection. Unfortunately, for want of solid evidence to inform policy makers, the policy debate is often driven by ideological, one-size- fits-all solutions. This book attempts to partially fill the void by systematically assessing health financing reforms in nine low- and middle-income countries that have demon- strated "good performance" in expanding their populations' health insurance coverage--to both improve health status and protect against catastrophic medical expenses. Good performance also includes average or better-than-average popula- tion health outcomes relative to resources devoted to health and to national income and educational levels. Among the low- and middle-income countries that are in the process of achieving high levels of population coverage and financial protection, nine were selected as examples of good performance by an expert steering committee repre- senting all six World Bank Regions. They are Chile, Colombia, Costa Rica, Estonia, the Kyrgyz Republic, Sri Lanka, Thailand, Tunisia, and Vietnam. Each country case is analyzed using a standardized taxonomy that captures the key health and nonhealth sector­specific factors affecting the performance of its health financing xv xvi Foreword system. Because this study relies only on post-reform information, experiments with control groups or randomization are not possible. Thus no effort is made to attribute causality. However, based on the country studies, an attempt is made to identify the common enabling factors evident in most of the cases. The factors identified in the study make sense on conceptual grounds and would be likely hypotheses for any rigorous evaluation of the impacts of health financing changes. While these findings are interesting, their more important con- tribution is demonstrating that the global community must do a much better job of defining and documenting "good performance" and evaluating and dissemi- nating the global evidence base. To do this, at least five important actions need to be taken. First, what "good performance" means in health financing must be well defined. Second, standard- ized and appropriate qualitative and quantitative information must be collected. Third, health system characteristics must be described in enough detail so that the critical components and their interactions can be identified and assessed. Fourth, rigorous evaluations must be undertaken. Fifth, these evaluations-- conducted in a policy-relevant and user-friendly manner--must be disseminated to all stakeholders. In all these areas, the global community has done a rather poor job. With the growing importance of this issue, it is essential that both national and interna- tional decision makers and health policy analysts take charge and move this agenda forward. It is hoped that this volume in some small way can contribute to focusing the global health financing agenda in this direction. Julian Schweitzer Director, Health, Nutrition, and Population The World Bank Acknowledgments Many individuals contributed to making this report a reality. Part I was written by Lisa K. Fleisher, Pablo Gottret, Adam Leive, George J. Schieber, Ajay Tandon, and Hugh R. Waters. Part II was edited by Pablo Gottret, George J. Schieber, and Hugh R. Waters. The country case studies were prepared by country-based experts who have first-hand knowledge of the health financing and health coverage arrange- ments--and the history of health policy reforms--in their respective countries. The work was guided by a Steering Committee representing the World Bank's six regions. The Steering Committee in conjunction with the Health, Nutrition, and Population (HNP) unit of the Human Development Network developed criteria for the definition of a "good practice" and provided recommendations on the choice of country cases as well as appropriate experts to prepare the cases. The HNP unit, under the guidance of George Schieber, Pablo Gottret, and Hugh Waters (Johns Hopkins University), conceived and carried out this study. Ajay Tandon, Adam Leive, Lisa Fleisher, Valerie Moran, Emiliana Gunawan, Bjorn Ekman (Health Economics Program, Lund University, Sweden), and Axel Rahola (now an Adviser to the Minister of Budgeting, France) assisted in this work. The peer reviewers were Ruth Levine (Director of Programs and Senior Fellow, Center for Global Development), John Langenbrunner (Lead Health Economist, EASHD), and Christoph Kurowski (Senior Health Specialist, LCSHD). Toomas Palu provided extremely useful comments on the Estonia case study. Maureen Lewis chaired the review meeting and provided valuable comments. Kathleen A. Lynch (Editorial Consultant) provided critical editorial support. The report was generously funded by the Swedish International Development Agency, which also provided for the technical support of Bjorn Ekman. Other key individuals contributing to this initiative include: Case Study Authors Chile--Ricardo D. Bitrán and Urcullo C. Gonzalo, Bitrán & Asociados Colombia--Diana Pinto, Fedesarrollo Costa Rica--James Cercone and José Pacheco Jiménez, Sanigest Health Solutions xvii xviii Acknowledgments Estonia--Triin Habicht, Head of Health Economics Department, Estonian Health Insurance Fund; and Jarno Habicht, Head of WHO Country Office, World Health Organization Regional Office for Europe The Kyrgyz Republic--Melitta Jakab, WHO Health Policy Advisor, the Kyrgyz Republic; and Elina Manjieva, WHO and Center for Health Systems Develop- ment, the Kyrgyz Republic Sri Lanka--Ravi P. Rannan-Eliya and Lankani Sikurajapathy, Institute for Health Policy, Colombo, Sri Lanka Thailand--Suwit Wibulpolprasert and Suriwan Thaiprayoon, Ministry of Health, Thailand Tunisia--Chokri Arfa, National Institute of Labor and Social Studies, Tunisia; and Hédi Achouri, Director General of Public Health Facilities, Ministry of Public Health, Tunisia Vietnam--Björn Ekman, Health Economics Program, Lund University, Sweden; and Sarah Bales, Health Policy Unit (HPU), Ministry of Health, Vietnam Steering Committee Paul Gertler (Former Chief Economist, Human Development Network, Professor of Economics, Haas School of Business, University of California, Berkley) Robert Holzmann (Director, HDNSP) Cristian Baeza (Lead Health Policy Specialist, LCSHD) Kei Kawabata (Manager of Social Sector, IDB) Akiko Maeda (Sector Manager, MNSHD) Eduard Bos (Lead Population Specialist, HDNHE) Alexander S. Preker (Lead Health Economist, AFTH2) Peter Berman (Lead Health Economist, SASHD) Adam Wagstaff (Lead Economist, ECSHD) Mead Over (Lead Economist, ECSHD) Abdo Yazbeck (Lead Health Economist, WBIHD) Logan Brenzel (Senior Health Specialist, HDNHE) John Langenbrunner (Lead Health Economist, EASHD) Paolo Belli (Senior Health Economist, SASHD) Mark Charles Dorfman (Senior Economist, HDNSP) Executive Summary This volume focuses on nine countries that have completed, or are well along in the process of carrying out, major health financing reforms. These countries have significantly expanded their people's health care coverage or maintained such cov- erage after prolonged political or economic shocks (e.g., following the collapse of the Soviet Union). In doing so, this report seeks to expand the evidence base on "good performance" in health financing reforms in low- and middle-income countries. The countries chosen for the study were Chile, Colombia, Costa Rica, Estonia, the Kyrgyz Republic, Sri Lanka, Thailand, Tunisia, and Vietnam. With health at the center of global development policy on humanitarian as well as economic and health security grounds, the international community and devel- oping countries are closely focused on scaling up health systems to meet the Mil- lennium Development Goals (MDGs), improving financial protection, and ensuring long-term financing to sustain these gains. With the scaling up of aid, both donors and countries have come to realize that money alone cannot buy health gains or prevent impoverishment due to catastrophic medical bills. This realization has sent policy makers looking for reliable evidence about what works and what does not, but they have found little to guide their search. In 2004, the Center for Global Development published a seminal report, Mil- lions Saved: Proven Successes in Global Health, which documented 17 public health interventions of "proven" effectiveness (Levine and Kinder 2004). A major contri- bution of the report was an explicit specification of criteria for defining "what works." The report was acclaimed as an important synthesis of the global evidence base on public health interventions. Concomitantly, the global community and individual donors are looking closely at "results-based" aid, holistic approaches to health systems, and long-term sustainable financing. These concerns have triggered a major debate about the need to define "results" and assemble appropriate data for decision making, mon- itoring, and evaluation. Central to this endeavor is the need for effective, evidence- based national policies. Seminal work in areas including human resources for health, health systems, and health financing has shown that the critical evidence bases are often lacking (Joint Learning Initiative 2004; Gottret and Schieber 2006; Levine and Kinder 2004). xix xx Executive Summary This study attempts to partially fill this void with assessments of reforms in low- and middle-income countries that have demonstrated good performance in expanding their people's health insurance coverage to improve their health status and protect them against catastrophic medical expenses. Coverage expansions are measured, to the extent possible, by the number of individuals covered, while financial protection is measured as the share of out-of-pocket payments in overall health spending or household consumption. Besides broadening coverage, these countries have achieved average or better-than-average population health out- comes--as measured, for example, by infant mortality rates, life expectancy, and maternal mortality rates--in relation to the resources devoted to health and their income and educational levels. Among the few low- and middle-income countries that have almost achieved universal coverage and effective financial protection, nine--Chile, Colombia, Costa Rica, Estonia, the Kyrgyz Republic, Sri Lanka, Thailand, Tunisia, and Vietnam-- were selected as examples of good performance by an expert steering committee representing all six World Bank Regions. Each country case is analyzed on a stan- dardized taxonomy that captures most of the key health and nonhealth sector­ specific factors affecting the performance of its health financing system. Because this study relies only on post-reform information, experiments with control groups or randomization are not possible. Thus, no attempt is made to attribute causality. In the discussion sections, some attempt is made to identify common enabling factors evident in most of the cases. However, as is noted subsequently, this is done more to provoke debate and generate hypotheses than to establish causality. Of the selected cases, two countries--the Kyrgyz Republic and Vietnam--are classified by the World Bank as low income (i.e., with 2005 per capita GNI below US$875). Four countries--Costa Rica, Sri Lanka, Thailand, and Tunisia--are middle income (GNI of US$876 to US$3,465). Three countries--Chile, Colombia, and Estonia--are classified as upper middle income (GNI greater than US$3,466). All nine countries have immunization coverage rates above 88 percent. All but the Kyrgyz Republic have a life expectancy at birth above 70 years, and infant mortal- ity rates below 25 per 1,000 live births. Most of them do well in terms of providing financial protection against high or catastrophic out-of-pocket payments. How- ever, four of the countries--the three poorest (Kyrgyz Republic, Vietnam, Sri Lanka1) and Tunisia--perform less well, with out-of-pocket payments higher than 40 percent of all health spending. Trying to generalize "enabling conditions" for good performance in health financing and coverage expansions from these nine case studies is difficult, partic- ularly because these cases are hardly representative of all global experience and are not a random sample. To help contextualize and validate the findings, countries would have to be examined that do not have the same "conditions" and have demonstrated poor performance. Nevertheless, the cases selected do represent Executive Summary xxi countries that have undertaken serious reforms and appear to have done well, based on the definitions of good performance used in this volume. The key question is: as health reforms are very country-specific, can a common set of enabling conditions be found in these nine good performers, or is the con- clusion simply that everything is so country-specific as to preclude generaliza- tions? The paths taken by all these countries are clearly different and heavily con- tingent on each country's political economy and institutional arrangements. However, an analysis of expenditure, outcome, and demographic performance for these countries against global trends and the detailed case studies of each country reveals a number of general "enabling" conditions common to almost all of them. General "Enabling" Conditions In most of the country cases several common institutional, societal, policy, and implementation characteristics are apparent. They are classified as follows: · Economic, institutional, and societal factors: Strong and sustained economic growth; long-term political stability and sustained political commitment; a strong institutional and policy environment; and a well-educated population. · Policy factors: financial resources committed to health, including private financing; commitment to equity and solidarity; health coverage and financing mandates; consolidation of risk pools; recognized limits to decentralization; and focus on primary care. · Implementation factors: carefully sequenced health service delivery and provider payment reforms; good information systems and evidence-based decision making; strong stakeholder support; efficiency gains and copayments used as financing mechanisms; and flexibility and mid-course corrections. These results, it might be argued, could have been posited without a study. However, the findings are based on a standardized definition of good performance in terms of expanded coverage and financial protection in countries that have achieved good health outcomes with average or below average spending and for most with little or no development assistance for health. All the factors identified above make sense on conceptual grounds and would be likely hypotheses for any rigorous evaluation of the impacts of health financing changes. These findings are interesting, but their more important contribution is the evidence they present that the global community must do a much better job in defining "good performance" and evaluating and disseminating the global evi- dence base. To do this, at least five important actions need to be taken. First, what good per- formance means in health financing must be rigorously defined. Second, standard- ized and appropriate qualitative and quantitative information must be collected. Third, health system characteristics must be described in enough detail so that the critical components and their interactions can be identified and assessed. xxii Executive Summary Fourth, rigorous evaluations must be undertaken. Fifth, these evaluations-- conducted in a policy-relevant and user-friendly manner--must be disseminated to all stakeholders. In all these areas, the global community has done a poor job. With the growing importance of this issue, it is incumbent on the global community to take charge and move this agenda forward. It is hoped that global stakeholders, through vari- ous aid-effectiveness forums, as well as the Group of Eight and Paris Declaration Process, will rise to the occasion and put this desperately needed work on track. Acronyms and Abbreviations ADB Asian Development Bank ARI acute respiratory infection BMI body mass index CCT conditional cash transfer CPI Consumer Price Index CPIA Country Policy Institutional Assessment Index, World Bank DALY disability-adjusted life years DDD defined daily dose DOTS directly observed treatment shortcourse DPT diphtheria, pertussis, tetanus DRG diagnosis-related group EAP East Asia and Pacific Region, World Bank ECA Europe and Central Asia Region, World Bank EPI expanded program of immunization EU European Union FDI foreign direct investment FSU Former Soviet Union G-8 Group of Eight GBD global burden of disease GFATM Global Fund to Fight AIDS, Tuberculosis and Malaria GHWA Global Health Workforce Alliance GNI gross national income HIPC highly indebted poor country HMN Health Metrics Network HNP Health, Nutrition, and Population HRH human resources for health IDB Inter-American Development Bank IEC information, education, communication IHP+ Scaling up for Better Health IMR infant mortality rate LAC Latin America and Caribbean Region, World Bank LIC low-income country xxiii xxiv Acronyms and Abbreviations MCH Maternal and Child Health MDGs Milennium Development Goals, United Nations MENA Middle East and North Africa Region, World Bank MHI mandatory health insurance MHIF Mandatory Health Insurance Fund MMR maternal mortality rate MOF Ministry of Finance MOH Ministry of Health NCD noncommunicable disease NGO nongovernmental organization NHA National Health Accounts NHS National Health Service ODA overseas development assistance OOP out of pocket PHC primary health care PPP purchasing power parity SHI social health insurance TFR total fertility rate U-5 children under five years of age UN United Nations UNDP United Nations Development Programme WDI World Development Indicators WHO World Health Organization WTO World Trade Organization Part I Assessing Good Practice in Health Financing Reform Lisa Fleisher, Pablo Gottret, Adam Leive, George J. Schieber, Ajay Tandon, and Hugh R. Waters 1 Introduction Many developing countries have recently undertaken ambitious health reforms to improve resource mobilization for health care. Their goals are universal health care coverage for their people and financial protection against impoverishment due to the costs of catastrophic illness. However, few low-income countries (LICs) or middle-income countries (MICs) have achieved universal coverage. The expe- riences of countries that have done so, though of great relevance to other low- and middle-income countries on the verge of similar reforms, have not been con- sistently evaluated and documented. To partially address this gap, the World Bank has prepared this report documenting experiences in financing significant expan- sions of health care coverage in low- and middle-income countries. The nine "good practice" countries selected as case studies are: Chile, Colombia, Costa Rica, Estonia, the Kyrgyz Republic, Sri Lanka, Thailand, Tunisia, and Vietnam. Building the evidence base on health financing reforms is particularly impor- tant now, with health at the center stage of global development policy on human- itarian as well as global economic and health security grounds. The international community and developing countries themselves are increasingly focused on scal- ing up health systems to meet the United Nations' Millennium Development Goals (MDGs) and on providing financial protection through various risk- pooling mechanisms. Three of the MDGs directly pertain to health outcomes, but the first goal of halving extreme poverty is also relevant to this study, considering that out-of-pocket health expenditures are a major cause of impoverishment. With the significant scaling up of aid, both donors and recipient countries have realized that money alone cannot buy results, and that development effectiveness matters. This realization has sent policy makers looking for reliable evidence about what works and what does not, but they have found little to guide them. The World Bank's new Health, Nutrition, and Population (HNP) strategy decries the lack of available evidence to assess whether billions of dollars of HNP lending over the previous decade have led to significant results (World Bank 2007b). A similar sentiment is expressed by Savedoff and Levine (2006) who argue that part 3 4 Good Practice in Health Financing of the reason for the lack of evidence is due to bureaucratic disincentives to con- duct evaluations of policies and programs. In 2004, the Center for Global Development published a seminal report on "Proven Successes in Global Health," which highlighted some 17 public health interventions of "proven" effectiveness (Levine and Kinder 2004). A major contri- bution of the report was an explicit specification of criteria for defining "what works," including the important element of large-scale implementation. The report was acclaimed as an important synthesis of the global evidence base on public health interventions.1 Concomitantly, the global community and individ- ual donors have been focusing closely on the need for results-based aid, financing, and lending. These concerns have triggered a major debate and actions with respect to the need to define "results" and assemble appropriate data for decision making, mon- itoring, and evaluation. The recent creation of the Health Metrics Network (HMN) provides one example of global action to ensure availability of appropri- ate health information. Central to this venture is countries' need for knowledge about the global evidence base to implement effective, evidence-based policies. Seminal work in a number of areas including human resources for health, health systems, and health financing has shown that the critical evidence bases are often lacking (Joint Learning Initiative 2004; Gottret and Schieber 2006; Levine and Kinder 2004). For example, where is the comprehensive compilation of "success- ful" health financing reforms in developing countries? The lack of such evidence-based policy collations reflects the complexities of the health sector, country and donor priorities, and the inherent difficulty of doing rigorous evaluations in the social sciences. Measuring health outcomes, other than sentinel events such as death, is methodologically difficult and complex, and col- lecting data is costly. Numerous health-related and non-health-related factors affect an individual's health status. Individual behavior is an important determi- nant of health outcomes--and difficult to measure and change (Schieber, Fleisher, and Gottret 2006). Figure 1.1 highlights the many different confounding factors that interact in the health-sector milieu. These complexities are exacerbated by the enormous number of interactive factors affecting most health policies. In developing countries, lacking sufficient resources to meet many basic and often competing needs, it is not surprising that rigorous policy evaluation has not been a high priority. Similarly, in light of these needs, donors have focused their efforts on getting money out the door to assist countries dealing with catastrophic and unpredictable situations such as a tsunami or an avian flu outbreak. As a result, much of what passes for "evidence" is anecdotal and small scale: a box in a report indicating that in a particular health center or district a particular policy resulted in a particular outcome. Few rigorous evaluations are national in scope.2 In the health financing area, remarkably few single-country or multicountry studies have defined what is meant by a successful reform. Nor have they described the countries in question in a standardized way and with sufficient detail so that Introduction 5 Figure 1.1 Determinants of Health, Nutrition, and Population Outcomes Achieving change in HNP behavior of individuals / households income education water performance of health system health status outcomes sanitation · clinical effectiveness · fertility nutrition · accessibility and equity · mortality · quality and consumer satisfaction · morbidity · economic efficiency · nutritional status macroeconomic health care system environment delivery structure institutional capacity · facilities (public and private) · regulatory and legal framework · staff (public and private) · expenditure and finance · information, education, and · planning and budgeting systems communication · client and service information / accountability · incentives governance projects and policy advice Source: Wagstaff, Yazbeck, and Claeson 2004. readers can understand the interplay of the many key factors, determine causality based on rigorous evaluation methods, and provide generalizable lessons for other countries. The present study attempts to partially fill this void; however, it does not attempt to rigorously define "success," because failures are not defined or exam- ined. Nor does it attempt to attribute causality in terms of what specific factors and interventions directly led to specific reform outcomes. Such comprehensive evalua- tions to determine successful reforms are not attempted for several reasons. The opportunity for randomized control trials to determine which policy reforms are best is rarely an option. Moreover, even good pre- and post reform evaluations are hampered by the lack of adequate baseline and follow-up data, especially since many reforms extend back for decades. The challenges involved in measuring inputs that are essential to such evaluations at the country level is underscored by the current debate over how to comprehensively institutionalize National Health Accounts (NHA) to monitor health spending patterns. In addition, measuring the breadth and depth of coverage and financial protection of the population is often 6 Good Practice in Health Financing difficult for lack of the necessary microdata. All of these conditions make system- atic definition and evaluation of successful health reforms exceedingly problematic. Instead, this study examines countries that have exhibited "good performance," defined in terms of coverage expansions, health outcomes, and financial pro- tection compared with other countries with similar health spending and income levels. Financing reforms are discussed in terms of the three principal health financing functions of revenue collection, risk pooling, and purchasing, which, by assumption if effectively performed, will help achieve those objectives. The study then attempts to assess common enabling conditions for good performance. However, definitively identifying enabling conditions would require at least exam- ining countries without such conditions, as well as those that have performed poorly with regard to the above measures, an evidence base that does not exist. As discussed in chapter 3, nine country cases were selected by a steering com- mittee from all six World Bank Regions. Good performers were defined as coun- tries that have significantly increased the number of individuals covered, the depth of coverage in terms of the richness of the benefits package, and the level of finan- cial protection. For countries that had universal coverage but have faced drastic political or economic shocks (e.g., former Soviet Union [FSU] countries), good performance is a reform that resurrects health sector revenues and expenditures and restores a benefits package to precrisis coverage and financial protection. Good performance is also implicitly based on the panel's normative judgment and on data demonstrating that these countries have attained average or better-than- average health outcomes (infant mortality, life expectancy, and maternal mortal- ity) for their health spending and income levels. Although there are many other inputs to producing health outcomes, and health status changes cannot be attrib- uted to health coverage expansions alone, such expansions are likely to have con- tributed to improvements in health status and financial protection. Because this study relies only on existing postreform, secondary data in the nine countries, no experiments with control groups or randomization can be done. Therefore, no attempt is made to attribute causality. Moreover, due to data limitations, comparing pre- and post reform situations is limited. Each country case is described in detail, based on a standardized taxonomy that captures most key health- and non-health-sector specific factors affecting the performance of a health financing reform. Each country chapter includes a description of impor- tant background information and is organized so as to systematically address a core set of issues. However the emphasis given to the different points may vary across cases depending on the nuances of that case. Focal Points of the Case Studies To facilitate understanding of the country-specific environments, detailed contex- tual information on underlying socioeconomic, institutional, political, and health sector characteristics are provided for each of the country cases. These include Introduction 7 · The country's economic, institutional, social, and political environments. The implications of individual country's institutional structures cannot be understated and should be carefully described as part of the analysis and lessons learned. · Descriptions, chronologies, assessments of reform motivations, and analysis of reforms to expand health care coverage. · Overview of the health care delivery system. The breadth and depth of the benefits package. · Financial protection and access to services for chronic and catastrophic conditions. · Revenue sources. · Health indicators, outcomes, and their distribution. · Spending effectiveness. · Financial sustainability of the reforms. · Other key issues. These potentially include: how risk adjustment and cross-subsi- dization are handled; regulation of the insurance schemes; policies for opting out of insurance schemes; need for government subsidies; financing compared with depth of package (focus on "best buy"); impact on labor markets; other links between political economy and the health sector; cost-sharing mechanisms; and sustainability of financing. · Scaling up the reforms. · Key conditions for good performance. · Lessons for other countries. In classifying countries by income, this study uses current World Bank definitions, based on 2005 estimates of gross national income (GNI) per capita (World Bank 2007a): low-income US$875 per capita or less; lower-middle- income, US$876 to US$3,465; upper-middle-income, US$3,466 to US$10,725; and high-income, US$10,726 or more. The report is organized into two parts. Part I describes key issues in expanding health care coverage in low- and middle-income countries, defines what is meant by "good" performance in expansions of coverage and financial protection, dis- cusses the criteria used to choose the country cases, and summarizes the findings from each of the nine country case studies. The concluding chapter of part I draws generalizable lessons from the country cases and provides some guidance for future efforts to enhance the global evidence base in health financing. The full set of country case studies is presented in part II. Audience Governments and international policy makers, donors, staffs of stakeholder orga- nizations, and health care analysts working in low- and middle-income countries are the target audience for this report. It is intended as a practical guide for deci- sion makers in countries considering health financing and coverage reforms, as 8 Good Practice in Health Financing well as for World Bank operations teams, multilateral development organizations, bilateral aid agencies, private foundations, and other development partners. It is hoped that this modest attempt to assess good performance in major health financing changes will lead to more rigorous evaluations of such changes and important, much needed, systematic, multicountry contributions to the global evidence base. Endnotes 1. A recent update of the report includes three additional cases (Levine 2007). 2. Notable exceptions are the national health policy evaluations in Cambodia and Mex- ico. In Cambodia, different modalities of the health system reform were randomly allocated across districts and evaluated (Bhushan, Keller, and Schwartz 2002). In Mexico, the impact of conditional cash transfer programs was evaluated using a national phase-in controlled randomized design (Gertler 2004). 2 Health Financing Functions The expansion of health financing coverage is a prime focus of this report. To bet- ter understand what this entails, this chapter defines and clarifies some of the key health financing functions and concepts such as revenue generation, risk pooling, and purchasing and their importance to financial protection. Health coverage has at least three separate and interrelated dimensions: (1) the number of people covered by organized (public and private) financing initiatives (breadth of coverage); (2) the extent (number and type) of services cov- ered (depth of coverage); and (3) the resulting impacts on health outcomes and financial protection against high out-of-pocket expenditures. In expanding cover- age to promote health outcomes and financial protection, countries need to 1. Raise sufficient and sustainable revenues efficiently and equitably to provide individuals with a basic package of essential services that both improves health outcomes and provides financial protection against unpredictable catastrophic or impoverishing financial losses caused by illness and injury. 2. Manage these revenues to pool health risks equitably and efficiently so that indi- viduals are provided with "insurance" coverage against unpredictable cata- strophic medical care costs. 3. Ensure the purchase of health services in an allocatively and technically effi- cient manner (Gottret and Schieber 2006; Mossialos et al. 2002). The organization of the case studies in this report focuses on key health financ- ing functions within country health care systems to meet these objectives.1 Revenue collection and risk pooling refer to the accumulation and management of sufficient and sustainable revenues to assure that all individuals have access to an essential package of basic services designed to improve health outcomes. These funds must be "pooled" so that the risk of large, catastrophic medical expendi- tures are borne collectively by all pool members and not by each member individ- ually, thus providing all pool members with "insurance protection." Each pool member's "contribution" is prepaid through premiums, payroll taxes, and/or gen- eral tax payments. 9 10 Good Practice in Health Financing The case studies span the range of generic health financing "models" such as national health services (NHS), mandatory health insurance (MHI) funds, and private health insurance. A simplified categorization of health insurance schemes is used here, including (1) general-tax financing, managed by an NHS or ministry of health (MOH); (2) payroll tax­financed MHI managed by a quasi-public entity; and (3) private sector­based health insurance financed by contributions to private voluntary insurers. Globally, some 100 countries have health financing systems that are predominantly financed from general taxes; another 60 have pay- roll tax­based MHI systems. Only a few countries have predominantly private health insurance financed systems (e.g., the United States). In practice most coun- tries have mixed models. Examples of high-income countries with MHI systems include Germany and Japan. Most systems based on an NHS have compulsory universal coverage financed from general government revenues, with provision also predominantly in the public sector; the United Kingdom and the Scandinavian countries are exam- ples. In this study, Sri Lanka is the only pure NHS system. The others represent largely MHI (e.g., Estonia) or combination models (e.g., Thailand). Chile has ele- ments of both an MHI and a private health insurance approach. Health financing coverage is typically defined in terms of the breadth and depth of coverage as well as the resulting level of financial protection. Financial protection in health is generally taken to broadly imply that households and indi- viduals: (1) obtain health care when needed and are not prevented from doing so by excessive costs; (2) do not incur costs when they do access health care that pre- vent them from obtaining other basic household necessities--including food, education, and shelter; and (3) do not fall into poverty due to excessive medical care costs and lost income resulting from illness. A plethora of empirical evidence exists concerning the economic and social impacts of adverse health shocks and the need for policies to provide everyone, but particularly the poor, with financial protection against such large and unpredictable costs (Wagstaff 2005; van Doorslaer et al. 2005; Baeza and Packard 2006). Financial protection can be measured in a number of ways. For the health sys- tem as a whole, out-of-pocket payment as a percent of total health spending offers a rough estimate of financial protection. Evidence shows that higher levels of out- of-pocket financing in the health system are correlated with greater incidence of catastrophic payments (Xu et al. 2003; van Doorslaer et al. 2005). However, the extent of out-of-pocket financing alone does not give a complete picture because the distribution of out-of-pocket payments among population income groups or the severity of catastrophic spending or the impoverishing effect of out-of-pocket payments on households are also important to assess. More revealing measures of financial protection therefore require individual or household-level analysis. Descriptions of such measures appear in this chapter's annex. Many of these measures are complementary and attempt to account for both the extent and the severity of out-of-pocket payments. Compared with the Health Financing Functions 11 simple aggregate of the out-of-pocket share of total health spending, these mea- sures highlight to various degrees the underlying notion that using health care often represents unanticipated financial shocks that may negatively impact house- hold consumption and welfare. Given data limitations, as the case studies are all based on ex-post evaluations, information on many of these financial protection measures are unavailable for several of the countries. An example of some of these indicators for Tunisia is presented in annex 2A. Another aspect of health financing has to do with the provision of demand-side incentives. A substantial body of evidence suggests that for governments to enhance coverage among the poor and other vulnerable populations, extensive outreach programs and/or demand-side subsidies are often required. Subsidized government health services often benefit primarily the better-off, rather than the poor for whom these services are intended. However, governments can adopt tar- geting measures to increase the proportion of public benefits that flow to the poor or to promote the development of separate, privately funded health care delivery mechanisms that serve the better-off (Gwatkin 2004). Demand-side approaches, such as conditional cash transfers (CCTs), have also been found to motivate the poor to seek necessary and often covered care (Rawlings 2004). CCTs provide financial support to relatively poor families contingent upon certain behaviors related to household welfare--such as sending children to school or bringing them to health centers for regular checkups. Purchasing, the final health financing function, refers to the process by which pooled funds are paid to providers in return for delivering services. Purchasing can be broadly classified as passive--spending according to a budget, for exam- ple--or strategic--continually seeking which reimbursement schemes will maxi- mize health system performance (WHO 2000). The types and mixes of provider payment methods constitute an important part of the purchasing arrangement. The main types of provider payment methods include capitation, fee for service; salary, global budgeting, line-item budgeting, case-based payment, and diagnosis- related groups (DRGs). Each mechanism has important incentives for providing health care, controlling costs, and improving service quality (McGuire 2000; Rice 2006; Ellis and McGuire 1996; Ellis and McGuire 1993; Jack 2005; Preker and Lan- genbrunner 2005). In conclusion, health financing involves providing individuals with a basic package of benefits that is designed to improve health outcomes and ensure finan- cial protection. Financing a basic package of health services is accomplished through revenue collection, pooling of revenue and risk, and purchasing services. Countries need to ensure that these financing mechanisms are efficient, equitable, and sustainable. In this study, a "good performance" in health financing is defined in terms of countries that have significantly increased their people's coverage for an essential package of benefits, with better-than-average health outcomes, reasonable financial protection, and better-than-average health care costs when compared with other countries with similar income. 12 Good Practice in Health Financing Annex 2A Definition of Financial Protection Indicators and Applications in Tunisia The analysis in this section employs the methodology outlined in Wagstaff and van Doorslaer (2003) and van Doorslaer et al. (2007). Definition of Catastrophic Health Spending Indicators The following example illustrates the catastrophic effect of out-of-pocket pay- ments in Tunisia. The data used are from the 2003 World Health Survey, in Tunisia (WHO 2003a). However, that survey has only a limited number of questions on B O X 2 A Measures of Financial Protection in Tunisia There are a number of ways to measure the measures the size of this gap compared with incidence and the severity of out-of-pocket multiples of household income. health payments on households, in terms of The combination of both of these the catastrophic impact (Wagstaff and van measures expresses both the incidence and Doorslaer 2003; van Doorslaer et al. 2007). the intensity of out-of-pocket payments Catastrophic payment headcount is one through the mean catastrophic payment gap measure often used to express the extent of or overshoot.This is the product of the share out-of-pocket payments. It is the number or of the population with catastrophic spending share of households with out-of-pocket pay- and the mean positive gap. ments exceeding some prespecified threshold To reflect normative concerns over the dis- of total, nonfood, or nonsubsistence consump- tribution of catastrophic payments among tion, expenditure, or income. Various thresh- rich and poor households (i.e., it may be more olds for the budget share of out-of-pocket socially acceptable for a rich household to payments are often used (van Doorslaer et al. exceed 40 percent of total spending than for a 2007). Some analysts consider out-of-pocket poor household to do so), both the headcount payments catastrophic when exceeding a 10 and mean positive gap can be weighted by percent threshold of total expenditure (Prad- the complement of their corresponding con- han and Prescott 2002; Wagstaff and van centration index (Wagstaff and van Doorslaer Doorslaer 2003; Ranson 2002). On the other 2003).This effectively uses a weight equal to 2 hand, Xu et al. (2003) label as catastrophic for the poorest individual and 0 for the richest, spending out-of-pocket payments exceeding with weights declining linearly in between. By 40 percent of a household's non-subsistence weighting based on the household's rank in spending, basing subsistence spending on a income distribution, the rank-weighted head- food-based poverty line relative to the survey. count and rank-weighted overshoot or gap A limitation of simply measuring the inci- account for whether it is the poor or rich who dence of such spending is that the statistic generally incur catastrophic spending. Since a does not reflect the severity of out-of-pocket negative concentration index indicates a dis- payments (Wagstaff and van Doorslaer 2003). proportionate distribution among the poor, Mean positive gap or overshoot is used to this will increase the value of the headcount identify how excessive out-of-pocket payments or gap. On the other hand, a positive concen- are. It is the average amount by which the tration index will reduce the associated value threshold is exceeded among those passing and make the headcount or gap look less the threshold. Additionally, a normalized gap severe. Health Financing Functions 13 Figure 2A.1 Payments as Share of Total and Nonfood Expenditure in Tunisia, 2003 1.0 0.9 0.8 0.7 0.6 0.5 expenditure 0.4 % 0.3 0.2 0.1 0 0 0.2 0.4 0.6 0.8 1.0 cumulative proportion of households (ranked in descending order of out-of-pocket share of total expenditure) out-of-pocket as % of nonfood expenditure out-of-pocket as % of total expenditure Source: Authors. household expenditure and does not impute the rental equivalent of housing or durable goods. Data necessary to construct a consumption aggregate, the pre- ferred measure of living standards (Deaton and Zaidi 2002), were not available, and the use of limited expenditure data almost certainly results in measurement error. The results are therefore meant only to be illustrative and should be inter- preted with caution. Data on out-of-pocket payments include expenditures on inpatient care; outpatient care; care by traditional healers or dentists; medication or drugs; health care products such as prescription glasses, hearing aids or pros- thetic devices; laboratory tests; and any other health care payments. The recall period for both total expenditure and out-of-pocket health expenditure is one month. This analysis indicates that out-of-pocket payments have catastrophic effects, at least in the short term. Figure 2A.1 graphically depicts the shares of total and nonfood expenditure per capita composed of out-of-pocket for those households having a positive expenditure (slightly less than half of all households). The inci- dence of these spending levels is much higher for nonfood spending and reflects that food accounts for a large share of many households' expenditure. A small but noticeable share of households appears to spend their entire monthly non-out-of- pocket consumption on food. 14 Good Practice in Health Financing Table 2A.1 Catastrophic Impact of Out-of-Pocket Payments in Threshold Expenditure Shares in Tunisia, 2003 Percent of Percent of total expenditure nonfood expenditure Measure 10 25 40 10 25 40 Catastrophic headcount 33.2% 13.6% 5.9% 49.8% 37.7% 27.2% Concentration index 0.023 ­0.077 ­0.126 0.014 ­0.043 ­0.112 Rank-weighted catastrophic headcount 32.4% 14.7% 6.6% 49.1% 39.3% 30.2% Mean positive overshoot 17.5% 17.6% 17.2% 36.8% 32.0% 27.5% concentration index ­0.047 ­0.091 ­0.071 ­0.072 0.530 0.060 Mean catastrophic payment overshoot 5.8% 2.4% 1.0% 18.3% 12.1% 7.5% Rank-weighted mean catastrophic payment overshoot 6.1% 2.6% 1.1% 19.7% 5.7% 7.0% Source: Authors. More detailed analysis of the catastrophic impact of out-of-pocket payments appears in table 2A.1. The headcount statistics indicate a sizable number of households spending large shares of total and nonfood expenditure on out-of- pocket payments. The severity of these payments, as measured by the overshoots, is also strikingly high. Moreover, above the 25 percent and 40 percent thresholds, the poorer households incur catastrophic spending for both total and nonfood spending. The rank-weighted headcounts, which are accordingly larger than the unweighted headcounts, indicate the increasingly higher concentrations among the poor as the threshold rises. In terms of total expenditure, the severity of out- of-pocket payments is also concentrated among the poor for all thresholds; how- ever, for the two highest thresholds of nonfood spending, the rich account for most of the excessiveness of out-of-pocket payments. In this case, more poor households surpass the thresholds, but the rich spend disproportionately more of their discretionary spending on out-of-pocket payments. Endnote 1. Although classifications of health systems and health financing systems perfor- mance criteria are couched differently by different organizations, all focus on the ultimate results of improvements in health outcomes and financial protection and impose effi- ciency, equity, and sustainable criteria as well. The World Health Organization (WHO) also included health systems' responsiveness to consumers. See for example Murray and Frenk (2000) and Gottret and Schieber (2006: chapter 2). 3 Criteria for Defining "Good Practice" and Choosing Country Cases Evaluating good performance involves a range of different components. The most important criteria for assessing good performance and selecting the country case studies can be classified in two groups, ranked by order of importance. First Tier Criteria 1. Improvements in health care coverage. The principal criterion for selection was a documented effort to increase the proportion of the population with formal health care coverage and to improve the level of financial protection. 2. Applicability and pertinence for other low- and middle-income countries. 3. Large-scale initiatives. The initiatives described and analyzed should be at a national scale, or at least involve significant national policy making. In most cases, the principal form of expansion of health care coverage is through insti- tutional arrangements--formal sector employment, tax-based government affiliation, or market-based private insurance.1 4. Availability of information and data. Information needed to analyze the case includes the measurable key outcomes and indicators. Data availability plays a critical role in whether a "story" can be told about the performance of a coun- try's reform efforts. Household-level data are particularly important for mea- suring coverage, access, and financial protection. However, this is not the most important determinant of a country's inclusion and was not a necessary pre- condition for selection. Second Tier Criteria 1. Health indicators and outcomes. It is extremely difficult to document a causal association between health system and financing characteristics and population-level health outcomes because many other inputs in producing health status, as well as the associated time lags, are also important. However, the ultimate goals of health coverage reforms are to improve health status and 15 16 Good Practice in Health Financing financial protection along with responsiveness (WHO 2000). Accordingly, key population-level outcomes--including life expectancy, infant and child mor- tality, and immunization coverage--were used to assess the long-term success of reforms attempted in a specific country. 2. Relation of expenditures to outcomes. At a national level, the amount of funds devoted to health care reflects a societal commitment to promoting popula- tion health and protecting people against impoverishment from catastrophic medical care costs. However, spending levels by themselves provide little infor- mation about equity, the efficiency of spending, or the extent of financial pro- tection for individuals. While many developing countries face severe resource constraints in mobilizing resources, and many low-income countries (LICs) cannot raise enough revenues to meet the Millennium Development Goals (MDGs), value for money in terms of health outcomes (i.e., allocative effi- ciency) and production costs (i.e., technical efficiency) are critical issues in all countries. Thus, it has become increasingly recognized that more money alone will not buy improved health care outcomes. Nevertheless, a direct comparison of health spending per capita across comparable income countries with health outcomes--while confounded by factors well outside the reach of the health system and often difficult to measure accurately--provides a crude compara- tive benchmark of how effectively a country spends its health care resources vis-à-vis its health outcomes. One critical aspect of this study is the definition used to determine good perfor- mance in terms of coverage expansions and financial protection. Breadth of cover- age is readily measured by the number of individuals formally eligible to receive benefits from a particular "insurance" (i.e., national health service [NHS], manda- tory health insurance [MHI], or private health insurance [PVHI]) mechanism. Measuring the depth of coverage--the type and number of services covered--is problematic due to a lack of information on the number of users. In practice, depth of coverage is often defined and measured as the actuarial value of the benefits package per enrollee (and can be approximated by dividing spending less user fees by the number of users). Even program eligibility is not a straightforward concept to measure because individuals may be eligible for benefits in a public system but not formally enroll.2 Even if people do enroll and become program beneficiaries, they still may not be able to access services due to both demand-side barriers (e.g., cultural or educational) and supply-side constraints (e.g., no providers in rural areas, private providers excluded from reimbursement by public programs). More- over, putting aside the difficulty of assessing supply- and demand-side access barri- ers, simply measuring the number of people entitled to benefits is no easy task. Even in developed countries the number of family members eligible for coverage through the household head's eligibility may be unknown, not to mention how many actually enroll and of these how many receive covered services. All of these factors determine whether a coverage expansion achieves its health outcome and financial protection objectives. Criteria for Defining "Good Practice" and Choosing Country Cases 17 For the purposes of this study, · Breadth of coverage is measured by the number of people with formal coverage. · Depth of coverage is not measured separately. · Financial protection is measured in terms of out-of pocket payments as a per- cent of total health spending or out-of-pocket spending as a share of house- hold consumption, when available. In principle, out-of-pocket payment shares reflect both depth and breadth of coverage and/or lack thereof. · Infant mortality, life expectancy, and maternal mortality (health outcome mea- sures); expenditures per capita and as a share of GDP (a crude measure of macroefficiency); and revenue to GDP and public spending on health relative to the total government budget (crude sustainability indicators) are assessed through international comparisons and relations to global averages for compa- rable income countries. While good performance is measured in terms of expansions in coverage and financial protection in the context of good health outcomes and average or below overall spending levels, it was decided that, to qualify as good performers, coun- tries did not have to attain each dimension of the first tier criteria outlined above. However, for a given country, the reforms implemented would constitute good performance overall, upon meeting the first tier criteria of achieving improve- ments in coverage and financial protection while also exhibiting reasonable (aver- age or lower) overall health spending levels and average or better-than-average health outcomes for the amounts spent and/or their income levels. For countries that previously had universal coverage but encountered drastic political or eco- nomic shocks (e.g., former Soviet Union countries), good performance in a reform also resurrects health sector revenues and expenditures and restores a sat- isfactory benefits package of pre-economic crisis dimensions. Based on these criteria, the study steering committee recommended the nine case study countries, a fairly heterogeneous group of low-, middle-, and upper- income countries. Tables 3.1 and 3.2 provide information for these nine countries on 15 measures of health system characteristics. Some of these statistics may differ from those provided in the individual country case studies because some of the case studies use data from different sources and years. In order to ensure compa- rability across countries, tables 3.1 and 3.2 provide data from the same source for the same years for all countries. Performance of Country Cases Globally Many inputs go into producing health outcomes. These may include income as well as factors directly related to a health system, including health spending and the supply of doctors and hospital beds. However, many other factors, such as education, infrastructure, and geography, also play critical roles. For instance, Filmer and Pritchett (1999) find that 95 percent of the variation in child mortality across countries can be explained by just five factors: income per capita, female 18 Good Table 3.1 Income and Health Spending, 2004 Pract Total Total Government Government OOP Household ice health health health health health OOP GDP Total spending spending spending spending spending payments in per health per per (% total (% general (% total (% of total H capita spending capita capita health government health household ealth Country (US$)a (% GDP) (US$) (PPP) spending) expenditures) spending) consumption) Financ Chile 5,894 6.1 359 720 47.0b 13.1 24.3 Colombia 2,155 7.8 168 570 86.0 20.9 6.9 ing Costa Rica 4,349 6.6 290 592 77.0 21.3 20.4 Estonia 8,328 5.3 463 752 76.0 11.5 21.3 10.6c Kyrgyz Republic 434 5.6 24 102 40.9 8.4 55.7 2.4d Sri Lanka 1,033 4.3 43 163 45.6 8.4 45.7 2.1d 6.3c Thailand 2,539 3.5 88 293 64.7 11.2 26.4 1.7d Tunisia 2,832 6.2 175 502 52.1 8.8 39.8 10.4c Vietnam 550 5.5 30 184 27.1 5.0 64.2 5.5d 6.8c Source: World Bank World Development Indicators; WHO 2007. Note: OOP = out-of-pocket. a. GDP per capita in current U.S. dollars. b. A large share of health expenditure in Chile counted by the WHO as private, based on national health accounts definitions, could potentially be considered public instead since coverage of formal sector workers is mandatory although many are covered by private insurers (ISAPREs). c. Source of household OOP payments and consumption is WHO 2003b. d. Source of household OOP payments and consumption is van Doorslaer et al. 2007. Criteria for Defining "Good Practice" and Choosing Country Cases 19 Table 3.2 Health Outcome and Delivery Indicators, 2004 Infant Immunization, Immunization, Self-perceived mortality DPT measles unmet need rate (per Life (% of (% of (% not receiving 1,000 expectancy Doctors/ Beds/ children ages children ages health care Country live births) (years) 1,000 1,000 12­23 months) 12­23 months) when needed)a Chile 8 78 1.09 2.5 94 95 n.a. Colombia 18 73 1.35 1.2 89 92 n.a. Costa Rica 11 79 1.32 1.4 90 88 n.a. Estonia 6 73 4.48 5.8 94 96 2.4 Kyrgyz Republic 58 68 2.51 5.3 99 99 n.a. Sri Lanka 12 75 0.55 2.9 97 96 0.4 Thailand 18 71 0.37 2.2 98 96 n.a. Tunisia 21 73 1.34 2.1 97 95 1.6 Vietnam 17 71 0.53 2.3 96 97 1.0 Sources: World Bank World Development Indicators; WHO 2007. Note: n.a. = not available; DPT = diptheria, pertussis, tetanus. a. WHO 2003b. educational attainment, extent of ethnic fractionalization, level of income inequality, and predominant religion. The importance of nonhealth systems fac- tors for attaining health outcomes has been documented by others.3 Nevertheless, it is instructive to situate the nine case countries by examining some of their health system indicators in a global context. Virtually all of the countries documented in these case studies have achieved remarkable health outcomes--life expectancy, infant mortality, and access to essential services--despite low or moderate levels of income and total health spending (figure 3.1).4 All but the Kyrgyz Republic have a life expectancy at birth above 70 years, and infant mortality rates below 25 per 1,000 live births. And, with the exception of Estonia for life expectancy and Tunisia for infant mortality, all case countries have above average outcomes for their income and health spending levels. Vietnam and Sri Lanka, in particular, are near the top of the global league. Additionally, all nine countries perform well with respect to indicators for health services delivery (figure 3.2). Each has immunization coverage rates of at least 88 percent; most are well above 90 percent. Almost all perform better than average against comparators with regard to the percentage of births attended by skilled health personnel. As mentioned above, all nine countries generally attain these positive health outcome and health service delivery outcomes at low levels of total health spend- ing; in per capita terms, all countries spend about average amounts on health for their income level with the exception of Colombia, which spends slightly higher 20 Good Practice in Health Financing Figure 3.1 Population Health Indicators Relative to Income and Spending a. Life expectancy b. Infant mortality average average above Kyrgyz above Republic Vietnam Sri Lanka income Colombia Vietnam to Colombia Sri Lanka income Kyrgyz Estonia Tunisia Costa Rica to Republic Chile Chile Costa Rica Thailand relative Thailand relative Estonia Tunisia attainment attainment average average below below below average above average below average above average attainment relative to health expenditure per capita attainment relative to health expenditure per capita c. Maternal mortality rate vs. total health spending per capita 2,000 births) live 100,000 (per 1,000 rate morality 500 Thailand Costa Rica Tunisia 250 Colombia Vietnam Chile Estonia maternal 50 Kyrgyz Republic Sri Lanka 10 100 250 1,000 10,000 25,000 total health spending per capita (current US$, log scale) Sources: World Bank 2007a; WHO 2007. than average. Thailand and Sri Lanka spend slightly below average for their income level (figure 3.3). For five of the nine case countries, health spending as a share of GDP is close to the average for their income levels; however, Estonia, Sri Lanka, and Thailand spend markedly less than their average share of GDP for their income, while Colombia spends far more (figure 3.4). Criteria for Defining "Good Practice" and Choosing Country Cases 21 Figure 3.2 Health Service Delivery Indicators Relative to Income and Spending a. Births attended b. Measles immunization rate by skilled health personnel (12­23 months) average average above above income Vietnam Kyrgyz Republic income Thailand Kyrgyz Republic to to Sri Lanka Vietnam Colombia Sri Lanka Tunisia Thailand Tunisia Chile Estonia Colombia Costa Rica Chile relative Estonia relative Costa Rica average average attainment attainment below below below average above average below average above average attainment relative to health expenditure per capita attainment relative to health expenditure per capita Sources: World Bank 2007a; WHO 2007. Figure 3.3 Total Health Spending Relative to Income (US$) 1,000 Estonia capita Tunisia Chile per Colombia Costa Rica 100 Vietnam Thailand spending 34 Kyrgyz Republic Sri Lanka health 10 total 100 250 1,000 10,000 25,000 GDP per capita (current US$, log scale) Sources: World Bank 2007a; WHO 2007. 22 Good Practice in Health Financing Figure 3.4 Health Spending as Share of GDP and per Capita vs. Income 15 GDP) (% 10 Kyrgyz Republic Vietnam Colombia Costa Rica spending Tunisia Chile Estonia health 5 total Sri Lanka Thailand 0 100 250 1,000 10,000 25,000 GDP per capita (current US$, log scale) Sources: World Bank 2007a; WHO 2007. Revenue-raising capacity also varies among the nine countries. Chile, Costa Rica, Sri Lanka, and Thailand raise less revenue than the average for their income level, while the remaining five countries each raises more than average (figure 3.5). Revenue-raising capacity, which tends to increase with income, is important because it provides governments with fiscal space that can be used for health spending. Similarly, the public-private financing mix varies widely across the nine coun- tries (figure 3.6). Colombia, Costa Rica, Estonia, and Thailand each have govern- ment shares of total health spending well above the average for their income level.5 Regarding financial protection, the nine countries are also mixed in terms of the shares of out-of-pocket payments comprising total health spending (figure 3.7). Five countries have average or below average levels. However, four of the countries--the three poorest, Kyrgyz Republic, Sri Lanka, Vietnam, as well as Tunisia--have relatively high out-of-pocket payments, at least 40 percent or more as a share of overall health spending.6 This also reflects the difficulty poorer coun- tries may have in being able to afford both a package of essential services and financial protection because some 70 percent of health spending is out of pocket in poor countries on average. Moreover, three of the countries with high out-of-pocket levels--Kyrgyz Republic, Tunisia, and Vietnam are also countries that raise above-average rev- enues. These countries also show lower-than-predicted government expenditure Criteria for Defining "Good Practice" and Choosing Country Cases 23 Figure 3.5 Revenue to GDP Ratio vs. Income 100 GDP) 75 (% 50 revenues Estonia Colombia Tunisia 25 Kyrgyz Republic Vietnam Chile government Costa Rica Thailand Sri Lanka 0 10 100 250 1,000 2,500 10,000 25,000 GDP per capita (current US$, log scale) Sources: World Bank 2007a; WHO 2007. Figure 3.6 Government Share of Health vs. Income a. b. 100 30 budget) spending) Colombia 80 Costa Rica Estonia Colombia Costa Rica health 20 government total (% Thailand total 60 (% Tunisia Chile spending Chile Thailand Sri Lanka Estonia spending 10 Kyrgyz 40 Kyrgyz Republic Republic Tunisia health Sri Lanka health Vietnam Vietnam 20 government government 0 10 100 250 1,000 10,000 25,000 10 100 250 1,000 10,000 25,000 GDP per capita (current US$, log scale) GDP per capita (current US$, log scale) Sources: World Bank 2007a; WHO 2007. 24 Good Practice in Health Financing Figure 3.7 Out-of-Pocket Spending Relative to Income 80 Vietnam 60 Kyrgyz Republic spending spending) Sri Lanka health 40 Tunisia health total Thailand Chile (% 30 Costa Rica Estonia Out-of-pocket Colombia 20 10 100 250 1,000 10,000 25,000 GDP per capita (current US$, log scale) Sources: World Bank 2007a; WHO 2007. shares for their income level. This may reflect the fact that health is a low priority and there are difficulties in terms of political economy in committing govern- ments to cover an essential package of services for their population, despite good efforts in revenue collection. However, out-of-pocket levels are very low in Colombia and Costa Rica, which also have high public health spending shares of the overall government budget (annex figure 3A.3). In these countries, more than 20 percent of the government budget is spent on health. In terms of other health service indicators, there is variation in hospital bed supply around the global average, but physician supply is above average for seven of the nine countries (figure 3.8). The former Soviet Union (FSU) countries of Estonia and the Kyrgyz Republic, despite significant reductions in recent years, still have over twice as many hospital beds and physicians compared with the other countries (table 3.2) and are far above the global averages. Sri Lanka and Vietnam also have high numbers of hospital beds, but nowhere near the levels of some of the FSU countries. Thailand is the only country among the nine cases with a significantly lower physician-to-population ratio than the global average. Perhaps a minimal physician-to-population ratio is a necessary, though not suffi- cient condition, for a successful expansion of health insurance coverage. Adult literacy is another important factor that distinguishes most of these countries from others. With the exception of Tunisia, each country has above- average rates of both total adult and female adult literacy compared with coun- tries of similar income levels (figure 3.9). This is important since the evidence of a Criteria for Defining "Good Practice" and Choosing Country Cases 25 Figure 3.8 Hospital Bed and Physician Capacity vs. Income a. Hospital bed supply b. Doctor supply 15 8 6 people 10 people 1,000 Estonia per 1,000 6 per beds Estonia 5 Kyrgyz Republic Kyrgyz Republic doctors hospital 2 Sri Lanka Tunisia Vietnam Chile Colombia Costa Rica Thailand Chile Tunisia Sri Lanka Costa Rica Colombia Vietnam Thailand 0 0 10 100 250 1,000 2,500 10,000 25,000 10 100 250 1,000 2,500 10,000 25,000 GDP per capita (current US$, log scale) GDP per capita (current US$, log scale) Sources: World Bank 2007a; WHO 2007. Figure 3.9 Literacy vs. Income a. Adult literacy b. Adult female literacy Kyrgyz Costa Rica Kyrgyz Costa Rica 100 Estonia Republic Colombia 100 Estonia Republic Colombia Vietnam Chile Sri Lanka Chile Sri Lanka Thailand Vietnam above) Thailand 75 75 Tunisia above) and Tunisia and 15 15 50 age 50 age females people (% (% rate rate 25 25 literacy literacy adult female adult 10 100 250 1,000 2,500 10,000 25,000 10 100 250 1,000 2,500 10,000 25,000 GDP per capita (current US$, log scale) GDP per capita (current US$, log scale) Sources: World Bank 2007a; WHO 2007. 26 Good Practice in Health Financing Table 3.3 Correlations between Population Health Outcomes and Income, Health Spending, and Literacy Levels Indicator Life expectancy Infant mortality Maternal mortality Adult female literacy rate 0.65 ­0.75 ­0.77 Adult literacy rate 0.64 ­0.71 ­0.73 GDP per capita 0.53 ­0.50 ­0.41 Health spending per capita 0.48 ­0.45 ­0.36 Public health spending per capita (US$) 0.47 ­0.43 ­0.34 Total health spending (% GDP) 0.30 ­0.34 ­0.29 Out-of-pocket/total health spending ­0.24 0.38 0.28 Sources: World Bank 2007a; WHO 2007; UNDP 2007. strong positive link between higher education and better health outcomes is well documented (Cutler and Lleras-Muney 2006). For the aggregate outcome and spending data used in the analysis in this chap- ter, the average educational attainment in the population is more highly corre- lated with health outcomes than other important factors, such as income and health spending. Endnotes 1. This focus omits countries with significant community health insurance develop- ment. However, these experiences have been well documented elsewhere (Preker and Car- rin 2003; Preker et al. 2000, 2002; Gottret and Schieber 2006). 2. For example, many poor people eligible for a health card, say in Indonesia or for the Medicaid program in the United States, do not bother to register). 3. See, for instance, Cutler and Lleras-Muney (2006); Bokhari, Gai, and Gottret (2007); and Soares (2007). 4. These graphs plot the residuals for each country from two regressions: the logarithm of the health outcome is regressed first on the logarithm of GDP per capita alone and then on the logarithm of health spending per capita alone. The scatter plot of the residuals from these regressions is one concise way to assess a country's performance on the outcome compared to the global average for a given income level and total health spending per capita. See ADB (2006). The distribution of residuals lies primarily in quadrants I and III because income and health spending per capita are closely correlated (0.95) and also posi- tively correlated with the health and delivery outcomes measured. The analysis contains observations for 178 countries for which data were available. This includes 51 low-income countries, 56 lower-middle-income countries, 36 upper-middle-income countries, and 35 high-income countries. 5. As income rises, there is a tendency for the public share of health expenditure to rise. This is often attributed to changes in relative prices and to changes in societal preferences as incomes rise. 6. In Sri Lanka out-of-pocket expenditure tends to fall on the rich, which poses less of a problem from an equity perspective. 4 Summaries of Country Cases The detailed reform experiences and lessons to be learned from each of these "good practice" cases are discussed in detail for each country in part II of this study.1 This chapter summarizes the main reform elements and lessons from each of the countries. Chapter 5, the concluding chapter of part I of this report, attempts to assess the generalizable lessons from these country cases. Chile Chile is an upper-middle-income country in the Latin America and Caribbean Region (LAC) with an average-sized population (16.3 million). Most people reside in urban settings; only 13 percent of the population lives in rural areas. In 2004, per capita GDP was US$5,894. Chile devoted some 6.1 percent of GDP (US$359 per capita) to health in 2004, a little below average for its income level. Some 53 per- cent of this amount is paid by private sources, including 24.3 percent paid directly by households as out-of-pocket expenditures. Health spending accounts for some 13.1 percent of the government's budget, which is above average for its income level. In terms of health outcomes, Chile ranks among the world's highest perform- ers. Life expectancy is 78.2 years, and infant mortality is 7.6 per 1,000 live births. In 2005, the number of physicians and hospital beds per 1,000 people were 1.09 and 2.5, respectively, slightly below average for a country of its income level. In Chile, the expansion of a fiscally sound social security system, including health care coverage, has required considerable economic resources and was made possible by an extended period of economic growth and an efficient tax collection system. Chile is an early reformer relative to many of the countries in this study. Its reforms can be traced back over a half century, culminating in the creation of the national health system. Key Reform Elements and Accomplishments The National Health Service was created in 1948 and originally provided health coverage only to public servants and private formal sector employees. In 1981, 27 28 Good Practice in Health Financing under Pinochet's military government, reforms followed the neoliberal princi- ples of smaller government, public subsidization of the poor, competition, and decentralization. The main structure of the insurance system, created in 1981, remains today. It consists of a public insurer (Fondo Nacional de Salud, FONASA) and private insurers (Instituciones de Salud Previsional, ISAPREs). By law, all formal sector workers who are not self-employed, retired workers with a pension, and self- employed workers with a retirement fund must enroll with the mandatory health insurance (MHI) system by making a monthly contribution equal to 7 percent of the first US$2,000 of their monthly income or pension. The required payroll con- tribution has risen over time, from 4 percent in 1981. Other individuals may enroll as well. They include independent workers, who can voluntarily enroll with FONASA or an ISAPRE conditional on their 7 percent contribution, and legally certified indigent citizens and legally unemployed workers, who are entitled to free coverage by FONASA. ISAPRE beneficiaries may voluntarily make an extra contri- bution to their insurer to purchase additional coverage. However, with the return of democracy in 1990, concerns about equity and underinvestment in government services dominated the policy debate; the legitimacy of the existing health system became an important and contentious issue. As a result, public health spending per capita doubled in real terms within a few years, and efforts to regulate private insurers were strengthened. Today, the system covers over 86 percent of the population, close to 70 percent through FONASA and some 16 percent through ISAPREs. Ninety-six percent of the poorest individuals are covered by FONASA. Most of the population above 50 years of age (even within the highest income quintile) is affiliated with FONASA. A complete subsidy (without copayments) is provided for the poorest. Financing of these subsidies requires a solid domestic tax base for sustainability. FONASA is also internally equitable--41 percent of its beneficiaries are indigent and make no copayments. The share of total health expenditures paid by the public sector through FONASA increased from 37 percent in 1998 to 49 percent in 2003. Health services are delivered by both public and private providers. Private providers sell their services to insurers and private consumers. The average number of services provided per beneficiary has also increased dramatically over the years both in FONASA and the ISAPREs. To contain public health spending, Chile has set bud- getary ceilings for FONASA. As part of the democratic reforms to improve private sector regulation, the Superintendent of Health (Superintendencia de Salud), was created in 1991 to reg- ulate the ISAPRES--and in 2001, the Superintendencia began regulating FONASA as well. More recently, a minimum benefits package for all health plans was man- dated in 2005. A set of new laws, collectively known as Explicit Health Guarantees (Garantías Explicitas de Salud, GES), requires both public and private insurers to provide coverage for 56 defined conditions--beginning with 25 conditions in July Summaries of Country Cases 29 2005 and culminating with all 56 in 2007. This effort has been the most recent con- certed policy effort to improve equity in health provision and financing. The reforms of 2005 and those over the previous half century were similarly motivated: (1) to improve the quality of public care, (2) to expand and consoli- date public provision of health services, (3) to foster health insurance coverage for all citizens, (4) to promote private investment in health infrastructure, (5) to pro- mote private-public competition in insurance, (6) to separate financing from pro- vision in the public sector, and (7) to decentralize public health care services at all levels. Key Lessons Some of the key lessons of the reforms are · Expansion of health care coverage requires sustained financial investments into the health sector. In particular, expanding coverage of social insurance sys- tems--either in terms of the population covered or the depth of benefits pack- age--is likely to increase health care spending substantially, partly due to moral hazard and the need to subsidize the poor and the vulnerable. Therefore, it is easier to expand social security coverage when public finances are adequate to bear these costs. Such systems are usually put in place during periods of robust economic growth. · Increased coverage for low-income groups requires subsidies from government. Subsidies provided to the overall population should also be progressive. · Cost containment measures are important. They can prevent rampant increases in public expenditures due to moral hazard, increased population coverage with subsidies, and the introduction of new technologies. FONASA has copay- ments that increase with the beneficiaries' income, and the public sector has put a hard expenditure ceiling on public sector providers. The budgetary ceil- ing was historically based on the previous year's spending. However, an addi- tional reform in Chile is the elimination of supply-side subsidies in the form of budget support to public sector providers. · Private funding support is essential for a sustainable health system. In Chile, pri- vate financing comes mainly from two sources--copayments, which vary with the services consumed, and monthly contributions, set at 7 percent of the ben- eficiary's income. The average monthly salary in Chile is US$490; countries with lower income levels will likely need to charge beneficiaries higher rates to offer a comparable package of benefits. To achieve sustainability, the required contribution has risen from its starting point of 4 percent in 1981. · Incomplete regulation and supervision in a competitive insurance system can lead to adverse risk selection, high marketing expenses, and difficulties for beneficiaries in selection of health plans. ISAPREs offer more than 10,000 health plans--gen- erating asymmetries of information and making it difficult for beneficiaries to 30 Good Practice in Health Financing make informed choices. The 2005 reform that mandated the same basic bene- fits package for both FONASA and ISAPREs sought to remedy this situation. · The success of the Chilean model is highly dependent on an effective public insur- ance system. If the Chilean social security system were dissolved in favor of a completely free-market system with only private insurers, it is likely that the process of adverse selection would leave only the lowest-risk, highest-income individuals insured. · Chile has benefited from credible, independent, and efficient institutions. The Superintendencia de las ISAPREs regulates the benefits packages, contribution rates, and services provided by the ISAPREs--providing transparent rules and fostering competition among private insurers. An efficient tax collection mechanism and a low rate of tax evasion have also helped to ensure that the necessary resources will be available to finance public spending. · Solid political economy and consistency are important for health reform. The ini- tial health sector reforms in Chile were introduced during a nondemocratic period. After 1990, successive democratic governments introduced reforms to increase investment in the private sector, improve public sector administra- tion, enhance coverage of the poor, tighten regulation and supervision, and promote consumer protection. However, the current system maintains many of the same financing, organizational, and functional characteristics of the model initially established. Moreover, the reforms were introduced only after attaining strong political consensus. Colombia Colombia is a lower-middle-income country located in the northwestern tip of South America. It is the continent's fourth largest country, covering 1,038,700 square kilometers. It has a population of 45.6 million and varied geography. In 2004, per capita GDP was US$2,155. In terms of health, total health expenditures were US$168 per capita in 2004--a relatively high 7.8 percent of GDP. Only 6.9 percent of total health spending comes from out-of-pocket expenditures by households, well below the average of countries with similar income levels. In terms of health outcomes, with life expectancy in Colombia at 73 years and infant mortality at 18 per 1,000 live births, Colombia performs better than other countries of its health spending levels and income category. In terms of its deliv- ery system, Colombia has a relatively high physician supply--1.33 doctors per 1,000 people. However, with 1.2 hospital beds per 1,000 people, Colombia ranks somewhat below average for its income level. Until the mid-1990s, Colombia had one of the most stable economies in Latin America. The economy grew until 1996, when a period of economic recession began. By 1999, unemployment had reached 20.1 percent, and 58.2 percent of the population was below the poverty line. In recent years, the economy has been recovering. If continued, this trend will favor the sustainability of current health insurance expansion policies. Summaries of Country Cases 31 Prior to 1993 Colombia had a three-tiered health care system. One tier con- sisted of the National Health System, with a network of public facilities for health care delivery--financed from national revenues and serving roughly 50 percent of the population. A private health care delivery market grew in parallel to the public network, targeting the population with ability to pay and meeting the demand for services and quality not provided by the public sector. Private providers supplied a large part of ambulatory care services in both urban and rural areas--42 percent and 36 percent, respectively, of all visits. A second tier was a mandated social insurance plan for workers in the public and formal private sectors, financed by employee and employer contributions and covering 20 percent of the population. The third tier consisted of private insur- ance or health care services paid out of pocket by higher-income groups. This market was covered by private commercial health insurers, prepaid group health organizations, and worker cooperative organizations. Under the Colombian health system prior to 1993, access to basic health services was limited for a large proportion of the population. Health expenditures were inequitable, spending was inefficient, and there were shortcomings in the quality of the health services delivered. About 75 percent of the population was uninsured. In 1992, one out of every six individuals in the poorest income quintile who fell ill did not seek medical care because the individual could not afford to pay for it. Accord- ing to national household surveys conducted in 1992, health care expenditures rep- resented 2.4 percent of total household expenditures--but the lowest-income decile spent about 10 percent of its income on health, while the upper-income decile spent less than 0.5 percent. The cost of health care was reported by more than 55 percent of the poor as a major barrier to accessing health care services. Public resources were poorly targeted: 40 percent of the subsidies to public hospitals benefited the wealthiest 5 percent of the population, and allocation of health care resources was biased toward more expensive, curative care, despite evi- dence of low productivity. Allocation of resources followed historical hospital spending trends rather than people's needs. The poor quality of health services was reflected by low use and acceptance of the network of public providers. Key Reform Elements and Accomplishments The health reform proposal generated a heated debate around its goals and means between two coalitions, one supporting a promarket model (government staff in charge of economic policy) and the other favoring a welfare state (unions and bureaucracy of social security agencies and public health sector providers). The main tensions lay between the ideas of solidarity and efficiency and the roles of the public and the private sectors. To enhance the political feasibility of the reform, the government team undertook a consensus-building process before the proposal reached Congress and paired the health reform with a pension reform in a single policy-reform package. The reform was approved in 1993, with enactment of Law 100 (Gonzalez-Rossetti and Bossert 2000). 32 Good Practice in Health Financing For more than a decade, Colombia's health care reform has expanded insurance coverage to 78 percent of the total population. In 1993, Colombia introduced an ambitious health care coverage reform, which created a national social health insurance (NSHI), organized as a model of "managed competition." The NSHI includes two main components. The contributory regime is the mandatory health insurance scheme for formal sector and informal (self-employed) sector workers with the ability to pay, as well as for retirees. Health insurance coverage includes immediate family members. By December 2005, 15.5 million people were enrolled in the contributory regime (37 percent of total population), including dependents. In principle, Law 100 established universal health insurance coverage for all Colombian citizens, to be provided by the contributory regime (CR) for those with ability to pay, and by the subsidized regime (SR) for the poor. Individuals in the formal sector must contribute 12 percent of their salaries as premiums to the CR--4 percent is contributed by the employee and 8 percent by the employer. The primary sources of financing for the SR are national government transfers to departamentos (departments) and municipalities earmarked for health (56.3 per- cent of total resources); a 1 percent solidarity contribution from the CR (34.4 per- cent of resources); contributions from revenues of family benefits funds or Cajas de Compensación Familiar, companies that manage health and other benefits pro- vided by employers, such as recreational services (0.5 percent of total resources); and local tax revenues earmarked for health, obtained from "sin" taxes (8.8 percent of total resources). Solidarity is a key principle of the system--leading to subsidies from the rich to the poor and from the healthy to the sick. The subsidized regime allocates public subsidies to individual insurance premiums for the poor, following a proxy- means testing index known as Beneficiaries Identification System (Sistema de Identificación de Beneficiarios, SISBEN). SISBEN scores are calculated on the basis of a number of dimensions of poverty, including labor market participation, income, educational attainment, family structure, assets, housing material and crowding, and access to water and sanitation. Health plans are free to establish prices for services they buy from providers, as well as the payment mechanisms. Fees for health services in the NSHI have been based on the fee schedules developed as benchmarks by public health plans prior to the reform, adjusted for inflation, but little updating for the "real cost" of ser- vices has been done. These schedules are used as ceilings for price negotiations between health plans and providers. Provider associations are currently insisting on price regulation (setting floors). Two patterns of payment are common to all health plans: (1) preventive and pri- mary care services are contracted mainly through capitation, and (2) most special- ist and hospital care is paid on a fee-for-service basis or by service packages. The effect of these payment mechanisms on access and quality has not been evaluated. The benefits packages-- the "compulsory health plans" (Planes Obligatorios de Salud, POS)--are grouped into three levels of ascending complexity that differen- Summaries of Country Cases 33 tiate interventions by the intensity of specialization and technology and the finan- cial resources required for their provision. The first level of complexity includes preventive and emergency care, plus basic medical, dental, and diagnostic services. The second and third levels include specialized and rehabilitation care, hospital- izations, and the corresponding diagnostic tests. Several key challenges remain in terms of closing some continuing health insurance gaps and ensuring the financial sustainability of the NSHI. A first and major task is finding mechanisms to enroll the informal sector, which accounted for 58 percent of the working population in 2005. About a third of this population is known as the "sandwich population," which can be sorted out in two groups. One group includes people who earn less than twice the minimum wage and who are not eligible for subsidies but for whom payment of the full contribution to the CR premium can be onerous. The second consists of workers who can contribute to the premium but who lack incentives to become affiliated. According to a household survey in 2003, the total sandwich population was about 7.3 million workers and their families, equivalent to 17 percent of the total population. Of these, 2.5 million had some capacity to pay. The administrative costs of enrolling, monitoring, and collecting contributions from this population can be substantial, and the potential for adverse selection is large. The government is currently pilot- ing a program to provide partial subsidies to individual contributions to the premium. A second challenge in the CR is tax evasion. To reduce evasion, Colombia has started to invest in improvements in information systems, establishing links between pension system data and health insurance contribution data. Increased enrollment of informal workers could complement these measures. Third, in the case of the SR, not only will it be necessary to finalize the transfor- mation of supply-side to demand-side subsidies by fully eliminating historical budgets to public providers, but it is also likely that alternative sources of financ- ing will have to be found. Under current growth, employment, poverty, and fiscal conditions, universal health insurance is not predicted to be reached until 2010. Key Lessons Some of the key lessons from the reforms are · The overall level of economic development is a key factor in determining the options for expanding health care coverage. Sustained economic growth for four years before and after the implementation of the 1993 reforms was critical for popular acceptance of increases in contributions to the contributory regime, as well as for the mobilization of general revenues to finance insurance expansion. · Expansion of health care coverage requires substantial infusions of money into the health sector. In particular, expanding coverage of social insurance systems-- either in terms of the population covered or the depth of benefits package--is likely to lead to large increases in health care spending, partly due to moral 34 Good Practice in Health Financing hazard but also due to the need to cover premiums for the poor. As Colombia's health care reforms have expanded insurance coverage from 20 percent to 78 percent of the total population, total health expenditures have grown from 6.2 percent of GDP in 1993 to 7.8 percent in 2004, mainly as a result of increases in public expenditures. · A clear legal and institutional framework is important in expanding formal sector insurance coverage. The 1991 Constitution sets the legal framework and pro- vides political legitimacy for the NSHI with the participation of the private sector. · In Colombia, the previous implementation of a program for the development of community-based health insurers in the rural areas allowed the NSHI to quickly enroll a large number of persons throughout the country. The availability of SIS- BEN has given the system with an instrument for identifying its target popula- tion and for targeting subsidies to the neediest. · Cost containment measures are needed to limit spending increases. These may include full separation of financing and provision by eliminating supply-side subsidies to public providers. · Inclusion of the informal sector is a challenge, especially people who are technically above the poverty line but for whom insurance premiums still constitute a heavy, if not impossible, expenditure. Costa Rica Costa Rica is a small (population 4.43 million) upper-middle-income country with a long tradition of political stability in the Latin America and Caribbean Region of the World Bank. It has a GDP per capita of US$4,349, about average for an upper-middle-income country. In terms of health, Costa Rica devotes some 6.6 percent of its GDP to health, which is about average for its income level, and spends about US$290 per capita on health, which is average for that income level. Some 77 percent of total health spending is public, a much larger share than in comparable countries. Only 20 percent of total health spending comes from out- of-pocket payments. In terms of health outcomes, with life expectancy at 79 years and an infant mortality rate of 11 per 1,000 live births, Costa Rica fares better than other coun- tries with similar income or health spending levels. In life expectancy, Costa Rica ranks with OECD countries. The country has 1.32 doctors per 1,000 people, slightly above average for its income level. However, Costa Rica is well below aver- age in terms of hospital beds, with 1.4 per 1,000 people. Both in terms of breadth and depth of the coverage, the health care system pro- vides significant protection to the population. The Costa Rican Social Security Fund (Caja Costarricense de Seguro Social, CCSS) is a public entity that insures 89 percent of the population and also administers the national pension system. CCSS is also the country's principal curative health care provider. Membership and financing for the public health insurance system is based on employment; Summaries of Country Cases 35 membership is mandatory for formal sector employees. Dependents and indigent persons are provided free care under special regimes. Physical access to primary health care services in Costa Rica is universal--99 percent of the population is able to access primary health services. The good outcomes observed in terms of coverage (both breadth and depth) are partly the result of a long history of efforts to improve the people's living con- ditions that date back to the 19th century. The CCSS was created in 1941. How- ever, before health care reforms in 1994, access to primary care services was restricted to approximately 25 percent of the population. Key Reform Elements and Accomplishments CCSS was set up in 1941 to oversee both the financial management and delivery of health services. The institution was originally established to protect public employ- ees and manufacturers against the risks of illness, maternity, and work-related injuries. In 1960, the CCSS initiated efforts to achieve universal coverage by expand- ing the scope of benefits to farmers, independent workers, poor households, and dependants. By the mid-1970s, health insurance covered 54 percent of the popula- tion. In 1973, further vertical integration of the health system was achieved as all public hospitals were transferred from the Ministry of Health (MOH) to the CCSS. These initial reforms were further consolidated in 1993 when the health sector was reorganized. The MOH became the regulator and steward of the sector, while pri- mary, secondary, and tertiary care was managed entirely by the CCSS. Costa Rica's economic crisis at the beginning of the 1980s and a wide range of cumulative structural problems shook the financial situation of the CCSS and put the system at a crossroads. Coverage under primary care programs had decreased from a high of 60 percent by the end of the 1970s and further declined to 45 per- cent by 1994 under the traditional primary care model. Furthermore, financial problems undermined the quality of the provider network, and satisfaction with the system fell to all-time lows by the mid-1990s. To address these issues, the government launched an ambitious reform agenda, with the financial and technical support of the World Bank and loans from the Inter-American Development Bank. These reforms aimed at correcting the struc- tural problems and further separating financing, purchasing, and provision within the vertically integrated CCSS model. In summary, the reform process aimed to address the following objectives: · Strengthening the primary health care network by creating new clinics and mobile teams based on a geographic model oriented to achieve full population coverage · Introducing performance contracts with all providers in the CCSS network, to clearly establish objectives in production, quality, user satisfaction, and clinical practice · Shifting resource allocation from historical budgeting to performance-based payments and capitation for primary health care 36 Good Practice in Health Financing · Improving revenue collection through the introduction of a single, unified, Internet-based payroll collection system The Ministry of Planning and the CCSS itself were heavily involved in the design and implementation of the reform. In addition, the two most important political parties of Costa Rica, the Partido Liberación Nacional and the Partido Unidad Social Cristiana, strongly backed the process, allowing the reform to con- tinue despite a change of government in 1994. Opposition to the reform existed, but it was more the result of sporadic groups than it was the collective opposition of key stakeholders. Transparency and continuous communication were critical to facilitate the implementation. For instance, mixed commissions were created to involve unions and hospitals in the discussion and the understanding of the process. Hospitals also supported a pilot test of the new resource allocation model, which helped facilitate the process. In summary, one of the main advan- tages of the Costa Rican case was general consensus among the main stakeholders: government/ CCSS, opposition, unions, and providers. Costa Rica represents a good example of a middle-income country with high health care coverage, considerable financial protection, and an extensive package of services. Both in terms of breadth and depth of the coverage, the health care system provides significant protection to the population. Evidence from national surveys shows that the average Costa Rican household allocated 2.6 percent of its income to health spending in 2005. Most of house- holds' health expenditures go for specialist and hospital services and drugs. The share of health expenditures in the richest quintile is approximately 6.6 percent of income, 3.7 times higher than the share of health expenses in the poorest quintile. In other words, most of the out-of-pocket expenditures are from wealthier fami- lies. Access to high-quality medicines is also universal and exceeds WHO basic drug list standards. The formulary includes 460 active principal drugs in 608 pre- sentations, with products for all types of pathologies. In Costa Rica health coverage is financed through a substantial commitment from both government and employers. Following a social insurance model, employers contribute to the CCSS 9.25 percent of the wages paid. Workers con- tribute an additional 5.50 percent of their wages, and the state contributes 0.25 percent of the total national wages. In sum, this amounts to 15 percent of workers' salaries. Self-employed and informal-sector workers are encouraged to join the CCSS voluntary plans where workers pay between 5.75 percent and 13.75 percent of their salaries depending on income. The health system in Costa Rica is based on solidarity--no matter the size of a person's contribution to the system, he or she has equal access to health care ser- vices in the public delivery system. The poor are covered by the "noncontribu- tory" and the "insured by state" regimes; an estimated 620,000 individuals belong to these two regimes, which are cross-subsidized by formal sector employees. According to CCSS actuarial studies, about 50 percent of the contributions from Summaries of Country Cases 37 formal sector employees are used to cover health expenditures of pensioners, independent workers, and poorer households. Contributions to social security are the most important source of financing, almost 60 percent of the health sector's total revenues. The second source--sales of goods and services--corresponds to out-of-pocket expenditures by house- holds. When out-of-pocket expenditures and employee contributions to the social security are considered together, households contribute about half of the total revenues. Government contributes 7.3 percent of its total revenues (5 percent in the form of taxes and 2.3 percent in the form of contributions to the CCSS). This figure does not, however, include the government's contributions to finance health insurance for poor households. The role of external funds is minimal--less than 4 percent of total health expenditures. These achievements are due to sustained public health expenditures and polit- ical commitment. Costa Rica has been able to apply adequate legislative and insti- tutional arrangements to achieve high rates of health care coverage and to ensure universal provision of health services over the long term. When Costa Rica encountered difficult obstacles to achieving its coverage objectives, it was able to renew the structure by redefining the provider network and by integrating private and nongovernmental partners in the delivery of health services. However, a number of issues pose challenges to the health system and the reform. These include (1) the monopoly of the CCSS in the administration of health insurance schemes; (2) limited choice of providers, especially of hospital services; and (3) the corruption scandal in 2004 involving the executive president and some managers of the CCSS who took advantage of the administrative restructuring, gave more power and autonomy to the latter, and severely damaged any future effort to implement further changes in the organizational, financing, and delivery structure. Key Lessons Some key lessons from the reform are · Political commitment: both political parties agreed on the need for change. Even though they did not agree on certain details, the common view that a health reform was required prevailed. The relatively rapid approval of the reform loans in the Congress signaled this bipartisan political support. · Transparency and accountability have been critical ingredients in successful expansion of health care coverage. Political parties, the CCSS, health care providers and other workers, and MOH staff all had an important voice in shaping the reforms that have resulted in the current health coverage system. Internal institutional reforms have also played a role. · Wide participation of main stakeholders in the reform process extended beyond political parties. It included CCSS workers (especially medical staff), MOH staff, medical doctors, and communities. Transparency in the flow of informa- 38 Good Practice in Health Financing tion to the stakeholders was a key aspect. Several groups opposed the reforms, as expected, but continuous negotiation and a consensus building allowed the implementation of the main reform components. It is important to recognize that the original reform agenda has still not been fully implemented, and some elements are missing. For instance, allowing patients the right to choose their provider is an issue not yet resolved. · The creation of a dedicated purchasing unit was the most important organiza- tional change introduced in the CCSS as part of the reforms. It is administratively and financially independent from the CCSS hospitals. Separation of purchas- ing from provision functions allows one department to concentrate fully on planning, negotiating, monitoring, and evaluating the performance of health providers. It removes conflicts of interest in the purchasing relationship that might compromise the efficiency of the purchasing process. · Information systems have played an important role in Costa Rica. Investments in computer systems have allowed the CCSS financial managers to monitor the flow of revenues daily, eliminating the previous 30-day delay. These systems have facilitated the implementation of new payment mechanisms between the CCSS and primary care providers to enhance provider efficiency and perfor- mance. Monitoring management agreements signed by both parties explicitly allows defining and linking coverage targets to payments. · These reforms take time and money to produce results. The health reforms in Costa Rica encompassed separating purchasing from delivery functions within the CCSS, reorienting a curative-based coverage system toward preventive care, and modernizing medical technology and information. While achieving full coverage with primary care services took the country less than 15 years, univer- sal health insurance coverage for a full benefits package will take much more time. · Important health outcomes were achieved by expanding primary care services as the main channel to provide universal access to the entire population. The share of the population with access to primary services jumped from 25 percent of the population just before the reform to almost universal coverage by 2005. · Sustainability of the reforms depends on continuous financial support. It is impor- tant to have a short-term budget to launch the program, such as the two reform loans approved in Congress that came from the World Bank (US$22 million) and the Inter-American Development Bank (US$42 million). A commitment to investing resources over the long run is also critical. Estonia Estonia is a small upper-middle-income country in the Europe and Central Asia (ECA) Region of the World Bank, with a population of 1.3 million and GDP per capita of US$8,328. It was part of the Soviet Union, gaining independence in 1991. The economy contracted initially following separation from the Soviet Summaries of Country Cases 39 Union but recovered soon thereafter. Since then, Estonia has been one of the fastest-growing economies among formerly socialist transition economies in Eastern and Central Europe. Estonia's population, like much of Europe's, is both declining in size and aging. In 2002, more than 20 percent of the population was above 60, and the age dependency ratio is 48 percent. Almost 70 percent of the population is urban. Total health expenditure is about 5.3 percent of GDP, lower than that of other countries at similar income levels. The public share of total health expenditure is high (76 percent), with 87 percent of funds coming through the health insurance system. About 11.5 percent of the government's budget (including MHI spending) is devoted to health, about the average for Estonia's income level. In 2004, 88.3 percent of private expenditure and 21.3 percent of overall health expenditure was out-of-pocket. Life expectancy in Estonia declined after independence, reaching a low point of 61 years for men and 73 for women in 1994. It has been on the rise ever since: in 2005, life expectancy was 66 for men and 77 for women. Average life expectancy (73 years) is still low by European Union (EU) standards, and remains lower than average for its income level. Estonia's infant mortality rate is very low: 6 per 1,000 live births in 2005. In keeping with the trend among ex-Soviet republics, the den- sity of physicians and hospital beds per 1,000 is high at 4.48 and 5.3, respectively. Literacy rates are close to 100 percent. Prior to reforms, Estonia's health system was that of the Soviet Union. Univer- sal health care was centralized and provided nominally free to everybody through a state-run national health service. However, technology and clinical methods were less advanced than those in Western countries. Prereform, as was the case with other former Soviet regions, the system was characterized by input norms and targets, oversupply of hospital beds, overspecialization, and an out-migration of health personnel. The rationale for reform was to introduce a system that would ensure secure sustainable and predictable financing for the health sector, especially given the precarious state of the economy. Following the breakup of the Soviet Union, there was a strong desire to move away from the input-based system to embrace market principles. However, it was understood that privatization and other reforms needed to be balanced by building up a sustainable health and pension system in the social sector. The first Health Insurance Act was approved by the Parliament even before political independence was achieved. Key Reform Elements and Accomplishments The reform of 1991 introduced a classic Bismarkian model of coverage with MHI and decentralization. A health insurance tax of 13 percent on employee salaries was introduced, paid fully by employers. In 1994, the health insurance tax was incorporated into the social tax, with an earmarked share for the health system. Initially, funds were not pooled across the 22 noncompeting, district-based funds. As a result, some of the more deprived areas had lower resources than others. 40 Good Practice in Health Financing However, this has changed. Revenue collection has been streamlined through the Government Tax Revenue Office, and since 2000, the Estonia Health Insurance Fund (EHIF), incorporated under a separate public law, has operated as a single pooled fund with selected administrative and contracting responsibilities dele- gated to four regional offices. The EHIF is legally obliged to balance yearly rev- enues and expenditures, a requirement fulfilled almost every year since the inception of the scheme. Initially, there were no out-of-pocket payments. In 1993, however, Estonia introduced the prescription pharmaceutical reimbursement system, based on some cost sharing. Some 23 percent of out-of-pocket spending goes toward dental care. Flat copayments are charged for some types of health services such as pri- mary care physician home visits, outpatient visits, and hospital-bed days. The health insurance system is mandatory without an opt-out possibility. Pri- vate insurance was allowed to be taken to defray expenses not covered by manda- tory health insurance. Contributions are related to being active in the workforce. Non-contributing individuals (e.g., children, pensioners) make up almost half (49 percent) of the insured population, and their expenses are implicitly subsidized by the others. The state officially contributes for only about 4 percent of the cov- ered population. By the end of 2003, 94 percent of the population was covered by Estonia's MHI scheme. The uninsured 6 percent are working-age individuals not employed in the formal labor market and ineligible under other criteria, such as being regis- tered as unemployed or disabled. The health insurance reform was accompanied by carefully phased changes in the service delivery system. Major delivery system changes included imple- mentation of the family practice model, rationalization of the hospital sector, modernization of the pharmaceutical sector, and the adoption of incentive-based provider payment and risk-sharing mechanisms. Primary health care is now pro- vided by private family practitioners, and hospital and specialist care predomi- nately by autonomous public hospitals incorporated under private law as stock companies or not-for profit foundations (trusts). However, out-of-pocket payments in Estonia have been rising. Habicht et al. (2006) found that between 1995 and 2002, the percentage of households who had out-of-pocket medical expenses exceeding 20 percent of their capacity to pay increased from 3.4 percent to 7.4 percent. Additionally, out-of-pocket payments pushed 1.3 percent of households into poverty. Key Lessons Some key lessons from the reforms are · The reform from a Soviet NHS to a MHI model was carefully planned. Major financing changes included a dedicated 13 percent payroll tax accompa- nied by carefully phased major changes in the delivery system and regulatory environment. Summaries of Country Cases 41 · Health system revenue collection should be in line with overall fiscal policy and should take into account labor market policies, future growth and labor force projections. · The reform benefited from solid economic growth--except at the very beginning. · Implementing the system through the formal employment sector limited corrup- tion and other distortions. · Streamlining revenue collection through the Government Tax Revenue Office allowed the EHIF to focus on purchasing health care for its beneficiaries. · A single risk pool and clear regulatory frameworks have allowed the EHIF to be an efficient administrator of MHI funds and perform as an effective purchaser of services. · Annual actuarial soundness is ensured by legislatively limiting MHI spending to available revenues. · The government has been adept at monitoring and undertaking mid-course cor- rections. For example, risk pools and certain delivery system responsibilities initially delegated to local governments were recentralized after some instances of their capture by providers and inability to deal effectively with certain cross- regional service delivery externalities. · Strategically designed out-of-pocket payments--in conjunction with social insur- ance--can play an important role in ensuring sustainability of health care financ- ing. They help moderate demand and ensure that poor and other vulnerable groups receive coverage for essential health care and protection from cata- strophic expenditures. · The reform enjoyed strong support from the medical community. Kyrgyz Republic The Kyrgyz Republic is a small (population 5.2 million), mountainous Central Asian country, with two thirds of its population rural. Beyond some gold and agriculture, it is poorly endowed with resources. In 2004, per capita GDP was US$434, about average for a low-income country. In terms of health, the country devotes some 5.6 percent of GDP, US$24 per person to health spend- ing. Some 41 percent of total health spending is public, and health accounts for 8.4 percent of the government budget. Out-of-pocket payments account for 55.7 percent of all health spending, a high level relative to comparable income countries. In terms of health outcomes with an infant mortality rate of 58 and life expectancy of 68, the Kyrgyz Republic compares favorably with other countries of its income or health spending levels. In terms of its delivery system, its high physician-to-population and hospital-bed-to-population ratios of 2.51 and 5.3 per 1,000, respectively, are typical of the FSU pattern. Educational levels are also high with adult literacy above 90 percent. 42 Good Practice in Health Financing Before the health sector reforms of the mid-1990s, the Kyrgyz Republic had a typical Soviet norm-driven, centrally planned, general revenue­financed health system in which free health care was every citizen's right. After the breakup of the Soviet Union in 1991, the Kyrgyz economy collapsed, and between 1991 and 1996, GDP fell by more than 50 percent. The Kyrgyz Republic found itself saddled with an over-resourced, unaffordable health system in which "fixed" salary costs and infrastructure accounted for 75 percent of all health spending, private sector pro- vision was almost nonexistent, and private payments were largely under the table "informal" payments to public providers for preferential treatment in the vast state-owned and state-managed health infrastructure. Theoretically, the Soviet system had provided both universal and deep cover- age in terms of financial protection, but in actuality supply-side rationing of technologies, diagnostic tests, and devices was rampant and manifested itself in waiting lists and informal payments. With the collapse of the Soviet Union and its drastic economic impact, the Kyrgyz Republic could no longer financially sustain this system and undertook a series of reforms to deal with both funding and delivery system improvements. Inevitably, some of the needed "financing" had to be derived from efficiency gains from streamlining the bloated health infrastructure. Key Reform Elements and Accomplishments The economic decline and severe fiscal contraction of the early transition period eroded the previously high levels of financial protection and coverage. They also revealed an inequitable distribution of public resources disproportionately favor- ing tertiary care facilities in the capital city, an inefficiently large service delivery sector, and low health care quality. The widespread availability of data and infor- mation showed that high out-of-pocket payments (both formal and informal) were a major financial burden for many households and constituted barriers to access. With the wider economic and social reform context in the mid-1990s, emphasizing poverty reduction and expanding coverage through health reform became important instruments of the poverty reduction strategy. Strong support for health reform in its early phase from the president of the Republic played a key role in introducing coverage reforms. Additionally, Kyrgyz policy makers were willing and open to discuss the issue of high out-of-pocket payments as an irrefutable symptom of a broken system, which facilitated productive discussion of the problems and constraints besetting the health system. The huge excess capacity and limited fiscal space meant that efficiency gains had to be achieved before equity could be directly addressed. However, the objectives of increased efficiency and equity were mutually reinforcing; provider payment and service delivery reforms relaxed the need for high out-of-pocket payments and in turn improved access and financial protection. In 1996, the Kyrgyz government undertook a two-phased health sector reform program. The first phase, from 1997 to 2001, was focused on obtaining additional Summaries of Country Cases 43 revenues through the introduction of the Mandatory Health Insurance Fund (MHIF) financed by a small complimentary payroll tax. The second phase of the reform, launched in 2001, was a complete reform of the funding flows through the system and purchasing mechanisms, explicit specification of the benefits package, and a restructuring of the service delivery system. MHIF, initiated in 1997, focused first on the economically active population who paid a 2 percent payroll tax contribution and on pensioners and the registered unemployed who were funded out of the pension and unemployment funds. Cov- erage, starting with 30 percent, reached 83 percent of the population in 2001 with the addition of children funded by the state and social welfare recipients funded by the social welfare funds (Meimanaliev 2003). Although the program did not sig- nificantly change the depth of coverage from Soviet times, the real benefit was to allow the step-by-step introduction of population and output-based purchasing mechanisms (Kutzin 2002). Phase 2, introduced in 2001, marked the introduc- tion of significant provider payment and service delivery reforms, including the explicit designation of the benefits package and a subsidized outpatient drug ben- efit program. The reform accomplished a number of objectives, including a focus on pri- mary care, service delivery rationalization (reform of the bloated hospital system, improved efficiency through provider payment mechanisms, updating of treat- ment protocols), development of national risk pooling, diversification of health sector financing, broadening of consumer choice, encouragement of private sec- tor provision, and the important national political benefit of clarity of entitle- ments to specific benefits. Although the reform has led to major improvements in efficiency and sufficient revenues to reach pre-breakup spending levels, the persis- tence of a high out-of-pocket share of total spending shows that the Kyrgyz Republic still faces significant financial protection and equity challenges. The average share of household consumption composed of out-of-pocket payments is 2.4 percent. Key Lessons The key lessons from the reform are · Successes are in part due to the comprehensive approach, not a single instruments or magic bullet. · Complex reforms require careful sequencing of various reform steps. · Paying attention to institutional aspects was important in order to ensure sus- tainable benefits. Creating the MHIF as a parastatal agency was crucial for the adoption of strategic purchasing and abandonment of inefficient input-based budgeting. · Phased implementation and careful sequencing were an effective implementation approach and helped build capacity and stakeholder support as well as learning by doing. 44 Good Practice in Health Financing · Strong collaboration of the development partners facilitated harmonized support for reform design and implementation. · Poor economic conditions during the implementation limited government's ability to achieve financial protection objectives. · The slow pace of reforms in public financial management created a challenge for achieving health sector reforms. Compared with other sectors in the Kyrgyz Republic, the health sector was revolutionary: it was the only one to move away from input-based line-item budgets and administrative control mechanisms toward performance-based management. · Elimination of copayments without commensurate increases in public funding is leading to the return of "informal" payments. · For a country of its income level, the Kyrgyz Republic has a well-developed health information system that facilitated policy development, especially prospective provider reimbursement, based on enrollment at primary care facilities, hospital admissions, and outpatient utilization. · The Kyrgyz reforms should be replicable in transition economies with excess capacity and reduced fiscal space. Sri Lanka With a population of 20 million and GDP per capita of just over US$1,000, Sri Lanka is a lower-middle-income country. The country is predominantly rural, with only 15 percent of the population living in urban areas. Total health spending--at US$43 per capita--is about 4.3 percent of GDP, below average for its income level. Most expenditure (54 percent) is private, 48 percent of all health spending is out of pocket. Inpatient provision is largely public (more than 95 percent). The largest part of private sector provision is ambulatory care. Government spending on health is about 8.4 percent of the overall budget. Sri Lanka's population health indicators are better than those of comparable income countries. In 2004, life expectancy was 75 years, and the infant mortality rate was 12 per 1,000 live births. In 2005, the number of physicians and hospital beds per 1,000 was 0.55 and 2.9, respectively, higher than in comparable income countries. Government health spending is financed exclusively from general tax revenues. There is no Mandatory Health Insurance. Private insurance coverage rates are low. Sri Lanka was an early reformer in relation to other countries included in this study. Its reforms can be traced back to 1931 and its first national elections with universal franchise. The elections resulted in the transfer of power to Sri Lankan chosen leaders while the country was still a British colony (it achieved full inde- pendence 17 years later, in 1948). The health system prior to reforms was colonial. The emphasis was on preven- tive care through public health programs. The few modern urban health facilities, funded by a mixture of user fees and general revenue, catered primarily to Dutch Summaries of Country Cases 45 and British residents. Most Sri Lankans depended on traditional healers. In the 1920s, the situation in the island differed little from that in most other British colonies. Government intervention in health was limited to providing health care for a small urban population that operated the colonial infrastructure and admin- istration, for an equally small workforce involved in export agriculture, and for a sanitary regime designed to control major epidemic threats such as cholera. Moti- vated primarily by economic and productivity considerations, some health care was also provided for plantation workers financed by fees on exports. This group generally had better health outcomes than their rural nonplantation counterparts. A severe malaria epidemic in 1934­35 devastated the rural population. Although by this time Sri Lanka was under elected self-rule, the new administra- tion did little to alleviate the problems created by the epidemic. Public resentment after the inadequate response forced policy makers to reprioritize the role of the state in ensuring the people's health. Key Reform Elements and Accomplishments Democracy has been the most important motivation for reforming health services in Sri Lanka. It was introduced expressly to empower the society's poorer groups and women to put pressure on the elites to pay attention to social and health con- ditions. After the 1931 elections, the political economy of the island changed irrev- ocably with the shift of the political power base from urban residents to the rural majority. From the perspective of the political economy of health, the impact of democracy was accentuated in Sri Lanka by the emergence of competitive politics along a left-right dimension, with two-party competition well embedded by the late 1950s; a rural bias in the delimitation of electorates; the single-member con- stituency system, which encouraged politicians to engage in parish-pump politics to maximize the government infrastructure built in their districts; and the small- ness of the country, where each national legislator represented fewer than 10,000 voters in the 1930s. The principal scaling up of health reforms occurred after the malaria epidemic, from the mid-1930s to the 1950s. Subsequent developments have been relatively minor and incremental to this initial fundamental shift in the structure of the health system. The introduction of democratic politics forced successive governments to continuously expand public free health services into rural areas, where the voters wanted the same standards of provision that had been estab- lished earlier for urban dwellers. Health sector managers were not averse to responding. The colonial health department was the first ministry to come under the control of local civil servants, who were imbued with a nationalist ethos of "serving the masses." A key aspect of the reform in Sri Lanka was the expansion of free health care provision to rural areas by building and staffing government hos- pitals and dispensaries, particularly in the 1940s. As a result of the reforms, the emphasis of the health system changed from one that was preventive and sanita- tion-oriented to one that emphasized universal access to curative health services 46 Good Practice in Health Financing through hospitals. This was driven by official recognition in the 1930s that risk protection against the impoverishing impacts of major illness was a necessary concern of public intervention. An important government commission in 1948 later elaborated on this by finding that MHI was not necessary, because direct government provision fulfilled the same insurance function. User fees, another remnant of the colonial health system, were removed in 1951. Additionally, when resource constraints began to bite, policy makers prioritized consumer access to services over service quality. As resource limits tightened, policy makers in the health sector learned to focus on improving productivity, rather than expanding budgets. The health system changed from one in which only the urban rich had access to modern medicine-- while the rural population relied primary on traditional healers--to one in which effectively the entire population has access to modern care. The density of cover- age is very high: most Sri Lankans live within 2 to 3 kilometers of a public health facility. Once democracy had served to establish a widely dispersed government health infrastructure, accessible by all, it then acted to ensure its survival under often dif- ficult fiscal conditions. Subsequently, successful market-oriented and reform- minded governments in Sri Lanka have generally understood that the cost of adequate public health services accessible to the poor was a small fiscal price to pay for the political support that they engender to enable other more important economic reforms. Other aspects of the Sri Lankan reform are noteworthy. The incidence of public health expenditure has been pro-poor, and the system of health financing is pro- gressive (Rannan-Eliya, 2001). Catastrophic health spending is low (van Doorslaer et al. 2007). Out-of-pocket health expenditure is incurred primarily by the rich who are more likely to seek private care due to its (modestly) higher quality and responsiveness. Efficiency has been an important and critical element in Sri Lanka's success: it enabled it to use a limited budget to reach the poor. Sri Lankan public hospitals deliver inpatient admissions and outpatient visits at a far lower cost per capita and health-to-GDP ratio than comparable income developing countries. This has been achieved by high bed turnover rates and short average lengths of stay. Labor productivity is also high: government doctors and nurses see, on average, more inpatients and outpatients than is the norm for developing coun- tries. As part of the reforms, public sector doctors have also been allowed to prac- tice privately after-hours, thereby increasing their income-generation abilities and providing an incentive for relocation to rural areas (Rannan-Eliya 2001). Key Lessons Some of the key lessons of the reforms are · Democratic accountability is important to ensure that the health system is responsive to the needs of the poor. A key enabling factor for Sri Lanka is Summaries of Country Cases 47 democracy, which has given the rural poor a voice. Anti-incumbency in a functioning democracy made the choice to pursue universal coverage an easy one (politically).2 · Countries can rely on effective (and free) public provision of health that is funded out of general taxation. Sri Lanka has no social insurance and relatively low labor market formality, again suggesting that neither is necessary for successful health financing reforms. Therefore, pushing social insurance policies in coun- tries that are not ready for them may not be prudent. · Sri Lanka is not exceptional in terms of its revenue generation but has used effi- ciency gains to finance increases in coverage. (At the same time, to ease the bur- den on the system, it has nudged the rich toward private care.) This suggests the importance of good governance in allowing for effective implementation of such policies. Sri Lanka has a history of good public administration with little corruption, both strong preconditions for its kind of NHS system. · The people's high educational attainment appears to have played a facilitative role in Sri Lanka. In terms of influencing demand, it has enabled the high priority placed on health to be sustained through changes in government. Thailand With a population of about 64 million and a GDP per capita of US$2,441, Thai- land is a relatively large lower-middle-income country in the East Asia and Pacific Region of the World Bank. It stands out as one of the few countries in Asia never to have been colonized. Thailand was battered by the 1997 Asian financial crisis but has since recovered. Sixty-eight percent of its population is rural. At 3.5 percent, the proportion of spending on health relative to GDP in Thailand is low for its income level. In 2004, the government's share of health expenditure was 64.7 percent. Of private spending, 74.8 percent was out of pocket. Although overall health expendi- tures are low, a large part (11.2 percent) of the government's budget is devoted to health. Health provision is mixed, with both public and private providers. Thailand has excellent population health outcomes. Life expectancy in 2005 was 71 years, and the infant mortality rate was 18 per 1,000 live births. There are 2.2 hospital beds and 0.37 doctors per 1,000, the latter somewhat low for a coun- try at Thailand's income level. Literacy rates, both for adults in general and for adult females, are higher than 90 percent. Key Reform Elements and Accomplishments Before the 2002 reforms expanding health coverage, Thailand sought univer- sal health coverage through four different schemes. Social movements and the development of democracy were key contributing factors toward the series of health reforms in Thailand. The first was the Medical Welfare Scheme (MWS), introduced in 1975 by the first democratically elected government after the mili- tary dictatorship had been replaced in 1973. The MWS aimed to provide free med- 48 Good Practice in Health Financing ical care to low-income groups and covered about one quarter of the population. The scheme was funded by general taxation and covered the poor, the elderly, the disabled, and children younger than 12 years. Under this scheme, an estimated 11 million people were eligible for a comprehensive package of services without user charges at public facilities. There were some problems related to funding and tar- geting, and the quality of care provided was questionable (Pannarunothai 2002). The second scheme, the Civil Servant Medical Benefit Scheme (CSMBS) was introduced in 1978 as a noncontributory regime to provide health benefits to civil servants and their dependents (a total of 7 million people). This scheme, too, was funded through general taxation. Beneficiaries were free to choose private or pub- lic providers, although only 50 percent of costs incurred not exceeding baht 3,000 were reimbursed at private health facilities. Compared with the other schemes, CSMBS had the highest expenditures as a result of its fee-for-service reimburse- ment model (Sriratanaban 2002). Under a new military-led government, little progress was made for more than a decade. The Social Security Scheme (SSS), launched in 1990, covered about 8 million formal sector employees (but not their dependents). It followed a capita- tion model, but concerns persist about the quality of care provided (Tangcharoen- sathien, Willbulpholprasert, and Nitayaramphong 2002). The Voluntary Health Card Scheme (VHCS) was introduced in 1993 to cover those ineligible for any of the other schemes. The VHCS collected premiums from households, the Ministry of Public Health, and from an Asian Development Bank loan. However, there were concerns that the scheme was not reaching the intended target population and suffered from a classic adverse selection problem. The VHCS was not successful in expanding coverage due to financial nonviability (Donaldson, Pannarunothai, and Tangcharoensathien 1999). By 1998, about 80 percent of Thailand's population was insured. In 2001, Thai- land introduced a Universal Coverage Scheme (UCS) (Tangcharoensathien, Srithamrongsawat, and Pitayarangsarit 2002), motivated by the 1997 Constitu- tion, which reintroduced a focus on populist policy. The rationale for introduc- tion was to deal with some of the problems of the earlier schemes and to increase coverage among the uninsured. The UCS began in 2001 as a mandatory scheme funded by general taxation and a baht 30 copayment. It merged the MWS and the VHCS and covered the remain- ing uninsured. Unlike many other countries of similar income levels seeking to expand coverage, Thailand's reform was rapidly implemented and scaled up nationwide within a year. All Thai citizens are eligible for the scheme, which pro- vides a standard benefits package, and all contract provider networks--both pub- lic and private--are required to provide these services to registered beneficiaries. The uninsured have decreased from 20 percent of the population in 1998 to 5 per- cent in 2003. The UCS alone covers 74.7 percent of the population (47.7 million people) (Vasavid et al. 2004). More recently, the poor, the elderly, children, and the disabled have been granted exemption from copayments. Summaries of Country Cases 49 The results of the reform have been remarkably pro-poor. Since the implemen- tation of the UCS, there has been a 25 percent increase in outpatient care and a 9 percent increase in hospitalizations, both concentrated among the poor. Addi- tionally, government health subsidies now reach the poor at health centers, district hospitals, and provincial hospitals for both outpatient and inpatient care and thus reduce inequality in living standards among the population (Limwattananon, Tangcharoensathien, and Prakongsai 2007). The UCS has improved financial pro- tection: the rates of catastrophic spending have decreased, and fewer people are impoverished due to out-of-pocket spending than before the reform (Limwat- tananon, Tangcharoensathien, and Prakongsai 2007). Regarding equity in the financing of health care, the better-off pay more than the worse-off in general taxes as a share of their income (O'Donnell et al. 2005). Accordingly, the UCS has achieved a strong pro-poor focus through improvements in the equity of utiliza- tion, public subsidies, and the health system's progressive financing structure. Although the UCS scheme is widely regarded as successful, especially in increas- ing access and providing financial protection, concerns persist that it has encour- aged an unsustainable increase in demand, resulting in a rapid increase in the workload of health personnel. In addition, there are no earmarked funds for the UCS, and every year it is vulnerable to budgetary competition. Key Lessons Some of the key lessons of the reforms are · Sustained growth may help lay the foundation for expansions in coverage, but major health reforms can still occur shortly after a recession if fiscal space exists. Even after GDP growth resumed in 1999 and 2000 after double-digit declines, total health spending still shrank slightly over this period. Significant reduc- tions in defense spending created budgetary room for the health sector's expansion in 2001. · Strong political and social support for universal coverage provided the necessary conditions for a rapid scale-up. The motto "30 baht treats all diseases" proved to be simple and popular, and the National Health Security Bill was the first bill sent to parliament because its proposal had been signed by more than 50,000 citizens. · The sustained development of primary care since 1980 and establishing its impor- tance under the UCS has supported efficiency. The gatekeeping role is especially salient because of concerns that the system is underfunded and capitation rates are too low. · Setting appropriate provider reimbursement through a combination of prospective payment methods has also improved efficiency. Thailand's experience with a variety of payment schemes over the last three decades helped inform the reimbursement design in the UCS. However, the Civil Servant Medical Benefit 50 Good Practice in Health Financing Scheme pays providers based on fee for service and has the highest per capita spending of any of the insurance schemes. The MWS had been conducting sys- tematic monitoring and evaluation on DRGs since 1998; capitation has been successful at containing costs for the SSS. As a result of these experiences, the UCS was designed to use a mix of capitation, global budgets, and DRGs to put financial risks on providers to contain expenditures. Reform of the payment mechanism of the Civil Service Medical Benefits scheme is an ongoing political challenge. · General revenue financing has been pragmatic and beneficial to the equity in health care financing. The VHCS suffered from adverse selection, did not expand coverage, and was not financially viable. By contrast, the UC system is funded by progressive general taxation and does not rely on household contri- butions, which were difficult to collect in the informal sector. · The rapid implementation of UCS and its achievement of national coverage within one year were possible partly because reforms were built on previous invest- ments in the delivery system. Thailand has had a sustained investment in the health system infrastructure for decades, including gradual development of facilities and human resources in rural areas. This network, coupled with some utilization of the private sector, greatly enabled the increase in access immedi- ately after the "big-bang" reform. · Good governance arrangements have built stakeholder participation. Partly due to the strong social movement, civil society has an influential position with the top decision-making authority for the UCS. The committee, comprising all key stakeholders and chaired by the deputy health minister, meets weekly to deal with policy implementation. · Excellent technical capacity and a strong evidence base for a country with Thai- land's income level were important to inform policy making and facilitate imple- mentation. Health systems and health financing research has been active in Thailand since the late 1980s, especially after the enactment of the Health Sys- tems Research Institute Act of 1992. This network built up enough capacity and a strong enough knowledge base to facilitate the formulation and imple- mentation of reforms focusing on universal coverage. Tunisia Tunisia is a small middle-income country in the Middle East North Africa region of the World Bank with a population of 9.91 million and per capita GDP of US$2,832. Total per capita health expenditures were US$175 in 2004--6.2 percent of GDP. About half of health spending comes from private sources, most of it (83 percent) from out-of-pocket expenditures by households. In terms of health status, Tunisia is one of the best performers in the MENA region and above average compared with other countries with similar income lev- els. Life expectancy is 73.5 years, and infant mortality is 21.0 per 1,000 live births. Summaries of Country Cases 51 With respect to the delivery system, Tunisia has 1.34 physicians and 2.1 hospital beds per 1,000 people, which places the country slightly above and below average, respectively, compared with other countries of its income category. Disparities in health status between urban and rural areas, as well as among different socioeco- nomic groups persist: the population's health status is significantly better in urban areas than in rural areas. However, these disparities are less pronounced in Tunisia than in other MENA countries, especially in terms of infant mortality and life expectancy. Key Reform Elements and Accomplishments Over the past three decades, Tunisia has developed a health system that covers nearly its entire population and which compares favorably with that of other mid- dle-income countries. During this period, Tunisia's population has undergone dramatic demographic, socioeconomic, and health status changes that motivated the reforms. The demographic and epidemiological transitions have been charac- terized by an expansion of the proportion of the elderly population, a reduction of the incidence of communicable diseases, a growing incidence of accidents (traffic and work-related), and an increase of the prevalence of chronic diseases. The improvement of economic conditions and the increased living standards of Tunisians generated new expectations from the population. Tunisians became more demanding of their health system, seeking better service quality, new med- ical technologies, improved access, and financial protection against out-of-pocket health expenditures for vulnerable households. Health care coverage has been expanded through substantial increases in health workers, health facilities, modern medical equipment, and health care cov- erage. Consequently, health care expenditures have surpassed 5 percent of GDP since the second half of the 1990s. Until the 1980s, the health care system was based on the colonial tradition of a hospital-centered health infrastructure, concentrated in large urban areas. In the past three decades, Tunisia has progressively put into place a health system that covers nearly the entire population. It has implemented mandatory health insurance for large parts of the population, including government and other for- mal sector workers. At the same time, the health system has strongly emphasized preventive programs, financed completely by the government. As a result, some infectious diseases have been eradicated, and the incidence of others has been sig- nificantly reduced. The Tunisian strategy for expanding coverage has emphasized the expansion of geographic coverage of the population--through primary health facilities that are geographically and financially accessible to most of the population. The strategy also focuses strongly on qualified human resources in the health sector. Tunisia has followed a consistent strategy in training and deploying medical personal 52 Good Practice in Health Financing since the 1960s--first concentrating on the capital, Tunis, and now decentralized, especially for nurses' training. More than 80 percent of Tunisians are now covered by health insurance through either a health insurance scheme or a medical assistance program. There are two main types of social security systems, both mandatory and together cover- ing about 7 million persons (71 percent of the population). The CNSS (Caisse Nationale de Sécurité Sociale) provides health care coverage for 1.2 million pri- vate sector enrollees and their families--including employees, independent work- ers, and other categories such as students, the disabled, nonsalaried agricultural workers, and Tunisians working abroad. The payroll contribution rate for this system is 4.75 percent. The CNRPS (Caisse Nationale de Retraite et de Prévoyance Sociale) covers about 0.6 million public sector employees and retirees and their families. The contribution rate for this scheme is 1 percent each from employers and employees. Relatively poor households are covered by two subsidized medical assistance programs provided to the enrollees and their families, together known as Assistance Médicale Gratuite (AMG). The poorest are exempt from all fees, while others who are eligible for a reduced fee scale contribute 10 Tunisian dinars annually. The expansion of insurance coverage in Tunisia has featured the early intro- duction of an MHI for civil servants and formal sector employees and its gradual extension to additional groups. At the same time, additional reforms have placed a strong emphasis on preventive programs and the extension of coverage through primary health facilities, geographically and financially accessible for the entire population. Private providers have been increasingly incorporated in the health care delivery system. Private hospital bed capacity has doubled in the past 10 years, and the number of medical examinations provided in the private sector approaches that of public providers. However, the public sector remains the main health care service provider--supplying about 85 percent of all hospital beds and more than 55 percent of medical personnel. Since 1990, the government has pursued a health sector strategy emphasizing: (1) the continuation and consolidation of gains in primary health care services, (2) improvements in hospital care--particularly in university hospitals--through structural and institutional reforms, and (3) legislative reforms to permit greater private investment in health service delivery. In terms of health financing, the extension of the health care coverage has been accompanied during the last 16 years by an increase in the relative household con- tribution, a decrease in the government contribution, and a slight increase in the Mandatory Health Insurance contribution. Out-of-pocket payments have high catastrophic and poverty impacts, and the catastrophic headcounts are concen- trated among the poor (see chapter annex 2A). The health status of the Tunisian population continues to improve with the further development of preventive and curative health care services and a contin- ued decrease in the birth rate. Summaries of Country Cases 53 Key Lessons Some key lessons from the reforms are · A strong initial commitment to primary health care contributed to the success of the Tunisian case. This concentration has been associated with the gradual development of more intensive and more costly hospital care, influenced by the regulated introduction of new medical technologies. · The extensive growth of the private sector, associated with the quantitative and qualitative change in health care demand, has contributed to a significant growth in household health expenditures and poses a potential burden for poorer house- holds. Since the end of the 1980s, the private health care delivery has developed rapidly in terms of number of providers and health care categories. Private hospitalization, especially for surgery, has developed strongly. · The dimensions of the large-scale mandatory health insurance reform should not be underestimated when initially implemented. Coverage expansions are com- plex and entail extensive financial commitments. · Regulatory and administrative capacity is critical to the successful expansion of health care coverage and must be backed up by a commitment to provide access to quality health care services. The development of public hospitals' management capacity and the promotion of participatory management will lead to improve- ments in hospital performance. · The development of information systems is essential. Despite substantial invest- ments in information technology, Tunisia currently faces extensive demands in this area. Vietnam With a population of about 84 million and a GDP per capita of US$550, Vietnam is one of the larger low-income countries in the East Asia and Pacific region of the World Bank. It is recognized as one of the most dynamic economies in the world, after a decade of economic growth averaging between 6 and 8 percent a year fol- lowing economic reforms in the mid-1980s. The country is predominantly rural: only 25 percent of the population lives in urban areas, but the ratio of the rural to urban population is rapidly falling. In 2004, Vietnam spent about 5.5 percent of its GDP on health, about US$30 per capita--about the average for its income level. Most health spending is private--72.9 percent of total health spending, and 74.2 percent of that is out of pocket. The government spends a relatively low 5.0 percent of its budget on health. Vietnam has excellent health outcomes for its income level. In 2005, life expectancy was 71 years, and infant mortality was 17 per 1,000. Hospital beds and physicians per 1,000 people are 2.1 and 1.34, respectively, somewhat higher than the average for other low-income countries. Both adult literacy and female literacy rates are high. 54 Good Practice in Health Financing Prior to the reforms, Vietnam's health system was geared toward providing "health for all" via an extensive network of community health services and inter- communal polyclinics for primary care and government hospitals for higher lev- els of care (MOH 2001). Coverage was fairly widespread. However, the quality of care was low primarily due to a lack of resources, which resulted in shortages, especially of modern medical drugs and equipment. By some estimates only 8 percent of perceived need for medical equipment was met. There were three main motivations for the health financing reforms of the past decades: (1) a historic emphasis on good health, (2) the ingrained acceptance of health as an important component for poverty eradication, and (3) a desire to make health care more affordable for the poor. Together with education and social development generally, health has historically been viewed as central to the well- being of the Vietnamese people and society. Writings of the founding father of modern Vietnam, Ho Chi Minh, emphasized good health among the population. Closely related is the general objective of poverty eradication, in which good health was and still is viewed as a cornerstone. Gradually, the Vietnamese govern- ment and local authorities became aware of the difficulties posed by the introduc- tion of user fees for health services in the late 1980s in terms of access to care, particularly for the poor. Health financing, already pinched, became tighter at the end of the 1980s after the collapse of the Soviet Union, a country to which Viet- nam looked for foreign aid. Alternate sources of finance had to be mobilized quickly if the health sector was to function effectively. These factors led the government to introduce targeted health financing sup- port. Initially, this support was part of other broader poverty eradication policies. Later, it took the shape of more stand-alone health financing programs aimed at the poor and other target groups, including students and children. Key Reform Elements and Accomplishments Health system reforms were undertaken in the backdrop of broad economic and social reforms (Doi Moi) that began in 1986 with a gradual transition from a cen- trally planned, Soviet-style economy to a socialist-oriented, market-based one. In 1987, private health and pharmaceutical sectors were officially sanctioned. A small-scale private health sector burgeoned in the urban areas, often staffed by moonlighting government health workers. Recent estimates suggest that almost 30 percent of all delivery is now private, concentrated in the urban areas. Partial service fees and charges for drugs and diagnostics were introduced in 1989 in all public facilities, with the revenues utilized to improve services. This "commercialization" of the health sector had an inimical impact on utiliza- tion rates among the poor, triggering the introduction of health insurance reforms in 1992: both as a means to reduce the burden on the government budget and to improve access among the poor. Insurance coverage rates increased among those with compulsory, employment-related coverage and voluntary student insurance. Summaries of Country Cases 55 However, extensions of coverage to the general population have not met expecta- tions. As a result, the government has made several attempts--such as provision of subsidized health insurance cards for the poor and mandating free health care for children--in an effort to improve coverage among vulnerable groups. The government's goal is to provide universal insurance coverage by 2010 through a mix of Mandatory Health Insurance, targeted health insurance for the poor and voluntary insurance schemes. In 2004, about 60 percent of the popula- tion was covered by health insurance. Compulsory insurance contribution rates are set at a modest level of 3 percent of contractual salary and basic allowances, pension, social insurance payments, scholarship or minimum wage depending on which entitlement group the insured individual belongs to. For the employed, workers pay 1 percent and their employers pay 2 percent of salary. For retirees and people receiving social insurance benefits, the contributions are paid by Vietnam social security. For the other groups the government budget pays the contribu- tions. For the poor and the elderly aged 90 and older, the contribution is a fixed amount of US$3.1 per person per year with the contribution paid from the state budget. The reforms in Vietnam have improved access to care and health outcomes. Inequalities remain an issue, and the government is making a special effort to tar- get the poor in order to mitigate those. The catastrophic impact of out-of-pocket payments remain high across different thresholds and are often concentrated among the poor (van Doorslaer et al. 2007). Key Lessons The key lessons from the reform are · Social insurance may not be a panacea, especially in a low-income setting. Viet- nam's experience indicates the difficulty of reaching high levels of coverage using social insurance alone. There is a large level of informality and, in order to cover informal workers and other vulnerable groups, the reforms had to be adjusted to incorporate special targeting programs to reach those not covered by social insurance. · Vietnam's and Sri Lanka's experiences contrast sharply, suggesting that many different paths can be followed to successful reform. Sri Lanka is a democracy; Vietnam, a one-party state. Sri Lanka financed universal coverage using tax- revenue-funded public provision; Vietnam favored an approach that is more reliant on use of cost-sharing, user fees, and social insurance with adjustment to reach the uninsured. Initially, the poor suffered in Vietnam as a result of the reforms (high out-of-pocket payments and catastrophic expenses) and the focus on equity came later in the reform process. Sri Lanka started with an ini- tial focus on equity. 56 Good Practice in Health Financing · Widespread literacy seems to have played an important role in Vietnam, as in Sri Lanka. The people's long-standing awareness of the importance of health appears to have helped facilitate the choice and implementation of reforms. · Good governance has been an important enabling factor. This has been true in terms of both implementation of social insurance and successful targeted pro- grams such as the special funds for the poor. Decentralization in the imple- mentation of the programs appears to have helped as well, especially in targeting. · Inheritance of a network of primary health care facilities was also an important enabler. Initial conditions do matter--this was also true for some of the other cases (e.g., the Kyrgyz Republic). · The Vietnam case also highlights the importance of strong economic growth in helping sustain the reforms. · The gradual implementation of reforms in Vietnam--as opposed to a "big-bang" approach--appears to have helped instill a learning-by-doing twist into policy making. Deficiencies in the initial reforms triggered adjustments to solve the problem. Endnotes 1. Income levels, health spending and delivery system, and outcomes data in the back- ground section of each summary are derived from the World Bank (2007a) and WHO (2007). 2. This is in line with other recent research that suggests that governments are more responsive in settings where the electorate is informed and politically active (see, for instance, Besley and Burgess 2002). Nobel laureate Amartya Sen has often made similar points about the importance of democracy and a free press in helping governments priori- tize health and education. However, cross-country evidence of the positive impact of democracy on health is weak and inconclusive. Safaei (2006) does find a small positive effect of democracies on health outcomes in a cross-country setting, but many other stud- ies have found no impact (ADB 2006). 5 Enabling Factors for Expanding Coverage The nine case countries described in this volume have all significantly expanded health care coverage, although each country is quite different and has a different story to tell. This chapter attempts to draw generalizable lessons across the nine countries and concludes by suggesting future directions for improving the global evidence base on health financing reforms. All nine countries have one common experience: they recognized the consider- able amount of time it takes to expand health care coverage and made a commit- ment to do it. Their good performance was assisted by political stability and economic growth and buttressed by sound planning, institutional strengthening, and financial investments in human resources, physical infrastructure, and infor- mation systems. Generalizable Conditions for Good Performance Trying to generalize about enabling conditions for good performance in health financing reforms from these nine case studies is daunting--they are not repre- sentative of all global experience and do not constitute a random sample. More- over, because failures are not assessed, the observations below are not based on the entire spectrum of financing reform outcomes. Nevertheless, all nine countries have undertaken serious reforms and by the study definitions of "good perfor- mance" have done well. These nine cases include two low-income countries and seven middle-income countries spanning the entire spectrum of middle-income countries. The structures of the health systems also vary across the nine countries. Sri Lanka has the only pure model of a national health service (NHS). Chile repre- sents a hybrid model of mandatory health insurance (MHI) and private health insurance. However, these categorizations are not straightforward. For example, Costa Rica's MHI system very much resembles an NHS with one fund, which owns its own facilities and is heavily subsidized from general revenues. The other countries represent MHI models, but with the exception of Estonia, each is heav- ily subsidized by general revenues. Estonia is one of the few countries in the world that funds virtually its entire MHI system exclusively from payroll taxes. 57 58 Good Practice in Health Financing However, a SHI/MHI financed by payroll taxes may not be a panacea, especially in a low-income setting. In these settings, reaching broad and deep coverage is dif- ficult, using social insurance alone (as illustrated by Vietnam); and providing health coverage to a broader segment of the population may be feasible using an NHS model (as in Sri Lanka). Labor market informality is pervasive. To cover informal workers and other vulnerable groups, the reforms had to be adjusted to incorporate special targeting programs and funding mechanisms to reach people not covered by social insurance. As argued in chapter 2, the model a country chooses is not of primary impor- tance. The design and implementation of the three key health financing functions are the likely determinants of the success or failure of a health financing system. Much of the detailed information is lacking to measure the equity, efficiency, and sustainability of revenue collection; the efficiency and equity of risk pooling; and the technical and allocative efficiency of purchasing. However, some indications have been presented about the sustainability of spending by looking at health shares of the national government budgets and revenue-to-GDP ratios. The extent of risk pooling has also been gauged by proxy through nonpooled out-of-pocket payments and, where possible, financial protection has been measured via the shares of household budgets absorbed by health spending. Allocative efficiency and technical efficiency have been crudely evaluated by comparing delivery system capacity, health outcomes, and overall spending levels to global trends. Good performance has been examined in terms of breadth of coverage--defined as the percentage of the population with insurance coverage and financial pro- tection--by looking at out-of-pocket spending as a share of total spending, and in some cases as a share of household income. More important, an attempt was made to see what common underlying factors appear to have enabled these results. Motivations for reforms were different for Estonia and the Kyrgyz Republic. Both countries faced major financial crises after the breakup of the Soviet Union. Both countries revamped their Soviet-style systems with MHI systems, but more funda- mentally made drastic changes in their inefficient delivery systems. Vietnam, ideo- logically a communist state, was affected financially in much the same way because of its heavy dependence on the Soviet Union for subsidies prior to the breakup. Other countries were motivated by humanitarian and solidarity considerations. In some cases, like Sri Lanka, important rural representation in the political process highlighted equity considerations related to both physical and financial access. The different paths taken by these nine countries were heavily contingent on historical factors, as well as political economy and institutional arrangements in the individual countries. A key question remains: since health systems and health reforms are so country-specific, can a common set of enabling conditions be found in these nine good performers or are generalizations impossible? From the lessons and cross-country analyses summarized in chapter 4 and the detailed studies in part II of this volume, a number of commonalities across all nine coun- tries emerge. The common factors identified across all or most of the nine coun- tries include Enabling Factors for Expanding Coverage 59 · Institutional and societal factors: strong and sustained economic growth long- term political stability and sustained political commitment, strong institu- tional and policy environment, and a well-educated population. · Policy factors: financial resources committed to health, including private financing; commitment to equity and solidarity; health coverage and financing mandates; consolidation of risk pools; limits to decentralization; and focus on primary care. · Implementation factors: carefully sequenced health service delivery and provider payment reforms, good information systems and evidence-based decision mak- ing, strong stakeholder support, use of efficiency gains and copayments as financing mechanisms, and flexibility to make mid-course corrections. These results, it might be argued, could have been posited without a study. While these case study-based findings are not the result of rigorous evaluations for the reasons previously stated, they are based on a standardized definition of good per- formance in terms of coverage expansions and financial protection in countries that have achieved good health outcomes with average or below average spending. Institutional and Societal Factors This chapter discusses these "enabling" conditions as potentially necessary con- comitants for good performance in health financing reforms. In this context, the unmet goals of these reforming countries are also highlighted. Strong and Sustained Economic Growth Considering the length of the reform periods, the nine case countries faced vary- ing economic circumstances. Virtually all of the countries, however, had long periods of favorable economic growth during much of their reform efforts (figure 5.1). Economic growth greatly facilitated health coverage expansion by bringing a greater share of the population into the formal employment sector, raising house- holds' ability to pay for health care, and providing governments with tax revenues to subsidize health services. Strong economic growth is particularly evident in the Latin American cases. Economic factors have been important in Colombia, where sustained economic growth for four years before and after the 1993 reforms was critical for popular acceptance of increases in contributions to the National Social Health Insurance, as well as the mobilization of general revenues to finance insurance expansions for the poor. The enrollment of a large fraction of the population through coverage of the families of contributing individuals allowed the public sector to free addi- tional resources to provide health care for the poor. The other countries documented in this volume also achieved substantial growth that facilitated reforms. Estonia's strong rebound after the breakup of the Soviet Union facilitated the creation of an MHI system that covers nearly the whole population funded from a 13 percent payroll tax. 60 Figure 5.1 Real GDP Trends per Capita, 1960­2005 Good Sri Lanka Vietnam Thailand 5,000 5,000 5,000 Pract capita capita capita 2,500 2,500 2,500 ice per per per in GDP 0 GDP 0 GDP 0 H ealth 1960 1970 1980 1990 2000 2010 1960 1970 1980 1990 2000 2010 1960 1970 1980 1990 2000 2010 Financ Estonia Tunisia The Kyrgyz Republic ing 5,000 5,000 5,000 capita capita capita 2,500 2,500 2,500 per per per GDP 0 GDP 0 GDP 0 1960 1970 1980 1990 2000 2010 1960 1970 1980 1990 2000 2010 1960 1970 1980 1990 2000 2010 Colombia Chile Costa Rica 5,000 5,000 5,000 capita capita capita 2,500 2,500 2,500 per per per GDP 0 GDP 0 GDP 0 1960 1970 1980 1990 2000 2010 1960 1970 1980 1990 2000 2010 1960 1970 1980 1990 2000 2010 Source: World Bank 2007a. Enabling Factors for Expanding Coverage 61 Continuous and steady real growth of 5.5 percent over the past 30 years has enabled Tunisia to increase coverage and at the same time develop the delivery capacity to accommodate these expansions. It has also allowed Tunisia to reduce poverty and have a continuous improvement in its people's living conditions. Although Thailand suffered a recession between 1996 and 1998, its economy experienced strong growth in the early 1990s. Thailand's economy also rebounded quickly from the shock and appears to have since recovered, perhaps enabling the dramatic increase in health spending after the introduction of universal coverage in 2001. The case study of Vietnam shows that sustained economic growth over the past 15 to 20 years has enabled the government to undertake ambitious health financ- ing reforms, currently reaching more than 60 percent of the population, with a goal of universal health insurance by 2010. In real terms, per capita spending on health has increased by two thirds during this time period. Moreover, increases in public revenues have enabled the government to significantly increase its share of health spending, creating the possibility of new programs of free health care for the poor and for children. Similarly to Vietnam, economic liberalization has helped spur economic growth in Sri Lanka. Such growth has been an important enabling factor for the government to expand coverage and enact its propoor reforms, particularly because sustained growth can provide the basis for enhanced government rev- enues. It also facilitates formalization of the labor market, which can further improve the collection of revenues in MHI-based (and indeed all) financing sys- tems. The Kyrgyz Republic is the poorest country documented in this book, but it has achieved notably improved economic circumstances since the breakup of the Soviet Union. As these countries have experienced strong economic growth, government per capita health expenditures have also been rising. The sustainability of government spending will be a key issue as they move forward with their reforms. For example, it is not clear whether the governments of Costa Rica and Colombia can continue to devote so high a high share of their budgets to the health sector. Some countries have demonstrated fiscal discipline, such as eliminating supply-side subsidies (e.g., historical budget support to public providers) and establishing budget ceil- ings, as in Chile, to contain costs. Others countries will soon need to address the sustainability of government spending through supply-side controls or demand- side cost sharing. Long-Term Political Stability and Sustained Political Commitment In all of the cases documented in this volume, governmental commitment to the reform has lasted for many years or decades. In no case presented here has the reform been ongoing for less than 10 years. Chile has been pursuing its reform for more than 25 years. Given the complexity of many of these reforms in terms of 62 Good Practice in Health Financing scope, sequencing and concomitant implementation of major delivery system changes, and the frequent development and adoption of modern health informa- tion and provider payment systems, it is not surprising that any success would require long-term government commitment. The adage that health reform is a perpetual process certainly comes to mind. Hsiao and Shaw (2007) make this same point with regard to MHI implementation experiences. Thailand, for example, introduced its health insurance scheme for the poor in 1975. The insurance system in place today in Chile was first implemented in 1981. Both of these systems have undergone important changes over time. The good outcomes observed in Costa Rica in terms of coverage--both breadth and depth--are the result of efforts begun in the 19th century to improve the living conditions of the population. Costa Rica has a long tradition of political stability. The Costa Rican Social Security Fund (CCSS) was created in 1941. How- ever, before health care reforms in 1994, only 25 percent of the population had access to primary care services. At that time, the major political parties agreed on the need for change. Transparency and accountability have been critical ingredi- ents for successful expansion of health care coverage. Political parties, the CCSS, health care providers and other workers, and MOH staff members all had an important voice in shaping the reforms that have resulted in the current health coverage system. Internal institutional reforms have also played a role. In Sri Lanka, the principal scaling-up of health reforms stretched from the 1930s through to the 1950s, following the malaria epidemic of the mid-1930s. Subsequent developments have been relatively minor and incremental to this ini- tial fundamental shift in the structure of the health system. One of the key aspects of the reform in Sri Lanka was the expansion of free health care provision to rural areas--through the building and staffing of government hospitals and dispen- saries, particularly in the 1940s. Strong Institutional and Policy Environment Institutional strength and political stability are also important enabling factors for a successful reform. Efficient tax collection and the ability to enforce mandatory membership in social insurance programs are essential functions for expanding coverage; both depend critically on institutional strength and stability. Likewise, the existence of competent, independent regulatory agencies is a prerequisite for the provision of both insurance coverage and health services. When all nine countries started their reforms, they had--and still have--better than average institutional and policy environments, as measured by the World Bank's Country Policy Institutional Assessment Index (CPIA). The index is made up of some 16 components measuring a wide range of institutional and policy management issues including macroeconomic and trade policy, financial sectors, public sector management, social safety nets, poverty focus, environmental man- agement, and other factors. Moreover, strong institutional environments were Enabling Factors for Expanding Coverage 63 found for both democratic and nondemocratic countries, as shown by the polity index of democracy in figure 5.2. Higher values of the polity index signify greater degrees of democracy, and changes over time reflect changes in the nature of the polity (Marshall and Jaggers 2005). Among the nine country cases, there are highly democratic countries like Costa Rica, Estonia, and Sri Lanka, and less democratic countries like Tunisia, the Kyrgyz Republic, and Vietnam. All the others have gone through periods of democracy and nondemocracy, although the recent trend appears to be toward democratization. These institutional features are country-specific. An important component of the Vietnamese experience has been great economic and political stability. The one-party structure of the Vietnamese governance system has enabled the country to develop policies and implement them over an extended period of time. This is a potential advantage compared with the complexity of brokering agreements on health policy reform measures that span the political divide in many multiparty democracies. Such political stability helps obviate the risk that health reform mea- sures may be precluded from running their course as a result of frequent changes in government. Competent regulatory bodies are an important institutional element. This is true for both public and private health insurance. Chile in particular has benefited from credible, independent, and efficient institutions. The Superintendencia de las ISAPREs regulates the benefits packages, contribution rates, and services provided by the ISAPREs--private health insurers--providing transparent rules of the game and fostering competition among them. An efficient tax collection mecha- nism and a low rate of tax evasion have also helped to ensure that the necessary resources will be available to finance public spending. At the same time, the success of the Chilean model is highly dependent on an effective public insurance system. Similarly, in Colombia a clear legal and institutional framework has proven important in expanding formal sector insurance coverage. The 1991 Constitution sets the legal framework and provides political legitimacy for the National Social Health Insurance scheme with the participation of the private sector. In Costa Rica, transparency and accountability have likewise been critical ingredients for successful expansion of health care coverage. Political parties, the Costa Rican Social Security Fund, health care providers and other workers, and MOH staff members all had an important voice in shaping the reforms that have resulted in the current health coverage system. Internal institutional reforms have also played a role. Separation of purchasing from provision functions has allowed a single department to concentrate fully on planning, negotiating, monitoring, and evaluating the performance of health providers--and removed conflicts of interest in the purchasing relationship that might had compromised the efficiency of the purchasing process. A strong institutional background has also marked Thailand's successful expansion of health insurance coverage. In Thailand, the proposal for universal coverage (the UCS) was the first bill sent to the national parliament because it had 64 Figure 5.2 Political Freedom Trends in Case Countries, 1900­2004 Good Sri Lanka Vietnam Thailand 10 10 10 50 50 50 Pract score 0 score 0 score 0 ice ­5 ­5 ­5 in Polity Polity Polity ­10 ­10 ­10 H ealth 1900 1920 1940 1960 1980 2000 1900 1920 1940 1960 1980 2000 1900 1920 1940 1960 1980 2000 Financ Estonia Tunisia The Kyrgyz Republic ing 10 10 10 50 50 50 score 0 score 0 score 0 ­5 ­5 ­5 Polity Polity Polity ­10 ­10 ­10 1900 1920 1940 1960 1980 2000 1900 1920 1940 1960 1980 2000 1900 1920 1940 1960 1980 2000 Colombia Chile Costa Rica 10 10 10 50 50 50 score 0 score 0 score 0 ­5 ­5 ­5 Polity Polity Polity ­10 ­10 ­10 1900 1920 1940 1960 1980 2000 1900 1920 1940 1960 1980 2000 1900 1920 1940 1960 1980 2000 Source: POLITY IV data set. Note: A polity score of +10 represents a fully functional democracy and ­10 represents an autocracy. Enabling Factors for Expanding Coverage 65 received more than 50,000 signatures from citizens. Thailand's experience with a variety of payment schemes over three decades helped inform the reimbursement system in its Universal Coverage Scheme (UCS), designed to use a mix of capita- tion, global budgets, and DRGs to put financial risks on providers to contain expenditures. Some of the country studies suggest that democratic accountability can be a key facilitating factor, encouraging health systems to be responsive to the needs of the poor. Democratic accountability, achieved through free and fair elections, should not be confused with political decentralization or community participa- tion. Sri Lanka has had little success with community participation in local gov- ernment and in the running of individual facilities--but the most basic decisions about the health system are not made at the local level, but nationally, where the poor do have a voice. Well-Educated Population Adult literacy is another important factor that distinguishes most of these coun- tries from others. One can also speculate that high literacy rates made it easier to sell the reform to the population, resulted in better demand-side response to coverage expansions with respect to health seeking behaviors, and enabled indi- viduals to be more informed consumers of both public and private care. High educational levels certainly appear to have played a facilitative role in Sri Lanka, for instance in terms of influencing demand for health by placing a high priority on this issue and sustaining it through changes in government. High literacy levels also seem to have played an important role in Vietnam. The population's aware- ness of the importance of health is high and that appears to have helped facilitate the choice and implementation of reforms. Policy Factors In addition to strong economic growth, expansion of health care coverage requires a sustained financial commitment and continuity in other policies. Financial Resources Committed to Health Sustained financial investments in the health sector are crucial. Expanding health insurance coverage--either in terms of the population covered or the depth of the benefits package--is likely to lead to substantial increases in health care spending, partly due to moral hazard and the need to subsidize the poor and the vulnerable. The insured tend to consume more physician visits, medications, and other services than the uninsured, necessitating budget ceilings for public programs. Implementing and expanding coverage is easier when public finances are adequate to bear these costs during periods of expansion and robust economic growth. Chile devoted some 6.1 percent of GDP to health in 2004--US$359 per capita--slightly below average for its income level. Health spending accounts for some 13.1 percent of the government's budget, above average for its income level. 66 Good Practice in Health Financing To contain public health spending, the Chilean reform set budgetary ceilings, based on previous year's spending. Similarly, as health insurance coverage in Colombia has expanded from 20 per- cent to 78 percent of the population, total health expenditures have grown from 6.2 percent of GDP in 1993 to 7.8 percent in 2004, mainly as a result in increases in public expenditures. Costa Rica has a GDP per capita of US$4,349, about average for an upper-middle-income country. In terms of health, Costa Rica devotes some 6.6 percent of its GDP to health, above average for its income level, and spends about US$290 per capita on health, which is average. Some 77 percent of total health spending is public, well above other countries with similar income levels. Only 20 percent of total health spending comes from out-of-pocket payments. In Estonia, 87 percent of total health expenditures come through the health insurance system, and the public share of health spending is 76 percent. Including MHI spending, 11.5 percent of the government's budget is devoted to health, about the average for Estonia's income level. Thailand, the Kyrgyz Republic, and Sri Lanka stand out as exceptions to the general rule of substantial investment in health to achieve strong health coverage results. All three countries spend less than US$100 per person per year on health care. In Thailand and the Kyrgyz Republic, health expenditures account for just 3.5 percent and 5.6 percent of GDP, respectively. However, health expenditures account for 11.2 percent of the government budget in Thailand, and 8.4 percent in the Kyrgyz Republic. In Sri Lanka, total health spending--at US$43 per capita--represents only 4.3 percent of GDP. Sri Lanka has been able to overcome some of the difficulties imposed by limited health financing by realizing improvements in efficiency-- achieving efficiency gains in the public health system of between 1 and 4 percent a year. Sri Lanka's life expectancy of 75 years suggests that health spending alone is not the most important factor in determining health outcomes. Private Financing With the exception of Sri Lanka, the cases all suggest that complementary private funding is an important factor for a sustainable health system. In Chile, the required contribution has risen over time, from 4 percent of the beneficiary's income in 1981 to 7 percent. Countries with lower income levels will likely need to charge beneficiaries higher rates to offer a comparable benefits package. Commitment to Equity and Solidarity Though difficult to measure, social solidarity is also an important facilitating fac- tor. In general, the greater the tradition of solidarity within a country--that is, the greater the number of systems and institutions working toward solidarity--the better are the chances of success for pooling arrangements, including social secu- rity and health insurance. Enabling Factors for Expanding Coverage 67 The governments of most of the nine countries have demonstrated strong fis- cal commitment to finance coverage expansions for the poor through a combina- tion of dedicated revenue sources and cross-subsidization through risk pooling. Many have subsidized both insurance contributions and copayment rates of the poor. For example, in Chile, the public insurer (FONASA) covers 96 percent of the poor and fully subsidizes health services for both this group and the unemployed. Colombia uses proxy-means testing to allocate subsidies to the poor in its social health insurance system. Thailand set up a new, completely revamped, general revenue funded program with no copayments for the poor, elderly, children, and disabled. Sri Lanka covers its entire population through general revenues. The poorest are also fully exempt from copayments in the Kyrgyz Republic and Tunisia. Even in Vietnam where targeting and implementation of pro-poor policies is weaker, there remains strong government support to subsidize the poor's con- tributions. Additionally, creating larger risk pools to facilitate cross-subsidization from rich to poor has also demonstrated the pro-poor fiscal commitment. The transition in the Kyrgyz Republic from a system characterized by fragmented rev- enue collection and pooling to one with larger risk pools--first at the regional level and in 2006 at the national level--has provided the foundation for pro-poor fiscal redistribution. This move offers the prospect to improve both vertical and horizontal equity in health financing, and perhaps especially so in a country with a small population. Estonia's MHI system uses premiums from the working pop- ulation to subsidize the remainder. In Colombia, the health system is largely publicly financed--with 86 percent of total health spending coming from public sources. Only 6.9 percent of health spending comes from out-of-pocket expenditures by households, well below the average of countries with similar income levels. Solidarity is a key principle of the system where public subsidies to individual insurance premiums for the poor are allocated on the basis of a proxy-means testing index. Health Coverage and Financing Mandates Solidarity--in terms of sharing both financing and risk broadly across a popula- tion requires coverage mandates. Chile and Estonia provide leading examples of mandates in the context of employment-based insurance. In Chile, health insur- ance is compulsory for workers in the formal sector and voluntary for those in the informal sector. Members of the health insurance system must contribute 7 per- cent of their earnings up to a ceiling of about US$2,000 as a premium, and may make additional voluntary contributions, in addition to copayments at the point of service. In Estonia, the health insurance system is mandatory without an opt-out possi- bility. Private insurance was allowed to be taken to cover expenses not covered by the mandatory health insurance. Contributions are related to being active in the workforce. Noncontributing individuals (e.g., children, pensioners) represent almost half (49 percent) of the insured population, and their expenses are implicitly 68 Good Practice in Health Financing subsidized by the others. The state officially contributes for only about 4 percent of the covered population. Having a single risk pool and clear regulatory frameworks has allowed the Estonian Health Insurance Fund (EHIF) to be an efficient adminis- trator of MHI funds and perform as an effective purchaser of services. Risk-Pool Consolidation The risk-pool consolidation trend, which is increasingly apparent in both high- income and middle-income countries, is also found here in several of the case studies. Estonia and the Kyrgyz Republic have consolidated geographically defined pools into single, national-level pools. Thailand has reduced its number of major risk-pooling schemes from four to three. Sri Lanka has a single national risk pool. Among the nine cases, there are no cases of expansion in the number of risk pools, only consolidations. Limits to Decentralization Estonia and the Kyrgyz Republic both established decentralized, regional risk pools then recentralized them at the national level to improve both efficiency and equity. Estonia likewise decentralized certain delivery system functions then recentralized some of them as some local governments had been captured by medical care providers and, from a provision perspective, certain economies of scale and scope required larger geographic entities to be fully exploited. Primary Care Focus Every country case but Sri Lanka mentions primary care, disease prevention, or both as an important delivery system change. In some cases like Estonia and the Kyrgyz Republic, this entailed a complete reorientation of the medical system toward a family physician model. While the Sri Lanka case argues that a key suc- cess factor was a hospital-based system geographically distributed in all areas of the country, without a specific focus on cost-effectiveness, hospital outpatient units likely proved effective purveyors of primary care. In other words, the real issue may be that appropriate care can just as readily be efficiently distributed in hospital-based settings as under the more usual rural clinic and community health worker approaches. Such a focus on primary care helps explain the good health outcomes found in these countries. Implementation Factors Political commitment is not a sufficient condition to ensure successful implemen- tation of a reform. Political ability and leadership to implement the reform are equally critical. Such implementation requires a careful calibration of revenue changes with changes in the organization, composition, and management of the delivery system; dealing with the major stakeholder groups; and ensuring uptake on the demand side. Enabling Factors for Expanding Coverage 69 Carefully Sequenced Health Service Delivery and Provider Payment Reforms Perhaps it is obvious that significant coverage changes would have to be accompa- nied by serious delivery system reforms. This is what happened in all nine case countries. The reforms were of two types (1) enhancements in delivery capacity to accommodate the increases in coverage; and (2) significant downsizing, rational- ization, and upgrading of medical practice standards in the former Soviet Union (FSU) countries. Sequencing of delivery system changes with financing reforms has been an important element in all the reforms. Dealing with physical infrastructure and human resources for health (HRH) have gone hand in hand. Several of the coun- tries, perhaps most notably Thailand, have implemented advanced HRH policies to staff rural areas. Implementation of modern contracting techniques, provider payment systems, and risk-sharing mechanisms has also been an important ele- ment in all the reforms. The financial resources devoted to health have increased in these countries, but most have inhibited cost escalation through prospective payment and other approaches to supply-side risk sharing. Pharmaceuticals have also been a reform target, given their high share of total and out-of-pocket health spending in terms of essential drug lists, pricing, and practice patterns (particu- larly in the FSU). Good Information Systems and Evidence-Based Decision Making Intrinsically related to evidence-based policy, information systems were high- lighted by a number of countries. Evidence-based decision making was a key enabling factor in the Thai reform. It was also important in Estonia and the Kyr- gyz Republic as they adopted modern treatment protocols and, along with several other countries, modern provider payment reforms. Estonia in particular has taken a lead in terms of investments in information technology. In the mid-1990s, bills from providers were transferred to the health insurance fund on paper. All health care providers are now equipped with personal computers; electronic com- munication links the Estonian Health Insurance Fund with its patients, employ- ees, and providers. In addition to facilitating efficiency in provision, electronic data transmitting has enabled the collection of high-quality data for detailed analysis such as benchmarking of providers and monitoring service utilization by specific vulnerable population groups. The Kyrgyz Republic also has a well- developed health information system that has facilitated policy development, especially the new provider payment mechanisms. Tunisia's reforms were embodied in its investments in administrative capac- ity--particularly related to management and information technology in the pub- lic hospital sector. Information systems also have played an important role in Costa Rica. Investments in computer systems have allowed the financial managers of the CCSS to monitor the flow of revenues on a daily basis, instead of with the 70 Good Practice in Health Financing previous 30-day delay. These systems have facilitated the implementation of new payment mechanisms designed to enhance provider efficiency and performance by monitoring management agreements between the CCSS and primary care providers that explicitly define coverage targets linked to payments. In Colombia, the previous implementation of a program for the development of community-based health insurers in the rural areas allowed the National Social Health Insurance system to quickly sign up a large number of enrollees through- out the country. The availability of SISBEN, an information system providing updated information on beneficiaries' economic status, has provided the system with an instrument for identifying its target population and directing subsidies to the neediest. Strong Stakeholder Support Strong stakeholder support from the bureaucracy, consumers, medical societies, and other stakeholders were identified as important elements of the reform processes in a number of the case countries. The Estonian reform was strongly supported by the medical association, which had much to gain from the move- ment away from the old Soviet-style system. The support and voice of the rural poor were highlighted as important factors in the success and sustainability of the Sri Lanka reform. Strong support from the people through their political leaders has been a hallmark of the reforms in Thailand. The Costa Rican reforms have had strong support from the political parties, the CCSS itself, medical care providers, workers, and MOH staff members as they all had a voice in shaping the reform. Efficiency Gains and Copayments Used as Financing Mechanisms Efficiency gains are an important means of both financing a health system and providing fiscal space for coverage expansions. Efficiency gains were a critical ele- ment in the expansions in the FSU countries and Sri Lanka. Thailand created fiscal space by reducing defense expenditures and has implemented modern provider payment systems to ensure efficiency in its new UCS program. Costa Rica has also focused on efficient purchasing arrangements to reduce outlays. Copayments have been used as both a financing mechanism and a means of containing moral hazard in Chile, Vietnam, Estonia, and to a lesser extent in Sri Lanka, the Kyrgyz Republic, and Thailand. Flexibility and Mid-Course Corrections Considering the long implementation periods of the reforms in all of these coun- tries and the careful sequencing of delivery system changes with the coverage expansions, it is not surprising that both evidence-based policies and mid-course corrections have affected implementation. Thailand prides itself on the use of evidence-based policies in all aspects of its reform efforts. It was one of the earliest developing countries to institute comprehensive national health accounts Enabling Factors for Expanding Coverage 71 (NHA), and it has both used and significantly contributed to the global HRH evidence base. Thailand has implemented modern provider payment and risk- sharing approaches for several of its insurance programs. It has undertaken detailed studies on equity and impoverishment due to high medical expenses. Estonia has exhibited marked flexibility in its reform and has, for example, reversed decentralization polices for risk pools and delivery system responsibili- ties. Colombia, Thailand, Sri Lanka, Vietnam, and the Kyrgyz Republic have made significant and explicit choices in the way that the care for the poor is subsidized through the MHI schemes. Next Steps for Enhancing the Global Evidence Base This study has attempted to assess "good performance" in the context of major health financing reforms. Defining "good performance" in terms of complex financing expansions involving revenue collection, risk pooling, purchasing, and related delivery system changes is, as discussed, a complex venture. This study might be viewed as a tentative first step to provide some guidance or indeed a straw man for the global community to consider in its quest to collate the global evidence base on health financing. These concluding observations from part I of this report attempt to lay out the necessary future steps in terms of measurement, data, evaluation, and dissemination to allow needed refinements in the global evi- dence base on health care financing reforms. One of the largest impediments in developing the global evidence base on health financing reforms is the difficulty of defining what is meant by "good per- formance" or "success." Second order questions of measurement and evaluation/ attribution then arise. As discussed, all of these are problematic. Nevertheless, the global community needs to reach some consensus in these areas if it is serious about the development and dissemination of global evidence as a guide to policy making. To do this, at least five important actions are needed: · Define rigorously what successes and good performance mean in health financing.1 · Collect standardized and appropriate qualitative and quantitative information. · Describe health systems characteristics in sufficient detail so that the critical components and their interactions can be clearly identified and assessed. · Undertake rigorous evaluations. · Disseminate these evaluations to all stakeholders in a policy-relevant and user- friendly manner. Definition of "Success" and "Good Performance" "Good performance" with respect to health financing reforms is defined in this study in terms of countries that have expanded health coverage and financial pro- tection in the context of reasonable and sustainable spending levels, equity, and 72 Good Practice in Health Financing good health outcomes relative to other comparable income countries. This defin- ition is derived from the definitions of basic health financing function definitions and objectives outlined in chapter 2. The presumption here is that if revenue col- lection, risk pooling, and purchasing functions are performed effectively, the result will be better health outcomes, financial protection, equity, efficiency, and financial sustainability of the health system. "Good practice" performers were chosen on the basis of both a priori judg- ments from an expert steering committee and comparative analyses of a number dimensions--including financial protection, health outcomes, equity, efficiency, and sustainability. Financial protection is explicitly defined in terms of expansions in the number of individuals who are formally covered under a publicly or pri- vately financed health program as well as various measures of financial protection for that individual/household.2 The serious data gaps are openly discussed, as well as the inherent causality attribution problems. Numerous interactive factors affect health outcomes, and the multiple enabling factors are hard to attribute in so complex an area due to the difficulties of performing case control studies, including randomized trials. From the detailed case studies, the authors have attempted to highlight the critical demographic, geographic, political, socioeconomic, and health and nonhealth sector-related enabling factors. Based on the case studies and comparative data, an attempt is then made to assess whether there are common enabling factors that are likely to account for the observed performance. The factors identified make sense on conceptual grounds and are useful additions to the sparse global evi- dence base. The more important contribution, however, is not the findings in themselves, but rather the illustration of the need for the international commu- nity to do a better job in measurement, data development, evaluation, and dis- semination of the evidence. Need to Collect Standardized and Appropriate Qualitative and Quantitative Information As vividly shown in chapters 3 and 4, as well as in the case studies in part II, base- line indicators of coverage, financial protection, equity, and efficiency are gener- ally lacking, often both pre- and postreform. Thus, first and foremost attention must be given to developing the necessary baseline and postreform information. Specifically, although most of the country cases contained information on num- bers of individuals eligible for various health financing programs, there was little information on enrollment and the breadth and depth of coverage both pre- and postreform. The lack of measures of financial protection--particularly for prere- form periods--is even more problematic. The postreform information came mostly from special surveys, either country-based household surveys or the "World Health Survey 2003 Results" (WHO 2003b). Definitions of total expendi- ture, income, and out-of-pocket spending, as well as financial protection, among Enabling Factors for Expanding Coverage 73 these two different sources are often not equivalent and sometimes are based on different recall periods, making comparisons difficult. Furthermore, attempts to measure changes in the aggregate spending (and rev- enue) accompanying these reforms (needed to assess macro efficiency and sus- tainability) are attenuated by the lack of reliable NHA time series information and information on revenues, particularly subnational revenues. Similarly, attempts to measure the equity aspects are limited by the lack of microdata on out-of-pocket spending, revenue incidence information, and utilization data both pre- and post- reform. Few studies bother to discuss spending efficiency in other than generic terms regarding mismatches of basic benefits packages with current and future projected disease burdens. If individual countries--and the international community--are serious about improving the global evidence base on health financing reforms, they will need to start by ensuring the development and institutionalization of basic information such as NHA, detailed national and subnational revenue, expenditure, and utiliza- tion, output, and outcome data necessary for benefit, revenue, and net incidence assessments, and accurate counts of health program­eligible and health program­ enrolled individuals. The recently established Health Metrics Network needs to give these areas appropriate focus and publicity, and countries need incentives to systematically develop and collect such information. This information will also be needed to assess aid effectiveness and the equity and efficiency of public spending. To measure health system performance fully, macrodata sets will need to be linked with microdata--not just from household surveys, but also from provider and epidemiological surveys. Need Health Systems Characteristics Information Effective health coverage is important to ensure good health outcomes and finan- cial protection. But coverage (even effective coverage in a WHO sense) is neither a necessary nor sufficient condition for reaching either of these two goals. Beyond their coverage status, numerous demographic, socioeconomic, geographic, cul- tural, and political factors affect individuals' behaviors, as do other complex insti- tutional interrelations that shape health system performance in terms of health status and financial protection. Understanding these factors and, to the extent fea- sible, controlling and assessing their impact requires both quantitative and qualita- tive information on these factors themselves and on their behavioral interactions. Although the growing health policy literature (e.g., WHO Commission on Macro- economics and Health, Millennium Project, WHO) provides important insights into the most critical factors (e.g., female education, good public sector manage- ment), detailed specification of these behavioral relationships is far from a science or complete. Nevertheless, to the extent feasible, it is important to both specify these factors upfront quantitatively and qualitatively and to control for them when attempting to draw conclusions about performance. 74 Good Practice in Health Financing Rigorous Evaluations All new global initiatives, including the new HNP Strategy of the World Bank (World Bank 2007b), have decried the insufficient attention given to monitoring and evaluation, as well as to the previously discussed collection of critical baseline data. Studies of health financing reforms are equally implicated. Considering the complexity of evaluating such reforms and the many other factors affecting health outcomes and financial protection, the gold standard of randomized control trials or even a silver standard of case control studies with a nonrandomized control group may be less feasible as techniques for evaluating large financing changes. Although "before" and "after" evaluation techniques are more the order of the day, they too can lead to spurious conclusions, given the difficulty of teasing out the impact of secular trends that may affect health system outcomes from those that result from specific health policy interventions. Nevertheless, it goes without saying that the global community must do a bet- ter job to develop the basic evidence and collate the existing base using the most appropriate evaluation tools. At the very least, key institutional and behavioral factors likely to affect health coverage, health status, and financial protection out- comes should be specified. The impact of these factors on the outcomes of the financing changes should then be evaluated as best as possible, qualitatively and quantitatively, through appropriate statistical and experimental methodologies. Evaluating failures as well as success would be another needed departure from much of the present effort. Success or good performance cannot be fully evaluated without understanding what constitutes failure or poor performance--as a great deal can also be learned from failures. However, little has been written on major reforms that have not lived up to expectations such as the adoption of regionally based Mandatory Health Insurance in Russia and much of the former FSU. Although the failure of the Clinton reform in the United States has been exten- sively analyzed and the British attempts to introduce market incentives into the NHS have attracted great interest, their lessons for developing countries have not been well teased out. Dissemination of Evidence Collation and dissemination of the global evidence is one of the key features of every new initiative--Health Metrics Network (HMN), Global Health Workforce Alliance (GHWA), Scaling Up for Better Health (IHP+), the Global Alliance for Vaccines and Immunization (GAVI), the Global Fund to Fight AIDS, Tuberculosis and Malaria (GFATM), and the Bank's new HNP Strategy. In health financing, this process is especially complex, because of the issues highlighted above, but also because of the behavior of donors and nongovernmental organizations (NGOs). Irrespective of the global evidence, donors and NGOs often push particular policy positions and/or one-size-fits-all solutions. No one is in charge of global health policy (Schieber et al. 2007), much less global health financing policy, so there is no venue for reaching agreement on appropriate indicators and responsibility for Enabling Factors for Expanding Coverage 75 development, institutionalization, and collation of data. One ancillary outcome of the IHP+ and its focus on health systems strengthening, where the World Bank and the WHO are the secretariat for six other major development partners, might be some agreement on the needed indicators and a division of responsibility to ensure consistent development and institutionalization. This is consistent with the objectives of the IHP+ and other initiatives, including those of the G-8,3 which has shown a strong interest in promoting risk pooling in low-income set- tings. The IHP+ initiative would therefore seem to present a good opportunity for collaborative development of the necessary information bases. Conclusions The global community has done a poor job so far in developing, collating, and disseminating the global evidence base on what works in health financing and what does not. The G-8, as well as many individual countries, attach growing importance to health financing and risk pooling, while some organizations are pushing a single magic-bullet approach. It is incumbent on the global commu- nity to take charge and move this agenda forward in an analytically sound and policy-relevant manner. It is hoped that global stakeholders, through their aid- effectiveness forums and the G-8, IHP+, and Paris Declaration Process, will rise to the occasion and put this desperately needed effort on track. Endnotes 1. For purposes of this study, only good performance (not success) can be discussed, because no failures were evaluated. 2. For purposes of this financing assessment "coverage" is defined in an insurance con- text, not the more clinical contexts of WHO, which relate to utilization conditional on need. See Tanahashi (1978). 3. G-8 members are Germany, France, the United Kingdom, Italy, Japan, the United States, Canada (since 1976), and Russia (since 1998). The European Commission is also represented at all the meetings. German Chancellor Angela Merkel, G-8 president, invited Algeria, Brazil, China, the Arab Republic of Egypt, India, Mexico Nigeria, Senegal, and South Africa to participate in the outreach meetings in Heiligendamm, Germany, in June 2007. 76 Good Practice in Health Financing References for Part I Abel-Smith, Brian. 1992. "Health Insurance in Developing Countries: Lessons from Experi- ence." Health Policy and Planning 7 (3): 215­26. Adams, Susan. 2005. "Vietnam's Health Care System: A Macroeconomic Perspective." Paper prepared for the International Symposium on Health Care Systems in Asia, Hanoi, Viet- nam, January 21­22, 2005, Washington, DC, International Monetary Fund. ADB (Asian Development Bank). 2006. Measuring Policy Effectiveness in Health and Educa- tion. Manila: ADB. Antioch, K. M., and M. K. Walsh. 2004. "The Risk-Adjusted Vision beyond Casemix (DRG) Funding in Australia--International Lessons in High Complexity and Capitation." Euro- pean Journal of Health Economics 5 (2): 95­109. Aris, B. 2004. "Slovenia's Health System Out-Performs EU Neighbours. Peace and Prosper- ity Have Helped Slovenia Ministers Push Through Effective Reforms." Lancet 363 (9427): 2143­46. Asfaw, A., and J. von Braun 2004. "Can Community Health Insurance Schemes Shield the Poor Against the Downside Health Effects of Economic Reforms? The Case of Rural Ethiopia." Health Policy 70 (1): 97­108. Baeza, C., P. Crocco, M. Núñez, and M. Shaffer. 2002. Toward Decent Work: Social Protection in Health for All Workers and Their Families. Strategies and Tools against Social Exclusion and Poverty (STEP). Geneva: International Labour Organization. Baeza, C., and T. Packard. 2006. Beyond Survival: Protecting Households from Health Shocks in Latin America. Washington, DC/Palo Alto, CA: World Bank/Stanford University Press. Baeza, Cristian, and Fernando Montenegro Torres. 2004. "Selecting Health Care Providers." Health, Nutrition and Population (HNP) Discussion Paper. World Bank, Washington, DC. Balabanova, D., and M. McKee. 2004. "Reforming Health Care Financing in Bulgaria: The Population Perspective." Social Science & Medicine 58 (4): 753­65. Barnum, Howard, Joseph Kutzin, and Helen Saxenian. 1995. "Incentives and Provider Pay- ment Methods." World Bank Human Resources Development and Operations Policy Working Paper No. 51, Washington, DC. http://www.worldbank.org/html/extdr/hnp/ hddflash/hcwp/hrwp043.html. Bassett, Mark. 2005. "Frameworks for Analyzing Health Systems, Health Financing and the Regulation of Health Insurance." Paper prepared for Conference on Private Health Insurance in Developing Countries cosponsored by the World Bank and the Wharton School of the University of Pennsylvania, at the Wharton School, Philadelphia, Pa., March 15­16, 2005. Besley, T., and R. Burgess. 2002. "The Political Economy of Government Responsiveness: Theory and Evidence from India." Quarterly Journal of Economics. 117 (4): 1415­51. Bhushan, I., S. Keller, and B. Schwartz. 2002. "Achieving the Twin Objectives of Efficiency and Equity: Contracting Health Services in Cambodia." ERD Policy Brief No. 6. Manila: Asian Development Bank. Blau, David M. 1985. "Self-Employment and Self-Selection in Developing Country Labor Markets." Southern Economic Journal 51 (2): 351­63. Enabling Factors for Expanding Coverage 77 Bokhari, F., Y. Gai, and P. Gottret. 2007. "Government Health Expenditures and Health Outcomes. Health Economics 16 (3): 257­73. Boncz, I., J. Nagy, A. Sebestyen, and L. Korosi. 2004. "Financing of Health Care Services in Hungary." European Journal of Health Economics 5 (3): 252­58. Bosch, X. 2004. "French Government Approves Unpopular Health Reforms." Lancet 363 (9427): 2148. Carrin, Guy, and P. Hanvoravongchai. 2003. "Provider Payments and Patient Charges as Policy Tools for Cost-Containment: How Successful Are They in High-Income Coun- tries?" Human Resources for Health 1 (6): downloaded from http://www.human- resources- health.com/content/pdf/1478-4491-1-6.pdf. Carrin, Guy, and Chris James. 2004. "Reaching Universal Coverage via Social Health Insur- ance: Key Design Features in the Transition Period." EIP/FER/DP.04.2, World Health Organization, Geneva. Castañeda, Tarsicio. 2005. "Targeting Social Spending to the Poor with Proxy­Means Test- ing: Colombia's SISBEN System." Social Protection Discussion Paper Series, No. 0529, Social Protection Unit, Human Development Network, World Bank, Washington, DC. Charoenparij, Sriracha, Somsak Chunharas, Dayl Donaldson, Daniel Kraushaar, Supasit Pannorunothai, Sutham Pinjaroen, Supattra Srivanichakorn, Paibul Suriyawongpaisal, Viroj Tangcharoensathien, and Aree Valyasevi. 1999. "Health Financing in Thailand: Final Integrated Report." Thailand/Boston: Health Management and Financing Study Project/ Management Sciences for Health. Cheng, T. 2003. "Taiwan's New National Health Insurance Program: Genesis and Experi- ence So Far." Health Affairs 22: 61­76. Chollet, Deborah, and Maureen Lewis. 1997. "Private Insurance: Principles and Practice." In "Innovations in Health Care Financing," ed. George Schieber, pp. 77­103. World Bank Discussion Paper 365. World Bank, Washington, DC. Clark, Mary. 2002. "Health Sector Reform in Costa Rica: Reinforcing a Public System." Paper prepared for Woodrow Wilson Center Workshops on the Politics of Education and Health Reforms, Washington, DC, April 2002. Colombo, Francesca, and Nicole Tapay. 2004. "Private Health Insurance in OECD Coun- tries: The Benefits and Costs for Individuals and Health Systems." DELSA/ELSA/WD/ HEA(2004)6, Organisation for Economic Co-operation and Development, Paris. Commission on Macroeconomics and Health. 2002. "Macroeconomics and Health: Invest- ing in Health for Economic Development." In Report of the Commission on Macroeco- nomics and Health, Commission on Macroeconomics and Health. Geneva: World Health Organization. Cornell, J., J. Goudge, D. McIntyre, and S. Mbatsha. 2001. "South African National Health Accounts: The Private Sector." National Health Accounts Research Team and National Department of Health, Pretoria. Csonka, A. 2004. "Hungary Makes Slow Progress in Health-System Reform. Private Funds, Which Could Rescue Hungary's Ailing Health-Care System, Have Been Slow to Materi- alise." Lancet 363 (9425): 1957­60. 78 Good Practice in Health Financing Cutler, D. M., and A. Lleras-Muney. 2006. "Education and Health: Evaluating Theories and Evidence." National Poverty Center Working Paper Series 06­19, University of Michigan, Ann Harbor. Davoodi, H., E. Tiongson, and S. Asawanuchit. 2003. "How Useful Are Benefit Incidence Analyses of Public Education and Health Spending?" IMF Working Paper, International Monetary Fund, Washington, DC. Deaton, A. 1997. The Analysis of Household Surveys: A Microeconometric Approach to Devel- opment Policy. Baltimore, MD: Johns Hopkins University Press. Deaton, A., and S. Zaidi. 2002. "Guidelines for Constructing Consumption Aggregates for Welfare Analysis." LSMS Working Paper 135, World Bank, Washington, DC. Docteur, E., and H. Oxley. 2003. "Health-Care Systems: Lessons from the Reform Experi- ence." DELSA/ELSA/WD/HEA(2003)9, Organisation for Economic Co-operation and Development, Directorate for Employment, Labour and Social Affairs, Paris. Doherty, Jane, Stephen Thomas, and Debbie Muirhead. 2002. "Health Financing and Expenditure in Post-Apartheid South Africa, 1996/97­1998/99." The National Health Accounts Project, Centre for Health Policy, University of the Witwatersrand and Health Economics Unit, University of Cape Town. Donaldson, D., S. Pannarunothai, and V. Tangcharoensathien. 1999. Health Financing in Thailand Summary Review and Proposed Reforms. Nonthaburi, Thailand: Health Systems Research Institute. Ellis, Randall P., and Thomas G. McGuire. 1993. "Supply-side Cost Sharing in Health Care." Journal of Economic Perspectives 7: 135­51. ------. 1996. "Hospital Response to Prospective Payment: Selection and Moral Hazard Effects." Journal of Health Economics 15 (3): 257­79. El-Saharty, Sameh, Axel Róala, and Hugh Waters. 2006. "Etude du secteur de la santé en Tunisie." Human Development Department, Middle East and North Africa Region, World Bank, Washington, DC. Ensor, Tim. 1999. "Developing Health Insurance in Transitional Asia." Social Science & Medicine 48 (7): 871­79. Ensor, Tim, and Stephanie Cooper. 2004. "Overcoming Barriers to Health Service Access and Influencing the Demand Side through Purchasing." Health, Nutrition and Popula- tion (HNP) Discussion Paper 31598, World Bank, Washington, DC. Ensor, Tim, and Robin Thompson. 1998. "Health Insurance as a Catalyst to Change in For- mer Communist Countries." Social Science & Medicine 43 (3): 203­18. Figueras, Joseph, Ray Robinson, and Elke Jakubowski, eds. 2005. "Purchasing to Improve Health Systems Performance." European Observatory on Health Systems and Policies Series. England: Open University Press. Filmer, D., and L. Pritchett. 1999. "The Impact of Public Spending on Health: Does Money Matter." Social Science & Medicine 49 (10): 1309­23. Frenk, Julio, Jaime Sepúlveda, Octavio Gómez-Dantés, and Felicia Knaul. 2003. "Evidence- Based Health Policy: Three Generations of Reform in Mexico." Lancet 362 (9396): 1667­71. Gertler, P. 2004. "Do Conditional Cash Transfers Improve Child Health? Evidence from PROGRESA's Controlled Randomized Experiment." Health, Health Care, and Economic Development 94 (2): 336­41. Enabling Factors for Expanding Coverage 79 Gertler, P., and J. van der Gaag. 1988. "Willingness to Pay for Social Services in Develop- ing Countries." Measurement Study (LSMS) Working Paper 45, World Bank, Washing- ton, DC. Gertler, Paul, and Jonathan Gruber. 2002. "Insuring Consumption against Illness." Ameri- can Economic Review 92 (1): 51­76. Gill, I., and N. Ilahi. 2000. "Economic Insecurity, Individual Behavior, and Social Policy." Background Paper for the Regional Study on Economic Insecurity, Office of the Chief Economist, Latin America and the Caribbean, World Bank, Washington, DC. Gill, I., T. Packard, and J. Yermo. 2004. Keeping the Promise of Social Security in Latin Amer- ica. Latin American Development Forum. Washington, DC: World Bank. Goglio, Alessandro. 2005. "In Search of Efficiency: Improving Health Care in Hungary." Organisation for Economic Co-operation and Development, Paris. Gomez-Dantes, O., J. Gomez-Jauregui, and C. Inclan. 2004. "La equidad y la imparcialidad en la reforma del sistema mexicano de salud." [Equity and fairness in the Mexican health system reform]. Salud Pública de México 46 (5): 399­416. González-Rossetti, A., and T. Bossert. 2000. "Enhancing the Political Feasibility of Health Reform: A Comparative Analysis of Chile, Colombia and Mexico." Harvard School of Public Health, Department of Population Studies and International Health, Data for Decision Making Project and the Latin American and Caribbean Health Sector Reform Initiative, Cambridge, MA. Gottret, Pablo, and George Schieber. 2006. Health Financing Revisited. Washington DC: World Bank. Government of Taiwan. 2003. Statistical Yearbook. http://www.gio.gov.tw/Taiwan-website/ 5-gp/yearbook/chpt02.htm#1. Gwatkin, Davidson. 2004. "Are Free Government Health Services the Best Way to Reach the Poor?" Health, Nutrition and Population (HNP) Discussion Paper 29185, World Bank, Washington, DC. Habicht, J., K. Xu, A. Couffinhal, and J. Kutzin. 2006. "Detecting Changes in Financial Protection: Creating Evidence for Policy in Estonia." Health Policy and Planning 21 (6): 421­31. Hauck, Katharina, Peter C. Smith, and Maria Goddard. 2004. "The Economics of Priority Setting for Health Care: A Literature Review." Health, Nutrition and Population (HNP) Discussion Paper 28878, World Bank, Washington, DC. Hlavacka, S., R. Wágner, and A. Riesberg. 2004. "Health Care Systems in Transition: Slova- kia." WHO Regional Office for Europe on behalf of the European Observatory on Health Systems and Policies, Copenhagen. Hsiao, W., and P. Shaw. 2007. Social Health Insurance for Developing Nations. Washington, DC: World Bank. Hsaio, William. 2000. "What Should Macroeconomists Know about Health Care Policy? A Primer." IMF Working Paper WP/00/136. International Monetary Fund, Washington, DC. http://www.imf.org/external/pubs/cat/longres.cfm?sk&sk=3701.0 Hu, T. W. 2004."Financing and Organization of China's Health Care." Bulletin of the World Health Organization 82 (7): 480. Ikegami, N., and J. C. Campbell. 2004. "Japan's Health Care System: Containing Costs and Attempting Reform." Health Affairs 23 (3): 26­36. 80 Good Practice in Health Financing Institute of Policy Studies. 2001. "Equity in Financing and Delivery of Health Services in Bangladesh, Nepal and Sri Lanka: Results of the Tri-country Study." Colombo: Institute of Policy Studies. Institute of Policy Studies of Sri Lanka. 2001. "Sri Lanka National Health Accounts--Sri Lanka National Health Expenditures 1990­1999." Ministry of Health. Health Policy Pro- gramme supported by the IDA/WB Health Services Project. Ministry of Health, Nutri- tion and Welfare and Institute of Policy Studies, Colombo. ILO (International Labour Office). 2002. "Extending Social Protection through Community-Based Health Organizations--Evidence and Challenges." Discussion Paper. Geneva: International Labour Office. Jack, W. 2005. "Purchasing Health Care Services from Providers with Unknown Altruism." Journal of Health Economics 24 (1): 73­93. Jegers, M., K. Kesteloot, D. De Graeve, and W. Gilles. 2002. "A Typology for Provider Pay- ment Systems in Health Care." Health Policy 60 (3): 255­73. Jesse, M., J. Habicht, A. Aaviksoo, A. Koppel, A. Irs, and S. Thomson. 2004. "Health Care Systems in Transition: Estonia." WHO Regional Office for Europe on behalf of the Euro- pean Observatory on Health Systems and Policies, Copenhagen. John, T., S. Charles, R. Nathan, K. Wilczynska, L. Mgalula, C. Mbuya, R. Mswia, F. Manzi, D. de Savigny, D. Schellenberg, and C. Victora. 2004. "Effectiveness and Cost of Facility- Based Integrated Management of Childhood Illness (IMCI) in Tanzania." Lancet 364 (9445): 1583­94. Joint Learning Initiative. 2004. Human Resources for Health: Overcoming the Crisis. Cam- bridge, MA: Harvard University Press. Karashkevica, J. 2004. "Latvia's Health Care System on the Move." Cahiers de sociologie et de demographie medicales 44 (2): 221­42. Kuszewski, K., and C. Gericke. 2005. Health Systems in Transition: Poland. WHO Regional Office for Europe on behalf of the European Observatory on Health Systems and Poli- cies, Copenhagen. Kutzin, Joseph. 2001. "A Descriptive Framework for Country-Level Analysis of Health Care Financing Arrangements." Health Policy 56: 171­204. ------. 2002. Health Care in Central Asia. Policy Brief. Copenhagen: European Observa- tory on Health Care Systems. La Forgia, Gerard, ed. 2004. "Health System Innovations in Central America--Lessons and Impact of New Approaches." World Bank Working Paper 57, World Bank, Washington, DC. Langenbrunner, John C., and Xingzhu Liu. 2004. "How to Pay? Understanding and Using Incentives." Health, Nutrition and Population (HNP) Discussion Paper 31633, World Bank, Washington, DC. Levine, R. 2007. Case Studies in Global Health: Millions Saved. Washington, DC: Center for Global Development. Levine, R., and M. Kinder. 2004. Millions Saved: Proven Successes in Global Health. Washing- ton, DC: Center for Global Development. Lim, M. K. 2004. "Shifting the Burden of Health Care Finance: A Case Study of Public- Private Partnership in Singapore." Health Policy 69 (1): 83­92. Enabling Factors for Expanding Coverage 81 Limwattananon, S., V. Tangcharoensathien, and P. Prakongsai. 2007. "Catastrophic and Poverty Impacts of Health Payments: Results from National Household Surveys in Thai- land." Bulletin of the World Health Organization 85 (8): 569­48. Liu, Xingzhu, and Sheila O'Dougherty. 2004. "Purchasing Priority Public Health Services." Health, Nutrition and Population (HNP) Discussion Paper 31593, World Bank, Wash- ington, DC. Londoño, J. 1996. "Estructurando pluralismo en los servicios de salud: la experiencia colombiana." Revista de análisis económico 11 (2): 37­60. Londoño, J., and J. Frenk. 1997. "Structured Pluralism: Toward an Innovative Model for Health System Reform in Latin America." Health Policy 41 (1): 1­36. Maceira, Daniel. 1998. "Provider Payment Mechanisms in Health Care: Incentives, Out- comes, and Organizational Impact in Developing Countries." Major Applied Research 2, Working Paper 2. Partnerships for Health Reform Project, Abt Associates Inc. Bethesda, MD. http://www.phrproject.com/publicat/arp/download/mar2/m2wp2fin.pdf. Marek, Tonia, Rena Eichler, and Philip Schnab. 2004. "Resource Allocation and Purchasing in Africa: What Is Effective in Improving the Health of the Poor?" Africa Region Human Development Working Paper Series, World Bank, Washington, DC. Markota, Mladen, Igor Svab, Ksenija Sarazin Klemenèiè, and Tit Albreht. 1999. "Slovenian Experience on Health Care Reform." Croatian Medical Journal 40 (2): 190­94. Marshall, M. G., and K. Jaggers. 2005. "Polity IV Project: Political Regime Characteristics and Transactions, 1800­2004." University of Maryland and George Mason University, College Park, MD, and Fairfax, VA. McGuire, T. 2000. "Physician Agency." In Handbook of Health Economics, ed. J. Newhouse and A. Culyer, 461­536. The Netherlands: Elsevier. Meimanaliev, Tilek. 2003. Kyrgyz Model of Health Care System. Bishkek: Uchkun. Meimanaliev, A. S., A. Ibraimova, B. Elebesov, and B. Rechel. 2005. "Health Care Systems in Transition: Kyrgyz Republic." WHO Regional Office for Europe on behalf of the Euro- pean Observatory on Health Systems and Policies, Copenhagen. Meng, Q., C. Rehnberg, N. Zhuang, Y. Bian, G. Tomson, and S. Tang. 2004. "The Impact of Urban Health Insurance Reform on Hospital Charges: A Case Study from Two Cities in China." Health Policy 68 (2): 197­209. MOH (Ministry of Health, Vietnam). 2001. 55 Years of Developing Vietnam's Revolutionary Health System. Hanoi: Medical Publishing House. MOH (Ministry of Public Health, Thailand). 2002. Thailand Health Profile 1999­2000. Thailand: Bureau of Policy and Strategy. Montenegro, F., and R. Nazareli. 2004. "Health and Household Income and Consumption: A Review of the Literature." Background Paper, World Bank, Washington, DC. Morlock, Laura, Alan Sorkin, and Hugh Waters, eds. Forthcoming. National Health Insur- ance in Taiwan: What Lessons May Be Learned? New York: Elsevier. Morlock, Laura, Hugh Waters, Alan Lyles, Haluk Özsari, and Göksenin Aktulay. 2004. "Health Care Reform in Turkey--Charting the Way Forward." Final report submitted to The Turkish Industrialists' and Businessmen's Association Turk Sanayici ve Isadamlari Dernegi (TÜSIAD). Baltimore, MD: Johns Hopkins University. 82 Good Practice in Health Financing Mossialos, Elias, Anna Dixon, Joseph Figueras, and Joe Kutzin, eds. 2002. "Funding Health- care: Options for Europe." European Observatory on Health Care Systems Series, Open University Press, Buckingham and Philadelphia. Murray, C., and J. Frenk. 2000. "A Framework for Assessing the Performance of Health Sys- tems." Bulletin of the World Health Organization 78 (6): 717­28. Musgrove, P. 1996. "Un fundamento conceptual para el rol público y privado en la salud." Análisis Económico 11 (2): 9­36. Musgrove, Philip, Riadh Zeramdini, and Guy Carrin. 2002. "Basic Patterns of Health Expenditure." Bulletin of the World Health Organization 80: 134­42. O'Donnell, O., Eddy van Doorslaer, Ravi P. Rannan-Eliya, Aparnaa Somanathan, Shiva Raj Adhikari, Deni Harbianto, Charu C. Garg, Piya Hanvoravongchai, Mohammed N. Huq, Anup Karan, Gabriel M. Leung, Chiu Wan Ng, Badri Raj Pande, Keith Tin, Kanjana Tisayaticom, Laksono Trisnantoro, Yuhui Zhang, and Yuxin Zhao 2007. "The Incidence of Public Spending on Health Care: Comparative Evidence from Asia." World Bank Eco- nomic Review 21: 93­123. O'Donnell, O., Eddy van Doorslaer, Ravi P. Rannan-Eliya, Aparnaa Somanathan, Charu C. Garg, Piya Hanvoravongchai, Mohammed N. Huq, Anup Karan, Gabriel M. Leung, Keith Tin, and Chitpranee Vasavid. 2005. "Explaining the Incidence of Catastrophic Expendi- tures on Health Care: Comparative Evidence from Asia." EQUITAP Project, Working Paper 5. Erasmus University, Rotterdam and Institute of Policy Studies, Colombo. OECD (Organisation for Economic Co-operation and Development). 2002. "The Reform of Health Care: A Comparative Analysis of Seven OECD Countries." Health Policy Stud- ies, No. 2. Paris. ------. 2003. OECD Reviews of Health Systems--Korea. Paris: OECD. ------. 2004. Proposal for a Taxonomy of Health Insurance. OECD Study on Private Health Insurance. Paris: OECD. ------. 2005. OECD Reviews of Health Systems--Mexico. Paris: OECD. Packard, T., and A. Barr. 2002. "Revealed Preference and Self-Insurance: Can We Learn from the Self-Employed in Chile?" Policy Research Working Paper 2754, World Bank, Washington, DC. PAHO (Pan American Health Organization). 1999. Cuba Profile of the Health Services Sys- tem. Program on Organization and Management of Health Systems and Services, Divi- sion of Health Systems and Services Development. Washington, DC: PAHO. ------. 2002. Profile of the Health Services System of Costa Rica, 2d ed. Program on Organi- zation and Management of Health Systems and Services, Division of Health Systems and Services Development. Washington, DC: PAHO. Pannarunothai, S. 2002. "Medical Welfare Scheme: Financing and Targeting the Poor." In Health Insurance Systems in Thailand, ed. P. Pramualratana and S. Wibulpolprasert, 62­78. Nonthaburi, Thailand: Health Systems Research Institute. Pannarunothai, S., D. Patmasiriwat, and S. Srithamrongsawat. 2004. "Universal Health Cov- erage in Thailand: Ideas for Reform and Policy Struggling." Health Policy 68 (1): 17­30. Pauly, Mark. 1986. "Taxation, Health Insurance, and Market Failure in the Medical Econ- omy." Journal of Economic Literature, 24 (2): 629­75. Phelps, C. E. 1978. "Illness Prevention and Medical Insurance." Journal of Human Resources 13: 183­207. Supplement. National Bureau of Economic Research Conference on the Economics of Physician and Patient Behavior. Enabling Factors for Expanding Coverage 83 POLITY IV dataset: http://www.cidcm.umd.edu/polity/data/. Pradhan, M., and N. Prescott. 2002. "Social Risk Management Options for Medical Care in Indonesia." Health Economics 11 (5): 431­46. Preker, A. S., G. Carrin, D. Dror, M. Jakab, W. Hsiao, and D. Arhin-Tenkorang. 2002. "Effec- tiveness of Community Health Financing in Meeting the Cost of Illness." Bulletin of the World Health Organization 80 (2): 143­50. Preker, Alexander S., Cristian Baeza, Melita Jakab, and John C. Langenbrunner. 2000. "Arrangements that Benefit the Poor and Excluded Groups." Concept Note. World Bank/ILO/WHO Resource Allocation and Purchasing (RAP) Project, Washington, DC. Preker, Alexander S., and Guy Carrin, eds. 2003. Health Financing for Poor People: Resource Mobilization and Risk Sharing. Washington, DC: World Bank. Preker, Alexander S., and John C. Langenbrunner, eds. 2005. Spending Wisely: Buying Health Services for the Poor. Washington, DC: World Bank. Rannan-Eliya, R. P. 2001. Strategies for Improving the Health of the Poor--The Sri Lankan Experience. Colombo: Institute of Policy Studies Health Policy Programme. Ranson, M. 2002. "Reduction of Catastrophic Health Care Expenditures by a Community- Based Health Insurance Scheme in Gujarat, India: Current Experiences and Challenges." Bulletin of the World Health Organization 80: 613­21. Ravallion, M. 1994. "Poverty Comparisons: A Guide to Concepts and Methods." Living Standards Measurement Study (LSMS) Working Paper LSM 88, World Bank, Washing- ton, DC. Ravallion, M., and B. Bidani. 1994. "How Robust Is a Poverty Profile? Vol. 1." World Bank Economic Review 8 (1): 75­102. Rawlings, Laura B. 2004. "A New Approach to Social Assistance: Latin America's Experience with Conditional Cash Transfer Programs." Social Protection Discussion Paper 0416, World Bank, Washington, DC. Rice, T. 2006. "The Physician as the Patient's Agent." In The Elgar Companion to Health E- conomics, ed. Andrew Jones, 261­68. Cheltenham, UK: Edward Elgar. Rodwin, V. G., and C. Le Pen. 2004. "Health Care Reform in France--The Birth of State- Led Managed Care." New England Journal of Medicine 351 (22): 2259­62. Rosa, R. M., and I. C. Alberto. 2004. "Universal Health Care for Colombians 10 Years after Law 100: Challenges and Opportunities." Health Policy 68 (2): 129­42. Safaei, J. 2006. "Is Democracy Good for Health?" International Journal of Health Services 36 (4): 767­86. Savedoff, W. D., and R. Levine. 2006. "Learning from Development: the Case for an Inter- national Council to Catalyze Independent Impact Evaluations of Social Sector Interven- tions." Center for Global Development, Washington, DC. Schieber, G., C. Baeza, D. Kress, and M. Maier. 2006. "Financing Health Systems in the 21st Century." In Disease Control Priorities in Developing Countries, 2d ed., ed. Dean Jamison, George Alleyne, Joel Breman, Mariam Claeson, David Evans, Prabhat Jha, Anthony Measham, Anne Mills, and Philip Musgrove, pp 255­42. Oxford and New York, NY: Oxford University Press for World Bank. Schieber, G., L. Fleisher, and P. Gottret. 2006. "Getting Real on Health Financing." Finance and Development 43 (4): 46­50. 84 Good Practice in Health Financing Schieber, G., P. Gottret, L. Fleisher, and A. Leive. 2007. "Financing Global Health: Mission Unaccomplished." Health Affairs 26 (4): 921­34. Schieber, George, ed. 1997. Innovations in Health Care Financing. Washington, DC: World Bank. Schieber, George, and Nicole Klingen. 1999. "Islamic Republic of Iran--Health Financing Reform in Iran: Principles and Possible Next Steps." Prepared for Social Security Research Institute, Health Economic Congress, October 1999. World Bank, Washington, DC. Schieber, George, and Akiko Maeda. 1999. "Health Care Financing and Delivery in Devel- oping Countries." Health Affairs 18 (3): 193­205. Soares, R. R. 2007. "On the Determinants of Mortality Reductions in the Developing World," NBER Working Paper 12837, National Bureau of Economic Research, Cam- bridge, MA. Somanathan, Aparnaa. 2003. "Beneficiaries from Public Health Care in Five Asian Coun- tries: Equity in Asia-Pacific Health Systems Network." Presentation at World Bank con- ference on Reaching the Poor, February 18­20, 2004, Washington, DC. Sriratanaban, J. 2002. "Civil Servant Medical Benefit Scheme: Unregulated Fee-for-Service and Cost Escalation." In Health Insurance Systems in Thailand, ed. P. Pramualratana and S. Wibulpolprasert, 43­50. Nonthaburi, Thailand: Health Systems Research Institute. Suraratdecha, Chutima. 2003. "Thailand's Universal Insurance Coverage Initiative." Presen- tation at World Bank conference on Reaching the Poor, February 18­20, 2004, Washing- ton, DC. Szende A., and Z. Mogyorosy. 2004. "Health Care Provider Payment Mechanisms in the New EU Members of Central Europe and the Baltic States: Current Reforms, Incentives, and Challenges." European Journal of Health Economics 5 (3): 259­62. Tabor, Steven R. 2005. "Community-Based Health Insurance and Social Protection." Social Protection Discussion Paper 0503, World Bank, Washington, DC. Tanahashi, T. 1978. "Health Service Coverage and Its Evaluation." Bulletin of the World Health Organization 56 (2): 295­303. Tangcharoensathien, V., S. Srithamrongsawat, and S. Pitayarangsarit. 2002. "Overview of Health Insurance Systems." In Health Insurance Systems in Thailand, ed. P. Pramualratana and S. Wibulpolprasert, 28­38. Nonthaburi, Thailand: Health Systems Research Institute. Tangcharoensathien, V., S. Wibulpholprasert, and S. Nitayaramphong. 2004. "Knowledge- Based Changes to Health Systems: The Thai Experience in Policy Development." Bulletin of the World Health Organization 82 (10): 750­56. Taylor, Rob, and Simon Blair. 2003. "Financing Health Care--Singapore's Innovative Ap- proach. Public Policy for the Private Sector." Note No. 261, World Bank, Washington, DC. Thomas, S., and L. Gilson. 2004. "Actor Management in the Development of Health Financing Reform: Health Insurance in South Africa, 1994­1999." Health Policy Plan 19 (5): 279­91. Towse, A., A. Mills, and V. Tangcharoensathien. 2004. "Learning from Thailand's Health Reforms." British Medical Journal 328 (7431): 103­5. Trejos, Juan Diego. 2002. "La Equidad de la Inversión Social en el 2000." Octavo Informe sobre el Estado de la Nación en Desarrollo Humano Sostenible, United Nations Devel- opment Programme, San José, Costa Rica. Enabling Factors for Expanding Coverage 85 Trujillo, A. J., and D. C. McCalla. 2004. "Are Colombian Sickness Funds Cream Skimming Enrollees? An Analysis with Suggestions for Policy Improvement." Journal of Policy Analysis and Management 23 (4): 873­88. UNDP (United Nations Development Programme). 2007. Human Development Database. http://www.undp.org. van Doorslaer, E., O. O'Donnell, R. P. Rannan-Eliya, A. Somanathan, S. R., Adhikari, C. C. Garg, D. Harbianto, A. N. Herrin, M. N. Huq, S. Ibragimova, A. Karan, T. Lee, G. M. Leung, J. R. Lu, C. W. Ng, B. R. Pande, R. Racelis, S. Tao, K. Tin, K. Tisayaticom, L. Tris- nantoro, C. Vasavid, and Y. Zhao. 2007. "Catastrophic Payments for Health Care in Asia." Health Economics 16 (11): 1159­84. van Doorslaer, E., O. O'Donnell, R. P. Rannan-Eliya, A. Somanathan, S. R. Adhikari, C. C. Garg, D. Harbianto, A. N. Herrin, M. N. Huq, S. Ibragimova, A. Karan, B. R. Pande, C. W. Ng, R. Racelis, S. Tao, K. Tin, K. Tisayaticom, L. Trisnantoro, C. Visasvid, and Y. Zhao. 2006. "Effect of Payments for Health Care on Poverty Estimates in 11 Countries in Asia: Analysis of Household Survey Data." Lancet 368(9544): 1357­64. van Doorslaer, E., O. O'Donnell, R. Rannan-Eliya, A. Somanathan S. R. Adhikari, B. Akkazieva, C, C. Garg, D. Harbianto, A. N. Herrin, M. N. Huq, S. Ibragimova, A. Karan, T. Lee, G. M. Leung, J. R. Lu, C. W. Ng, B. R. Pande, R. Racelis, S. Tao, K. Tin, K. Tisayati- com, L. Trisnantoro, C. Visasvid, and Y. Zhao. 2005. "Paying Out-of-Pocket for Health Care in Asia: Catastrophic and Poverty Impact." Equitap Project: Working Paper 2. Eras- mus University, Rotterdam, and Institute of Policy Studies. Colombo. Vasavid, C., K.Tisayaticom, W. Patcharanarumol, and V. Tangcharoensathien. 2004. "Impact of Universal Health Care Coverage on the Thai Households." In From Policy to Implementation: Historical Events during 2001­2004 of Universal Coverage in Thailand, ed. V. Tangcharoensathien and P. Jongudomsak, 129­149. Nonthaburi, Thailand: National Health Security Office. Wagstaff, Adam. 2005. "The Economic Consequences of Health Shocks." Policy Research Working Paper 3644, Development Research Group, World Bank: Washington, DC. Wagstaff, Adam, and Menno Pradhan. 2005. "Health Insurance Impacts on Health and Nonmedical Consumption in a Developing-Country." Development Research Group and East Asia Human Development Unit, World Bank, Washington, DC. Wagstaff, Adam, and Eddy van Doorslaer. 2003. "Catastrophe and Impoverishment in Pay- ing for Health Care: With Applications to Vietnam." Health Economics 12: 921­34. Wagstaff, Adam, Abdo Yazbeck, and Mariam Claeson. "Poverty Reduction Strategy Frame- work." Adapted from Adam Wagstaff and Mariam Claeson. 2004. "The Millennium Development Goals for Health Rising to the Challenges." Washington, DC: International Bank for Reconstruction and Development/World Bank. Waters, Hugh. 1999. "Measuring the Impact of Health Insurance with a Correction for Selection Bias--A Case Study of Ecuador." Health Economics 8 (5): 473­83. Waters, Hugh, Gerard Anderson, and Jim Mays. 2004. "Measuring Financial Protection in Health in the United States." Health Policy 69 (3): 339­49. Waters, Hugh, and Peter Hussey. 2004. "Pricing Health Services for Purchasers--A Review of Methods and Experiences." Health Policy 70 (2): 175­84. 86 Good Practice in Health Financing Waters, Hugh, Laura Morlock, and Laurel Hatt. 2004. "How Healthcare Purchasers Can Influence Quality--A Conceptual Framework and Comparative Analysis of Contextual Factors." International Journal of Health Planning and Management 19 (4): 365­81. Waters, Hugh, Axel Rahola, and Jennifer Hulme. 2005. "Health Insurance Expansion in Morocco." World Bank, Washington, DC. Watson, N. 2004. "Poland's Broken Promises." Lancet 364 (9430): 235­38. Webster, P. 2004. "Reforms Mean 25 Million Russians Lose Free Health Care." Canadian Medical Association Journal 171 (10): 1157. Williamson, O. 1985. The Economic Institutions of Capitalism. New York, NY: Free Press. WHO (World Health Organization). 2000. World Health Report 2000--Health Systems: Improving Performance. Geneva: WHO. ------. 2003a. World Health Survey, Tunisia. Geneva: WHO. ------. 2003b. "World Health Survey 2003 Results"; accessed at http://www.who.int/ healthinfo/survey/whsresults/en/index.html. ------. 2007. WHO Statistical Information System (WHOSIS). Online database. http:// www.who.int/whosis/database/core/core_select.cfm. ------. 2007. "WHO World Health Statistics." Geneva: WHO. World Bank. 2002. "HNP Strategy Paper." World Bank, Washington, DC. ------. 2003. "Project Appraisal Document--Republic of Ghana Health Sector Program Support Project." Report No: 24842-GH. Human Development II. Africa Regional Office, Ghana. World Bank: Washington, DC. ------. 2004a. World Development Report 2004: Making Services Work for Poor People. Washington, DC: World Bank. ------. 2004b. "Croatia--Health Finance Study. Report No. 27151-HR." Human Develop- ment Sector Unit, Europe and Central Asia Region. ------. 2007a. World Development Indicators 2007. Washington, DC: World Bank. ------. 2007b. "Healthy Development: The World Bank Strategy for Health, Nutrition, and Population Results." Washington, DC: World Bank. ------. 2007c. World Development Report 2007: Development and the Next Generation. Washington, DC: World Bank. Wouters, Annemarie. 1999. "Alternative Provider Payment Methods: Incentives for Improv- ing Health Care Delivery." Partnerships for Health Reform, Primer for Policymakers 1. Bethesda, MD: Abt Associates Inc. http://www.phrproject.com. Xu, Ke, D. Evans, K. Kawabatta, R. Zeramdini, J. Klavus, and C. Murray. 2003. "Household Catastrophic Health Expenditure: A Multicountry Analysis." Lancet 362 (9378): 111­17. Yepes, F. 2001. "Health Reform in Colombia." In Health Care Reform and Poverty in Latin America, ed. P. Lloyd-Sherlock, 163­77. London: Institute of Latin American Studies, University of London. Zurn, Pascal, and Orville Adams. 2004. "A Framework for Purchasing Health Care Labor." Health, Nutrition and Population (HNP) Discussion Paper 31597, World Bank, Wash- ington, DC. Part II Nine Case Studies of Good Practice in Health Financing Reform 6 Chile: Good Practice in Expanding Health Care Coverage--Lessons from Reforms Ricardo D. Bitrán and Gonzalo C. Urcullo Chile, is a tricontinental country, with land in Latin America (756,950 sq. km), Antarctica, and Oceania. It is in the Latin America and Caribbean Region (LAC) of the World Bank. Chile has a population of 16.3 million, a total fertility rate of 1.98, and per capita income of $12,365 (PPP-adjusted). Total health expenditures equaled $282 per capita in 2005, 51.2 percent of this amount paid by private sources, includ- ing 23.7 percent paid directly by households as out-of-pocket expenditures. Health outcomes are among the best in the world: life expectancy is 78.2 years, and infant mortality is 7.6 per 1,000 live births. In Chile, the expansion of a fiscally sound social security system, including health care coverage, has required considerable economic resources. An extended period of economic growth and an efficient tax collection system made this expansion possible. Chile also has a high degree of formality in employment. Chile's experience suggests that registration in social security systems and other formal labor market mecha- nisms is perhaps the single most important facilitator for extending health care cover- age. The insurance system in place today in Chile, first implemented in 1981, covers more than 90 percent of the population. Although some questions remain regarding the equity and efficiency of the system, proposals for reform are aimed at improving the current system rather than replacing it. Background Sustained economic growth and a fiscally sound social security system provided the resources needed to extend health care coverage. Economic Context Economic growth in Chile has, on the whole, been positive throughout its history (figure 6.1). Figure 6.2 provides a more detailed look at real GDP growth rates 89 90 Good Practice in Health Financing Figure 6.1 Chile: Economic Growth, 1810­2005 8 6 (%) 4 rate 2 0 8­02 1810­191820­291830­391840­491850­591860­691870­791880­891890­991900­091910­191200­291930­391940­491950­591960­691970­791980­841985­97199 2003­05 Source: Authors' elaborations based on data from the Central Bank of Chile. from 1997 through 2005. As shown, only one year showed an economic downturn As a result of its sustained economic growth, Chile's GDP in international dollars has become one of the largest in Latin America (figure 6.3). Chile has reported significant budget surpluses in recent years and continues to project surpluses for the future. These surpluses are due mainly to record prices for copper, Chile's main export product. As the global economy has grown, demand for minerals and metals has increased, especially in India and China. Chile's internal economy has also performed well. In 2005, for example, internal demand rose 10 percent and investments grew by 26 percent, representing a record 26 percent of the GDP. In 2005, projected central government revenues were US$28.6 billion. The size of this figure is attributable in part to growing income from copper exports, but it is also due to increased tax revenues. Tax revenues, not including those related to copper exports, represent about 70 percent of the total central government rev- enues, while revenues related to copper exports represent about 10 percent of the total. The tax revenue structure is shown in table 6.1. The fiscal surpluses noted above have gradually reduced the relative weight of the public debt and increased the resources available to the Collective Capitaliza- tion Fund of the social security system. Figure 6.4 shows the evolution of external debt in recent years. Debt has increased from year to year, although the rate of increase has decelerated. According to available reports,"the central government's debt has decreased from 11 percent of the GDP to around 8.4 percent, a record low."1 Most of Chile's external debt is owed to private creditors (figure 6.5). Chile: Good Practice in Expanding Health Care Coverage 91 Figure 6.2 Chile: Growth of Real GDP, 1997­2005 8 6 4 (%) rate 2 0 ­2 1997 1998 1999 2000 2001 2002 2003 2004 2005 Source: Authors' elaboration based on data from the Central Bank of Chile. Figure 6.3 Chile: GDP per Capita, 2004 Peru Venezuela, R.B. de Colombia Brazil Mexico Chile Argentina 0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 GDP per capita (international $) Source: World Health Organization 2006. Chile enjoys striking macroeconomic stability, in terms of economic growth, external debt sustainability, internal capacity for savings, attractiveness to inves- tors, and high import and export levels due to growing commercial and financial integration with the rest of the world. Table 6.2 shows the evolution of the major macroeconomic variables of the Chilean economy from 2000 to 2005. The years 92 Good Practice in Health Financing Table 6.1 Chile: Net Tax Revenue Structure, 1996­2004 (percent) Type of tax 1996 1997 1998 1999 2000 2001 2002 2003 2004 Income tax 23.59 22.95 24.05 22.60 24.97 26.70 27.74 27.07 25.72 Value-added tax 47.90 48.07 47.81 48.44 48.46 47.12 48.37 50.48 51.50 Taxes on specific 11.03 11.79 12.43 14.09 13.72 13.74 13.67 13.41 12.17 products Taxes on legal 3.81 4.12 3.85 4.22 3.64 4.10 4.39 4.46 4.36 documents Import taxes 11.86 10.86 10.30 9.22 8.28 7.02 5.67 3.91 2.85 Other 1.81 2.20 1.56 1.43 0.92 1.32 0.17 0.68 3.40 Source: Author's elaboration based on data form the Central Bank of Chile. Figure 6.4 Chile: External Debt, 1996­2005 50,000 40,000 30,000 millions US$ 20,000 10,000 0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Source: Central Bank of Chile. 2004 and 2005 were especially outstanding, particularly in terms of gross capital as a percentage of GDP and import and export levels. The increased trade is due in part to increases in the price of copper. Furthermore, the budget moved from deficit to surplus. Demographic, Epidemiological, and Social Context Until 1930, Chile's population lived mainly in rural areas. By 2002, the 2,026,322 people living in rural areas represented only 13.40 percent of the country's total population. Chile: Good Practice in Expanding Health Care Coverage 93 Figure 6.5 Chile: Composition of External Debt, 2004 Short term US$7.707 17% IBRD US$441 1% Other multilateral: US$ 573 1% Bilateral: US$ 321 1% Private: US$35.016 80% Source: World Bank 2004. Table 6.2 Chile: Macroeconomic Performance, 2000­05 Macroeconomic variable 2000 2001 2002 2003 2004 2005 GDP (% real annual variation) 4.5 3.4 2.2 3.9 6.2 6.3 Internal demand (% real annual variation) 6.0 2.4 2.4 4.9 8.1 11.4 National income (% real annual variation) 4.4 2.4 2.8 3.9 8.6 9.1 Investment (% GDP) 21.9 22.1 21.7 22.0 21.4 23.0 National savings rate (% GDP) 20.6 20.6 20.7 20.7 23.0 23.6 Inflation (December of each year) 4.5 2.6 2.8 1.1 2.4 3.7 Unemployment rate (%) 9.2 9.1 8.9 8.5 8.8 8.0 Real salaries (% annual variation) 1.4 1.6 2.1 1.0 1.8 1.9 Exports (US$ billions) 19.2 18.3 18.2 21.7 32.2 40.6 Imports (US$ billions) 17.1 16.4 15.8 18.0 23.0 30.4 Budget deficit or surplus (US$ millions) ­897 ­1,100 ­580 ­964 1,586 702 Price of copper (US$ cents /pound) 82.3 71.6 70.7 80.7 130.0 166.9 Source: Central Bank of Chile, INE Cochilco. Two features define the Chilean demographic situation, categorized as in the advanced transition stage. First is the projected decrease in gross reproductive rate from 18.2 per 1,000 population in 2000­05 to 13.4 per 1,000 population in 2045­ 50. Second is the growth of the elderly population. In 2000, 36.9 percent of 94 Good Practice in Health Financing Chileans were under the age of 19; by 2050, only 22.6 percent of the population will be under 19. During the 20th century Chile experienced a 45-percent decrease in birth rate and an 85-percent decrease in the mortality rate. In the first half of the century, population growth increased at a moderate pace, accelerating in the 1950s. After 1960, the birth rate quickly began to decline, finally stabilizing in the 1980s. This shift was due to changing marriage trends, as well as increased individual control over family planning. This latter phenomenon occurred more rapidly than in developed countries. Between 1971 and 1992, the number of children per woman dropped from 3.80 to 2.39. The decrease was most pronounced in rural areas, even though historically this population has had higher fertility rates than urban areas. Figure 6.6 illustrates the aging of the population structure. In the middle of the twentieth century, Chile was a young country, with 40 per- cent of its population under the age of 15 and only 3 percent over the age of 65. By the end of the century, only 29 percent of the population was younger than 15, and 7 percent were 65 or older, marking it as an aging country. In a half century, the average age of the population increased by nearly five years, from 26 in 1952 to 31 in 2002.2 Other markers of the changing structure of the Chilean population are the indices of dependency and of aging.3 In the last 50 years, the dependency index has decreased by nearly a fourth, reaching 50 percent. Meanwhile, the index of aging has increased by over half, reaching nearly 20 percent. Furthermore, the lat- est census data indicate that there are more women than men in Chile and that the gender ration (number of males per 100 females) is greater in rural than in urban areas. Epidemiological context. The latest available study on Chile's disease burden was published in 1996 using 1993 data. At the risk of ignoring epidemiological changes that might have occurred after 1993, the authors selected this study as the most reliable reflection of Chile's epidemiological context (Chilean Ministry of Health 1996). The results of this study showed that Chile lost 1,769,557 years of healthy life (DALYs), 128.5 DALYs per 1,000 population. The disease burden varies by age and gender. The disease burden in Chile was 1.25 times higher for men than for women. In men, most of the disease burden was attributable to premature death, while for women most was attributable to disability. In both genders, the disease burden was highest for the very young and very old, and men lost more DALYs than women at all ages. Disease burden was mainly attributable to premature death among the very young and very old, and mainly to disability among those of intermediate ages. Table 6.3 lists the 15 diseases the produced the greatest death and disease burden in Chile in 1993, by DALYs lost, the percentage of lost DALYs attributable to that disease versus the total DALYs lost in Chile that year, and the DALYs lost attributable to that disease per 1,000 population.4 Chile: Good Practice in Expanding Health Care Coverage 95 Figure 6.6 Chile: Population, 1990, 2005, and 2020 a. 1990 b. 2005 100+ 90­94 90­94 male female male female 80­84 75­79 70­74 60­64 60­64 50­54 age 45­49 age 40­44 30­34 30­34 20­24 15­19 10­14 0­4 0­4 15 10 5 0 5 10 15 15 10 5 0 5 10 15 population (%) population (%) c. 2020 90­94 male female 75­79 60­64 45­49 age 30­34 15­19 0­4 15 10 5 0 5 10 15 population (%) Source: National Statistical Institute of Chile (http://www.ine.cl). These rankings vary by gender and age. In men, alcohol dependence places sec- ond, followed by traffic accidents and ischemic heart disease. In women, depres- sion places second, followed by bile duct disorders and acute lower respiratory infections. Neoplasms do not make the top 15 due to low life expectancy. Cervical uterine cancer produces the most lost DALYs. However, among those over the age of 45, various cancers begin to rank higher on the list, including gallbladder, breast, and stomach cancer in women, and stomach, lung, and prostate cancer in men. Among the very young and the old--those under the age of 5 or over the age of 60--the disease profile is similar among males and females. The largest gender differences in disease profile occur between the ages of 15 and 44, mainly due to the higher rates of injuries and alcohol dependence and abuse among men. 96 Good Practice in Health Financing Table 6.3 Chile: Disease Burden, 1993 (DALYs lost) Illness DALYs % Ratio/1,000 Congenital anomalies 103.654 5.86 7.53 Acute lower respiratory infections 73.234 4.14 5.32 Ischemic heart disease 67.534 3.82 4.90 Hypertensive disease 60.172 3.40 4.37 Cerebrovascular disease 57.700 3.26 4.19 Asthma 55.118 3.11 4.0 Traffic accident 53.692 3.03 3.90 Alcohol dependence 53.498 3.02 3.88 Bile duct disorders 53.361 3.02 3.87 Major depressive disorder 53.279 3.01 3.87 Arthritis and similar diseases 48.452 2.74 3.52 Alzheimer's and other dementias 42.889 2.42 3.11 Perinatal infections 41.710 2.36 3.03 Psychosis 32.474 1.84 2.36 Cirrhosis 32.172 1.82 2.34 Total 828.941 46.84 60.19 Source: MOH. Health indicators. Infant mortality has gradually decreased. The most dramatic advances occurred between 1940 and 1950, when the infant mortality rate decreased by 58 percent. The rate continued to decrease from 1960 to 2002, although at a decelerated pace (figure 6.7). The Chilean infant mortality rate decreased from 234 to 8 per 1,000 live births from 1930 to 2003, a 96.6 percent reduction. The decreased infant mortality rate correlates with a number of other factors: improved living con- ditions, especially in terms of basic sanitation, availability of potable water, and san- itary waste disposal systems; better delivery of public services; better education, greater availability of food; and better access to medical care. In 2000­05, life expectancy was 74.8 years for males and 80.8 years for females. This represents a 21.9-year increase in life expectancy for males and 24.03-year increase for females since the mid-20th century. The difference in life expectancy by gender held constant throughout all periods studied (figure 6.8). Chile enjoys one of the lowest infant mortality rates and highest life expectan- cies in Latin America. Figure 6.9 shows 2004 data for infant mortality and life expectancy for several Latin American countries, listed in order of GDP per capita in international dollars. The two trend lines show that, as economic development increases, infant mortality decreases and life expectancy increases. Chile, the second-most-developed country, has the longest life expectancy and the second- lowest infant mortality rate of the Latin American countries shown here. Chile: Good Practice in Expanding Health Care Coverage 97 Figure 6.7 Chile: Infant Mortality, 1960­2002 140 120 100 births live 80 1,000 60 per 40 deaths 20 0 2 2 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 199 1994 1996 1998 2000 200 Source: National Statistical Institute of Chile. Figure 6.8 Chile: Life Expectancy, by Historical Period and Gender, 1950­2025 90 85 80 75 years in 70 age 65 60 55 50 1950­55 1960­65 1970­75 1980­85 1990­95 2000­05 2010­15 2020­25 both genders men women Source: National Statistical Institute of Chile. 98 Good Practice in Health Financing Figure 6.9 Poverty Compared with Other Latin America Countries, 1999 Honduras Ecuador Bolivia Colombia Mexico Brazil Chile 0 10 20 30 40 50 60 70 80 90 % of poor population Sources: Universidad Emilio Máspero (http://www.utal.org/economia/hogaresypoblafcion.htm). Social context. Chile's solid economic performance throughout its history, and especially in recent years, has allowed the country to make significant gains in alle- viating poverty. According to data from the 2003 socioeconomic census (CASEN), 80 percent of the population is classified as "not poor." The number of people in Chile living on less than US$1 per day was 843,000, about 5.6 percent of the total population. Between 1987 and 2003, the population living below the poverty line in Chile decreased from 45.1 percent to 18.8 percent. Poverty in Chile is much less severe than in most other Latin American countries. Figure 6.10 shows that in 1999, only 20 percent of the Chilean population was poor, as compared with 40 percent in most other Latin American countries, and 43 percent for the region. Although Chile has made important strides against poverty, the country's wealth remains inequitably distributed. Chile is considered one of the most inequitable countries in Latin America. According to 2003 CASEN census data, average income in the richest quintile was 16 times higher than that of the poorest quintile. The richest 10 percent of the population earned 41.2 percent of the country's total income, while the poorest 10 percent earned only 1.2 percent, 34.3 times less. Political Context During the 19th and 20th centuries, Chile slowly consolidated a highly stable democratic political system (box 6.1). Chile: Good Practice in Expanding Health Care Coverage 99 Figure 6.10 Chile: Infant Mortality and Life Expectancy Compared with Other Latin American Countries, 2004 85 80 80 70 life 60 75 (%) expectancy rate 50 70 40 65 (years) mortality 30 child 60 20 10 55 0 50 zil de u la Per Haiti entinaChile guay Rica Bra mbia Cuba Mexico Panama , R.B. Bolivia Ecuador Arg Uru Republic Salvador Costa Colo Honduras El GuatemaParaguay Nicaragua ican Venezuela Domin infant mortality rate life expectancy life expectancy infant mortality rate Source: National Statistical Institute (www.ine.cl). B O X 6 . 1 Chile: Key Political Milestones 1810 First elections held 1925 Fourth Constitution drafted 1822 First Constitution drafted 1935 Women given right to vote in municipal 1828 Second Constitution drafted elections 1833 Third Constitution drafted 1949 Women given right to vote in 1871 Presidents prohibited from running for presidential and parliamentary elections a second consecutive term 1972 Illiterates given right to vote 1874 Voting rights expanded 1890 Secret ballot established Source: http://www.memoriachilena.cl. From 1835 to 1888, candidates put forth by the State Party regularly won presi- dential and parliamentary elections. After 1888, the electoral process became more transparent. However, while government intervention in elections decreased, irregularities in the electoral process remained common, and political power remained concentrated in one party. After 1920, social movements began to erupt; more of the population began to participate in elections; and leftist political 100 Good Practice in Health Financing parties developed, including a Communist Party (1922) and a Socialist Party (1933). The political mobilization of the population in the 1960s and early 1970s, in a highly polarized political climate, came to an abrupt end after the 1973 coup d'état. The autocratic government ended in 1990, followed by a series of democra- tically elected administrations. The latest election was in 2006.5 Chile's current system of government is tripartite. The executive branch is directed by the president of the republic, who is also the chief of state; the legislative branch consists of a two-chamber congress (representatives and senators); and the judicial branch is led by the Supreme Court. Functions of each branch are clearly delineated among the three branches, each of which operates independently. Chile is divided into 13 regions, 51 provinces, and 342 communities. Each region is led by a superintendent, who represents the president of the republic in that region. The regional governments also include a regional council, which approves regional development plans and allocates regional investment resources, among other duties. Within the regions, each province is led by a governor, who is subordinate to the superintendent, but functions in a decentralized manner. The governor pre- sides over an economic and social provincial council. Finally, at community level, a mayor presides over the economic and social municipal council, which allocates resources and makes community-level deci- sions. Local leaders are chosen by popular election to serve four-year terms.6 Overview of Health Care Coverage and Financing in Chile The defining feature of Chile's health care system is the mandatory social security system, which covers about 90 percent of the population. This section provides an overview of the Chilean social security system, as well as a discussion of health care financing, social security coverage, basic package coverage, and the equity and efficiency of the Chilean social security system. Chile's Mandatory Health Insurance System Chile's mandatory health insurance (MHI) system consists of a single nonprofit public insurer (Fondo Nacional de Salud, FONASA) and multiple for-profit or non- profit private insurers (Instituciones de Salud Previsional, ISAPREs), all operating in competition (figure 6.11). By law all formal sector workers who are not self- employed, retired workers with a pension or self-employed workers with a retire- ment fund must enroll with the MHI by making a monthly contribution equal to 7 percent of their income or pension, up to a monthly ceiling of US$2,000.7 Other individuals may enroll as well. They include independent workers, who can volun- tarily enroll with the FONASA or an ISAPRE conditional on their 7 percent contri- bution; and legally certified indigent citizens and legally unemployed workers, who are entitled to free coverage by the FONASA. ISAPRE beneficiaries may voluntarily make an extra contribution to their insurer to purchase additional coverage. Until Chile: Good Practice in Expanding Health Care Coverage 101 Figure 6.11 Chile: The Mandatory Health Insurance System, 2006 government's subsidy (general taxation) insurers public insurer, national private insurers public health health fund (ISAPRES) goods (FONASA) complementary reimbursements payment optional mandatory 7% mandatory 7% reimbursements reimbursements population population in public sector population in private sector A B C D E indigent contributing affiliates fee-for-service copayments copayments providers public private providers providers free-choice modality: copayments financing health care services Source: Authors. 2005 no such thing as a basic benefits package was required of the FONASA or ISAPREs, although the latter could not provide less financial coverage than the for- mer. Starting in 2005, a set of new laws collectively known as Explicit Health Guar- antees (Garantías Explícitas de Salud, GES) mandated the public and private MHI insurers incrementally to provide coverage for 56 legally defined health problems, starting with 25 of them in July 2005 and encompassing all 56 problems in 2007. The FONASA must by law purchase most of its covered health services from public hos- pitals and health centers, although it also provides a modest subsidy to its beneficia- ries willing to purchase private health care. Public health care providers must by law sell most of their services to the FONASA and are subject to tight limits on the kinds and volume of services they may sell to private patients and ISAPRE beneficiaries. 102 Good Practice in Health Financing The National Health Superintendence is responsible for regulating the FONASA and the ISAPREs. Main regulatory functions include beneficiary protection, finan- cial solvency of ISAPREs, and compliance by both the FONASA and ISAPREs with the provision of benefits stipulated in the GES law. Chile's MHI system is also characterized by a split between financing and delivery. In the public sector, the FONASA acts solely as a financing agent, and delivery is left in the hands of the Ministry of Health (MOH). In the private sector, by law ISAPREs are not allowed to provide health services directly (i.e., they cannot be vertically integrated). This split has widened beneficiaries' choice of provider alternatives. FONASA beneficiaries may choose to seek care from any provider, public or private, as long as they are reg- istered with the FONASA. If the provider is public (called Modalidad de Atención Institucional, MAI) copayments are small or nil. If the provider is private (called Modalidad de Libre Elección, MLE) copayments are larger. ISAPRE beneficiaries have similar choices but almost always opt for private care. Key Indicators Health care spending in Chile amounts to 6 percent of the GDP and is financed by both public and private sector entities.8 As shown in table 6.4, public sector health spending has increased, but, until 2004, was lower than private spending. Spend- ing decreased between 1999 and 2003 and then increased again in 2004 to over the 1998 amount. Spending as a percentage of GDP dropped about 1 percent from 1998 to 2004. Table 6.4 Chile: Key Health Spending Indicators, 1998­2004 Indicator 1998 1999 2000 2001 2002 2003 2004 Health spending (US$ billions) 5.6 5.2 4.6 4.3 4.2 4.5 5.7 Health spending as a percentage of GDP 7.1 7.1 6.1 6.2 6.2 6.1 6.1 Public health spending (US$ billions) 2.1 2.2 2.1 2.0 2.0 2.2 2.7 Private health spending (US$ billions) 3.6 3.2 2.5 2.2 2.2 2.3 3.1 Percentage of health spending by public sector 36.7 39.0 46.4 48.1 48.0 48.8 46.4 Percentage of health spending by private sector 63.3 61.0 53.6 51.9 52.0 51.2 53.6 Out-of-pocket spending (US$ billions) 2.2 1.9 1.2 1.1 1.0 1.1 1.4 Out-of-pocket spending as a percentage of total health spending 38.9 37.0 25.2 24.9 24.6 23.7 24.6 Spending on private health plans (US$ billions) 1.4 1.2 1.3 1.1 1.1 1.2 1.7 Spending on private health plans as a percentage of total health spending 24.4 24.0 28.4 27.0 27.4 27.5 29.0 Total per capita health spending (US$) 375.0 342.0 299.0 272.0 265.0 282.0 355.0 Source: Authors' elaboration based on World Bank data (http://www.worldbank.org). Chile: Good Practice in Expanding Health Care Coverage 103 Figure 6.12 Chile: Structure of Health Spending, by Source, 1998­2004 100 80 60 percentage 40 20 0 1998 1999 2000 2001 2002 2003 2004 public expenditure out-of-pocket payment private prepayment Source: Authors' elaboration based on World Bank data. Private health spending consists of out-of-pocket payments on health care and payments for health insurance premiums. In 1998, out-of-pocket spending was 59 percent higher than spending on premiums, but this trend had reversed by 2000. The decrease in private health spending relative to public spending, therefore, can be explained as a decrease in out-of-pocket spending on health care. Figure 6.12 shows how public spending and private spending on health insur- ance premiums have increased relative to out-of-pocket spending on care. Private spending on health as a percentage of total health spending is greater in Chile than in most other similarly developed Latin American countries. Com- pared with Brazil, Costa Rica, and Mexico, Chile has the lowest level of public spending on health as a percentage of GDP and the highest level of private spend- ing (table 6.5). In 2002, public spending on health as a percentage of total public spending was 10.2 percent. From 2000 to 2005, public spending on health increased relative to total public spending on social welfare programs by about 2 percent, reaching almost 15 percent. This amounted to an average annual increase of US$100 million. Table 6.5 Chile: Health Spending as a Percentage of GDP, 2004 (percentage) Source Brazil Costa Rica Mexico Chile Public 3.6 4.9 2.8 2.6 Private 3.6 3.4 3.2 3.8 Total 7.2 8.3 6.0 6.4 Source: PAHO (http://www.paho.org). 104 Good Practice in Health Financing Table 6.6 Chile: Composition of Spending on Social Welfare Programs, 2000­05 (percentage) Function 2000 2001 2002 2003 2004 2005 Health 12.8 13.3 13.4 14.0 14.5 14.9 Education 16.7 17.4 17.8 17.8 18.5 18.8 Social protection 35.5 35.5 34.7 34.5 34.1 33.6 Other social welfare programsa 2.2 1.9 1.9 1.8 1.5 1.6 Total spending on social welfare programs 67.2 68.2 67.8 68.1 68.6 68.9 Other 32.8 31.8 32.2 31.9 31.4 31.1 Source: Ministry of the Interior, Budget Office (http://www.dipres.cl). a. Includes environmental protection, housing and community services, recreational activities, culture, and religion. Public spending on education also increased over this period, while spending on social protection and other social welfare programs decreased (table 6.6). The Basic Benefits Package In 2005, Chile developed a benefits package with four basic guarantees: guaranteed coverage for eligible public and private sector beneficiaries; guaranteed quality of care, achieved through accreditation of participating health providers; guaranteed access to care, achieved by establishing maximum wait times and processes for enforcement; and guaranteed financing for care, achieved through a government fund to cover specified serious health conditions.9 These promises, "the Explicit Health Guarantees" (GESs), contain the following specific provisions10: · Primary care. The GESs follow the Integrated Family Care Model, continuing the coverage provided as part of the previous Family Health Plan. Coverage includes certain preventative and curative care provided by a family medicine team led by a physician. The GES covers acute illnesses, preventative health care, health screenings, and special programs for mental health, cardiovascular health, specialist referrals, and home visits. · Emergency care. The GESs guarantee an emergency care network, covering the entire country, that provides prehospital care, transportation, diagnostic care, stabilization, and treatment for life-threatening or serious emergencies. · Targeted diseases. The GESs guarantee effective treatment, according to current standards of practice, for certain targeted diseases in Chile. These diseases are selected according to the seriousness of the illness, the disease burden pro- duced, the cost of care, considerations regarding equity, and other factors. The GESs began by guaranteeing coverage for 25 health problems in 2005 (table 6.7, col. 1). Coverage was expanded to 40 health problems on July 1, 2006, and as of July 1, 2007, coverage was set to expand to 56 problems. Chile: Good Practice in Expanding Health Care Coverage 105 Table 6.7 Chile: Health Problems Covered under the GES, 2005­07 2005 2006 2007 No. Health problem No. Health problem No. Health problem 1 End-stage renal disease 26 Preventative cholecystec- 41 Hearing loss in individuals 2 Operable congenital cardiopathies tomy for gallbladder over 65 years in children under 15 years cancer 42 Leukemia in adults 3 Cervical uterine cancer 27 Gastric cancer 43 Eye trauma 4 Pain relief and palliative care for 28 Prostate cancer 44 Cystic fibrosis advanced cancer 29 Refractive disorders in 45 Severe burns 5 Acute myocardial infarction individuals over 65 years 46 Drug and alcohol depen- 6 Type I diabetes mellitus 30 Strabismus in children dence in adolescents from under 9 years 7 Type II diabetes mellitus 10 to 19 years 31 Diabetic retinopathy 8 Breast cancer in individuals over 47 Complete prenatal and 15 years 32 Detached retina delivery care 9 Spinal defects 33 Hemophilia 48 Rheumatoid arthritis 10 Surgical treatment for scoliosis in 34 Depression in individuals 49 Mild and moderate individuals under 25 years over 15 years osteoarthritis of hip in indi- viduals over 60 years; mild 11 Surgical treatment for cataracts 35 Benign prostatic hyperplasia and moderate osteoarthri- 12 Total hip replacement for tis of knee in individuals advanced osteoarthritis in individ- 36 Acute cerebrovascular over 65 years uals over 65 years accident 50 Ruptured aneurysms; 13 Cleft palate 37 Chronic obstructive pul- ruptured ateriovenous monary disease 14 Cancer in children under 15 years malformations 38 Bronchial asthma 15 Schizophrenia 51 Central nervous system 39 Infant respiratory distress 16 Testicular cancer in individuals tumors and cysts syndrome over 15 years 52 Herniated disks 40 Orthotics and technical 17 Lymphoma in individuals over 15 53 Dental emergencies support for individuals years over 65 years 54 Dental care for adults over 18 Acquired immunodeficiency Syn- 65 years drome (AIDS) / HIV 55 Multitrauma 19 Outpatient treatment for acute res- 56 Traumatic brain injury piratory infection in children under 5 years 20 Walking pneumonia in individuals over 65 years 21 Primary (essential) arterial hyper- tension in individuals over 15 years 22 Nonrefractory epilepsy in children 1 to 15 years 23 Complete oral health care for chil- dren under 6 years: prevention and education 24 Prematurity ­ Retinopathy of Prematurity ­ Hypoacusia Prematurity 25 Major conduction disorders requiring a pacemaker in individu- als over 15 years Source: Bitrán & Asociados 2005. 106 Good Practice in Health Financing Social Security Coverage The social security system covers a large proportion of the population, 90 percent in recent years. Over two thirds of the population is covered by public health insurance (FONASA); a fifth to a sixth, by ISAPREs. A tiny percentage is covered by private insurance through the armed forces or universities. Table 6.8 shows the Chilean social security system's health coverage in 2005, by insurer.11 The public system is the main provider of health insurance in Chile. All the ISAPREs combined have never covered more than 26 percent of the total insured population. Average ISAPRE membership from 1984 to 2005 was 17.8 percent of the population. Figure 6.13 shows the evolution of coverage under the Chilean social security system. As seen, the public system has many more beneficiaries than the private. For a time, the number of ISAPRE beneficiaries was increasing and the number of FONASA beneficiaries was decreasing. This trend peaked in 1995­97 then reversed. The number of beneficiaries of other systems peaked in the late 1980s, then dropped by about half in succeeding years. When the system was established in 1981, there were only five participating ISAPREs. Many more ISAPREs quickly entered the market, peaking at 35 in the 1990s. Later, the number of participating ISAPREs began to drop back down. In 2006, there were 15 ISAPREs, 7 of them closed and 8 open. Figure 6.14 shows the number of beneficiaries (subscribers plus dependents) in the ISAPRE system in April 2006. Over two-thirds of the total beneficiaries belong to three ISAPREs: Banmedica, Consalud, and ING Salud. Type care provided. The most common type care provided to both FONASA and ISAPRE beneficiaries were diagnostic exams. Table 6.9 shows the services provided to FONASA beneficiaries. As seen, although diagnostic exams and clinical support services are the most common types, physician visits and days hospitalized account for a larger proportion of the cost. Table 6.8 Chile: Health Coverage Provided by Social Security System, 2005 Individuals Population Insurer covered covered (%) FONASA 11,329,481 69.65 Open ISAPREs 2,521,444 15.50 Closed ISAPREs 138,894 0.85 Uninsured 1,701,648 10.46 Others 575,771 3.54 Total population 16,267,278 100.00 Source: Superintendence of Health (http://www.supersalud.cl). Chile: Good Practice in Expanding Health Care Coverage 107 Figure 6.13 Chile: Coverage of Social Security System, 1984­2005 90 80 70 60 50 40 percentage 30 20 10 0 2 0 3 1984 1985 1986 1987 1988 1989 1990 1991 199 1993 1994 1995 1996 1997 1998 1999 200 2001 2002 200 2004 2005 FONASA ISAPRES others Source: Superintendence of Health (http://www.supersalud.cl). Figure 6.14 Chile: Coverage of Open ISAPREs, 2006 (number of beneficiaries and % of total beneficiaries) Normédica Sfera 49,594 32,469 Vida Tres (2%) (1%) Banmédica 135,440 610,250 (5%) (25%) Más Vida 192,967 (8%) Colmena 373,453 (15%) Consalud 563,509 (22%) ING Salud 547,653 (22%) Source: Superintendence of Health. 108 Good Practice in Health Financing Table 6.9 Chile: Services Provided by the FONASA and Average Annual Spending per Beneficiary, 2004 Number of services Spending Type of service provided provided (%) (US$ millions) (%) Physician visits 5.14 33.7 20.15 24.0 Diagnostic testing and clinical support services 6.65 43.6 16.79 20.0 Diagnostic and therapeutic procedures 1.38 9.0 7.97 9.5 Hemodialysis and peritoneal dialysis 0.16 1.0 5.74 6.8 Surgical interventions 0.10 0.7 10.16 12.1 Days hospitalized 0.14 0.9 1.88 2.2 Deliveries and obstetric surgeries 0.65 4.3 19.21 22.9 Other services 1.04 6.8 1.97 2.3 Total services 15.26 100 83.87 100 Source: FONASA. Table 6.10 shows the evolution of services provided by ISAPREs. Most striking is the nearly 50-percent increase in average annual services per beneficiary from 1996 to 2004, especially the number of laboratory tests. Table 6.11 shows the number of hospital admissions by type of insurance. In 2003, FONASA patients accounted for 78 percent of hospital admissions. Given that in 2003 the FONASA covered 68 percent of the population, FONASA benefi- ciaries seem more likely to be hospitalized than the rest of the population. This may be because the FONASA has a greater proportion of elderly and poor benefi- ciaries than do the ISAPREs. For example, according to 2003 CASEN census data, 18.1 percent of ISAPRE beneficiaries were 20 to 39-year-old females and only 5.9 percent were females over the age of 65. Furthermore, ISAPREs cover only 3.2 per- cent of the poorest quintile of the population. Table 6.11 also shows that the num- Table 6.10 Chile: Average Number of Health Services Provided by ISAPRES per Beneficiary, 1996­2004 2004 Type of service provided 1996 1997 1998 1999 2000 2001 2002 2003 2004 (%) Physician visits 3.34 3.42 3.62 3.96 4.13 3.82 3.97 4.00 4.53 26.8 Diagnostic testing 3.74 3.84 4.01 4.68 5.35 5.25 5.39 5.54 6.48 40.9 Clinical support services 1.87 1.86 1.97 2.44 2.80 2.98 2.70 2.72 3.22 20.3 Surgical interventions 0.09 0.09 0.09 0.10 0.11 0.11 0.11 0.11 0.12 0.8 Other services 1.26 1.36 1.47 1.41 1.17 0.91 0.88 0.85 0.85 5.4 Not classified n.a. n.a. n.a. n.a. n.a. 0.96 0.69 0.57 0.66 4.2 Total services 10.30 10.56 11.16 12.58 13.56 14.03 13.74 13.78 15.86 100 Source: Superintendence of Health. Note: n.a. = not applicable. Chile: Good Practice in Expanding Health Care Coverage 109 Table 6.11 Chile: Number of Hospital Admissions, 2001­03 Insurer 2001 2002 2003 FONASA 1,117,826 1,139,727 1,248,869 ISAPRE 245,553 261,579 252,879 Private 62,884 62,875 28,201 Other 121,343 128,926 68,599 Unknown 18,581 5,968 732 Total 1,566,187 1,599,075 1,599,280 Source: Authors, based on information from the MOH (http://www.misal.cl). ber of hospital admissions for individuals with unknown insurance status dropped significantly from 2001 to 2003, likely due to improved registration systems. Finally, the table shows that the number of individuals belonging to other insur- ance companies dropped by nearly half between 2001 and 2003. Funding Sources In Chile, the entire population is guaranteed access to the public health system, whether or not they have the resources to pay premiums. By law, dependent work- ers must enroll in the FONASA or an ISAPRE by contributing 7 percent of their income to the health system, up to a ceiling of 60 UF (an inflation-indexed unit) per month.12 These monthly contributions account for about a third of FONASA funding. About half of FONASA funding comes from the state. The ISAPREs, in contrast, are funded mainly by their beneficiaries' monthly contributions. Fund- ing for social security health benefits in Chile, not including copayments, comes from contributions to ISAPREs (35 percent), contributions to the FONASA (24 percent), and government subsidies (41 percent). Funding for health care in public facilities comes from four main sources: gov- ernment subsidies, monthly contributions, operational income, and copayments. The percentage of funding for health care in public facilities from government subsidies increased by 10 percent from 1990 to 1993. In the following years, this figure remained constant at about 55 percent, before dropping to about 50 percent by 2002 and 2003. Operating income and copayments represented a low propor- tion of funding throughout the period studied, and the proportion represented by copayment decreased slightly (figure 6.15). For example, in 2003, the public sector spent US$2.3 billion on health care, 51 percent of it from government subsidies, 35 percent from monthly contributions, 6.4 percent from operating income, and 7.3 percent from copayments. Equity The following series of characteristics show the existing equity or inequity in the Chilean health system. 110 Good Practice in Health Financing Figure 6.15 Chile: Structure of Financing for Public Health Spending 100 90 80 70 60 50 percentage 40 30 20 10 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 public expenditure contribution operative income copayment Source: Superintendence of Health. FONASA and ISAPRE membership vary by income. Most lower-income individ- uals subscribe to the FONASA; those with higher incomes, an ISAPRE. Within the FONASA, beneficiary services also vary according income. FONASA Group A contains the poorest beneficiaries, while those with progressively higher incomes belong to FONASA Groups B, C, or D. Figure 6.16 shows the different levels within the system. Some of the key attrib- utes of this system include · The social security system in Chile provides broad coverage (9 out of 10 Chileans are covered), which is uniform across levels of income. Only 2 percent of individuals in each decile of income are not covered. · Likelihood of belonging to an ISAPRE increases with income. Only within the highest income decile do more individuals belong to an ISAPRE than to the FONASA. · Low-income individuals are unlikely to subscribe to an ISAPRE, because FONASA A beneficiaries do not have to pay monthly contributions or copayments. · Only 1 in 10 of individuals in the poorest decile is not covered by FONASA A or B (not including the uninsured). · The percentage of high-income individuals covered by FONASA A is low. Chile: Good Practice in Expanding Health Care Coverage 111 Figure 6.16 Chile: Social Security System Beneficiaries, by Income Decile and Insurance Type, 2000 100 80 60 population 40 of % 20 0 1 2 3 4 5 6 7 8 9 10 (lowest) income decile (highest) FONASA A FONASA B FONASA C FONASA D FONASA group unknown ISAPRE Other None Source: CASEN Census 2000. Across all income levels, the FONASA and the ISAPREs cover a large propor- tion of the population. The FONASA covers about two-thirds of the population. The FONASA and the open ISAPREs together cover 86 percent of the total popu- lation, or 96 percent of individuals with health insurance. Ninety-six percent of the poorest individuals are covered under FONASA A and B. The FONASA is internally equitable. According to a study of equity within the FONASA in 1995 (Bitrán, R. 1997), nearly 41 percent of FONASA beneficiaries are indigent (Group A) and make no payments. Government subsidies for health are well targeted, with 90 percent of funds reaching the indigent and 7.5 percent reaching low-income individuals in Group B. Between 32 percent and 40 percent of the contributions of higher-income FONASA beneficiaries cross-subsidize care for the poorer beneficiaries, indicating that the FONASA's internal funding struc- ture is progressive. Indigents receive the most benefits per capita each year (US$116.63), followed by Group B beneficiaries (US$52.80). Public subsidies targeted to public health measures amounted to US$475.6 million, while government spending on public health measures was US$127.5 million, US$9.20 per capita. Groups C and D ben- eficiaries derive negative net benefits from the system, their contributions helping to subsidize those in Group B, and elderly beneficiaries receive a positive cross- subsidy from younger beneficiaries. 112 Good Practice in Health Financing Table 6.12 Chile: Contributions, Benefits, and Subsidies within the FONASA, 1995 (1995 Chilean pesos, millions) Group Contributions (%) Benefits (%) Net benefit A 27,344 10.26 202,594 44.35 175,250 B 111,664 41.88 172,408 37.74 60,744 C 60,915 22.85 41,661 9.12 ­19,254 D 66,707 25.02 40,139 8.79 ­26,568 Total 266,630 100.00 456,802 100.00 190,172 Source: Bitrán 1997. Table 6.12 presents more data from this study, showing the contributions made and net benefits received by beneficiaries of each FONASA group. Benefits consid- ered include primary, secondary, and tertiary levels of care, services provided under the free choice modality, the maternity subsidy, and health benefits pro- vided through contracts with the private sector. The contributions considered include monthly premiums, copayments for publicly provided services, and copayments for services provided under the free choice modality. The lowest- income groups (A and B) received a positive net benefit; the higher-income groups derived negative net benefits. Group A received the greatest net benefit, followed by Group B, indicating that net benefits are distributed equitably within the FONASA. Group B contributed nearly double the amount contributed by Groups C and D, but received four to five times the amount in benefits. Forty-two percent of the benefits were subsidized by the government. ISAPREs discriminate among beneficiaries. According to one study, most of the population above 50 years of age (even within the highest quintile of income) subscribes to the FONASA. This phenomenon is attributed to discrimination by the insurance companies against older beneficiaries, either by increasing prices or reducing coverage. Risk discrimination also affects the low-income individuals at high risk of illness (Titelman 2000). This type of discrimination can occur because the ISAPREs are able to decide whether to accept an individual as a bene- ficiary after gathering information about the person. The FONASA, in contrast, accepts anyone paying the monthly contribution or classified as poor. Figure 6.17 shows ISAPRE and FONASA membership by age group at two time points, illus- trating how the ISAPREs discriminate against the elderly. As shown, the benefi- ciary population of the two types of insurers has changed little in 15 years, and the FONASA's beneficiary population is "older" than that of the ISAPREs. The benefits derived by ISAPRE members are proportional to their contributions. The variety of plans offered by the ISAPREs allows them to assign beneficiaries to a plan that corresponds to their risk and income. The GES defines a basic benefits Chile: Good Practice in Expanding Health Care Coverage 113 Figure 6.17 Chile: FONASA and ISAPRE Beneficiaries, by Age, 1990 and 2005 45 40 35 30 25 20 percentage 15 10 5 0 FONASA ISAPRE FONASA ISAPRE 1990 2005 < 20 20­29 30­39 40­49 50­59 60­69 70­79 80­84 > 84 Source: Bitrán & Asociados 2006. package that all plans must provide. Therefore ISAPRE beneficiaries receive all the benefits covered under the GES, as well as any provided by additional coverage that they may choose to purchase under their individual plan. The benefits pro- vided by these individual plans are proportional to the cost of the plan. Premiums are equal to the amount specified by law plus the cost of any additional coverage. Efficiency Some aspects that determine the degree of efficiency of the Chilean health system are described next. Moral hazard. Having insurance coverage can influence an individual's behavior. First, knowing that their financial losses will be covered, insured individuals may be less careful to avoid injury or illness. Second, as the insured do not pay for the full cost of their treatment, they may visit the doctor more frequently and con- sume more medications than they would were they not insured. This phenome- non, called moral hazard, occurs with all types of insurance. To reduce the impact of moral risk, Chile has implemented copayments. Furthermore, the ISAPREs require a waiting period before insurance becomes active and before certain ben- efits are available. The waiting lists for certain services provided FONASA benefi- ciaries also reduce problems associated with moral hazard. 114 Good Practice in Health Financing Risk selection. There is an asymmetry of information between the insured and the insurer. When the beneficiaries possess better information about their own risk than do the companies, given that the premiums charged do not necessarily reflect individual risk, the beneficiaries may choose not to subscribe (adverse selection). In other cases, companies may be better able to discriminate between high- and low- risk individuals than the individuals themselves and selectively enroll low-risk individuals to maximize profits (cream skimming).13 In Chile, the mandatory sys- tem and set premiums protect against risk selection and allow greater diversifica- tion of risk. Similarly, the basic package is a tool to minimize cream skimming. The health authorities also work to reduce asymmetries of information. The system requires a minimum health expenditure, reducing the efficiency of society. Chilean law requires all formal sector workers to contribute 7 percent of their income to enroll in an individual insurance plan. Since beneficiaries have different personal and family characteristics and preferences, some beneficiaries are compelled to purchase more insurance than they need. This diminishes the efficiency of the economy. A counterargument to this position is that, if free mar- ket forces are allowed to govern the social security system, a lower-than-optimal amount of insurance will be consumed. In general, individuals underestimate their future health risks and fail to take externalities into account. Fixed premiums based on income also pose a problem in that plans for high-income and low-risk beneficiaries include benefits that emphasize comfort and amenities in the set- tings in which the interventions are carried out. A wide variety of health plans are available, which minimizes some losses of effi- ciency but generates others. The ISAPREs have created a wide variety of plans to fit individuals at various income and risk levels. This minimizes efficiency losses, because beneficiaries can choose the plan that best fits their needs and prefer- ences. However, the inability of much of the population to determine which plan offers them the most advantages limits this potential advantage. High spending on sales and marketing. Because the private insurance market is highly competitive, the ISAPREs spend a great deal of resources to attract mem- bers. For example, the companies mount intense marketing campaigns to position their products and deploy a large network of salespeople. The FONASA rations the services it provides. The FONASA allocates health ser- vices using quantity rationing (lines and waiting lists) rather than price rationing. The entity has also tended to fund standard care without regard to health out- comes or member preferences. The salary scheme does not encourage efficiency. Salaries for health professionals in public facilities are predetermined according to certain variables (professional degree, years of experience, location) rather than performance, effort, or results. Chile: Good Practice in Expanding Health Care Coverage 115 Overview of the Health Care System The supply capacity, regulatory features and the social security system of the Chilean health system are presented below. Public Hospitals, Physicians, and Health Programs The supply capacity of the Chilean health system is described in terms of the available number of hospitals, beds, doctors, nurses, drugs and medical equip- ment, and the main health programs being implemented. Hospitals. The public health sector provides outpatient facilities and various types of hospitals. In 2003, of the 191 public hospitals, 56 of them provided pri- mary and secondary-level care, 31 tertiary-level care, and 104 quaternary level care. In addition, more than 500 outpatient facilities and nearly 2,000 rural health posts provided primary care. Beds. According to the WHO, Chile had 25 hospital beds per 10,000 population in 2003, one bed for every 400 people. Eighty percent of these were public beds. Fig- ure 6.18 shows the average number of beds available within the National Health Services System (SNSS, Sistema Nacional de Servicios de Salud). The department of social assistance and investment at the Chilean Ministry of Health estimates that about 80 percent of these beds are used for primary and secondary-level care. The MOH also estimates that existing hospital stock in 2003 represented an investment of US$1.8 billion, of which US$1.1 billion was attributable to infrastructure and Figure 6.18 Chile: Available Beds in the SNSS System, 1990­2002 (annual average) 32,000 31,088 31,000 29,950 30,000 29,630 29,340 29,18529,110 28,999 beds 29,000 28,704 28,641 of 28,53428,471 27,881 28,000 27,687 number 27,000 26,000 25,000 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Source: Superintendence of Health. 116 Good Practice in Health Financing US$700 million to equipment. The projected investment for 2004­08 was US$523 million (US$523 million for equipment and US$296 million for infrastructure). Physicians and nurses. According to the 1992 Population and Housing Census, there were 13,897 physicians in Chile. Based on this statistic, as well as data on degrees earned, net income of foreign-born physicians in Chile, mortality rates, and retirement rates, one study estimated that there were 18,548 physicians in 2000 and a projected 25,704 physicians for 2007 (Bastías, Marshall, and Zúñiga 2000). MOH estimates put the number of physicians at 15,000 and medical sup- port staff members at 60,000 in 2003. According to 2003 WHO data, there were 17,250 physicians, 10,000 nurses, and 6,750 dentists in Chile. Medications. Data for 1999 indicated 16,000 registered pharmaceutical products. However, companies sometimes register a brand-name and generic version of the same product. Medications sold without a prescription represented 14 percent of total sales, while generic medications represented 38 percent of the pharmaceuti- cal market. The total medications sold amounted to US$632 million, US$42 per capita each year. Medical equipment. Investment in the public network has been heavy, especially during the 1990s. Thirteen hospitals were built or renovated during that time, at a cost of US$260 million. An additional 53 hospitals and 13 specialist facilities were modernized, at respective costs of US$180 million and US$105 million. Regional studies indicated that the care network acquired equipment valued at US$571 million in 1999. Thirty-two percent of the installed equipment is obsolete. Programs. In July 2006, the MOH was executing various programs targeting spe- cific segments of the population (programs for adolescents, older adults, women, and children), preventing illnesses (immunization campaigns, campaigns on pre- venting and managing acute respiratory infections in children), fighting diseases (programs for cancer, cardiovascular disease, schizophrenia, mental health, and tuberculosis), rehabilitation (integrated treatment for depression, integrated health and human rights services), and nutrition (free meal program).14 Regulatory Issues In a perfect market, no regulation is required. Market failures in the health insur- ance and health care markets in Chile, however, have led to a series of regulations. Health services are provided by the Regional Secretaries of the MOH (Secretarías Regionales Ministeriales de Salud, SEREMIS). When the ISAPRE system was launched in 1981, the insurance market was unregulated. Regulation of private insurers began in 1991, and in 2005 regulations were developed for public insur- ance providers. This has made the process more symmetrical, although different regulations apply to public providers and to private providers. For example, the Chile: Good Practice in Expanding Health Care Coverage 117 Superintendent of Health requires private insurers to provide extensive informa- tion regarding their financial stability, which is not required of the public insurer. However, many other requirements are symmetrical, such as compliance with the GES. The earlier absence of insurance regulations for part of the market created two types of problems. First, beneficiaries had trouble understanding and comparing the products offered by the FONASA with the ISAPRE products. Second, the ISAPREs had complete freedom in screening their beneficiaries and dismissing them when they become too costly. Beneficiaries had no recourse. The Superintendent of Health is responsible for: authorizing the creation of new ISAPREs, applying the laws and regulations on ISAPREs, specifying the guidelines and standards for the application of regulations and reporting proce- dures, arbitrating conflicts between ISAPREs and beneficiaries, and monitoring the financial solvency of ISAPREs. The number of accountability mechanisms in the hands of the SIS (Superin- tendencia de Salud) has grown over time. At present, the SIS is responsible for car- rying out the following supervisory procedures: inspecting all operations, goods, accounting books, files, and documents from ISAPREs and requesting any neces- sary clarifications; accessing ISAPRE financial balances at any time; and defining what information should be always available at the ISAPRE central office. Health system regulations are intended mainly to protect consumers and create transparency of information. Asymmetries of information and adverse selection are the biggest and most complex problems within the health insurance market, and regulation is aimed at reducing the attendant risks. Systemic Strengths and Weaknesses Below is an evaluation of the regulatory framework under which the health social security system operates and some observations about the main characteristics of the ISAPRE market. Market concentration. About 15 ISAPREs, covering a quarter of the population, operate in Chile. The three ISAPREs with the largest memberships (Banmedica, Consalud, and ING) cover about 60 percent of this market, meaning that there are high levels of concentration within the ISAPRE market. However, concentration is not necessarily a negative characteristic as long as competition is sufficient to min- imize service costs. Some authors argue that in risk administration markets (insur- ers) concentration is beneficial, because risk pooling reduces volatility and minimizes the size of the assets that the insurer must maintain. Other authors feel that high market concentrations lead to anticompetitive practices. Product homogeneity. The health care sector offers diverse plans and products. Currently about 10,000 plans are available. The mandatory 7-percent-of-salary contribution means that a wide range of premiums are paid, so insurers offer a 118 Good Practice in Health Financing wide range of products. This feature of the system also creates an additional asym- metry of information that the insurer could exploit by appropriating excess contributions. Market transparency. Market transparency has improved with increased regula- tion of ISAPREs. However, the heterogeneity of the insurance products offered makes it impossible to guarantee adequate market transparency. The way that prices and benefits are listed, along with the fee schedules and coverage limits, makes it difficult to compare health plans. These features vary not only from ISAPRE to ISAPRE but also within the same insurer. Furthermore, the fee schedules that the ISAPREs use as a reference in creating their health plans do not necessarily reflect real market prices for the services covered, making it difficult for consumers to evaluate health plans. --Superintendent of ISAPREs (1999) Barriers to entry. There are no legal restrictions to make entry to the insurance market difficult. Legally, capital requirements are relatively low: companies must have at least US$160,000. The main barriers to entry are the accumulated power of established brands and the investment of physical capital. The first factor is important because of both product heterogeneity and asymmetries of informa- tion in the insurance market. Confidence in the ISAPRE is a major factor in choosing an insurer. Such confidence is created by establishing powerful brands, meaning expensive marketing and sales campaigns. The second barrier is the sheer amount of office space and the large sales, administrative, and specialized staff necessary for interacting with consumers and health providers. Finally, the economies of scale enjoyed by existing insurers could be considered a barrier to new entrants. Market behavior. This market is characterized by clear segmentation or target marketing of high-income, low-risk members. Price is not the endogenous vari- able in the competition model for this market. In fact, the 7-percent-of-salary pre- mium represents an exogenous variable determined by government regulations. Equilibrium should be achieved by offering the best plans in terms of coverage and quality in exchange for the above premium. Yet selecting a health plan is problematic. The information regarding the con- tent of the programs offered is difficult to process, and therefore some uncertainly is involved in selecting the best plan. Health Insurance Reforms In the history of health care in Chile, there have been two pivotal moments :15 the 1952 reform, which culminated in the creation of a National Health System, and the 1980s reform, which redefined the role of the state within the health care framework, created the ISAPREs, and transferred responsibility for primary care services to the municipal corporations. Chile: Good Practice in Expanding Health Care Coverage 119 Figure 6.19 Chile: Chronology of Health Reforms, 1917­2006 National Health ISAPRES creation GES Service (NHS) and primary level implementation creation municipalization financial public and NHS expenditure management consolidation reduce reforms 1917 1952 1964 1973 1981 1990 2005 maturing of the strong development health system expansion of the new and social of financing, development health provisional supply and health care model Source: Authors. Figure 6.19 provides a timeline for the major developments in the history of the Chilean health system, including the two landmark moments described above fol- lowed by the establishment of the GES in 2005. Developments that did not occur on a specific date are depicted as occurring over a continuum. Six major periods of reform can be identified over the time frame examined here (1917­2006). The 1952 Reform The health and social security system entered a period of maturation in 1917. That year, all the social welfare agencies and insurance providers participated in a national congress convened to improve organization within the health and health insurance system. In 1924, Chile established a Ministry of Hygiene, Social Assis- tance, and Social Protection; Mandatory Worker Insurance; and a Workers' Insur- ance Fund to cover risks of disability, aging, and death. In 1938, the pioneer Law of Preventive Medicine was drafted. SERMENA (Employees' National Medical Ser- vice) was created, covering public and private sector employees, and in 1948, the Medical Association of Chile was established. Creation of SNS. The National Health Service (SNS) was created in 1952, uniting various public entities that provided health services. These included the General Health Office, hospital services provided by the Central Association of Benefi- cence and Social Assistance; the medical department of the Workers' Insurance Fund; the Office of Protection for Mothers, Children, and Adolescents; and certain municipal medical services. With the creation of SNS, nearly 90 percent of public 120 Good Practice in Health Financing resources for health were concentrated in one entity. Other coverage was provided by universities, public companies, the armed forces, and police departments. SNS was led by a general director, with two suboffices (one technical, the other administrative) and emphasized centralized planning. SNS divided its efforts across three geographic territories, each covering one or more provinces (Palma, Fried- mann, and Heyermann 1995). Evolution of the reform. In 1964, Chile initiated a variety of social reforms, includ- ing investment in health centers, training for medical personnel, and extension of geographic coverage. SNS eventually employed 120,000 functionaries, in a health network covering a large part of the geographic terrain and population. Central- ized planning dominated political, economic, and organizational development in the health sector. From 1964 to 1973, health care was closely associated with social development by Chile's social and political leadership. From 1973 to 1980, social spending and funding for SNS declined significantly. The organizational frame- work of SNS was maintained but was controlled by the authoritarian regime. After 1979, the state health sector was restructured. The Ministry of Health and related institutions were reorganized, and the National Health Services System (SNSS) was established. The 1980s Reform In 1981, the private sector entered the health insurance market. ISAPREs and decentralization. In 1985, the Health Law created a framework for health care provision, including the current model for financing, insurance, and health services. Between 1981 and 1986, various legal initiatives created the ISAPREs and transferred responsibility for most primary care facilities to the municipalities. Validation of the model with minor adjustments. Since 1990, the elected Coali- tion of Parties for Democracy administrations has assumed responsibility for the health system they inherited. The current health system maintains many of the same financing, organizational, and functional characteristics of the model estab- lished during the military period, especially in terms of health care provision (governed by the Health Law), the legal-normative configuration of the SNSS, and fiscal financing of the public system, the municipal administration of health facil- ities, the role of the ISAPREs in insurance provision, and the model for worker compensation claims. However, recent administrations have implemented some reforms mainly to: increase investment in the health sector, with the support of international loans and researching proposals to improve and reform the sector; to improve public sector administration; and to accelerate the decentralization process. Some changes have been made in the primary care statute and the laws governing med- Chile: Good Practice in Expanding Health Care Coverage 121 icine, ISAPREs, and the FONASA. The major reform has been the introduction of the GES, as described above. Evaluation of the Insurance Reforms Many of the changes in the health system were closely tied to the economic, social, and political context in which they occurred. Therefore, to evaluate the principal Chilean health system reforms, a chronological analysis covering three periods is provided: centralized planning under the SNS (1952­73); 1980s reform under the military government (1973­90); and adjustments during the current democratic period (1990­2006). Centralized planning under the SNS, 1952­73. The SNS operated for 27 years and developed five important national programs: Program for Maternal Health, Program for Child and Adolescent Health, Program for Senescent Adults, Pro- gram for Social Health, and Program for Environmental Health. Most of the country and population groups were covered by these programs.16 In the mid-1960s and early 1970s, society became increasingly polarized, divid- ing the political-social arena into two diverging approaches. From 1964 to 1970, the state was involved in the market, but social policy was secondary to other con- cerns. Between 1970 and 1973, the market was closely regulated by the state, with the explicit goal of achieving certain social objectives. Public spending on social programs as a percentage of gross global product (GGP) increased during both periods, from 16.3 percent in 1963 to 25.8 percent in 1971. Public spending on health also increased as a percentage of GGP, from 3.1 percent to 4.0 percent over the same period. From 1964 to 1970, the SNS attempted to cover as broad a segment of the pop- ulation as possible, with the goal of having only one health provider. Health pro- vision entities specific to certain sectors or unions were dissolved. In the early 1970s, the government created the Single Health Care Service (Ser- vicio Único de Salud, SUS), in an attempt to consolidate health care provision into one egalitarian, effective, coherent, and free service provider. However, these changes did not have the intended effect, and the SUS's functions were eventually absorbed by other public and private institutions. To expand coverage, in 1968 the Curative Medicine for Employees law was drafted, bringing 2.5 million public and private employees into the socialized health system. By the mid-1970s, the health system had broadened its coverage, continuing the process initiated 20 years earlier. By 1920, the beneficiary popula- tion of the main health providers included the SNS--dependent workers and laborers (both active and retired), the indigent, the unemployed, and their fami- lies; SERMENA--employees (both active and retired) enrolled in various insur- ance funds (state and private) and their families; and Security Assistance Institutions (Mutuales de Seguridad)--active workers whose companies carry workers' compensation insurance. 122 Good Practice in Health Financing The insured population consisted of workers covered by SNS (50 percent), employees covered by SERMENA (20 percent), and independent workers who paid for coverage through either SNS or SERMENA (7 percent). Between 18 and 20 percent were not covered by any of those institutions. According to Palma, et al (1995), global health indictors continuously improved under the SNS, and faster than elsewhere in Latin America. The number of pro- fessional medical services provided also increased--as did health spending. However, the SNS had a number of flaws: excessive centralism and a swelling bureaucracy; inefficiency in the use and distribution of resources; administrative inefficiencies; inequitable payment and benefit schedules; and inadequate target- ing of subsidies. 1980s reform under the military government. The military government's health policies were based on three principles: a limited role for the state; free market policies; and support for private business as the key to economic development.17 The military regime developed new health policies in four stages: · Initial attempts to restructure the public health sector (1974­78). This phase was characterized by strong inertia within the public health system, punctuated by the announcement in 1975 that services would no longer be provided free of charge, except in cases of extreme indigence. · Definitive restructuring of the system (1979­80). SNS beneficiaries were allowed free choice among public and private providers. The Ministry of Health was reorganized into three branches with clearly defined responsibilities: the Min- istry of Health (regulation of the health system); Health Services (execution of health services); and the FONASA (financing of health services). · Participation of the private sector and municipalization of primary-level care (1980­84). In 1981 the private sector entered the health insurance market in the form of ISAPREs. These companies collected workers' mandatory health premiums and in exchange provided the services that would have fallen to the Health Services department and the FONASA. Service contracts covered inpa- tient care, intensive care, and emergency care. As part of the decentralization process initiated in 1981, responsibility for primary-level care (general urban and rural health posts and facilities) was transferred to the municipalities. · Conclusion of the health sector reform (1985­90). The fourth stage ended with the Health Law, which had the following goals: equity in the health system, eliminating the disparities in quality of care; distributive justice, establishing a mechanism through which the consumer pays according to ability; and improved access, guaranteeing free or egalitarian opportunities for health care through a free choice system. Health indicators measured during the military regime showed that the system was effective; however, there were gaps in efficiency and equity. In terms of effi- Chile: Good Practice in Expanding Health Care Coverage 123 ciency, the main problems were ineffective management of resources by health facilities and providers and failure to fully implement certain reforms (especially the restructuring and decentralization of the system). In terms of equity, the reforms channeled Chileans by income into one of two health systems. Spending per capita and services provided per beneficiary were quite different under each system (Oyarzo, César 1993). Adjustments during the democratic period. After the military government ended, the new administrations tried to accelerate decentralization of the Chilean health system. However, the SEREMIS were weak, had poorly defined responsibilities, and had many organizational flaws, including lack of collaboratively generated plans, objectives, and goals, lack of internal coordination, and poor flow of information. Despite these weaknesses, the public sector sought to transform its health ser- vice providers and hospitals into autonomous entities through a series of actions intended to strengthen local administration. The government also sought to cre- ate technical, legal, and administrative conditions at other levels of the state to support this process. During this period, the state increased the MOH budget as a percentage of total fiscal budget. Along with the international loans, this increase represented a strong investment in health services.18 It also established free health care for all FONASA beneficiaries (about 70 percent of the population at that time) and supported primary care facilities by establishing Regional Funds, creat- ing a network of emergency primary-level services and increasing the real salaries of public health sector workers. At the same time, it explored new ways of financ- ing the health system at the operational level. Despite these measures, the CEP-Adimark survey carried out in the mid-1990s indicated that the public perceived a health sector in crisis. Some of the perceived flaws included insufficient decentralization of the system, inefficiency of the bureaucratic-administrative system, poor management of the public health sys- tem, and inadequate funding (Palma, Friedmann, and Heyermann 1995). Aedo (2000) reported in the late 1990s: Chile has achieved better global health indicators than predicted for a country at this level of development. However, the public indicates low levels of satisfaction with their health system. The public sector suffers from funding and administrative limitations. The ISAPRE system is perceived as providing good health services, but under an inequitable funding model. There are few good alternatives, then, for the elderly or chronically ill. Table 6.13 summarizes Chile's global health indicators. Aedo (2000) also identified some factors that have limited the development of the post-reform Chilean health system. They include · Imperfections in the health insurance market. The FONASA charges a monthly premium on a sliding scale according to beneficiaries' income and provides net 124 Good Practice in Health Financing Table 6.13 Chile: Global Health Indicators: 1960 and 1995 Indicator 1960 1995 Infant mortality 120/1,000 live births 11.1/1,000 live births Maternal mortality 3/1,000 live births 0.4/1,000 live births General mortality 12/1,000 population 5.5/1,000 population Mortality from tuberculosis 53/1,000 cases 5/1,000 cases Mortality from infections 193/1,000 cases 5/1,000 cases Life expectancy 58 years 74.8 years Source: MOH http://www.minsal.cl. benefits inversely proportional to income. ISAPRE beneficiaries, in contrast, receive net benefits that are proportional to their contribution. Furthermore, price schedules, promptness of delivery of care, and level of income vary between the two systems. This disparity has created tension between the two systems. · Mandatory contributions for health care and the design of the mandatory insur- ance policy. Making participation mandatory reduced opportunism (deriving unearned benefits from the system) and partially counterbalances the informa- tion asymmetries. However, the mandatory 7-percent-of-salary payment cre- ates two problems. First, a number of people will be overinsured relative to their risk and income. Second, allowing individuals to design their own insur- ance plan can lead individuals to focus on outpatient rather than hospital care. This thwarts the function of insurance and of mandatory participation. · Spending on health care for older adults. Chile is in a transitional phase in which the population is aging. Some studies have indicated that there should be ade- quate funds to finance care for the older adult, both because they have fewer dependents than younger people and because many will have a pension. How- ever, social security coverage has decreased steadily since the 1980s, which may open funding gaps. · Increasing costs in the health sector. The cost of medical services and products has increased for various reasons. Although the population is healthier than ever before, individuals are demanding better quality of care. New technologies and developments in health care mean more effective treatment for illnesses but require more highly trained personnel; and new medications are more sophisti- cated and effective, but also more expensive. Moreover, increased life expectancy increases not only the length of time for which a persona will require medical care, but also increases the likelihood of developing a chronic disease. In the mid-1990s, the only changes were modifications in financial and admin- istrative aspects of the FONASA. This was not because the government failed to Chile: Good Practice in Expanding Health Care Coverage 125 recognize the profound epidemiological changes that had occurred over the past 20 years, but because a political decision had been made to postpone major changes until a minimum consensus could be reached. In any case, these changes represented important advances in the operational and conceptual framework of the system, for example, the separation of functions within the public sector, the purchase of services, and the establishment of new payment mechanisms to pro- gressively replace the old procedures for allocating resources. The administrations were also successful in increased public sector coverage of the population. These were the highlights (Acuña 1998): · Increased coverage for low-income groups, accomplished by providing a full subsidy (no copayments) for specific health services. · Establishment of a catastrophe insurance fund for public system beneficiaries, designed to cover all the costs of treating the most common catastrophic ill- nesses (including cervical uterine cancer, breast cancer, prostate cancer, coro- nary disease, renal insufficiency, bone marrow transplants for children under 16 with leukemia, and kidney, liver, and heart transplants). · Establishment of a special program to provide 100 percent coverage for health problems common among adults over 65, including joint replacement surgeries. · Drafting of a Beneficiaries' Bill of Rights defining the rights and responsibilities of public sector consumers. This Bill of Rights was well received by the popula- tion but created problems for public hospitals in that it imposed 25 demands in terms of access to information and standards of care, some of them beyond the facilities' capabilities. The final important reform in the Chilean health sector has been the establish- ment of the Explicit Health Guarantees System Act by which the state guarantees the entire population coverage, access, financing, and quality of care for a variety of health problems, without discriminating by insurance plan, gender, or income. The GES Act represents an important change in the health system, making benefi- ciaries consumers with enforceable rights. According to FONASA data, some public providers cannot satisfy all the GES requirements. In fact, the FONASA estimates that the first 25 health problems covered under the GES created a supply deficit among public providers that com- pelled the system to purchase services from public providers at a cost of about $35 billion in 2005 (about US$60 million). The FONASA estimates that services purchased from public providers in 2006 will cost about $50 billion; no estimates are yet available for 2007. The cost of observing the GES is highly concentrated in specific health prob- lems. Figure 6.20 displays the cumulative cost of the GES, starting with the most expensive health condition covered (arterial hypertension, type II diabetes melli- tus, depression in individuals over the age of 15, and so forth) (Bitrán & Asociados 2005). 126 Good Practice in Health Financing Figure 6.20 Chile: Cumulative Cost of the 56 GES-Covered Health Conditions, estimates for 2007 Primary Arterial Hypertension (persons >15) End Stage Renal Disease Depression (persons >15) Type II Diabetes Mellitus HIV/AIDS Hypoacusia (persons >65) Traumatic Brain Injury Benign Prostatic Hyperplasia Dental Emergencies Diabetic Retinopathy Acute Cerebrovascular Accident Rheumatoid Arthritis Surgical Treatment of Cataracts Cholecystectomy to prevent Gall Bladder Cancer Prematurity Bronchial Asthma Moderate Arthritis of Hip (persons >60) Moderate Arthritis of Knee (persons >55) Respiratory Distress Syndrome, newborns Conduction Disorders (persons >15) requiring a pacemaker Total Hip Replacement (persons >65) with Severe Arthritis Breast Cancer (persons >15) Drug/alcohol dependence, adolescents Gastric Cancer Operable Congenital Cardiopathies (persons <15) Refractory Disorders (persons >65) Acute Myocardial Infarction Ruptured Aneurysm + Intracranial Vascular Malformations Type I Diabetes Mellitus Cervical / Uterine Cancer Cancer (persons <15) Major Burn Hemophilia Testicular Cancer (persons >15) Leukemia in Adults Dental Care (persons >65) Chronic Obstructive Pulmonary Disease Central Nervous System Tumors and Cysts Strabismus (persons >15) Lymphomas (persons >15) Integrated Oral Health for Children Prostate Cancer Schizophrenia Acute Respiratory Infection (persons <5) Herniated Nucleus Pulposus Integrated Care: Pregnancy, Childbirth Surgical Treatment of Scoliosis (persons <25) Cystic Fibrosis Ocular Trauma Ambulatory Pneumonia (persons >65) Nonrefractory Epilepsy Retinal Detachment Polytrauma Orthosis and Technical Aids (persons >65) Pain relief from Advanced Cancer + Palliative Care Cleft Palate 0 10 20 30 40 50 60 70 80 90 100 accumulated cost (%) Source: Bitrán & Asociados 2005. Chile: Good Practice in Expanding Health Care Coverage 127 Future Reforms: Characteristics and Conditions for Success Chile has reached a socioeconomic level that allows it to provide a large segment of its population with care comparable to that available in developed nations, while also providing at least basic health coverage for practically its entire popula- tion. Within this framework, the major challenge is to keep the cost of health care from growing at a rate faster than the country's economic growth can sustain. Below is a summary of priorities identified by health sector specialists and 10 proposals that a group of analysts have suggested as approaches to achieving these priority goals (Health Engineering Commission 2005). Priorities identified by health sector specialists include · Increase funding for investment in hospital equipment for public facilities. Government studies indicate a US$600 million deficit in funds to cover invest- ment needs over the next 10 years. · Contain the growing cost of medical leave claims. Currently the system reim- burses for 100 percent of lost salary, with no maximum length of leave. · Spread the economic risk of health care among insurers and providers. · Allow beneficiaries to allocate their own subsidies. · Improve public services or lose beneficiaries, who, now free to choose providers, seek care in the private sector. · Increase investment in research and development. · Move from a supply-side subsidy to a demand-side subsidy. The 10 proposals identified by analysts are · Establish an explicit minimum level of coverage for the entire population, identifying the specific benefits an individual insurer may exclude. Priority should be given to preventive and coverage for catastrophic illness. · Improve management of public hospitals, introducing competition and flexi- bility. Also, raise private funds for hospitals requiring major investments, contract services out to private facilities, including the Security Assistance Institutions, and change the Worker Compensation law so that employers rather than the public sector provide this coverage. · Incorporate information technology tools to support health sector administra- tion, for example, by incorporating the use of open communications standards between information systems and incorporating systems that allow for the development of networks. · Shift from supply-side to demand-side subsidies; that is, give subsidies directly to the individuals that need them and allow the individuals to freely select the health provider that they prefer. · Encourage the use of managed care, standardizing diagnostic and treatment procedures and establishing protocols for care "packages" at set prices and other mechanisms to transfer per capita risks. 128 Good Practice in Health Financing · Open construction projects and administration of primary care facilities to private companies. · Create special programs for the elderly and chronically ill. The health land- scape is changing, and the system is experiencing more and more pressure from older adults. Priority should be given to primary care and chronic and degenerative diseases that affect older adults. To cover the increased costs that their care represents, insurers should be permitted to charge higher premiums than the 7-percent-of-salary contribution. · Control the cost of paid medical leave. Suggested changes to this subsidy include: establish a maximum duration of paid leave; provide support to work- ers without replacing their entire salary (a form of copayment); establish sanc- tions for workers who abuse the system; allow for some flexibility in the levels of benefit paid and conditions for receiving the benefit. · Update prevention programs to fight the spread of modern diseases such as obesity, diabetes, hypertension, tobacco addiction, high cholesterol, AIDS, drug addiction. · Link premiums to coverage and cost of care. The mandatory contributions for health care were established in 1986 and have never been modified, even though health care has changed significantly in the intervening years. Potential Problems in the Short and Medium Term Some potential short- and medium-term problems that the Chilean health system may present are described next. Public insurer, public providers. The FONASA provides most of its beneficiaries' care in public facilities, which operate on fixed budgets. This situation has allowed the system to provide adequate coverage and quality of care, because operating and management costs are more moderate for public than private facilities (Bitrán & Asociados, 2005). If the FONASA had to buy all its services from private pro- viders, its costs would increase. Asymmetrical regulation. The Superintendent of Health regulates the ISAPREs and FONASA asymmetrically. For example, the information that the two types of providers must report is different. Moreover, the Superintendent of Health may be more lenient with the FONASA than with the ISAPREs in terms of enforcing the GES. Asymmetrical obligations. The ISAPREs and FONASA are not required to follow the same rules in terms of enrollment. ISAPREs are permitted to practice risk selection, but the FONASA must accept any applicant fulfilling its requirements. The risk, due to adverse selection, is that the highest-risk beneficiaries may flock to the FONASA. Furthermore, moving to another ISAPRE is difficult for high-risk Chile: Good Practice in Expanding Health Care Coverage 129 ISAPRE beneficiaries. They remain captive members, facing ever-increasing pre- miums, which makes them more likely to leave the private system and enroll in the FONASA. Financing the GES. The GES was drafted after reviewing studies estimating its implementation costs. However, the cost may be higher than anticipated, leaving the system inadequate resources to finance the guarantees. If funds do not cover the cost of the GES, what will happen is unclear. The GES and politics. It is tempting for politicians to add illnesses to the list of covered conditions as a way of gaining popularity, but adding new guarantees without conducting cost studies could jeopardize the sustainability of the system. Another latent risk is that that so many conditions will be added to the GES that the cost of health spending will escalate, and the guarantees will be impossible to meet. The system and politics. The ISAPREs are perceived as elite institutions by the common citizen. Therefore, there is a risk that, in the name of equity, the system will be dismantled in favor of channeling more resources to the public system. Changes in public hospital administration. There have been trends toward reforming public hospital administration and governance, for example by intro- ducing competition for contracts or providing facilities with greater autonomy to produce an atmosphere of greater accountability. However, these changes will probably occur slowly, if at all, because of resistance from various interest groups. Lack of coordination and moral hazard. Primary care is managed at municipal level, while hospital care is administered centrally. At times, coordination among levels has been poor, reducing the effectiveness of care. In other cases, this discon- nect has meant that a public hospital provides care that could have been provided at a primary facility. This is significant, because the municipalities receive capi- tated payments (according to the municipality's population). Focus of the GES. The GES does not give preventive care a priority, nor does it necessarily encourage the most cost-effective public health measures. This decreases the efficiency of health spending in Chile. Price increases. Traditionally, the ISAPREs have reimbursed providers per service provided. This system is inflationary. Now, the system is making a transition to a managed care scheme that will better contain costs. Health Coverage Reforms: Lessons for Other Countries The Chilean health system is characterized by broad coverage under its social security system, with only 10 percent of the population not covered by any 130 Good Practice in Health Financing insurer. Some of the factors that have allowed this system to function adequately are described below. Conditional Factors Implementing a social security system requires considerable economic resources, especially in the early years, because of moral hazard, or the likelihood that an insured population will use more health services than the uninsured population. Therefore, implementation is easier when public finances are adequate to bear these costs. Implementation of such systems typically occurs during periods of expansion, when economic growth is strong. Furthermore, because the state sub- sidizes the health social security system for about a quarter of its population, it must have adequate resources to do so. The greater the economic growth, the more taxes the state can collect to fund these costs. The greater the concentration of the population in urban areas, the broader is the coverage of a social security system, due to reduced administrative costs. In Chile, the rural population amounts to just over 10 percent of the total popula- tion, a facilitating factor in coverage reforms. The higher the income level within an economy, the greater is the probability of its being able to bear the high administrative and operational costs of a social secu- rity system with wide coverage. In Chile, the income level is considered medium- high. Income per capita was about US$7,700 in 2005, which partially explains the success of the social security system. Necessary Institutional Arrangements "Institutional agreements are the overall rules that allow the actors, through polit- ical negotiation, to make economic and political deals," according to José Ayala Espino (2001).19 Chile has always had credible, independent, efficient institutions, and it has one of the lowest international risk ratings among Latin American countries. Its strong institutions have played a vital role in its health reforms. Existence of a competent, independent regulatory entity. Because the private sec- tor provides services within an imperfect market (e.g., with information asymme- tries), a regulatory entity must ensure that the market behaves as if it were perfectly competitive. This entity must enforce general and sector-specific legal regulations and protect the rights of beneficiaries. Although the Superintendent of Health did not begin to regulate private insurers until 10 years after the system had been implemented, and did not begin to regulate the FONASA until 10 years after that, the creation of this entity at the start of the system provides a lesson learned for other countries. Clear rules of the game. The three branches of government--executive, legisla- tive, and judicial--must provide regulatory tools and mechanisms that allow Chile: Good Practice in Expanding Health Care Coverage 131 insurers to understand their rights and obligations and protect their investments. At the same time, the state needs to provide the Superintendent of Health with a framework that will allow it to carry out its fiscal responsibilities. Chile's institu- tional solidity has fostered competition among private insurers and encouraged high standards to maintain their solvency and prestige. A predominantly formal labor market. A health social security system is easier to implement where the institutional and legal framework fosters formality, particu- larly a formal labor market. Administrative processes are simplified in such a situ- ation, as is the collection of contributions. The Chilean economy in general and the labor market in particular are fairly formal in comparison with those in other Latin American countries. Competition among insurers. A pillar of the Chilean social security system's effi- ciency is competition among insurers, not least because markets in other sectors of the Chilean economy are competitive. In Chile, one public insurer and various pri- vate insurers, in theory, compete for beneficiaries. In practice, however, the only true competition within the health social security system takes place at the level of certain socioeconomic groups. The great majority of the poor are FONASA bene- ficiaries; most of the rich are ISAPRE subscribers. Existence of a public insurance system. If the social security system were dissolved in favor of a completely free market system with only private insurers, it is likely that the adverse selection process would leave only the lowest-risk, highest- income individuals insured (with complete coverage). To forestall this problem, the public insurer in Chile is not permitted to employ discriminatory practices, making it possible to achieve broad coverage of the population. Key Factors in Financing a Social Security System for Health Care Cost, equity, efficiency, and sustainability are the key factors in financing a social security system for health care. Escalating costs. Providing such health care coverage usually triggers a burst of consumer spending, partly due to moral hazard. For example, the insured consult a physician more frequently, take more medicines, and generally use more health care the uninsured. To contain public health spending, the Chilean reform set bud- getary ceilings, based on previous years' spending. This has provided a constant supply of health services, with demand controlled via rationing mechanisms. Adequate supply of health services. The greater the supply of public and private health services, and the greater the efficiency of those provider facilities, the bet- ter an extensive social security system will be able to function. Chile had the 132 Good Practice in Health Financing advantage of adequate health services from the beginning, although there have been difficulties in supplying less-populated and less-accessible areas with services. Adequate funding sources. A social security system should be financed through various sources. In Chile, these sources include monthly contributions from ben- eficiaries, public funds (used mainly to cover the medical costs of the indigent), copayments from beneficiaries, and other income from health services. Private funding support is essential for a sustainable system. In Chile, private support comes from two main sources: copayments, which vary according to ser- vices consumed, and monthly contributions, which are set at 7 percent of the ben- eficiary's income. The average salary in Chile was US$490. Countries with lower income levels will likely need to charge the beneficiaries higher rates to beneficia- ries to offer a comparable benefits package. The required contribution in Chile has risen over time, from 4 percent in 1981 to its current rate of 7 percent. Efficient tax collection agencies. A third to a half of the health care provided in Chile is funded by public spending. Efficient tax collection and low evasion help ensure that the money will be there to pay for public spending. Mandatory membership. Membership in the Chilean social security system is mandatory for formal workers and voluntary for independent and informal workers. The problem of adverse selection, present in any insurance system, is minimized when membership is mandatory. Risk diversification--spread across a broad and diverse membership--reduces the costs of providing insurance. Key Political and Economic Factors in Reform Success Many political and economic factors influence the outcome of reform. They include a country's form of government, social solidarity, subsidy system, social and political accord, administrative system, and governmental credibility. Type of government system enacting the reform. The 1980s health sector reform transferring many health functions from the public to the private sector in Chile was carried out under an autocratic government, reducing the potential for resis- tance from involved parties. Other countries enacting such a reform under a democratic government would likely have to contend with social conflicts. Such resistance might jeopardize the success of the reform or result in only partial implementation. Solidarity. As a rule, the greater the tradition of solidarity within a country--that is, the more accustomed systems and institutions are to working together toward social consensus--the better are the chances of a successful social security system founded on risk pooling. In Chile, however, it may be argued, social security sys- Chile: Good Practice in Expanding Health Care Coverage 133 tem relies less on solidarity than other systems, because the cross-subsidies pro- vided are smaller than in other systems. Subsidy system. A quarter of the Chilean population receives free social security benefits through subsidies provided mainly by the state, supplemented by small cross-subsidies contributed by other members. Therefore, although health care is ostensibly provided under the social security modality, the system functions by financing health care through tax revenues. Social and political accord. The Chilean system includes two administrators, one public and one private. These administrators are linked to public and private providers. To work, this system depends on a certain level of social and politi- cal accord. Many other countries prefer an entirely public or an entirely private system. Capable administrators. The more capable the system administrators and the better the public's confidence in them, the more efficient and reliable is the system. A capable and credible government. As with the administrators, the more capably the government manages the system, the better is the system's chance of success. Replicability in Low-Income Countries In light of the lessons noted above, replicating the Chilean model in low-income countries may be difficult. Many low-income countries also have additional chal- lenges: a highly informal labor market, large rural population, too small a part of the population with sufficient income to cofinance the costs of a system offering an attractive benefits package, public lack of confidence in the reforming authori- ties and institutions, insufficient human resources and institutions capable of managing the changes, high resistance from interest groups, inability to limit fraud due to information gaps, shortages of health services to satisfy increased postreform demand, and insufficient budget to cover implementation costs. Replicability in Middle-Income Countries Middle-income countries are better situated to implement a social security system similar to the Chilean system. Where lower-income countries have challenges, middle-income countries usually have advantages: a larger formal labor market, a smaller rural population, a broader population base of people with enough income to bear the costs of a system offering an attractive benefits package, better credibility on the part of the authorities and institutions that would execute the reforms, more human resources and institutions capable of managing the changes, experienced coalitions to negotiate with interest groups to achieve the required reforms, better ability to limit fraud within the system, a better supply of health services to meet increased postreform demand, and enough money to pay for it all. 134 Good Practice in Health Financing Endnotes 1. In 2005 the Collective Capitalization Fund amounted to US$1.1 billion. 2. The data used to describe Chilean demographics in the mid-20th century were taken from the 1952 census, and the data to describe the end of the 20th century were taken from the 1992 census. 3. The dependency index is expressed in terms of unproductive population (those under 15 and over 65 years of age) divided by the productive population (those between the ages of 15 and 65). The index of aging represents the number people over 65 divided by the number of children under 15. 4. A DALY, disability-adjusted life year, represents one lost year of healthy life, due to premature death or disability. 5. This historical timeline is based on http://www.memoriachilena.cl "Elecciones, sufragio y democracia en Chile (1810­2005)." 6. See http://www.gobiernodechile.cl. 7. The maximum legal monthly contribution to FONASA or an ISAPRE is therefore US$140. 8. Some experts feel that official data on health spending as a percentage of GDP underestimate out-of-pocket spending on medications and that the true figure is closer to 8 percent of GDP. 9. Before the GESs were established, there was no national benefits package. FONASA did not guarantee specific rights to its beneficiaries. 10. Before the GESs were established, a Family Health Plan covered certain preventative and curative care in municipal facilities. This plan became part of the GHS. 11. There are two types of ISAPREs, open and closed. Open ISAPREs are available to anyone working in a given sector, while closed ISAPREs are available only to workers in cer- tain companies, both public and private. 12. UF [unidades de fomento] are units of account linked to the Chilean consumer price index. UF are used to set a price that will not devalue over time as a result of inflation. One UF is about US$35. 13. In Chile, the ISAPREs allow members to self-select by offering a variety of health plans. The ISAPREs also screen members for preexisting conditions, whereas by law FONASA may not reject applicants. 14. For more information, visit the Chilean Ministry of Health's Web page: http://www. minsal.cl. 15. Source: http://www.colegiomedico.cl. 16. This section is based on findings reported by Palma, Friedmann, and Heyermann (1995). 17. A limited role for the state means here that the state should assume direct control only of functions serving the common good and not addressed by the private sector. This principle leaves the free market to allocate economic resources through supply and demand, assigning the state the role of regulating the markets as needed. 18. Investment in health services increased from US$10 million in 1986­90 to US$540 million in 1990­97. References Acuña, Cecilia. 1998. "Evolución y reforma de los sistemas de protección de la salud, en los países del MERCOSUR y en Chile." http://www.fes.org.ar/. Chile: Good Practice in Expanding Health Care Coverage 135 Aedo, Cristian. 2000."Las reformas en la salud en Chile." In La transformación económica de Chile. Felipe Larraín and Rodrigo Vergara, ed. Santiago: Centro de Estudios Públicos. http://www.cepchile.cl/dms/archivo_3267_1620/ 14_aedo.pdf Ayala Espino, José. 2001. "Políticas de estado y arreglos institucionales para el desarrollo de México." Revista Planeación y Desarrollo 8/9 http://dialnet.unirioja.es/servlet/ articulo?codigo= 298394. Bastías, G., G. Marshall, and D. Zúñiga. 2000. "Número de médicos en Chile: estimaciones, proyecciones y comparación internacional." Revista médica de Chile 128(10). http://www. scielo.cl/scielo.php?pid=S0034-98872002000500015&script=sci_arttext. Bitrán & Asociados. 2006. "Coyuntura del sistema de salud en Chile: oportunidades y desafíos para clínica Las Condes." Documento elaborado para la Clínica Las Condes. ------. 2005. "Verificación del costo esperado por beneficiario del conjunto priorizado de problemas de salud con garantías explícitas 2005­2007." Documento elaborado para el Ministerio de salud de Chile. http://www.minsal.cl/ici/destacados/estudio_verificacion/ Resumen_Ejecutivo_GES_30_Noviembre_2005_Corregido.pdf. Bitrán, Ricardo (1997). "Equity in the Financing of Social Security for Health in Chile." Partnerships for Public Health, Abt Associates, Inc., Bethesda, MD. http://ideas.repec. org/a/eee/hepoli/v50y2000i3p171-196.html. Health Engineering Commission. 2005."El sistema de salud de Chile: Análisis, 10 propues- tas de cambios." Instituto de ingenieros de Chile, Santiago, Chile. Documento elaborado para el Ministerio de Planificación de Chile. Melis Jacob, Fernanda. 2000. "La situación de la mujer en Chile." Revista Occidente 56 http://www.revistaoccidente.cl/374/reportajes/SituacionMujer/lasituacion374.act. Concha, Marisol. 1996. "La carga de la enfermedad." Documento elaborado para el MOH. http://epi.minsal.cl/epi/html/sdesalud/carga/Inffin-carga-enf. pdf. Oyarzo, Cesar. 1993. "La mezcla pública privada: una reforma pendiente." Estudios Públi- cos 55, Santiago. http://www.cepchile.cl/dms/archivo_1635_169/rev55_oyarzo.pdf. Palma, E., R. Friedmann, and B. Heyermann. 1995. "El sistema de salud en Chile ante la descentralización política." Latin American and Caribbean Institute for Economic and Social Planning, World Health Organization, and Ministry of Health, Santiago. Superintendent of ISAPREs. 1999. "Sistema de salud privado en Chile." In L. Figueroa, and V. Lazén."Propuestas de políticas de salud privada en Chile, 2001." http://www.perseo.cl/ Politicas_de_salud.pdf. Titelman, Daniel. 2000. "Reformas al sistema de salud en Chile: desafíos pendientes." En Reformas, crecimiento y políticas sociales en Chile desde 1973. Ricardo French-Davis, ed. Santiago: Comisión Económica para América Latina. 7 Colombia: Good Practices in Expanding Health Care Coverage Diana Masis Pinto Background Colombia is a 1,038,700-km2, lower-middle income country in the Latin America and Caribbean Region of the World Bank. Its population numbers 41.2 million, with a per capita income (PPP-adjusted) of US$7,769.The country is divided into four capital districts corresponding to major cities and into 32 administrative depart- ments, subdivided into 1,098 municipalities. The concentration of more than 60 per- cent of the population in the six largest urban municipalities facilitates health insurance coverage in those areas. Seventy percent of the other municipalities are rural and have fewer than 20,000 inhabitants, posing challenges for insurance expan- sion and implementation of the managed competition model. In 1993, Colombia introduced one of the most ambitious health care reforms in Latin America, creating a National Social Health Insurance (NSHI), organized as a model of "managed competition." Two insurance schemes were implemented, the contributory regime (CR), for people who could pay, and the subsidized regime (SR), for those who could not. To expand health insurance coverage to all Colombians, the reforms mobilized a substantial amount of resources and undertook simultaneous changes in the organization of the health care delivery system and the administrative structure. Total health expenditures were $136 per capita in 2003--7.8 percent of GDP. The health system is largely publicly financed--with 15.9 percent of this amount paid by private sources, including just 7.5 percent paid directly by households as out-of-pocket expenditures. Life expectancy is 72.5 years, and infant mortality is 17.5 per 1,000 live births. 137 138 Good Practice in Health Financing Introduction Economic Environment Until the mid-1990s Colombia had one of the most stable economies in Latin America. Economic growth was good until 1996, when a period of economic reces- sion began. In 1999, unemployment hit 20.1 percent, and poverty reached a peak of 58.2 percent of the population. Social public expenditure as percentage of GDP decreased from 15.3 percent in 1997 to 13.6 percent in 2000 (Lasso 2004), which had a negative impact on enrollment in the National Social Health Insurance (NSHI). Since 2000, the economy has been on a path to recovery, which, if contin- ued, will be likely to favor current health insurance expansion policies. Table 7.1 presents current selected macroeconomic indicators. Demographic and epidemiologic profile The key demographic indicators for Colombia are shown in table 7.2. Total life expectancy at birth in Colombia was 72.5 years in 2004. Life expectancy for males was six years less than for females, a likely result of male mortality due to violence. Colombia has moderate birth, mortality, and fertility rates. Large disparities in child mortality persist between urban and rural areas. According to the 2005 census, the population is 49 percent male and 51 percent female. Age-specific fertility rates have been decreasing for all age groups since 1985 (figure 7.1). The population pyramid (figure 7.2) shows fewer people in the younger age categories as the population has shifted to more conservative birth rates. Table 7.1 Colombia: Selected Economic Indicators Indicator Value GDP per capita, Colombia, 2004 (constant 2004 US$)a 2,099 GDP per capita, Latin America, 2004 (constant 2004 US$) 2,835 GDP growth, annual, 2004 (%)a 7,256 Annual GDP growth rate (2004)b 4.3 Population under living under poverty line, 2005 (%)b 49.2 Population living under US$1 per day, 2004 (%)b 2.4 Population living under US$2 per day, 2004 (%)b 7.6 National unemployment rate, 2006 (%)c 11.3 Revenue, excluding grants (% of GDP)a 17.1 Tax revenue (% of GDP)a 13.8 External debt (% of GDP)a 37.1 Aid (% of GNI)d 0.5 Sources: a. World Bank 2006; b. MERPD 2006; DNP 2006; c. DANE 2006; d. IMF 2006. Colombia: Good Practices in Expanding Health Care Coverage 139 Table 7.2 Colombia: Demographic Profile Indicator Value Total population (2005)a 41,200,000 Percentage urban population (%)a 75 Life expectancy at birth, total, 2004 (years)b 72.5 Life expectancy at birth, females, 2004 (years)b 75.7 Life expectancy at birth, males, 2004 (years)b 69.6 Mortality rate for children under five years per 1,000 live births, 2005c Urban 19 Rural 30 Infant mortality rate per 1,000 live births, 2005c Urban 17 Rural 24 Total mortality rate per 100,000, 2002d 438.6 Male total mortality rate per 100,000, 2002d 535.5 Female total mortality rate per 100,000, 2002d 343.7 Maternal mortality rate per 100,000 newborns, 2002d 99.8 Total fertility rate per woman, 2005c Urban 2.1 Rural 3.4 Total dependency rate dependents to working-age populationa 57.6 Age dependency ratio (dependents to working-age population), 2004b 0.57 Sources: a. DANE 2005; b. World Bank 2006; c. Profamilia 2005; d. DANE 2002. Colombia has a double burden of disease: an increasing incidence of chronic and degenerative diseases typical of developed countries and persistence or resur- gence of infectious and parasitic diseases such as malaria and tuberculosis. Main causes of mortality for 2002 are summarized in table 7.3. Tables 7.4 and 7.5 show the top five conditions by contribution to the total burden of disease (DALYs). This profile poses several challenges for the National Social Health Insurance System (NSHIS). For example, the contents of the benefits packages need to be revised and updated to include the key interventions for the prevention and con- trol of these diseases. The system also needs to anticipate measures to deal with a trend toward rising costs for treatment of diseases of aging. Provision of incen- tives for preventing these diseases by health insurance plans and articulation of private and public programs is another key implication. Government and Political Environment Colombia has been a republic since 1810, when it declared its independence from Spain. The government structure is divided in three branches: the executive, the legislative, and the judicial. The executive branch is represented at the national level by a president elected by popular vote for four-year terms with the possibil- ity of reelection and at the departmental and municipal levels by governors and 140 Good Practice in Health Financing Figure 7.1 Colombia: Age-Specific Fertility Rates 180 160 140 120 women 100 1,000 80 per 60 births 40 20 0 15­19 20­24 25­29 30­34 35­39 40­44 45­49 Age 2000­2005 1995­99 1990­94 1985­89 Source: ORC Macro, 2006. MEASURE DHS STATcompiler (http://www.measuredhs.com). Figure 7.2 Colombia: Population Pyramid, 2005 85 y 80­84 75­79 male female 70­74 65­69 60­64 55­59 50­54 45­49 age 40­44 35­39 30­34 25­29 20­24 15­19 10­14 5­9 0­4 10 8 6 4 2 0 2 4 6 8 10 percent Source: DANE 2005. Colombia: Good Practices in Expanding Health Care Coverage 141 Table 7.3 Colombia: Main Causes of Mortality, by Age Group and Gender, 2002 Top three causes of mortality Age (years) Males Females < 1 Specific perinatal respiratory diseases Same as males Genetic and congenital malformations Other perinatal conditions 1­4 Acute respiratory infections Nutritional deficiencies and anemia Nutritional deficiencies and anemia Acute respiratory infections Accidental drowning Intestinal infections 5­14 Motor vehicle accidents Motor vehicle accidents Homicides Homicides Accidental drowning Lymphoma and leukemia 15­44 Homicides Homicides Motor vehicle accidents Motor vehicle accidents Accidental drowning Pregnancy, childbirth, and puerperium 45­64 Homicides Ischemic heart disease Ischemic heart disease Cerebrovascular disease Cerebrovascular disease Diabetes > 65 Ischemic heart disease Same as males Cerebrovascular disease Chronic respiratory disease Source: DANE 2002. Table 7.4 Colombia: Top Five Health Conditions, by Burden of Disease and Cause, 2002 (DALYs) Rank Health condition Total estimated DALYs (thousands) 1 Intentional injuries 1,882 2 Neuropsychiatric conditions 1,825 3 Infectious and parasitic diseases 800 4 Unintentional injuries 651 5 Cardiovascular diseases 575 Source: WHO 2002. Table 7.5 Colombia: Burden of Disease, by Disease Category and Cause, 2002 (DALYs) Total estimated DALYs Rank Disease category (thousands) % total 1 Communicable, maternal, perinatal, and nutritional conditions 1,499 17.8 2 Noncommunicable diseases 4,380 52.1 3 Injuries 2,533 30.1 Total 8,412 100.0 Source: WHO 2002. 142 Good Practice in Health Financing majors, respectively, both elected for three-year terms. The legislative branch is a bicameral congress, composed of the Senate and the House of Representatives. Colombia's 1991 Constitution assigned the state the responsibility for direct- ing, coordinating, and regulating a universal social security service, thus establish- ing a legal framework for National Social Health Insurance. In 2006, a center-right president was reelected, whose agenda for the next four years regarding health policy is centered on achieving universal health insurance coverage (Departa- mento Nacional de Planeación 2006). The head of the health sector is the Ministry of Social Protection (MPS), created in 2003 by merging the former Ministries of Health and Labor. Colombia has had a long history of armed conflict between guerrilla groups and military forces, initially motivated by leftist ideology. In the past decade, con- flict has been increasingly led by economic interests in control of drug traffic zones, with increased participation of paramilitary groups. Armed conflict has implied large losses for the Colombian economy--estimated at between one and two percentage points a year through losses in human and physical capital and productivity (Cárdenas 2006). Additionally, the opportunity costs of expenditures on public and private defense, repair of losses in infrastructure, and compensation of victims could amount to 2 percent of GDP per year, the approximate current total cost of NSHI for the poor. The health toll of armed conflict is also high (Pinto, Vergara, Lahuerta 2005). Armed conflict is likely to have: reduced life expectancy (by an estimated 1.7 years in life expectancy at birth for all Colombian citizens); increased mortality (10 times greater probability of dying for populations displaced by conflict than for the general population and a 6 times higher probability of dying for men than for women); reduced fertility (due to premature female deaths and widowhood); and increased disease (reemergence of vector-transmitted diseases such as malaria, dengue, and yellow fever in conflict zones due to security obstacles for reaching this population by government programs). Besides the expenditures on medical care necessitated by violence-related events, additional resources have had to be allo- cated to attend to the special needs of the growing numbers of displaced people. To end this problem, Colombia has implemented a combination of strategies and policies. They include cease-fires and negotiations with armed groups, invest- ments in national security and drug traffic control, and implementation of pro- poor economic policies such as provision of the NSHI as a strategy to protect vulnerable groups. Overview of Health Financing and Coverage The NSHI was introduced by the 1993 health care reforms as the key instrument for financing health care. It was designed as two separate insurance schemes according to ability to pay. This section describes each scheme in terms of its tar- get population and coverage, sources of financing, and payment mechanisms for insurers and providers, and compares the benefits packages. It also explains how Colombia: Good Practices in Expanding Health Care Coverage 143 services not covered by the benefits package and services for the uninsured are financed. Finally, key health expenditure indicators and indicators of access and financial protection are presented. National Social Health Insurance Schemes The contributory regime (CR) is the mandatory health insurance scheme for the formally employed (formal sector workers) and the self-employed1 (informal sec- tor workers) who can pay, and pensioners. The subsidized regime (SR) allocates public monies to pay individual insurance premiums for the poor. The contributory regime. CR health insurance covers all first-degree family mem- bers as beneficiaries of contributing individuals. In the case of working couples, both must contribute to a premium, called the Unidad de pago por capitación (UPC, per capita payment unit). By mid-2006, 15.9 million people were enrolled in the CR (34 percent of total population); 46 percent were individuals contribut- ing to the premium and 53 percent were their beneficiaries (MPS 2006). Individuals contribute 12 percent of their total salary to the UPC. In the case of formal workers, 4 percent is contributed by the employee and 8 percent by the employer. Independent workers contribute the full 12 percent of their salary, start- ing from a floor of two minimum wages (around US$355 in 2006). Individuals choose to enroll in health plans2 that serve the CR, called Empresas Promotoras de Salud (EPS, Health Promoting Companies), which collect the wage contributions and transfer them to a central fund that pools and distributes revenues among subfunds. This administrative fund is called the Fondo de Seguridad Social y Garantía (FOSYGA, Social Security and Guarantees Fund). The flow of funds of the CR takes place in the following way (figure 7.3): · FOSYGA receives total wage contributions collected by health plans from employers and from individuals. · To finance the SR, 1 percent of total revenues is transferred to a solidarity subfund. · To finance health promotion and prevention activities, 0.41 percent of total revenues is transferred to a promotion and prevention subfund. · For sickness leave payments, 0.25 percent of total revenues is set aside. · For maternity leave payments, 0.25 percent of total revenues is set aside. The remaining revenues are used for payment of UPCs to the EPS and for pay- ment of claims filed by these health plans to the FOSYGA for medications and procedures not included in the benefits package. Surpluses are invested in govern- ment bonds and reserved for future contingencies. The yearly final balance of the SSGF is determined mainly by the number and size of the wage contributions by enrollees (revenues), family density, and the level of the UPC (expenditures). The financial balance of the CR is closely related to macroeconomic and labor market conditions, as will be illustrated. 144 Good Practice in Health Financing Figure 7.3 Colombia: Flow of Funds in the Solidarity and Guarantees Fund total revenues = salary contributions sick leave 0.25% preventive payments care fund 0.41% maternity leave 0.25% solidarity payments 1.00% fund total for health payment = UPC* premiums total # of enrollees final (UPCs) balance (10.9) payment of claims for items not covered by benefits package Source: Author. The subsidized regime. Colombia allocated 1.1 percent of GDP to the SR in 2005, nearly US$1.39 billion. The SR subsidizes individual insurance premiums for the poor according to a proxy-means testing index known as Sistema de Identificación de Beneficiarios (SISBEN, the Beneficiaries Identification System). SISBEN scores are calculated on a number of dimensions of poverty, including labor market par- ticipation, income, educational attainment, family structure, assets, housing mate- rial and crowding, and access to water and sanitation. Data for calculating scores are obtained through a household survey designed for this purpose and adminis- tered at the municipal level. The total SISBEN score takes values between 0 and 100 and is divided into six categories. The people with the lowest scores, 1 and 2, are eligible for the SR, as long as they are not enrolled in the CR. According to house- hold survey data from 2003, about 28 percent of Colombia's population falls into SISBEN categories 1 and 2 (table 7.6). Because of resource constraints, not all persons eligible for the SR can be given health insurance. Priority enrollment is given to the people with the lowest scores and within vulnerable groups such as pregnant and lactating women, children under five (U5), the handicapped, and displaced populations, according to rules set at the national level (Acuerdo 244). By mid-2006, 18.3 million persons (39 per- cent of total population) were enrolled in the SR. The SR is financed from national government transfers (56.3 percent of total resources),3 a 1 percent solidarity contribution from the CR (34.4 percent of total resources), local tax revenues from "sin taxes" (8.8 percent of total resources), and contributions from family benefits funds4 (0.5 percent of total resources). Figure 7.4 depicts these sources of finance. Colombia: Good Practices in Expanding Health Care Coverage 145 Table 7.6 Colombia: Total Population, by SISBEN category, 2003 SISBEN category % of total population 1 7.2 2 21.1 3 32.1 4 22.5 5 15.9 6 1.3 Total 100.0 Source: Calculations by Panagiota Panopolou with data from DANE 2003. Figure 7.4 Colombia: Funding for the Subsidized Regime, by Source, 2005 Solidarity fund 34.4% National transfers for health (SGP) 56.3% Benefits funds 0.5% Lower tax revenues earmarked for health 8.8% Source: Data from Grupo de Economia de la Salud 2006. Note: Total resources in 2005 = Col$3.2 billion (US$1,390,000,000); 1 billion = 1x1012. The allocation formula for the national government transfers includes popula- tion size, risk of malaria and dengue, immunization rates, and geographic disper- sion of the population and local administrative capacity. The breakdown of total national government health transfers earmarked for health in 2005 was: 55.2 per- cent for health insurance coverage by the SR; 45.0 percent for the provision of health care services by public hospitals, or supply subsidies; and 11.9 percent for provision of collective public health activities (CONPES 2006). Funds flow in the following way. National transfers for the SR are allocated yearly to municipalities for payment of their current SR enrollees' premium and for programmed gradual expansion of insurance coverage to the eligible popula- tion.5 The allocation formula for the latter is defined by the National Department 146 Good Practice in Health Financing Table 7.7 Colombia: UPC Premium Value, 2006 Value per year Regime/age group Col$ US$ CR < 1 year 955,597 415 1­4 years 495,208 215 5­14 years 263,079 114 15­44 years, males 232,129 101 15­44 year, females 479,733 209 45­59 years 313,374 136 > 60 years 882,089 384 SR 215,712 94 Source: Acuerdo 322 de 2005, Consejo Nacional de Seguridad Social en Salud. of Planning, as a function of a municipality's total uninsured population weighted by the municipality's share of uninsured with respect to total uninsured popula- tion in the country. Total funds for health insurance subsidies are channeled to each municipality, where the local health authority directly contracts health plans6 serving the SR. Payment to insurer. The EPSs and the ARS receive a UPC with a fixed value for each enrollee. The value of the UPC is adjusted by age group only in the case of the CR; higher premiums are paid for young children, women of childbearing ages, and the elderly. Values of the UPC for 2006 are shown in table 7.7. The value of the premium is around 20 percent higher for dispersed geographical areas for both the CR and SR (not shown). As additional sources of revenue, health insurers can charge copayments for ambulatory and hospital services, set at the national level on a sliding, income- based scale. SR copayments apply only for hospital services. Ceilings are set for all copayments, and some services, such as preventive care, are excluded from cost- sharing. Colombia's health system has not developed mechanisms for monitoring and adjusting the premium value. Decisions on the premium have been guided by the annual increments in the minimum wage summed to considerations regarding the FOSYGA balance at the end of the fiscal year. After 13 years of the reform, no actuarial studies have been done of the real cost of providing the benefits pack- ages, mainly due to a lack of reliable information for these calculations. Acosta (2005) estimates that the UPC has lost its real value over time, about 20 percent for the SR and 10 percent for the CR. Poor information on the cost of the benefits package has serious implications: if the premiums are set above true costs, the sys- Colombia: Good Practices in Expanding Health Care Coverage 147 tem is wasting resources. If premiums are set below true costs, consumers will suf- fer, because health plans will have incentives to cut quality and access. Payment to providers. Health plans are free to establish payment mechanisms and payment levels for services they purchase from providers. As benchmarks for its fees, the NSHI has used the fee schedules developed by prereform public health plans, adjusted for inflation. In actuality, few adjustments have been made for the real cost of services. These schedules are used as ceilings for price negotiations between health plans and providers. Provider associations are aggressively seeking price regulation by the MPS (setting floors), and several proposals for fee sched- ules have been issued. Two patterns of payment are common to all health plans: preventive and pri- mary care services are contracted mainly by capitation, and most specialist care and hospital care is paid on a fee-for-service basis or by service packages. The effects of these payment mechanisms on access and quality have not been evaluated. Other insurance schemes. Although there were efforts to unify the insurance schemes existing in 1993, exceptions were made for the military and police forces, the education sector, and employees of the Colombian oil company (Empresa Colombiana de Petróleos, ECOPETROL), which remain autonomous in the orga- nization and provision of health benefits. About 5 percent of Colombia's popula- tion is covered by these independent schemes. In 2004 a subscheme known as Subsidios Parciales (Partial Subsidies, PS) was introduced in the SR. This subscheme provides temporary health insurance cover- age to the urban SISBEN 2 and 3 populations not covered by the SR because of lack of funds to provide full subsidies. Financing for the program comes from local resources of departments and municipalities that opt to implement the program. The national government supplements those funds. Wealthier municipalities can receive up to 40 percent of the total value of the local program and the less wealthy up to 50 percent. The insurance premium for this subscheme is about half the "full" SR premium, because it provides a small subset of services covered by the SR benefits package. By December 2005, 2.8 million people were covered by PS. Supply-side subsidies. To provide health care services for the uninsured or ser- vices not covered by the SR benefits package, there is a network of public pro- viders. These services are financed mainly by supply subsidies from national tax revenue transfers to municipalities earmarked for health and local tax revenues from "sin taxes." Supply-subsidy expenditures represented about 0.5 percent of GDP in 2005. To receive these funds, public providers establish service provision contracts with the local health authority. Public providers are allowed to charge user fees, priced nationally on a sliding scale according to income. The uninsured population needing such services can also pay private providers out of pocket. 148 Good Practice in Health Financing Table 7.8 Colombia: Summary of NSHI Schemes, December 2005 People Percent covered of total Scheme Target population (millions) population Financing source Contributory Formal sector 15.5 37 Employee and employer workers and people contributions with ability to pay Subsidized Poor population 16.5 40 National and local tax (SISBEN 1 and 2) revenues earmarked for health, solidarity contribution from CR, contribution from benefits funds Partial subsidies Urban SISBEN 2 and 2.1 5 Local tax revenues with 3 populations matching funds from government Supply subsidies Uninsured and 5.1 13 National and local tax insured requiring revenues earmarked for services not included health in benefits packages Special regimes Military and police 2.1 5 Combinations of employer / forces, education employee contributions sector, oil company workers Source: MPS 2006. The target population, coverage, and financing sources of the different schemes in Colombian Social Health Insurance are summarized in table 7.8. The Benefits Packages The benefits packages--the "compulsory health plans" under the CR and the SR (Planes Obligatorios de Salud, POS and POS-S, respectively) are detailed lists of the covered interventions, issued in 1994. Interventions are grouped into three levels of ascending complexity of specialization, technology, and financial resources required for their provision. Table 7.9 summarizes these interventions by cate- gories of medical care and levels of complexity. The first level of complexity includes preventive and emergency care, basic medical, dental, and diagnostic ser- vices. The second and third levels include specialized and rehabilitation care, hos- pitalizations and their corresponding diagnostic tests. For example, in the case of a person with diabetes, a general consultation would be provided at the first level, hospitalization for an infected diabetic foot at level 2, and diabetic coma at level 3. Catastrophe care is a subcategory for which interventions at all levels are included. Partial subsidies provide coverage only for prenatal and maternal services, services for children aged less than one, medications, and orthopedic and catastrophe care. Table 7.10 shows a comparison of coverage breadth, by population group, of the benefits package under each regime. The CRS benefits package, POS, is the Table 7.9 Colombia: Classification of Medical Care Types, by Complexity Coverage description and examples Type of care Level 1, Low complexity Level 2, Intermediate complexity Level 3, High complexity and specialized centers Preventive care Age and risk groups specific Not eligible Not eligible screening, visual and ear screening, immunizations Ambulatory care (consultations) General practitioner Includes family planning, nurse Not eligible Not eligible Colombia: practitioner Basic medical specialties Prenatal care by ObGyn Pediatrician, ObGyn, Internal medicine, Orthopedics Visual care Ophtalmology and optometry Not eligible Not eligible Good Other medical specialties and subspecialties Not eligible Neurologist, ORL, Cardiologist, Infectious disease Dental Fluoridation, prophylaxis, Not eligible Not eligible cavity occlusion Pract Emergency care Provided at all levels ices Hospital care Medical conditions Conditions managed by general Conditions requiring specialist or subspecialist care in practitioner Ex Obstetric Uncomplicated birth attendance Cesarean section High risk obstetric care panding Surgical conditions Skin sutures, abscess drainage General surgery (herniorraphy, Neurosurgical, cardiovascular surgery appendectomy, cholecystectomy, histerectomy) H Rehabilitation Not eligible Physical, respiratory, occupational and language therapy, ealth cardiovascular rehab, neurologic procedures Lab tests Basic tests (hemoglobin, urine Bacterial cultures, pathology Spirometry, immune function, hormones, toxicology, Car analysis, renal and liver function medication levels tests, glycemia, lipid profile, STDs), pap smear, biopsies eCover Radiology Bone and chest x-rays, obstetric Mammography, special x-ray projec- Arteriography, hemodynamic tests, MRI echography tions, endoscopies, ecographies age Catastrophic care Treatment with radiotherapy and chemotherapy for cancer, dialysis and organ transplant for renal failure; surgical treatment of heart, cerebrovascular, neurological and congenital conditions; treatment of major trauma, intensive 149 care unit, hip and knee replacement, major burns, treatment for AIDS Sources: Elaborated by author based on Acuerdos CNSSS 72, 74, and 83; Resolución 5261 of 1994. 150 Table 7.10 Colombia: Comparison of Breadth of CR, SR, and PS Benefits Packages Good TYPE OF BENEFIT Transportation Pract } Medications for referrals, ice Age/Population Level Level Level Catastrophic in National catastrophic Excluded in Group Preventive care 1 care 2 care 3 care care formulary care cases interventions H < 1 year Neonatal care and All All All All All All ealth screening (Vit K, anemia, } TSH), immunizations, well child care Financ 1­4 Well child care, All Not eligible All All All Aesthetic surgery immunizations, anemia ing Cataract and Infertility treatment screening strabismus Treatment for sleep disorders 5­19 years Well child care, All surgery, Not eligible All All All immunizations, anemia herniorraphy, Organ transplants screening appendectomy, (except renal, heart, cholecystectomy, chornea and bone 20­60 years Cardiovascular and renal All orthopedics, Not eligible All All All marrow) disease risk screening, rehabilitation cervical and breast services and Psychotherapy and cancer screening procedures psychoanalysis > 60 years Cardiovascular and renal All Not eligible All All All Treatments for end disease risk screening, stage disease cervical and breast cancer screening Pregnant High risk screening, STD, All Same as above Obstetric All All All women prenatal care plus obstetric care care Sources: Elaborated by author based on Acuerdos CNSSS 72, 74, and 83; Resolución 5261 of 1994. Colombia: Good Practices in Expanding Health Care Coverage 151 Table 7.11 Colombia: CR and SR Enrollment Rules, Copayments, and Choice, 2006 Contributory Subsidized Subsidized regime, Contents regime, POS regime, POS-S partial subsidy Enrollment rules Minimum weeks of Minimum period of 2 None enrollment required years enrollment before before full coverage of switching plans higher complexity care takes place; minimum period of 2 years' enrollment before switching plans No denial of coverage No denial of coverage if No denial of coverage if or preexclusions of eligibility criteria are eligibility criteria are conditions met or preexclusions of met or preexclusions of conditions conditions Risk adjustment of No No premium by age and gender Cost sharing No cost sharing for emergency care, preventive services, and services for conditions of public health interest (e.g., hypertension, diabetes). Sliding-scale copayments for ambulatory care and hospitalization up to yearly ceiling Choice of insurer Free choice within set Free choice within set Free choice within set of plans available in of SR plans operating in of SR plans operating in city where job is municipality municipality and located offering partial subsidy product Choice of provider Free choice within plan Choice restricted to Same as POS-S network, use of public network for low- gatekeepers complexity care, free choice within plan network; other care managed by gatekeepers Sources: Elaborated by author, based on Acuerdos CNSSS 72, 74, and 83; Resolucion 5261 of 1994. most comprehensive, covering interventions at every level and complexity, a wide national listing of medications, and medical transportation expenses. It has few explicit exclusions. The SR benefits plan, POS-S, is equal to the POS in terms of coverage of first-level interventions, catastrophe care, medications, and trans- portation but less comprehensive in terms of preventive, ambulatory care, and levels 2 and 3 interventions. The PS covers orthopedic and catastrophe care and the necessary medications to render these services. Coverage for children aged less than one year and for pregnant women is equivalent in all plans. Other services. Coverage of interventions that complement individual-level med- ical services included in the POS and POS-S is provided through different mecha- nisms (table 7.12). 152 Good Practice in Health Financing · A set of community-level public health interventions, such as health educa- tional and risk-prevention campaigns and control of vector-transmitted dis- eases are provided by local governments as part of a program known as the Plan de Atención Básica (PAB, Basic Care Plan). The Ministry of Social Protec- tion issues national priorities and guidelines for interventions to be included in PAB which are adopted and adapted locally. PAB is financed from transfers ear- marked for health from the national level to departments and municipalities. · Medical interventions for motor vehicle­related accidents are financed through compulsory insurance policies for all vehicles, known as Seguro Obligatorio de Accidentes de Transito (SOAT, Mandatory Insurance for Motor Vehicle Accidents). · Contributions by employers to insurance policies for work-related injuries, medical conditions, and disabilities are mandatory. Services are provided by specialized insurance carriers known as Administradoras de Riesgos Pro- fesionales (ARP). As of March 2006. 5.2 million employees were enrolled in ARPs (MPS 2006). · Private supplemental insurance policies can be purchased by individuals who want medical services not covered by the POS or more amenities (table 7.12). About 5 percent of the population purchases these policies. Appropriateness of coverage. With the exception of psychiatric disorders for which coverage is very limited, the benefits packages of both regimes (except PS) include minimum preventive and curative interventions for uncomplicated cases. This coverage should address the most common conditions in Colombia (table 7.13). Neurological conditions such as epilepsy are covered by POS and POS-S. However, as mentioned, there are no systematic mechanisms for revising the contents of the benefits packages. Few modifications have been made, and deci- sions to add medical interventions have frequently been a response to pressure by interest groups. Epidemiological information has not been updated systemati- cally. Thus, further action is needed to evaluate the appropriateness of the benefits package to address Colombia's changing health priorities. Table 7.12 Colombia: Coverage of Services to Supplement the Benefits Packages Type of health services Provision Individual POS, POS-S Public health, educational outreach PAB Motor vehicle accidents SOAT Work-related injuries and conditions ARP Additional medical services and amenities Private supplemental health insurance Source: Author. Colombia: Good Practices in Expanding Health Care Coverage 153 Table 7.13 Colombia: Coverage of Minimum Interventions to Address 2002 Burden of Disease Coverage of minimum interventions to control condition in uncomplicated cases Health condition by rank order (total DALYs) POS POS-S PS PAB SOAT Intentional injuries Neuropsychiatric conditions Infectious and parasitic diseases a Unintentional injuries Cardiovascular disease a Source: Author. a. Includes medications. Key Health Expenditure Indicators Selected health expenditure indicators for 2003 are shown in table 7.14. Financial Protection and Access Indicators There is evidence that NSHI has decreased financial barriers to access to health care (figure 7.5). According to a national household survey in 1992, the main rea- son for nonuse of health care in the lowest quintiles of the population was the cost of services. Responses to the same question in a subsequent household survey in 1997 showed a reduction in the percentage of people reporting nonuse of health care services for this reason, regardless of income level. There are large differences by insurance status in the percentage of the popula- tion that reports out-of-pocket payment as the source of payment for consulta- tions and hospitalizations (figure 7.6). Table 7.14 Colombia: Selected Health Expenditure Indicators, 2003 Indicator Value Absolute health expenditure (current US$) 6,250,797,175 Health expenditure, total (% of GDP) 7.8 Health expenditure per capita (current US$) 136 Health expenditure per capita (current international $) 514 Public health expenditure as % total health expenditure 39.6 Private health expenditure as % total health expenditure 15.9 Out of pocket health expenditure as % total health expenditure 7.5 CR expenditures as % of total expenditures in National Health Insurance 58.5 SR expenditures as % of total expenditures in National Health Insurance 14.5 Supply subsidy expenditures as % of total expenditures in National Health Insurance 26.9 Source: Baron 2006. 154 Good Practice in Health Financing Figure 7.5 Colombia: Main Reasons for Not Using Health Care Services, by Income Quintile, 1992 and 1997 a. 1992 b. 1997 100 100 90 90 80 80 70 70 60 60 50 50 respondents 40 respondents 40 of of % 30 % 30 20 20 10 10 0 0 1 2 3 4 5 1 2 3 4 5 income quintile (1 = lowest) too costly other no provider care denied Source: Escobar 2005. Note: "Other" includes perception. Figure 7.6 Colombia: Reported Sources of Payment for Consultations and Hospitalizations, by Insurance Status, 2000 a. Consultations uninsured insured 0 20 40 60 80 100 percentage b. Hospitalizations uninsured insured 0 20 40 60 80 100 percentage NSHI out of pocket private insurance other Source: calculations by Panagiota Panopolou, with data from DANE 2000. Colombia: Good Practices in Expanding Health Care Coverage 155 Table 7.15 Colombia: Households Experiencing Catastrophic Payments or Impoverishment Due to Ambulatory or Hospital Health Shock, by Insurance Status, 2003 Percent of households Insurance status Type of health shock and catastrophic health expenditure Contributory Subsidized Uninsured Total Ambulatory shock Impoverishment 0.4 0.9 0.7 0.6 Catastrophe (5%) 4.6 7.5 8.8 6.4 Catastrophe (10%) 2.1 3.9 4.6 3.3 Catastrophe (20%) 0.9 1.6 1.7 1.3 Catastrophe (30%) 0.3 0.8 0.9 0.7 Hospital shock Impoverishment 1.5 1.2 2.0 1.6 Catastrophe (5%) 8.2 11.6 12.1 10.0 Catastrophe (10%) 5.5 9.1 9.3 7.5 Catastrophe (20%) 3.5 5.8 6.1 4.9 Catastrophe (30%) 2.4 4.5 4.8 3.7 Source: Calculations by Rodrigo Muñoz, based on DANE 2003. Note: Poverty line based on definition by CEPAL 2002, adjusted for inflation. Overall, 0.6 percent and 1.6 percent of households were impoverished7 by a cat- astrophic ambulatory or a hospital health shock, respectively (table 7.15). Health expenditures greater than 30 percent of total consumption occurred in 0.7 percent of households experiencing an ambulatory shock and 3.7 percent of households experiencing a hospital shock. Overall, differences in rates of catastrophic expendi- tures and impoverishment by insurance status are observed for both types of shocks, and in every case the insured fare better than the uninsured. Insurance coverage has made a difference in access to care for the Colombian population, according to data from household surveys. Figure 7.7 presents a com- parison of treatment rates (the percentage of the population that had a health prob- lem and were seen by a doctor) between 2000 and 2003. Being insured seems to be positively related to being seen by a doctor. In that period in both urban and rural areas, not only are treatment rates higher for the insured than for the uninsured, but they also increased more. The rural-urban gap in the percentage of the insured population with a problem and seen by a doctor also seems to have decreased. Reports of use of preventive services are more frequent among the insured than the uninsured, although among the insured (figure 7.8), use of preventive services is less frequent among SR enrollees. The percentage of the insured popu- lation reporting use of preventive services also seems to have increased between 1997 and 2003, a trend not seen among the uninsured. These use patterns could be explained by the emphasis placed by the benefits packages on preventive ser- vices and also by differences in the breadth of coverage between the CR and SR. 156 Good Practice in Health Financing Figure 7.7 Colombia: Respondents with a Health Problem and Seen by a Doctor, by Insurance Status and Location, 2000 and 2003 a. 2000 b. 2003 70 80 60 70 60 50 50 40 respondents respondents of of 40 30 30 percent percent 20 20 10 10 0 0 urban rural urban rural Source: Author's calculations with data from DANE 1992 and 2003. Figure 7.8 Colombia: Population Reporting Use of Preventive Services, by Insurance Status, 1997 and 2003 100 90 80 70 60 50 respondents of 40 % 30 20 10 0 contributive subsidized uninsured total 1997 2003 Source: Ramírez et al. 2005. Colombia: Good Practices in Expanding Health Care Coverage 157 Figure 7.9 Colombia: Population Reporting Hospitalization in Last Year, by Insurance Status, 1997 and 2003 100 90 80 70 60 50 respondents 40 of % 30 20 10 0 contributive subsidized uninsured 1997 2003 Source: Ramírez et al. 2005. Hospitalization rates decreased for the entire insured population between 1997 and 2003 (figure 7.9). The decreases were highest for the CR insured and lowest for the uninsured. Although these differences could be due to increased access to hospital services through insurance, they might also be due to differences in pop- ulation health status. Overview of the Health Delivery System The 1993 reforms organized the CR and the SR health delivery systems following a two-market, managed-competition model that operates on two levels. On the first level, the insurance market, consumers have freedom to choose from among a set of public or private health insurance plans offering services covered by the POS/ POS-S in exchange for a fixed premium, the UPC described above. Because the pre- mium is fixed, theoretically health plans should compete for enrollees not through price, but through their distinctive service and quality features. Health plans act as group purchasers for their enrollees by arranging a network of providers they select based on the best price and quality. On the second level, the provider market, health care providers compete for inclusion in health plan provider networks and selec- tion by enrollees on the basis of price and quality. The Health Insurance Market Colombia's NSHI market is segmented into health plans serving the CR (EPS) and the SR (ARS). As of December 2005, 21 EPS plans operated in the CR, which are 158 Good Practice in Health Financing Figure 7.10 Colombia: Total Enrollees, by Type of Health Plan and Regime, 2005 a. contributory b. subsidized indigenous NP 6% NP private private 12% 16% public 18% public 14% FP private 70% community FP 36% private 28% Source: Enrollment data obtained from MPS 2005. private for-profit and nonprofit, and public. Of the total enrollees, 82 percent belonged to private plans (for-profit, 70 percent, nonprofit 12 percent); the rest (18 percent) belonged to public plans. The ARS insurance market has two other types of plans: community-based health plans (Empresas Solidarias de Salud, ESS) and health plans serving indige- nous populations. In December 2005, ARSs totaled 43--44 percent of them pri- vate (28 percent for-profit and 16 percent nonprofit), 14 percent public, 36 percent community-based, and the remainder (6 percent) for indigenous populations. Fig- ure 7.10 shows the distribution of health plans by percentage of total of enrollees in both regimes. Private health plans can choose their own provider network structure. Most use private providers for first- and second-level care and a mix of public and pri- vate for tertiary care. In the last five years, there has been a trend toward vertical integration (ownership of first-level providers by insurance plans). The main administrative strategies used by health plans to control demand are the use of gatekeepers and utilization management for specialist, hospital, and diagnostic care (Restrepo, Arango, and Casas 2001). The characteristics of public health plans differ from those of private plans. Under their original organizational struc- ture, financing and service provision were integrated, setting limitations on how much choice of provider they could offer and on their capacity to provide effi- ciency and quality incentives. Measures were taken to separate financing from provision in these organizations. Health plans in the SR are required by law to contract at least 40 percent of their network with public providers, but they are free to structure the rest of their Colombia: Good Practices in Expanding Health Care Coverage 159 network. This has led the ARSs to contract with public institutions for most pri- mary care and, for more complex levels of care, with public and private institu- tions They use payment methods and administrative strategies similar to those used by the EPS. Although there are several health plans in most municipalities, most municipal markets have a dominant health plan, with more than 73 percent of enrollees. Only in major cities does the market appear to be competitive (Restrepo, Arango, and Casas 2001). The Superintendencia Nacional de Salud (National Health Superintendency) is the government body in charge of health plan monitoring and oversight. Mini- mum quality, financial, and administrative standards for health plan operations have been defined by the Ministry of Social Protection, and health plans are required to demonstrate fulfilment of these requirements through a certification process begun by the superintendencia in 2005. The Health Care Provider Market At the time of the reform, Colombia had developed an extensive public network as a result of policies to improve access to health services implemented in the late 1980s. A supply of private providers had grown to meet demand for services not satisfied by the public sector. The 1993 reform integrated the supply of public and private providers into the national health insurance regimes, allowing health plans to include both public and private institutions in their provider networks. The supply of hospital services is organized by levels of care. First-level care comprises health posts, centers, and hospitals that offer general medicine. This level is expected to provide the most services needed. Second-level care includes providers of basic specialized medical and surgical services, both ambulatory and hospitalizations, and third-level care includes institutions that provide specialty and subspecialty care and high-complexity hospitalization. In 2000, about 43,200 physicians were practicing in Colombia, a ratio of 10.4 physicians per 10,000. The average ratio for the Latin America and Caribbean Region that year was 13 (Ascofame 2000). Of these physicians, 57 percent were general practitioners (5.9 general practitioners per 10,000); the rest were special- ists (4.4 per 10,000). General practitioners were concentrated in municipalities with more than 500,000 people, 23 times the number of physicians in municipal- ities with populations smaller than 20,000. With respect to other health care pro- fessionals, in 2003 there were about 24,800 nurses and 32,600 dentists, ratios of 0.56 and 0.80 per 10,000 residents, respectively (Cendex-Minsalud 2001). In 2004, the MPS provider registry reported a total of 13,804 health care facili- ties of different levels of complexity, 68 percent of them private (MPS 2005). There are 1,039 public institutions.8 More than half of all facilities, most of them private, are located in the Colombia's five major cities. Rural areas are served chiefly by public providers. Among public providers, 85 percent deliver primary care services, 13 percent deliver secondary care, and 2 percent deliver third-level 160 Good Practice in Health Financing Table 7.16 Colombia: Services Offered by Health Care Facilities, by Type, 2004 Type of service % of total Dental care 18.9 General medicine 17.4 Surgery 15.0 Physical therapy 11.2 Nursing 10.7 Internal medicine and subspecialties 7.8 Mental health 5.7 Emergency care 5.1 Other (pediatrics, immunization, anesthesia, intensive care unit) 8.2 Source: MPS 2005. care. Forty-three percent of hospital beds are devoted to primary care, 36 percent to secondary care, and 21 percent to tertiary care. With respect to the type of services offered by both private and public pro- viders, half are dental care, general medicine, and surgery. Types of services as per- cent of total services are shown in table 7.16. To guarantee that providers meet minimum quality, financial, and administra- tive standards, the MPS issued a set of certification requirements in 2002. To oper- ate, providers must register at the local health authority (Secretaría de Salud), which verifies that the requisites are met through an inspection visit and issues a three-year certification. In 2004, a hospital accreditation system was established. Accreditation is carried out by an independent organization, on a voluntary basis. So far five private and three public facilities have qualified. The 1993 Health Care Reforms Law 100, the legal foundation of the Colombian health care reform, was approved in 1993. The main goals of this reform were to improve access, efficiency, and ser- vice quality, and equity. The key strategy of the reform--improving access to health care--is NSHI, together with changes in the financing, organization, and administration of health service delivery. Beginning with a description of the Colombian health sector before reform and its problems, this section explains how the reform strategies to increase health insurance coverage were designed and implemented and assesses their achievements. The milestones in the reforms between 1993 and 2006 are presented in table 7.17. The Prereform Health System Prior to 1993, Colombia had a three-tiered health care system9. One tier consisted of the National Health System, designed after the primary care models in vogue in Colombia: Table 7.17 Colombia: Milestones toward Achievement of Universal Health Insurance Coverage Pre-reform period 1993 1994 1995­1997 1998­2001 2002 2003 2004 2005 2006 Good Decentrali- Law 100 Operation of Application of Further Quality New SISBEN Partial National- Universal zation enacted CR begins; SISBEN survey; development of assurance targetting subsidies subnational health legislation operation of operation of SR rules and legislation instrument program government insurance plan Pract (Laws 10 and National Health begins regulations; issued; issued; initiated; health SR cofinancing announced in 60); New Superinten- hospital subnational certification of plan strategies government ices Constitution dency and restructuring government providers certification launched; agenda in 1991; SISBEN National Social program resource begins; begins regional Ex targetting Security launched allocation Ministries of operation of instrument; ESS Council rules and Health and SR legislation panding development responsibilities Labor merged passed program reform (Law to become 715) Ministry of Social H ealth Protection Source: Author. Car eCover age 161 162 Good Practice in Health Financing the late 1960s and early 1970s. A network of public facilities for health care deliv- ery financed from national revenues offered services to about half of the popula- tion. A private health care provider market grew in parallel to the public network, targeting the population that could pay and meeting the demand for services and quality not provided by the public sector. Analyses of the 1992 Household Survey showed that public hospitals dominated the provision of inpatient services in rural areas; in urban areas private hospitals were used for about a third of admis- sions, regardless of insurance status. In both urban and rural areas, private providers supplied a large proportion of ambulatory care use (42 percent and 36 percent of all contacts, respectively), regardless of income group (Harvard School of Public Health 1996). A second tier was a mandated social insurance plan for workers in the public and formal sectors.10 Financed by employee and employer contributions, this plan covered about 50 percent of salaried workers and about 18 percent of their fami- lies, for a total of 20 percent of the population. A few public agencies11 held a monopoly for the provision of health insurance for this population group. The third tier consisted of private insurance or health care services paid out of pocket for the upper-income group, about 11 percent of the population. This market was covered by private commercial health insurers, prepaid group health organizations, or worker cooperative organizations. In the early 1980s, Colombia began to implement a nationwide fiscal, political, and institutional decentralization that sought to reassign functions and responsi- bilities between the national, departmental, and municipal governments. Although decentralization was not an explicit component of the 1993 health care reform, legislative mandates between 1990 and 1993 had introduced additional territorial functions and responsibilities and defined new sources of financing for health ser- vice provision and their respective allocation formula.12 Administrative procedures to certify local governments as "decentralized" were established, which if met, shifted authority, responsibility, and budgetary control over these resources to departmental and municipal governments.13 Within these laws, the legal basis for institutional decentralization of public hospital facilities was laid through pro- visions mandating separation of hospitals from administrative dependency on departments and conversion of these facilities to semipublic entities. Thus, they received the financial and managerial autonomy necessary to prepare for competi- tion with the private sector under the new health insurance scheme. Motivation for the Reforms In the pre-1993 Colombian health system, access to even basic health services was limited for a large part of the population, health spending was inequitable and inefficient, and quality was uneven. About 75 percent of the population had no health insurance, and one out of every six individuals in the first income quintile who fell ill in 1992 did not seek medical care because they could not afford to pay for it (Escobar 2005). About 19 percent of the population had no access to health Colombia: Good Practices in Expanding Health Care Coverage 163 services. According to national household surveys in 1992 health care expendi- tures were 2.4 percent of total household expenditures. The lowest income decile spent about 10 percent of their income on health, while the upper-income decile spent less than 0.5 percent. Cost was cited by more than 55 percent of the poorest people as a major access barrier to health care services. Public resources were poorly targeted, as evidenced by the fact that 40 percent of the subsidies to public hospitals benefited the wealthiest 50 percent of the pop- ulation, and allocation of health care resources was biased toward more expensive, curative care, despite evidence of low productivity. Allocation of resources fol- lowed historical hospital spending rather than people's needs. Large investments in hospital infrastructure with small correspondence to health service demand resulted in low occupancy rates and underutilization of services in secondary and tertiary hospitals. Finally, poor service quality was reflected in low use and accep- tance of the public provider network. Key Strategies for Broadening Health Insurance Coverage Law 100 established universal health insurance coverage for all Colombian citi- zens, to be provided by the CR for people who could pay and by the SR for the poor. Overall, the system sought solidarity between different population groups, from the rich to the poor, from the healthy to the sickly. The NSHI and coverage expansions, 1993­2005. Social health insurance was ini- tially designed as two separate schemes with target populations, financing sources, and benefits. This was to allow time for funds to become available to gradually expand insurance coverage and benefits for the population eligible for subsidies. Universal insurance and convergence of the two regimes in terms of benefits cov- erage was planned for 2001. First, universal coverage for the CR-eligible population (about 70 percent of total population) was to be reached by 2001 through an increased number of mandatory affiliations of individuals who could contribute. This would be possible by an annual increase in formal and informal sector workers (3.5 and 2.1 percent, respectively), favored by projected annual economic growth of 5 percent and a 1.8 percent increase in average income. CR enrollment would also increase by extend- ing coverage of health insurance to family members of contributing individuals. Second, achievement of universal coverage for the poor (30 percent of total population) was sought by adding new sources of financing and directing existing public resources toward demand subsidies for the SR. The eligible population would be identified using SISBEN, an instrument designed to target social pro- grams designed earlier. Application of SISBEN was initiated in 1994 in larger and more developed municipalities, under the responsibility of local majors and was extended to the rest of the country in 1997.14 SISBEN was increasingly criticized as a targeting instrument because of its potential for classification error and because of many operational problems, including evidence of corruption in the calculation 164 Good Practice in Health Financing Figure 7.11 Colombia: Expansion of CR and SR Insurance Coverage, 1992­2006 100 90 80 70 insured 60 50 40 population of 30 % 20 10 0 1992 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 total CR SR Sources: 1992­2004 (calculations with enrollment data provided by the MPS, percentage based on 1993 census projections); 2005 (MPS 2006, percentage based on 2005 census data). of scores to favor enrollment of ineligible individuals for political purposes (BDO and CCRP 2000; DNP and MS 2003; Fresneda 2003). This motivated changes in the design of the original instrument in 2002 and of the processes and rules to identify, select, and prioritize health subsidy beneficiaries. Implementation of these changes began in 2003.15 In the last 13 years Colombia's health care reform has expanded insurance cov- erage from 20 percent to 74 percent of the total population.16 Total growth of enrollment in each regime in 1992­2006 is shown in figure 7.11. Growth of enrollees of the CR has been largely a result of affiliation of con- tributing individuals' family members (figure 7.12). Between 1993 and 1997, about 5 million beneficiaries entered this regime, for a total of about 7.3 million. Since then, the total number of beneficiaries has remained about the same. The number of contributors increased from 5 million in 1993 to 7.5 million in 1997, fell in 2000, and slowly recovered to 7.2 million in 2005. The percentage of contributors belonging to the informal sector increased little, from 2.5 percent in 1993 to 4 per- cent in 2005 (figure 7.13). The growth of enrollment in the SR was amazing, from 0 percent to almost 40 percent of the total population in 13 years, 1993 to 2006. SR insurance coverage has reached the people most vulnerable to economic shocks. For example, a com- parison of the proportion of insured individuals by income quintile in 1992 and 2003 shows an increase of 37 percentage points or a variation of 444 percent for the first quintile (9 percent in 1992 to 48 percent in 2003), whereas in the fifth Colombia: Good Practices in Expanding Health Care Coverage 165 Figure 7.12 Colombia: Total CR enrollment, Contributors, and Beneficiaries, 1993­2005 14 12 10 people of 8 6 millions 4 2 0 1993 1997 2000 2003 2005 total contributors beneficiaries total enrollees Sources: For 1993, Ministerio de Salud 1994; for 1997­2005, author's calculations with data provided from FOSYGA. Figure 7.13 Colombia: Employment of SR Contributors, 1993­2006 1993 1997 2000 2003 2005 0 20 40 60 80 100 % of total contributors independent formal Sources: For 1993, Ministerio de Salud 1994; for 1997­2005, author's calculations with data provided from FOSYGA. quintile the increase was 21 percentage points (60 percent in 1992 to 81 percent in 2003), as seen in figure 7.14. Figure 7.15 illustrates how rural and urban disparities in insurance coverage have been progressively bridged. The top panel shows a rural-urban insurance coverage gap of 26 percent in 1993 that was narrowed to 13 points in 2003. The 166 Good Practice in Health Financing Figure 7.14 Colombia: Insurance Coverage, by Income Quintile, 1992­2003 90 80 82 75 74 70 66 68 60 58 58 61 quintile 53 53 50 45 48 by 49 41 40 44 40 insured, 30 31 % 20 21 10 9 0 1992 1997 2000 2003 quintiles 1 2 3 4 5 Sources: Based on MPS calculations DANE 1992 and 2003. bottom panel, covering the same period, shows that health insurance coverage in rural areas is now mainly at the expense of the SR. Changes in financing. These SR coverage trends were a response to the availabil- ity of resources as new sources were found and existing resources reallocated. The 1993 legislation opened new funding sources for each regime. Funds for the CR were obtained by mandating wage contributions for everyone who could pay and by increasing the of the contribution rate from 11 to 12 percent of salary. With the initially favorable macroeconomic and labor market assumptions, these resources would provide for CR expenditures and for SR solidarity funds. To finance the SR, national and local resources allocated to health were to be increased and directed toward demand subsidies. For example, between 1994 and 2000, the share of total national revenues for health transferred to local govern- ments would be raised from 8.7 to 10.5 percent, and supply subsidies to public providers would be gradually transformed into demand subsidies. New sources of funding for the SR were bundled into FOSYGA's solidarity subfund, which included: the solidarity point from the CR and contribution from Cajas, to be matched one to one by resources from the national budget; financial returns on FOSYGA surpluses; and expected revenues from taxes on oil exploitation. Together, national and local resources represented 40 percent of funding for the SR and solidarity resources 60 percent. Colombia: Good Practices in Expanding Health Care Coverage 167 Figure 7.15 Colombia: Growth of Insurance Coverage, by Regime and by Residence, 1993­2003 70 60 50 insured 40 30 population 20 of % 10 0 1993 1997 2000 2003 rural urban 60 50 insured 40 30 population 20 of % 10 0 1993 1997 2000 2003 urban CR rural CR urban SR urban SR Sources: 1993: Ministerio de Salud (1994); 1997­2003 based on calculations by Panagiota Panopolou with DANE 1997, 2000, and 2003. These changes had an impact on health spending and the CR financial balance. Total health expenditures as a share of GDP grew from 6.2 percent in 1993 to 7.8 percent in 2003, mainly as a result in increases in public expenditures (1.4 percent in 1993 to 3.1 percent in 2003). Social security expenditures grew from 1.6 percent in 1993 to 3.7 percent until 1999, when they began to decrease, to 3.0 percent in 2001. Total private expenditures remained around 3 percent from 1993 to 1997 and then decreased, to 1.2 percent in 2003. Direct out-of-pocket expenditures also decreased, from 2.7 percent to 0.6 percent over the period, most likely as a result of health insurance. Figure 7.16 summarizes these health spending distribu- tion and trends. 168 Good Practice in Health Financing Figure 7.16 Colombia: National Health Spending Distribution and Trends, 1993­2003 12 10 8 GDP 6 of % 4 2 0 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 total health expenditure total public expenditures social security expenditure private expenditures Source: Barón 2007. Note: According to NHA methodology, total expenditure includes direct public expenditures and public expenditures on insurance (subsidized regime), social insurance expenditures, which are essentially contributory regime expenditures, and private expenditures (including both voluntary private insurance and families´ out of pocket expenditures). The financial equilibrium of the CR is reflected in the balance of the FOSYGA account, obtained by subtracting all yearly payments to EPS17 from wage contri- bution revenues. Figure 7.17 shows the positive balance between 1996 and 1998, the deficits between 2000 and 2001, and the subsequent recovery (Restrepo 2006). In 1996 to 2005, macroeconomic conditions had a direct impact on FOSYGA revenues and expenditures. Figure 7.18 shows unemployment rates, GDP real growth, and population in poverty between 1991 and 2005. The period began with favorable real GDP growth rates around 5 percent and unemployment rates below 10 percent, which explain health reformers' optimistic projections. However, between 1996 and 2001 average real GDP growth was only 1.1 percent, hitting negative values in 1999, and unemployment soared to 20 percent. By 2002, the economy began to recover and, although unemployment rates have decreased, employment has not returned to prerecession levels. The percentage of the popu- lation in poverty rose from values below 50 percent to a peak of 57 percent between recession years and has also been recovering. Table 7.18 presents the values of key determinants of the FOSYGA balance for 1998 to 2002. Enrollment in the CR decreased, together with the ratio of contrib- utors to beneficiaries. Family density increased, and average salaries decreased. The behavior of these variables are related not only to low employment during Colombia: Good Practices in Expanding Health Care Coverage 169 Figure 7.17 Colombia: Balance of FOSYGA Compensation Fund, 1996­2005 700,000 600,000 average balance 500,000 400,000 300,000 millions 200,000 $Col. 100,000 2005 0 ­100,000 ­200,000 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Source: GES 2006. this period, but also to increased informality. All this resulted in an increase in payments to health plans that surpassed revenues from wage contributions, and in negative balances of the FOSYGA account, requiring use of the accumulated sur- plus. In subsequent years, the balance recovered, but its dependency on the behav- ior of factors not under control of the health sector shows the fragility of the balance. Evasion of premium contributions, particularly by independent workers has also had a negative impact on FOSYGA revenues. Health plans collect premium contributions directly from this population and have no incentives to promote accurate income reporting or income-verification mechanisms. Bitrán (2002) esti- mates that, in 2002, only 65 percent of potential contributors actually paid their obligations, and that the contributors who did pay contributed much less than they owed. Evasion and elusion together decreased CR revenues, possibly by as much as 35 percent. Without this problem in 2000, resources would have been available to enroll an additional 1.5 million people in the SR. As it was, the available funding amounted to only a third of the projected total needed to insure all eligible people in 2000 (figure 7.19). The budget constraints to reach universal health insurance have several expla- nations. First, unfavorable macroeconomic conditions after the reform reduced the availability of SR funding from other sources, such as solidarity contributions and national tax revenues (figure 7.20). The solidarity point reached only 70 per- cent of its expected level in 2000, not only because contributions to the CR fell short of projections, but also because, for political reasons and fiscal pressures, other projected sources of funding such as matching government funds and oil revenues were not allocated to the SR. 170 Good Practice in Health Financing Figure 7.18 Colombia: Unemployment, GDP Real Growth, and Percentage of Population in Poverty, 1991­2005 a. Unemployment and GDP growth 20 15 Unemployment 10 5 percentage 0 GDP growth ­5 ­10 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 b. Population in poverty 60 58 56 54 52 percentage 50 48 46 44 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Sources: Unemployment and growth, Lasso 2004 (data for years 1991­2000) and DANE (2000); population in poverty, calculations by DNP-SISD. In addition, the transformation of supply subsidies into demand subsidies has taken place more slowly than expected. Political opposition has come from public hospitals, and the persistence of uninsured groups and the design of the POS-S thrust the public network into the role of satisfying demand for a large part of second- and third-tier care, which has been used justify payment of supply sub- sidies to public hospitals (Sáenz 2001; Giedion, Lopez, and Marulanda 2000; Giedion, Morales, and Acosta 2000). Despite these constraints, the growth of resources for the SR has been substan- tial. Resources to finance SR premiums increased from 0.75 percent of GDP in 1995 to 1.1 percent in 2005. Growth of resources for the SR, halted during the eco- Colombia: Good Practices in Expanding Health Care Coverage 171 Table 7.18 Colombia: Selected Determinants of FOSYGA Balance, 1998­2002 Key determinant 1998 1999 2000 2001 2002 % total population enrolled in CR 39 33 32 30 29 Ratio of beneficiaries / contributors 1.27 1.46 1.47 1.49 1.51 Family density 2.27 2.46 2.47 2.49 2.51 Average salary (number or minimum wages) 2.18 2.09 2.07 2.10 2.00 Total revenues ($Col. millions) 3,314,565 3,841,614 4,096,909 4,539,282 4,741,443 Total expenditures ($Col. millions) 2,980,795 3,845,181 4,159,935 4,497,739 4,594,324 FOSYGA balance 333,770 ­3,567 ­63,026 41,543 147,119 Source: Acosta, Ramírez, and Cañon 2004. Notes: total revenues = contributions + financial returns, millions of pesos; total expenditures = premiums, maternity and sickness leaves, 1% solidarity, and prevention fund. Figure 7.19 Colombia: SR Projected Resources versus Actual Expenditures, 1994­2000 a. Projected resources b. Actual expenditures 3,500,000 3,500,000 3,000,000 3,000,000 2,500,000 2,500,000 2,000,000 2,000,000 millions millions Col$ 1,500,000 Col$ 1,500,000 2000 2000 1,000,000 1,000,000 500,000 500,000 0 0 1994 1997 2000 1994 1997 2000 solidarity national and municipal resources Source: Based on data in Martinez, Robayo, and Valencia 2002. Note: No data are available for actual expenditures in 1994. nomic crisis, have resumed growth. Sources of financing for the SR have also been changing, largely in response to the behavior of the solidarity account of the FOS- YGA. Also, in 2001, Law 715 was enacted, which sought to correct problems of pre- vious decentralization policies and stabilize national resources available to local 172 Good Practice in Health Financing Figure 7.20 Colombia: SR Projected Resources versus Actual Expenditures of Solidarity Funding Sources, 2000 450,000 400,000 350,000 300,000 millions 250,000 Col$ 200,000 2000 150,000 100,000 50,000 0 1% solidarity cajas matching financial contribution funds returns projected actual expenditures Source: Based on data in Martinez, Robayo, and Valencia 2002. governments. This law reset the amount of national fiscal resources for health and replaced the parameters used for their distribution with territorial indicators of equity and efficiency. It also redefined and aligned responsibilities with territorial capacity, although it allows decentralized municipalities to maintain competences and authority over resources and service provision as long as they meet prescribed performance indicators. Table 7.19 shows how the solidarity contribution, origi- nally the main source of financing, was gradually replaced with resources from the Table 7.19 Colombia: SR Funding Sources, 1995­2005 (percent of total resources) Source 1995 1996 1997 1998 1999 2000 2001 2002 2003 2005 FOSYGA 85.7 63.1 46.7 51.3 38.0 34.6 36.0 29.4 31.70 34.40 National transfers 14.3 36.9 44.0 37.2 50.0 51.9 51.0 67.9 62.0 56.3 Local resources -- -- 3.8 4.5 3.9 5.1 3.0 2.0 4.2 8.8 Sin taxes -- -- 1.8 2.2 3.4 2.1 9.0 -- -- -- Cajas -- -- 3.7 4.8 4.8 6.2 2.0 0.6 1.8 0.5 Source: GES 2006. Note: -- = not available. Colombia: Good Practices in Expanding Health Care Coverage 173 national budget, amounting to 56 percent of total funding sources in 2005, about US$783 million. In 2005, the MPS began to explore providing central government matching funds as an incentive for departmental and municipal governments to finance insurance premiums for the poor from their own resources. That year, 6 of the 32 departments, which include 250 municipalities and a total SISBEN 1 and 2 popu- lation of about 2 million, signed three-year commitments to provide resources to attain universal health insurance for SR-eligible people. Most of the participating local governments are the wealthiest ones or have just acquired new sources of revenues from exploitation of natural resources such as oil and coal. Due to the complexity of the system, the administrative costs of health plans can be large and variable, ranging from 4 percent to 60 percent of the value of the premium (Cendex 2000). More than 50 percent of the administrative costs are spent on support activities for daily operations (financial, personnel, and infor- mation management) and enrollment processes; little is spent on risk manage- ment and quality assurance (Cendex 2000). Changes in the organization and administration of health care delivery. The development of insurance markets, the integration of public and private provi- ders, and the preparation of public providers to compete are the main changes in the organization and administration of public health care delivery. Before 1993, public insurers, almost exclusively, provided employees with health insurance, mainly through the ISS, the largest insurer. Private insurance offered supplemental coverage or coverage for employees' family members. The reform ended the public health insurance monopoly by promoting the creation and entry of health plans (EPS and ARS) as organizations that would articulate the financing and delivery of services for CR and SR enrollees. The CR health insurance market developed rapidly, because there was already a supply of commercial health insurers, prepaid group health organizations, and benefits funds, with infrastructure, established provider networks, and experience managing health care services that were ready to enter the CR as EPS. Between 1995 and 1998, the market share of private health plans increased from 11 percent to 30 percent of the population in the CR, a net increase of 261.1 percent (Cés- pedes, Ramírez, and Reyes 1998). Competition by private health plans dramati- cally reduced the public sector's CR market shares in terms of CR population covered, from 100 percent in 1993 to 18 percent in 2005. Development of the ARS health insurance market was assisted by the imple- mentation of a government program to support the creation of community-based insurance organizations, ESSs, which began before the passage of Law 100. Non- governmental organizations (NGOs) were contracted to provide technical assis- tance to help local governments establish ESSs, and central government and local municipal funds were dedicated for this purpose. Between 1993 and 1994, 172 ESS contracts were established, covering 1.48 million people (Harvard School of Public 174 Good Practice in Health Financing Health 1996). ARSs were authorized to operate in 1995, and by December 1999, 239 ARSs had entered the market, with a total of 9 million enrollees, 22 percent of the total population (Cardona 1999). Participation was not restricted to ESSs, and other types of health plans also entered the market. In 1999, SR enrollment was distributed among public (10 percent), private for-profit (30 percent), ESS (40 per- cent), and private not-for-profit (20 percent) firms. ARS size ranged between a couple hundred to a million enrollees. However, there were too many ARSs to permit adequate risk pooling, their large transaction costs generated large inefficiencies, and oversight was difficult. Furthermore, some ARSs did not operate in some parts of the country. This situ- ation motivated the government to start implementing measures to consolidate the ARS market in 2001, first by requiring a minimum membership size and then by requiring regional operation of ARS by November 2005 (MPS 2005b; 2005c). Through a public competitive process, ARSs were assigned to five different regions of the country to operate during a four-year period. Regions were defined to per- mit a maximum of 15 ARS, each serving a population of 200,000, depending on the potential number of enrollees. At the end of the four years, the results will be evaluated in terms of efficiency, financial stability, and improved access. At the time of the reform, Colombia had a well-developed private supply of health services and a wide network of public providers set up during the 1970s as part of a national strategy to increase health service coverage. The NSHI model required the participation of both public and private providers. As part of the reforms, public institutions were to transform their management and budgeting structure to improve efficiency and service quality. For this pur- pose since 1999 the MPS has been implementing a program to guide and support public institutions' efforts to improve productivity and sustainability.18 So far 172 public institutions of all levels of complexity have voluntarily participated in the restructuring program. Among the observed program results, total expenditures and personnel expenditures have been reduced, and the number of services pro- vided and bed-occupancy rates have increased. Some institutions had been on the verge of closure, but they managed to recover and maintain their operations. However, many of these providers are still struggling with liquidity due to cash- flow problems and have not been able to pay off previously acquired pension debts to their employees (MPS 2005). In the managed competition model, the role of the government and its affiliated regulatory bodies is to provide information, to formulate, monitor, and enforce regulations to minimize market failures, and to devise mechanisms to guarantee equitable access to health care services. Thus, the 1993 reforms defined and assigned functions to the following governing institutions of the health sector: · Ministry of Health: formulation and monitoring of national policies, regulations · Consejo Nacional de Seguridad Social en Salud (CNSSS, National Health and Social Security Council): inspection and oversight of health plans and providers Colombia: Good Practices in Expanding Health Care Coverage 175 · FOSYGA: pooling and management of the system's revenues · Consejo Nacional de Seguridad Social en Salud: created and given regulatory and policy-making authority over benefits packages, premium payments, copayments and measures to avoid adverse selection, tariffs, SR operation rules, and management of the FOSYGA These institutions have had to face the typical problems of public, bureaucratic institutions, including lack of qualified personnel and frequent rotation of upper- level management. These inconveniences affected the speed with which they were able to focus on priority tasks and develop the skills to carry out their assigned functions. For instance, many of the foundations of a quality assurance system were outlined in the reform, and the regulatory basis was formulated in 2002, but only in 2004 did the provider certification process begin. Thus, evaluation of results is in the early stages. Lack of information systems has been a major handicap for all these institu- tions attempting to carry out their responsibilities. For example, it took FOSYGA almost 10 years and countless pesos to refine the procedures for detecting dupli- cate enrollees and multiple affiliations in its databases. There is still no reliable information for calculating the costs of providing the benefits packages to enable the CNSSS to make sound decisions about the affordability of premium adjust- ments and updates to the benefits packages. Though not a formal part of the 1993 reform, since the transformation of the former Ministry of Health into the Ministry of Social Protection in 2003, NSHI has acquired new significance as a policy for achieving goals beyond the scope of the health sector, such as poverty alleviation. Challenges and Lessons for Other Countries As a result of the 1993 health care reform, Colombia has achieved one of the most extensive health insurance systems in Latin America. This case study concludes by discussing major future challenges for the NSHI in Colombia, deriving lessons for other countries from the circumstances that contributed to the reform achieve- ments, and discussing replicability in other low- and middle-income countries. Future Challenges Multiple challenges still face Colombia in closing the remaining health insurance gap and achieving NSHI financial sustainability. First, with respect to the CR a major task is finding mechanisms to enroll the informal sector (58 percent of the working population in 2005) (DNP 2006). About a third of this population is known as the "sandwich population," which can be sorted into two groups: one includes people who earn less than two minimum wages, who are ineligible for sub- sidies, but for whom payment of the full contribution to the CR premium could be difficult; the other is people able to contribute to the premium but who lack incentives to become affiliated. According to a household survey in 2003, the total 176 Good Practice in Health Financing Figure 7.21 Colombia: Projections for Universal Health Insurance, 2005­2010 50,000,000 45,000,000 40,000,000 35,000,000 30,000,000 25,000,000 20,000,000 15,000,000 10,000,000 5,000,000 0 2005 2006 2007 2008 2009 2010 Uninsured 9,843,944 10,195,313 8,180,622 6,823,269 5,376,947 5,015,667 PS 2,142,715 2,136,436 2,763,723 3,209,770 3,674,393 3,835,849 SS 16,479,759 16,462,532 18,183,393 19,407,048 20,681,666 21,124,593 CS 15,533,582 15,906,388 16,288,141 16,679,057 17,079,354 17,489,259 Source: MPS calculations, 2005. sandwich population numbered about 7.3 million (workers and their families), 17 percent of the population (calculations by Panopolou 2004). Of these, 2.5 mil- lion had some capacity to pay (Bitrán, Giedion, and Muñoz 2004). The administra- tive costs of enrolling, monitoring, and collecting contributions from this population can be substantial, and the potential for adverse selection is large. Among the alternatives being explored by the MPS are providing partial subsidies to individual contributions to the premium. A pilot program will be launched in 2007 to evaluate the feasibility of this strategy. Another alternative is to devise tem- porary income protection mechanisms, within the framework of an integrated social protection system that the MPS seeks to develop. A second challenge for the CR is reduction of tax evasion. One of the anti- evasion strategies Colombia has started to implement is investment in improve- ments in the information system by establishing cross-links between pension system data and health insurance contribution data. Increased enrollment of informal workers could complement these measures. Third, in the case of the SR, not only will it be necessary to finalize the transfor- mation of supply subsidies, but it is also likely that alternative sources of financing will have to be found. For example, under current growth, employment, and poverty conditions and with the available resources, universal health insurance would not be reached by 2010 (figure 7.21). As mentioned, reaching universal health insurance by 2010 is in the govern- ment's agenda, but at the time of writing, stable sources of funding for this pur- pose have not been lined up. For example, although there has been some success Colombia: Good Practices in Expanding Health Care Coverage 177 in engaging local governments to use their own resources to cofinance demand subsidies, issues have been raised regarding the risk of local government defaults and the subsequent need for central government bailouts. Fourth, much still has to be done to design and activate of mechanisms for obtaining accurate information on enrollment, population eligibility, and cost and effectiveness of covered services. Evaluation of service quality is another pri- ority. The affordability and sustainability of the NSHI could be improved by reducing the inefficiencies generated by these information gaps. Finally, the equity and efficiency implications of maintaining two separate insurance schemes have to be evaluated and consideration given to feasible alter- natives toward convergence. For example, due to the gaps in SR service coverage, the poor are likely to experience discontinuities in care that may reduce the effec- tiveness of treatment.19 Also, since this population has access to health services both from health plans (covered services) and from local governments and public hospitals, the monitoring of access, quality, and efficiency has become especially cumbersome. Lessons from the Colombian Reform: Enabling Factors Prescience, in the form of legal and institutional groundwork, good economic conditions, and political support combined with a generous measure of luck to make a success of the 1993 reform in Colombia. Prior legal framework. The prereform decentralization laws, plus the formulation of the 1991 Constitution provided a legal framework for many of the strategies of the 1993 reform. Decentralization laws did the job of mobilizing national resources for health, which were then to be reallocated by Law 100 for the purposes of the reform. Decentralization increased local government autonomy to manage resources and set the stage for public hospitals to respond to market mechanisms, conditions that were necessary for insurance contracts. The 1991 Constitution gave political legitimacy for the provision of health services under a National Social Health Insurance (NSHI) scheme with the participation of the private sector. Institutional arrangements. The NSHI was able to quickly sign up a large number of enrollees throughout the country because a reasonable number of private and public institutions already could meet the organizational requirements for deliv- ering services under the managed competition model or could adapt to the fulfill these conditions. In addition, a program for the development of community- based health insurers in the rural areas had been implemented. Many of these net- works already had a public-private mix of providers, which facilitated the NSHI response to demand for health care services. The availability of the SISBEN pro- vided the reform with an instrument for identifying its target population and pro- viding subsidies for the neediest. 178 Good Practice in Health Financing Financial aspects. Positive economic growth conditions prior to and during the first four years of the reform smoothed acceptance of increases in contributions to the NSHI and mobilization of large sums of general revenues to finance the insur- ance expansions. The enrollment of a large part of the population through cover- age of families of contributing individuals allowed the public sector to free additional resources to provide the poor with health care. Political support. Anecdotal evidence suggests the potential to use the SR to gain electoral support played a role in local governments' uptake of the SR. Expansion of the NSHI has been a hot topic on the presidential agenda in the last two elec- tions. There is also evidence from household surveys that people see the value of being enrolled in the SR, very likely a result of the expected learning curve about eligibility and benefits after 13 years of experience with it (Fresneda 2002). Replicability in Other Countries The probability is low that Colombia's coverage expansion experience can be replicated in poor countries. The main obstacle would be that a large portion of the population would have to be subsidized and large sums of national resources would have to be mobilized, but these countries have small tax bases. This group of countries may have supply constraints that would have to be addressed, and the breadth of the SR benefits package would have to be reduced. The administrative costs of running the system might be better invested in improving the quality of public services. Middle-income countries considering national social health insurance should evaluate whether they meet key conditions to replicate the Colombian experience successfully and avoid some of its problems. The key conditions that assisted Colombia include · Favorable economic growth and a reasonable volume of general revenues and potential enrollee contributions to finance the desired benefits · Existence of quality assurance regulations guaranteeing minimum safety and quality standards · The existence of, or the possibility of setting up in a parallel with the reform, accurate and reliable information systems with population, financial, account- ing, clinical, and service-use data · Well-developed insurance and provider markets · Institutional capacity for oversight and enforcement Endnotes 1. Among informal sector workers are salespeople, small business owners, taxi drivers, and agricultural and construction workers. 2. Health plans are health insurance companies that articulate financing and delivery of services, modeled after health maintenance organizations in the United States. Colombia: Good Practices in Expanding Health Care Coverage 179 3. These resources are part of the social investment budget known as the Sistema Gen- eral de Participaciones (SGP). 4. Cajas de Compensacion Familiar [family benefits funds] are companies that man- age employer-provided benefits in addition to health care, such as recreational services. 5. Allocation rules for resources of the SR are established in Law 715, 2001; Law 812, 2003; Decree 159, 2002; and Decree 177, 2004. 6. These plans are called Administradoras del Régimen Subsidiado (ARSs, Subsidized Regime Administrating Companies). 7. Impoverishment is defined as falling below Colombia's poverty line. 8. The provider registry counts the number of services reported, so single facilities may appear more than once, inflating the number of facilities. 9. Most of the information in this subsection is derived from Ministerio de Salud (1994), unless specified otherwise. 10. At the time of the reform 46 percent of the working population belonged to the for- mal sector and the rest to the informal sector. 11. The largest was the Instituto de Seguros Sociales (Social Security Institute). 12. Ley 10 of 1990; Constitución Política de 1991; Ley 60 of 1993 (Competencias y recursos) y decreto reglamentario 1757; Ley 100 of 1993. 13. Ley 10 of 1990, Art. 37; Ley 60 of 1993 Arts. 14 and 16; Decreto 1770 of 1994. 14. For targeting purposes, municipalities where SISBEN was not available were allowed to use census listings of households with unsatisfied basic needs (another index of poverty used in Colombia) or socioeconomic categories used for definition of public ser- vice tariffs. 15. Acuerdo 244 of 2003. 16. This total percentage does not take into account that 5 percent of the population receives partial subsidies, which are a temporary measure, and another 5 percent is enrolled in exception regimes. This implies that 87 percent of the total population is covered by some type of insurance. A change in the denominator is important to mention between 2004 and 2005: the total population in 2004 was based on the 1993 census projection of 45 million; in 2005, on the 2005 census of 41.2 million. 17. Premium payments, maternity and sickness leave and payments to cover expenses of medications and procedures not covered by the POS. 18. This is known as "Programa de modernización de hospitales." 19. For example, a woman with a positive PAP smear for cervical cancer--covered by the POS-S--has to pay out of pocket to have the diagnosis confirmed by a colposcopy, which is not covered. References Acosta, O. L., M. Ramírez, and C. I. Cañon. 2004. "Principales estudios sobre sostenibilidad financiera del SGSSS." Working paper, Corona Foundation [Fundación Corona], Bogotá. Ascofame (Asociación Colombiana de Facultades de Medicina). 2000. Recurso Humano en Medicina: formación, distribución y bases para una propuesta política. Bogotá: Ascofame. Barón, G. 2007. "Cuentas de salud de Colombia 1993­2003. El gasto nacional en salud y su financiamiento." Programa de Apoyo a la Reforma de Salud del Ministerio de la Protec- ción Social, Bogotá. 180 Good Practice in Health Financing Bitrán and Associates, Econometría and Superior School of Public Administration [Bitrán & Asociados-Econometría and Escuela Superior de Administración Pública]. 2002. Propuestas de reestructuración de los procesos, estrategias y organismos públicos y privados encargados de la afiliación, pago y recaudo de aportes al SGSSS. Health System Reform Support Program of the Ministry of Social Protection [Programa de Apoyo a la Reforma de Salud del Ministerio de la Protección Social.] Project 03­99, Bogotá. Bitrán, R., U. Giedion, and R. Muñoz. 2004. Risk pooling, ahorro y prevención: Estudio regional de Políticas para la protección de los más pobres de los efectos de los shocks de salud. Washington, DC: World Bank. BDO and CCRP (Corporación Centro Regional de Población). 2000. "Evaluación del Sis- ben. Eficiencia, Eficacia Institucional de los proceso de clasificación y selección de ben- eficiarios." Informe final, Ministerio de Salud, Bogotá. Cárdenas, M. S. 2006. Introducción a la Economía Colombiana. Bogotá: Alfaomega- Fedesarrollo. Cardona, A. 1999."Participación de las entidades promotoras de salud (EPS) en el mercado de aseguramiento de salud." Revista Facultad Nacional de Salud Pública 17 (1): 52­62. Cendex-Ministerio de Salud. 2001. Los Recursos Humanos de Salud en Colombia: balance, competencias y prospectiva. Bogotá: Editorial Javegraf. Céspedes, J. E., M. Ramírez, and A. Reyes. 1998. "Análisis de la encuesta de calidad de vida del DANE." Santafé de Bogotá: Econometría S.A. CONPES (Consejo Nacional de Política Económica y Social). 2006. "Distribución del Sis- tema Nacional de Participaciones Vigencia 2006." Documento 97, Departamento Nacional de Planeación. Bogotá. DANE (Departamento Nacional de Estadística). 1992. Encuesta de Hogares, 1992. ------. 1997. Encuesta Nacional de Calidad de Vida, 1997. Bogotá: DANE. ------. 2000. Encuesta de Hogares 2000, modulo de salud. Bogotá: DANE. ------. 2003. Encuesta Nacional de Calidad de Vida, 2003. Bogotá: DANE. ------. 2002. Vital Statistics Registry for 2002. Bogotá: DANE. http://www.dane.gov.co. ------. 2005. Census Results 2005. Bogotá: DANE. http://www.dane.gov.co. ------. Labor Statistics 1990­2005. Bogotá: DANE. http://www.dane.gov.co. DNP (Departamento Nacional de Planeación). 2006. Bases del Plan Nacional de Desarrollo 2006­2010. Bogotá: DNP. http://www.dnp.gov.co/paginas_detalle.aspx?idp=890. DNP and MS (Ministerio de Salud). 2003. "Quien se beneficia del SISBEN. Evaluacion Integral." Government document. Bogotá. Escobar, M.L. 2005. "Health Sector Reform in Colombia." Development Outreach, May 2005. World Bank, Washington, DC. http://www.worldbank.org/devoutreach/may05/ article.asp?id= 295 - 28k Fresneda, Óscar. 2003."Focusing on the Subsidized Health Regime in Colombia." Rev. salud pública 5(3):209­45. ------. 2002. "La visión de la población sobre el sistema de salud. Ha mejorado el acceso en salud?" In Evaluacion de los procesos del regimen subsidiado, ed. D. Arevalo and F. Martinez. Bogotá: Universidad Nacional de Colombia. Colombia: Good Practices in Expanding Health Care Coverage 181 González-Rossetti, A., and T. J. Bossert. 2000. "Enhancing the Political Feasibility of Health Reform: A Comparative Analysis of Chile, Colombia, and Mexico." Contract No. HRN- 5974-C-00-5024-00. Harvard School of Public Health, Latin America and Carib- bean Regional Health Sector Reform Initiative. http://www.hsph.harvard.edu/ihsg/ publications/pdf/lac/PolicyProcessSynEng2.PDF. GES (Grupo de Economia de la Salud). 2006. "Resultados financieros del seguro público de salud en Colombia 1996­2005." Observatorio de la seguridad social 13. Medellín, Colombia: Universidad de Antioquia, Facultad de Ciencias Económicas. Giedion, U., H. R. Lopez, and J. A. Marulanda. 2000. "Desarrollo institucional del sector salud en Bogotá." Coyuntura Social 23: 57­81. Giedion, U., L. Morales, and O. L. Acosta. 2000. "Efectos de la reforma en salud sobre las conductas irregulares en los hospitales públicos." Coyuntura Social 23: 98­126. Harvard School of Public Health. 1996. "Colombia Health Sector Reform Project." Final Report, Boston, MA. Lasso, F. 2004. "Incidencia del gasto público social sobre la distribución del ingreso y la reducción de la pobreza." Working paper, Mission for the Design of a Strategy for Poverty and Inequality Reduction, National Department of Plannning [Misión para el diseño de una estrategia para la reducción de la pobreza y la desigualdad, Departamento Nacional de Planeación], Bogotá. IMF (International Monetary Fund). 2006. "Article IV Consultation, Poverty Reduction and Growth Facility Reports." IMF, Washington DC. Martínez, F., G. Robayo, and O. Valencia. 2002. "Porqué no se logra la cobertura universal de la seguridad social en salud?" Bogotá, Fundación para la Investigación y Desarrollo de la Salud y la Seguridad Social (FEDESALUD). MPS (Ministerio de la Protección Social). 2006. Informe de Actividades al CNSSS 2005­2006. MPS, Bogotá, Colombia. ------. 2005. Politica Nacional de Prestacion de Servicios de Salud. MPS, Bogotá. MERPD (Misión para el Diseño de una Estrategia para la Reducción de la Pobreza y la Desigualdad). 2006. "Metodología de Medición y Magnitud de la Pobreza en Colom- bia." Departamento Nacional de Planeación, Bogotá. ORC Macro. 2006. MEASURE DHS STATcompiler. http://www.measuredhs.com. Pinto, M. E., A. Vergara, and Y. Lahuerta. 2005. "Cuanto ha perdido Colombia por el Con- flicto?" Estudios Económicos, Documento 277, Departamento Nacional de Planeación, Bogotá. Profamilia. 2005. "Salud Sexual y Reproductiva en Colombia. Informe de resultados de la Encuesta Nacional de Demografía y Salud." Profamilia, Bogotá. Ramírez, M., A. Zambrano, F. J. Yepes, J. Guerra, and D. Rivera D. 2005."Una Aproximación a la Salud en Colombia a Partir de las Encuestas de Calidad de Vid." Borradores de Investigación No. 72. Bogotá: Universidad del Rosario. Restrepo, J. H. 2002."El seguro de salud en Colombia. Cobertura Universal?" Revista Geren- cia y Políticas de Salud 2: 25­40. Restrepo, J. H., M. Arango, and L. Casas. 2001. "Estructura y conducta de la oferta del seguro de salud en Colombia." Observatorio de la Seguridad Social 1: 1­10. 182 Good Practice in Health Financing Sáenz, L. 2001. Modernización de la Gestión Hospitalaria Colombiana: Lecciones aprendidas de la transformación de los hospitales en Empresas Sociales del Estado. Regional Initiative of Health Sector Reform in Latin America and the Caribbean, Document No. 46. http://www.lachsr.org/documents/187-modernizaciondelagestionhospitalaria colombianaleccionesaprendidasdelatransformac-ES.pdf. WHO (World Health Organization). 2002. "Global Burden of Disease (GBD) Estimates, 2002." WHO, Geneva. http://www.who.int/healthinfo/bodestimates/en/index.html. World Bank. 2006. World Development Indicators. Washington, DC: World Bank. 8 Costa Rica: "Good Practice" in Expanding Health Care Coverage--Lessons from Reforms in Low- and Middle-Income Countries James Cercone and José Pacheco Jiménez Background Costa Rica is a small (51,100 sq. km) republic in Central America in the Latin Amer- ica and Caribbean Region (LAC) of the World Bank with a population of 4.3 million and per capita income of US$9,985. Total health expenditures equaled $305 per capita in 2005. In 2004, life expectancy was 78.7 years, and infant mortality was 11.3 per 1,000 live births. Costa Rica is an outstanding example of a middle-income coun- try that provides its people, including the poor, with broad and deep health care cov- erage, considerable financial protection, and an extensive package of services. These good outcomes are the result of a century of efforts to improve living conditions. In 1941, with the creation of the Caja Costarricense del Seguro Social (CCSS), a health insurance system began operations, one of the first such systems in Latin America. By 2006, 88 percent of the population was affiliated with the health insur- ance scheme, and around 93 percent of the people had adequate access to primary care services. Before the 1994 health care reform, access to primary care services was restricted to 25 percent of the population. To surmount obstacles to achieving its coverage objectives, Costa Rica revamped its organizational structure by redefining the network and bringing private and nongovernmental partners into the delivery system. These reforms were aimed at correcting the structural problems and further separating financing, purchasing, and provision within the vertically integrated CCSS model. 183 184 Good Practice in Health Financing Introduction This section provides a brief overview of Costa Rica in three main areas: econom- ics, demographics and epidemiology, and the political environment. The main trends of the last 15 years are described as background information for issues addressed later in the chapter. Economic Profile Between 1990 and 2005, the Costa Rican economy was characterized by a stable macroeconomic context with persistent problems that weakened over time, a stag- nant poverty rate after the mid-1990s with increasing inequality after 1998, and decreasing dependence on external funds and declining external debt. Costa Rica has been classified as an upper-middle-income economy, with a GDP per capita of US$4,327 in 2004 (US$9,481 in PPP terms). Between 1990 and 2004, the GDP per capita grew at a 4.7 percent rate a year, but growth between 2000 and 2004 deceler- ated to an average rate of 3.3 percent. Between 1991 and 2005, real disposable income per capita grew 2 percent a year, but between 2000 and 2005, the annual rate fell to just 1 percent. Although the unemployment rate has risen during the last decade, it is still low, compared with regionally common rates close to 10 percent. The part of the population that is poor declined from 32 percent in 1990 to 21.7 percent in 2004. Despite substantial progress on this issue, in the last 10 years poverty has been stuck at around 20 percent of the population. Fiscal problems are constant in the Costa Rican economy. The gap between revenues and expenditures has been constant, with the fiscal deficit averaging 3 percent of GDP (table 8.1). The ratio of tax revenues to GDP has remained at between 12 percent and 13.5 percent. The ratio of expenditures to GDP has stayed in a 14 percent to16 percent range. Recurrent deficits are often blamed for causing higher inflation rates than international averages. For almost two decades, infla- Table 8.1 Costa Rica: Economic Indicators, 1990­2004 Indicator 1990 1995 2000 2004 GDP per capita PPP (current international US$) 5,219.6 6,630.2 8,621.3 9,481.4 GPD growth rate (%) 3.6 3.9 1.8 4.2 Headcount ratio 31.9 20.4 20.6 21.7 Revenue to GDP ratio (excluding grants) 23.1 20.3 20.9 22.4 Tax revenue to GDP ratio 13.1 12.1 12.1 13.4 Expenditures to GDP ratio 14.2 16.1 15 15.9 External debt to GDP ratio n.a. 36.6 19.9 21.1 ODA as % of GDP 4.2 0.3 0.08 0.07 Unemployment (% of total labor force) 4.5 5.1 5.2 6.5 Inflation, consumer prices (annual %) 19 23.2 11 12.3 Source: World Development Indicators 2006. Revenue-to-GDP ratio comes from MIDEPLAN. Costa Rica: Good Practices in Expanding Health Care Coverage 185 tion in Costa Rica has been higher than 10 percent; in 2005 it was the second high- est rate in Latin America. External debt has shown steady signs of decline. By 2005, the ratio of external debt to GDP was less than half the ratio in 1990, not a significant threat to macro- economic stability. Similarly, official development assistance (ODA) decreased in the overall context, to just 0.07 percent of GDP in 2004. Demographic and Health Profiles The Costa Rican population is aging, a trend expected to continue for the next 15 years (figure 8.1). The number of children between 0 to 4 years old is expected to decrease from 13.4 percent of the population in 1990 to 7.2 percent in 2020. Similarly, the 0 to 14-year-old group is expected to shrink from 36 percent to Figure 8.1 Costa Rica: Population Pyramids, 1990, 2005, and 2020 a. 1990 b. 2005 90­94 90­94 male female male female 75­79 75­79 60­64 60­64 age 45­49 age 45­49 30­34 30­34 15­19 15­19 0­4 0­4 15 10 5 0 5 10 15 15 10 5 0 5 10 15 % of population % of population c. 2020 90­94 male female 75­79 60­64 age 45­49 30­34 15­19 0­4 15 10 5 0 5 10 15 % of population Source: United Nations Population Division 2004. 186 Good Practice in Health Financing Table 8.2 Costa Rica: Population Growth Rates and Projections, 1990­2020 Period Annual average ( %) 1990­95 2.4 1995­2000 2.4 2000­05 2.0 2005­10 1.5 2010­15 1.3 2015­20 1.1 2020­25 1.01 Source: United Nations Population Division 2004. 22 percent of the total population. Thus, older groups will make up an ever- increasing share of the population. For instance, while in 1990, women 60 years of age and older made up 7.2 percent of the female population, in 2020, women of this age will make up 14.2 percent of the female population. In other words, the number of older people will practically double between 1990 and 2020. By resi- dence, the urban population accounts for 61 percent of the total population. Accelerating the demographic transition, which will double the old age popu- lation by 2020, is the decline in fertility: total fertility rates declined from 3.5 in 1980-85 to 2.5 during the second half of the present decade. These two trends will decrease the population growth rate and produce dramatic changes in the child and old-age dependency ratios. As shown in table 8.2, the population growth rate in the early 1990s was 2.4 percent but then began declining at the turn of the century. By 2020, the growth rate is expected to slow to 1 percent per year. The child dependency ratio will decline from 61 percent in 1990 to 31 in 2025 (figure 8.2). At the same time, the old-age dependency ratio will increase from 8 to 16 percent. Overall, life expectancy and infant mortality rates have improved since 1990 (table 8.3). Although infant mortality has decreased, the maternal mortality rate has nearly doubled, partially explained by improvements in maternal death recording and registration. Almost 98 percent of all births were attended by skilled staff. Immunization rates showed a U-shaped pattern, falling from 95 percent cov- erage of children aged 12 to 23 months to 85 percent in 1995, and then rising to 90 percent in 2004. The mortality and morbidity profile of Costa Rica shows the characteristic pat- tern of a developed economy (table 8.4). Half of the deaths in 2002 were due to cir- culatory diseases and cancers; noncommunicable diseases represented 76 percent of all deaths. In terms of disability-adjusted life years (DALYs), the top five causes accounted for 60 percent of all DALYs, led by neuropsychiatric conditions and unintentional injuries (38 percent of total DALYs). By disease category, around 70 Costa Rica: Good Practices in Expanding Health Care Coverage 187 Figure 8.2 Costa Rica: Basic Demographic Indicators 70 18 16 60 old 14 50 age ratio 12 dependency 40 10 30 8 dependency 6 ratio 20 child 4 10 2 0 0 1990 1995 2000 2005 2010 2015 2020 2025 Child dependency ratio Old-age dependency ratio Source: United Nations Population Division 2004. Table 8.3 Costa Rica: Basic Health Information Infant mortality Maternal mortality Immunization, Life expectancy (per 1,000 (per 10,000 (% of children Year at birth live births) live births) ages 12­23 months) 1990 76.8 16.0 1.5 95.0 1995 76.7 14.3 2.0 85.0 2000 77.7 12.5 3.6 88.0 2004 78.7 11.3 3.0 90.0 Sources: World Development Indicators 2006; for maternal mortality, MIDEPLAN (http://www.mideplan.go.cr/sides/social/ index.html). Table 8.4 Costa Rica: Leading Causes of Morbidity and Mortality, 2002 (percent of total) Top five burden of disease causes (DALYs) Top five causes of mortality Cause Percent Cause Percent Neuropsychiatric conditions 28 Cardiovascular diseases 31 Unintentional injuries 10 Malignant neoplasms 21 Infectious and parasitic diseases 9 Unintentional injuries 8 Cardiovascular diseases 7 Infectious and parasitic diseases 7 Malignant neoplasms 7 Digestive diseases 7 Source: World Health Organization 2004. 188 Good Practice in Health Financing percent of the burden of disease was caused by noncommunicable diseases. Com- municable and materno-infant diseases represented 14.5 percent of total episodes; injuries, one in seven cases of illness. Political and Administrative Profile Costa Rica is administratively organized into seven provinces, 81 cantons, and 459 districts. The government is democratic and structured in three branches: execu- tive, legislative, and judicial. The presidential term lasts four years, and reelection has been permitted since 2003. Costa Rica has a long tradition of political stabil- ity, and since 1889, elections have taken place normally except for two instances: the 1917 to 1919 tyranny and the 1948 to 1949 transitional period when a new Constitution emerged and the army was abolished. The last two elections (2002 and 2006) broke the hegemony of the two main political forces, and new actors emerged with solid voter support. The political, administrative, and fiscal systems are centralized, and resources are allocated centrally, usually following historical patterns. The excessive number of regulations makes for a rigid budget and leaves the Ministry of Finance (MOF) little freedom to reallocate resources according to needs or to control the fiscal deficit. High centralization is evident from the low participation of the 81 munic- ipalities in the total budget: just 2 percent of the funds are municipally managed. The National Development Plan (NDP) elaborated by each new government is the main political, programmatic, and budgetary tool. The Ministry of Planning is in charge of monitoring plan implementation through the National Evalua- tion System. Social policies are also part of the NDP. Besides the three branches, the institutional framework is complemented by three key entities: the Controller General Agency, the Constitutional Court (Sala IV), and the Ombudsman's Office. The main function of the Controller General, the public auditing institution, is to evaluate administrative contracting processes (like purchasing). The Constitu- tional Court is empowered to revoke any unconstitutional decision by the Con- gress. The Ombudsman supervises the quality of public services and protects citizens against any abuse of power. Health Financing and Coverage This section analyzes the main financing and coverage characteristics of the Costa Rican health system--health expenditure trends and distribution in the last decade. A brief analysis of the benefits package and payment mechanisms is also included, as well as an evaluation of the equity issues. Health Expenditures Total health spending in Costa Rica amounted to US$1.42 billion in 2005, 65 per- cent higher than in 1998 (table 8.5). In 2005, 7.1 percent of GDP was allocated to finance health services, a proportion that has grown over the years. Private health Costa Rica: Good Practices in Expanding Health Care Coverage 189 Table 8.5 Costa Rica: Health Expenditure Indicators 1998, 2000, and 2003 Sector 1998 2000 2005 Total health expenditures (US$ millions) 859.85 1,004.62 1,415.00 Health expenditures (% of GDP) 6.1 6.3 7.1 Private health expenditures (% of GDP) 1.3 1.3 1.7 Public health expenditures (% of GDP) 4.8 5.0 5.4 Per capita health expenditures (current US$) 230 258 327 Public health expenditures (% of total health expenditure) 77.9 79 76 Public health expenditure (% of total government expenditure) 21 21.7 21 Private health expenditures (% of total health expenditure) 25 28 24 Participation of hospitals in public health expenditures (%) 51.1 49.6 51.6a Sources: World Development Indicators 2006; MOH-PAHO 2003a; CCSS 2005. a. This figure is for 2003. outlays increased and then decreased compared to public health spending, allow- ing private expenditures to fall from 25 percent to 24 percent of total health expen- ditures. Despite this upward trend in private health expenditures, Costa Rica has one of the lowest private participations and one of the lowest proportions of out- of-pocket expenditure--19 percent of total health spending in 2005--in Latin America (World Development Indicators). Per capita expenses also rose, by 42 percent between 1998 and 2005, from US$327 to US$305 ($684 in current interna- tional U.S. dollars). Per capita public health expenditures exceeded US$200 in 2003. Public health expenditures accounted for more than a fifth of all government spending, and 52 percent of it was allocated to finance hospital activities in 2003. Benefit-Incidence Analysis In 2001, the distribution of public health spending by income category reflected a positive redistribution of resources in favor of poorer families (figure 8.3). For instance, regarding the CCSS spending, the poorest 20 percent of the population (receiving 4.7 percent of national income), got almost 30 percent of CCSS health- related expenditures. The wealthiest 20 percent of the families (representing 48 percent of the national income) received 11.1 percent of social security resources. Trejos (2002) estimated the proportion of public health expenditures associated with the different income groups, by program. On average, the poorest 20 percent of families received 29 percent of public health spending, in contrast to the 11 per- cent received by the richest 20 percent. Individually, poor families (first quintile) benefited from nutritional programs (55 percent of the corresponding budget), primary care (37 percent), outpatient services (26.5 percent), and hospitalization (26.3 percent). In no case did the poorest quintile receive less than 22 percent of the budget. 190 Good Practice in Health Financing Figure 8.3 Costa Rica: National Distribution of Public Spending on Health, by Income Category, 2001 60 50 40 30 participation % 20 10 0 1 2 3 4 5 quintile public health expenditures (%) national income (%) Source: INEC 2001. Benefits Package The benefits package, defined and provided by the CCSS, includes: general med- ical assistance, specialized services, and surgery; hospital assistance; pharmacy ser- vices; dental services; a cash subsidy for direct affiliates; a funeral stipend in cash; and social provisions. The benefits package is comprehensive; only a few interventions (e.g., some plastic surgeries) are excluded from the list. The Constitutive Law of the CCSS empowers the board of directors (BOD) to define the components of the package. However, recent resolutions of the Constitutional Court suggest that other entities (like the Court itself) are able to modify the package. The absence of a specific ref- erence regarding the right to health in the Constitution has created a vacuum in Costa Rican legislation. Should conflicts arise, the Constitutional Court, at the last minute, will rule on the specific situation. Rulings by the Constitutional Court are binding on the CCSS and create juris- diction for future cases, which has an enormous impact on the definition of the package of services. Some rulings have been broad, such as the one establishing that the benefits package cannot explicitly restrict access to health services, and narrow, such as the recently issued resolution forcing the CCSS to provide a spe- cific drug to a breast cancer patient. In practice, the BOD has only partial control over the definition of the package of services and drugs; it cannot exclude health services or drugs from the package because of a patient's contribution status or type of disease. Costa Rica: Good Practices in Expanding Health Care Coverage 191 Financing and Payment This section describes the financing scheme that operated in the Costa Rican mandatory health insurance system. The section covers two issue: the sources of financing and the payment mechanisms as tools for resource allocation. Financing sources. Revenues to finance health care and health-related activities come from three sources: taxes, as in the case of the Ministry of Health; contribu- tions (especially payroll contributions) to finance health insurance (CCSS); and fees, as in the case of the water and sewage entity (the AyA). The CCSS is the most important health institution, handling 76 percent of public health expenditures. The contributory base of the CCSS is the result of three parties: workers, employers, and the state (table 8.6). Legislation also regulates revenues collected from independent workers. The BOD determines premiums according to actuar- ial analyses, and under no circumstances are the premiums paid by the workers allowed to exceed the premiums paid by employers. The mandatory health insur- ance financing relies on a 15 percent payroll tax for formal employees, 10.25 per- cent of the income reported by independent workers, and 14 percent of the pension received by a retiree. As expected, the largest source of financing for the CCSS--88 percent of total revenues--is contributions from employers, workers, and the state contributions. Contributions from workers and employers increased from 83 percent in 1990 to 87 percent in 2004, and revenues collected from workers grew faster than revenues from employers. Poor households are covered by the state. The Ministry of Finance (MOF) col- lects revenues for this purpose through taxes on luxury goods, liquors, beer, colas, and other similar imports. The procedure is complex for allocating budget to cover expenditures by poor households. Based on the Household Survey, the MOF esti- mates the number of poor people. Then, the average "contribution" per poor per- son is calculated by applying a 14 percent rate to the average premium paid by the Table 8.6 Costa Rica: Payroll Fees, by Insurance Scheme Contribution by source (%) Health insurance type Employee Employer State Pension regime Total Salaried 5.50 9.25 0.25 n.a. 15.00 Independent 4.75 n.a. 5.50 n.a. 10.25 Voluntary 4.65 n.a. 5.50 n.a. 10.15 Contributory pensioner 5.00 n.a. 0.25 8.75 14.00 Noncontributory pensioner n.a. n.a. 0.25 13.75 14.00 Insured by state n.a. n.a. 14.00 n.a. 14.00 Source: CCSS. Note: n.a. = not applicable. 192 Good Practice in Health Financing other contributors at the national level. The projected number of poor people, multiplied by the average contribution, is the budget the MOF plans to transfer to the CCSS. In practice, however, the budget figures diverge significantly from the actual numbers. The CCSS does not have a program for identifying and registering poor households as health service beneficiaries, which means that affiliation of poor families mostly follows an as-you-go basis. Although no person can be denied health care in a public facility in Costa Rica, current regulation establishes that unaffiliated persons must pay for health care at subsidized rates. To avoid paying for care, poor persons must register at the CCSS or at the facility when services are provided. Exemption is effective when the CCSS validates the socioeconomic con- ditions of the person. However, due to shortages of staff and resources to track all the requests, the CCSS has problems detecting fraudulent cases of nonpoor peo- ple registering as state beneficiaries. Payment mechanisms. In the mid-1990s, the CCSS board introduced the man- agement agreement (Compromisos de Gestión), a legal and managerial tool for planning and improving resource allocation. Basically, the agreement is a means of regulating relations between the CCSS and health service providers, involving both primary care facilities and hospitals. These contract-type agreements include key indicators for assessing quality (e.g., intra-hospital infection, mortality, and readmission rates) and utilization (waiting lists and the average length of stay, ALOS). The agreements are also intended to improve coverage and health service quality by linking resource allocation with the achievement of prenegotiated health outcomes. Cost containment is an implicit objective. The management agreement is also a regulatory instrument for the CCSS. It defines, for instance, the health services to be supplied by providers throughout the country. Services included are integral care for children under 10 years old, teenagers 10 to 19 years old, women 20 to 64 years old,1 adults 20 to 64 years old, and the elderly (over 65 years old). For each area, the management agreement also incorporates prenegotiated health goals in terms of accreditation, quality, effi- ciency, and coverage. Hospital performance contracts were implemented in 1997 and, by 2005, 29 hospitals and 104 health regions had signed one with the CCSS. The CCSS has also signed agreements with nonprofit associations, universities, and cooperatives that provide services to more than 400,000 persons, one of the mechanisms used to increase coverage starting in the 1990s. Finally, payment mechanisms differ for primary and tertiary care facilities. Pri- mary care services are paid on a historical capitation basis. For hospitals, a pay- ment unit was defined, known as the "hospital production unit" (UPH in Spanish), for setting rates for hospital activities--in "patient days,""consultations," or "emergencies." The original payment scheme considered a variable component, generally 10 percent of total budget, as an incentive to reward the achievement of Costa Rica: Good Practices in Expanding Health Care Coverage 193 objectives and to cover unexpected events. The entire hospital payment system was originally expected to evolve into a results-based payment model similar to the related diagnostic groups. However, despite some significant progress on this issue, resource allocation has not yet been fully linked to performance. In fact, the incen- tives fund has not been activated. Equity: Health Indicators, Outcomes, and Their Distribution This section presents the most important health outcomes achieved by the Costa Rican health system between 1990 and 2004. Then, the section assesses the exist- ing gaps in terms of access and health outcomes by income quintile. Health Outcomes. Table 8.7 presents a list of key health indicators related to health processes and health outcomes; the information is presented on an averaged biennial basis from 1990 to 2004. In general terms, health outcomes improved in mortality-related indicators and life expectancy; the balance in morbidity diseases and immunization is not fully satisfactory. Life expectancy, for instance, increased by almost two years, one of the highest jumps in Latin America, while infant mor- tality rates fell 32 percent during the assessed period. Universal access to potable water was nearly achieved, and measles disappeared. However, maternal mortality increased (mostly attributable to improvements in recording systems), vaccination rates remained the same or decreased slightly, and the incidence of some diseases like dengue and HIV rose significantly. Table 8.7 Costa Rica: Health Indicators, 1990­2004 Situation Indicator 1990­91 1995­96 2003­04 1990­2004 Gross birth rate 26.4 22.7 17.3 Decrease Infant mortality rate (per 1,000 live births) 14.3 12.5 9.7 Decrease Life expectancy at birth 76.7 76.5 78.6 Increase Men 74.7 74.3 76.4 Increase Women 78.9 78.8 80.9 Increase Maternal mortality rate (per 100,000 live births) 19.8 22.0 25.3 Increase Children with low birth weight (%) 6.3 7.0 6.5 Increase Dengue per 100,000 inhabitants 9.5a 109.7 347.0 Increase Measles per 100,000 inhabitants 103.2 1.4 0.0 Decrease AIDS per 100,000 inhabitants 2.9 4.7 3.7 Increase Vaccination SRP-measles (% 1 year) 91.0 88.0 89.0 Decrease Vaccination VOP3-poliomyelitis (% 1 year) 92.0 86.0 89.0 Decrease Total population served by water system n.a. 95.8% 99.0% Increase Sources: Based on Ministry of Planning data; Proyecto Estado de La Nación 2005; and OPS data. Note: n.a. = not available. a. This rate corresponds to 1992­93. 194 Good Practice in Health Financing Equity problems: gaps. As mentioned, Costa Rica's outcomes were mixed during 1990 to 2004. Some major gains were in line with the health reform objectives. The new delivery model emphasized primary care and prevention at community level. The significant decline in infant mortality and even the improvements in recording and registering maternal mortality can be explained, at least partially, by the new policies accompanying the reform. In other areas, the outcomes are not satisfactory. For example, the outcomes for morbidity-related conditions and vaccination coverage are not the best, despite the fact that vaccination is part of the management agreements between primary care providers and the CCSS. On balance, however, there is a perception that overall living conditions have improved in Costa Rica. One of the main achievements of the reform was improvements in health out- comes, though not uniform, for practically every population group, income group, and region. In some cases, poorer regions got most of the benefits, but in others the gap between the best performer and the rest of the regions or groups widened. This section evaluates equity gaps in four areas: health insurance coverage, health service utilization, infant mortality rates, and per capita budget allocation. Concerning coverage, by 2004, 88 percent of the Costa Rican population was covered by social health insurance (table 8.8). Significant differences persist, how- ever, by income quintiles (table 8.8). For instance, while 85 percent of the Q-5 pop- ulation was covered by health insurance, 70 percent of the Q-1 population was in the same condition. More important, only 10 percent of the poorest 20 percent group was a direct contributor, a number that increased to 49 percent among Q-5. Gaps by region existed, but differences were small, in the range of 2 to 4 percentage points. This can be considered a major strength of the health system: disparities by region have been narrower than in other countries. Additionally, the contribution results reveal the existence of a cross-subsidy from richer groups to poorer families. This situation is clearer from the different use patterns by income group. Regarding health service utilization, poorer groups generally use public facilities more than richer groups (figure 8.4). The better-off tend to use private services instead of public ones, although richer groups presented higher rates of affiliation. Only hospital services showed similar rates of utilization between poorer and richer groups. Health outcomes have improved for almost the entire population regardless of place of residence. Regionally, except in Chorotega, child mortality rates and the under-five (U-5) mortality rates have decreased, and the gaps between the highest and the lowest rates have narrowed (table 8.9). At the canton level, differences in infant mortality rates were up to 5.5 times higher between Talamanca (17.8 deaths per 1,000 live births), with the highest rate, and Turrubares (3.2 deaths), with the lowest; however, dispersion is falling. In total, 95 percent of the cantons had infant mortality rates below 20 deaths per 1,000 live births. Per capita budget allocation, the last equity issue, deals with the distributional problems arising from the payment schemes.Although the reform has modified the way health services are paid and goes a step farther toward a performance-based Costa Rica: Good Practices in Expanding Health Care Coverage 195 Table 8.8 Costa Rica: Health Insurance Coverage, by Income Quintile, 2005 (percent of total population) Indicator Q-1 Q-2 Q-3 Q-4 Q-5 National Insured 70.7 77.6 82.5 86.3 84.7 Direct contributors 10.6 21.3 31.1 39.5 49.2 Insured as family dependents 34.4 46.1 45.2 44.0 33.5 Insured by the state 19.3 6.2 3.7 1.9 0.7 Insured by other plans 6.4 3.9 2.5 1.0 1.4 Urban Insured 72.0 77.9 83.8 86.5 87.8 Direct contributors 13.2 26.1 34.5 42.1 54.0 Insured as family dependents 35.0 42.6 45.6 41.5 35.6 Insured by the state 16.2 4.3 1.8 1.9 0.3 Insured by other plans 7.6 5.0 1.9 0.9 1.0 Rural Insured 68.2 75.3 81.1 85.5 82.4 Direct contributors 8.0 16.6 23.8 33.8 39.0 Insured as family dependents 30.2 47.4 49.2 46.1 40.6 Insured by the state 25.0 8.4 4.9 4.5 1.5 Insured by other plans 5.0 2.9 3.2 1.0 1.3 Source: Authors' estimations based on the INEC 2005 Household Survey. Figure 8.4 Costa Rica: Outpatient Consultations per Inhabitant, by Income Decile, 1998 and 2001 a. 1998 b. 2001 3.0 3.0 2.5 2.5 total total 2.0 2.0 inhabitant public inhabitant public per 1.5 per 1.5 1.0 1.0 consultations consultations 0.5 0.5 private private 0 0 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 income decile income decile Sources: Author estimations based on the Costa Rican Household Surveys for 1998 and 2001. Note: Income deciles (Ds) are defined using the 1998 and 2001 Household Surveys. 196 Good Practice in Health Financing Table 8.9 Costa Rica: Child and Infant Mortality, by Health Region 2001 2004 Under five Under one Under five Under one Region years of age year of age years of age year of age Chorotega 9.5 8.2 12.8 10.7 Central 11.5 9.7 9.4 7.8 Central Pacific 12.0 9.8 8.8 8.2 Brunca 16.9 12.2 12.0 10.0 Atlantic Huetar 16.7 12.7 13.1 10.0 North Huetar 14.6 9.7 10.7 9.1 Source: Authors' estimates, based on Centro Centroamericano de Población database. system, the allocation of funds is creating enormous gaps between regions. The MOH-PAHO-CCSS (2004) reported that some regions receive up to 60 times more funds (per capita) than others for no particular reason (i.e., no significantly differ- ent health profiles). This situation is especially true under the capitation formula used in primary care, which still bases allocations of funds on historical trends, not needs. Thus, because the formula does not fully take into account differences between regions (such as demographic profiles, distance, higher administrative costs), the mechanism perpetuates the perverse incentives generated by input- based models. Efficiency This section on efficiency deals with four main topics: the structure of health spending by use; the structure of health financing by source; an equity analysis of the distribution of public health expenditures. It also briefly assesses health care financial sustainability. Spending by Source and Use. The CCSS, Costa Rica's biggest health institution, contributes two out of three every colones spent in the health sector.2 Households are the second most important source of financing, almost one-fourth of total spending. Most family expenditures on health go to drug purchases and outpa- tient visits. Families use public hospitals in preference to the more costly private hospitals. Nongovernmental organizations (NGOs) and private insurance repre- sent a very low share of total health expenditures (less than 1 percent each). The composition of public health expenditures, by level of care, has two main features. First, the share of hospital spending was practically unchanged between 1997 and 2004 (table 8.10). About half of public health outlays goes to hospital services, the most dynamic of the three levels of care. In the postreform period, hospital expenditures have been growing 5 percent a year in real terms. Arce (2001) estimates that, if this trend continues, in 20 years the system will be forced Costa Rica: Good Practices in Expanding Health Care Coverage 197 Table 8.10 Costa Rica: Health Spending, by Level of Care (percentage) Level of care 1997 2000 2004 Primary 18.8 22.4 21.8 Outpatient 30.1 28.0 26.6 Tertiary 51.1 49.6 51.6 Total 100.0 100.0 100.0 Source: CCSS 2005. to allocate three times more resources to cover all the expenses generated by the hospital sector. For Arce and Saenz (2002), this behavior reflects the failure of the current hospital payment mechanism to control costs because the model incen- tives extended lengths of stay, and the reform has not been able to tighten the link between hospitals and primary care centers. The second relevant feature is the increasing share of primary care services in total public expenditures, a good sign that the system is moving in the right direction. However, if hospital reforms are not completed, the full benefits of incorporating more primary care services in the health sector will be hard to realize. Sources of Financing. Social security contributions are the most important source of financing, almost 60 percent of all health sector revenues (table 8.11). The sec- ond source, sales of goods and services, corresponds to out-of-pocket expenditures by households. If out-of-pocket expenditures and social security contributions are considered together, households contribute half of total revenues. The government contributes 7.3 percent of total revenues (5 percent in the form of taxes and 2.7 percent in the form of contributions to the CCSS). This figure does not, however, include government contributions to pay for health insurance for poor house- holds. The role of external funds is minimal, not more than 4 percent of total health revenues. External grants and donations mainly finance purchases of med- ical equipment and salaries of staff working in the reform program. Table 8.11 Costa Rica: Sources of Health Financing (percentage) Source Participation Taxes 5.0 Contributions to Social Security 59.7 Contributions to INS 4.8 Grants and Donations 3.4 Sales of Goods and Services 27.1 Source: MOH-PAHO-CCSS 2004. 198 Good Practice in Health Financing Figure 8.5 Costa Rica: Income and Social Expense Distribution by Function, 2000 100 education 90 health 80 social security housing 70 income 60 culture 50 percent 40 30 20 10 0 0 10 20 30 40 50 60 70 80 90 100 accumulated % of families Source: Trejos 2004. Health Spending, Equity, and Sustainability. In Costa Rica, public health spend- ing is the most progressive social spending component: the poorest families receive more resources than proportionate to their weight in the population. Because health expenditures favor poorer families more than nonpoor families, health spending is a key factor in improving living conditions of the poorest and reducing income inequality (Trejos 2004). The Lorenz Curve in figure 8.5 illustrates this conclusion.3 Curves above the 45° line are said to be progressive. In these circum- stances, only two sectors, health and housing, appear above the income line. Because health is above housing, it is evident that the health sector is, of all the expenses evaluated, the most progressive within the social area. The same pattern is also observed among regions. In all but the Central region, the poorest 20 percent of families receive at least 29 percent of the public resources invested in health (figure 8.6). In some cases, like the Brunca and the Chorotega regions (the two poorest regions of the country), the families in the first quintile receive between 38 percent and 42 percent of the health expenditures. The Lorenz Curve in figure 8.7 shows the distribution of health insurance con- tributions and the associated level of spending by income level. The principles of solidarity and equity in access are clearly presented. First, the expenditure curve is similar to the 45° line representing high equity in access by all income groups. Soli- darity is seen in the differences in expenditure and contribution participation. D-1, for instance, absorbs 12 percent of health insurance expenditures, but its contribu- tions represent just 0.6 percent of total contributions. On the other hand, D-10 Costa Rica: Good Practices in Expanding Health Care Coverage 199 Figure 8.6 Costa Rica: Distribution of Public Spending on Health Nationally and Regionally, by Income Group, 2001 National Central Chorotega Central Pacific Brunca Atlantic Huetar North Huetar 0 10 20 30 40 50 60 70 80 90 100 income quintile 1 2 3 4 5 Sources: Based on data from INEC2001. Figure 8.7 Costa Rica: Health Expenditures and Contributions to CCSS, by Income Decile 100 90 80 and 70 (%) 60 50 expenditures of 40 contributions 30 share 20 10 0 0 10 20 30 40 50 60 70 80 90 100 share of families (%) expenditures contributions Source: Duran 2005. 200 Good Practice in Health Financing families (richest 10 percent of all families) make 35 percent of the contributions but account for only 8 percent of the expenditures. A final element that deserves special attention is the financial sustainability of the health sector. At the present time, an important list of factors jeopardizes its financial sustainability: · Progressive aging of the population. People over 60 years of age represent a sig- nificant financial burden for the population. In 2002, health care for a person in this group was 3.4 times more expensive than a person aged 15 to 59 years and 4.4 times more expensive than a person 0 to 14 years old. · Low rates of affiliation among economically active population (EAP). Only 52 percent of total EAP (61 percent of the salaried EAP and 38 percent of the non- salaried EAP) are affiliated with the system. · Evasion and default revenues are still high. Despite recent measures to control these two problems, evasion and default are still high. According to the CCSS, evasion in the service sector is 35 percent of potential revenues; 27 percent in commerce; and 12 percent in agriculture (CCSS 2007). In total, 29 percent of potential revenues are evaded by employers. An estimated 40 percent of salaried workers and 60 percent of nonsalaried workers under-report their wages. · Government continuing default conditions. The government usually pays a frac- tion of mandatory contributions for public employees and contributions to protect poor families. A growing backlog of debt owed by the state is one of the financial limitations faced by the CCSS. By 2005, 45 percent of the total debt to CCSS had a public origin (SUPEN 2006). · Incentives for evasion and inconsistencies between contributions and costs. Differ- ential rates favoring independent workers are an incentive for evasion. Lower contributory rates are also dissociated from the real costs of carrying these contributors, creating a permanent "deficit" in special groups like pensioners, independent workers, and others. · Poor mechanisms to verify the socioeconomic conditions of state-subsidized families. For want of resources to check the veracity of statements filed by state-subsidized beneficiaries, the CCSS usually approves payments without further ado. · Increasing hospital costs. Hospital costs (in real terms) are growing faster than any other spending component, as seen above. Duran (2005) presented the expected financial balance of the CCSS by 2050, based on the occurrence of certain revenue and expenditure conditions. The main lesson to be learned, in spite of the assumptions about future expenditure behav- ior, is that the financial sustainability of the CCSS depends on how well the insti- tute manages to increase revenues. Even under an optimistic expenditure scenario (table 8.12), if revenue collections do not grow appropriately, the expected deficit would be 8.1 percent of total expenditures. Costa Rica: Good Practices in Expanding Health Care Coverage 201 Table 8.12 Costa Rica: Estimated Health Insurance Financial Balance in 2050 under Different Scenarios (percent of total expenditures) Revenues Expenditures Pessimistic Base Optimistic Pessimistic ­31.8 ­6.1 2.3 Base ­21.4 8.0 17.8 Optimistic ­8.1 26.6 37.8 Source: Duran 2005. Other Issues This section discusses a wide range of complementary issues about the health sys- tem in Costa Rica. The topic ranges from risk-sharing mechanisms to the role of labor within the sector and relationships with other entities. Risk sharing and cross-subsidization. The health system in Costa Rica is based on solidarity. No matter what a person contributes to the system, he or she will have equal access to health care services. This condition where risks, benefits, and contributions are not aligned one-to-one with contribution size allows cross- subsidization in favor of lower-income or higher-risk individuals. Poor people, an estimated 620,000 individuals, are covered by the "noncontributory" and the "insured by state" regimes. Cross-subsidization goes from contributors (formal employees) to the other groups. According to actuarial studies of the CCSS, about 50 percent of the contributions from formal employees are used to cover health expenditures of pensioners, independent workers, and poorer households. Insurance regulation. The system is not overseen by a formal health insurance reg- ulator. In effect, regulatory tasks are undertaken by the Ministry of Health (health issues) and the Office of the Controller General (budgetary and financial manage- ment). The controller, the most powerful supervisory entity in Costa Rica, has extensive power to force changes or refuse decisions and uses it frequently. The controller can, for instance, approve or reject the institutional budget or recom- mend modifications in contracts the CCSS signs with private firms. Additionally, the controller is the final approver of any investment project (especially those involving private firms), drug purchasing, and acquisition of medical equipment. Last but not least, the controller is empowered to recommend changes in the inter- nal processes of the CCSS (e.g., contracting, purchasing, planning). Although most of this agency's resolutions are "suggestions" to the CCSS, in the final analysis the incentive to obey is implicit because approvals are generally linked to compliance. 202 Good Practice in Health Financing Centralization and decentralization. In 1998, the MOH launched a decentraliza- tion process to improve its stewardship functions at local level. The structure con- sisted of one ministerial office, one general directorate, six central directorates, nine health regions, and 81 individual health stewardship areas. The CCSS also led a decentralization process, because excessive centralization was one of Costa Rica's key problems before the health care reform. Several mea- sures were undertaken in the last decade to dilute CCSS centralization. For exam- ple in 1998, the Congress approved Law 7852 on Decentralization, which enhanced the autonomy of hospital directors in budgetary procurement and administrative and human resource management. The act also set up local health boards to enhance community participation in local decisions. Some administrative changes were also made, such as collecting revenues at regional centers so that contributors pay in local CCSS headquarters. Currently, the CCSS is organized around six man- agement centers, seven sanitation regions, and 103 health care areas. Cost-sharing mechanisms. Costa Rica does not have a copayments system or any other cost-sharing mechanism. Informal copayments, however, are known to be common, especially in areas with long waiting lists or in cases of urgently needed care. Unions in the health sector. Medical labor unions are strong. Six labor unions have a direct relationship with the CCSS. The Sindicato de Profesionales en Cien- cias Medicas (SIPROCIMECA) and the Unión Medica Nacional are the most rep- resentative of the six. The influence of medical unions can be seen in many areas. For instance, doctors have a seat on the CCSS Board of Directors and a quite dif- ferent monetary incentives structure from the payment scheme applicable to all other public employees. Freedom of choice. In Costa Rica, users have no freedom to choose their health care providers within the mandatory health insurance (MHI) system. The CCSS assigns a provider by a person's geographic location (residence). However, Arce and Muñoz (2001) established that there is a "de facto, unregulated freedom of choice," allowing some socially prominent or wealthy individuals to get health ser- vices in a public facility where they are not registered. This system has some char- acteristics of a black market. In 1992, the CCSS introduced a program that offers its affiliates a chance to choose--and pay--a private provider while the institution partially covers selected services. Services may range from surgeries to childbirth, but the final approval of the service and the economic subsidy is in the hands of the Medical Division. Interaction with other programs. The Costa Rican health sector should be viewed as a health system in the sense that it is composed of many other institutions with health-related functions. One of these institutions is the National Insurance Insti- tute, which administers the state private insurance monopoly. The institute par- Costa Rica: Good Practices in Expanding Health Care Coverage 203 ticipates in the health sector by protecting workers against the risks of labor acci- dents and traffic accidents, for instance. The Costa Rican Institute of Water and Sewer Systems provides potable water supply services, collects and treats sewage and industrial liquid waste, and sets the operating norms for rainwater systems in urban areas. To finance its activities, the institute charges user fees. Finally, the University of Costa Rica contributes to the improvement of the health sector in three ways: by educating and training health professionals, by hosting and participating in research and social action projects in the health set- ting, and by providing health services to the CCSS under management agreements between the two institutions. The university's health professional teaching and training actions are financed from the general budget, in turn financed through direct annual transfers from the national budget. In addition, the CCSS is in charge of administering one of Costa Rica's pension regimes (Invalidez, Vejez y Muerte). For many years, the institution managed the health insurance system and the pension system in a single administrative body, and revenues collected from both systems were commingled. There was no trans- parency in the use of these resources, and how much each system collected individ- ually was not clear. The health reform separated the two systems and created the a Pension Division, to deal exclusively with this issue. Revenue collection is still in the hands of one entity, the Financial Division, which collects funds through the Centralized System of Revenue Collection (Sistema Centralizado de Recaudación, SICERE). Overview of the Health Delivery System This section presents a brief characterization of the network of providers involved in the Costa Rican health system. It describes the availability of health-related inputs, the regulatory framework governing providers and the insurance market, and the role of other components of the delivery system. The most important weaknesses and strengths of the system are also assessed. Supply and Organization of Hospitals, Physicians, and Public Health Programs For the understanding of the supply-side structure, this section discussed a series of critical elements regarding the organization of the network of providers and the availability of resources. Organization of the health sector and the network of providers. Institutionally speaking, the Costa Rican health system is organized around the MOH (the health policy governing, regulating, formulating, oversight, and regulatory body) and the CCSS, in charge of administering health insurance. Health insurance, as a vehicle for attending to people's health, is complemented by occupational health risk insurance and by compulsory automobile liability insurance, administered by the INS, the second institution that finances, contracts and provides health services in 204 Good Practice in Health Financing Costa Rica. The private sector,4 though relatively small, has been growing. Its ser- vices are primarily concentrated on delivering ambulatory care and marketing pharmaceuticals. System oversight and regulation have been assigned to the MOH, currently reorganizing to take on these tasks. According to García G. (2004: 2) the health sector in Costa Rica ". . . is made up of a group of institutions and organizations that are part of the public and private sectors whose direct or indi- rect purpose is to contribute to improving the health of people, families, and com- munities, whether the institutions are from the health or other sectors." The proposed distinction is founded in two executive decrees: Executive Decree No. 14313 SPPS-PLAN, dubbed "health sector constitution," which formally created the health sector in Costa Rica and which regulates its structure and organization; and Executive Decree No. 19276-S of November 9, 1989, which created the NHS and establishes the General System By-laws, in which the MOH became the gov- erning institution. The NHS groups together the health sector institutions and four additional types of organizations, such as municipalities, communities, and institutions in charge of training health staff, and the public or private institutions providing health services. The Costa Rican provider network is mainly publicly owned. With the health sector reform that began in the mid-1990s, the MOH became the sectoral over- sight and regulatory institution, while the CCSS gained control of administering the three levels of care. The network was consolidated in one institution after many years of effort. Until the CCSS took over in 1977, hospitals and secondary- level facilities had been run by the MOH or charitable organizations. Then, the reform completed the process by integrating primary care centers under the CCSS umbrella. The public provider network is organized around three levels of care. The first level of care, the gateway of the system, is organized around 103 health areas that provide primary care services in five programs of integral care to children, teen- agers, women, other adults, and elderly people. Primary care services are provided in health centers and clinics where 895 Basic Teams for Integral Health Care (Equipos Básicos de Atención Integral en Salud, EBAIS) work, delivering health services to an average 4,705 persons per EBAIS.5 According to the CCSS, 99 per- cent of the population was covered by primary care services in 2005, making the establishment of the EBAIS one of the most successful coverage-oriented policies of the reform process. The second level of care supplies outpatient services, medical treatments, surg- eries, and hospitalization in a broad range of areas including internal medicine, pediatrics, gynecology and obstetrics, and surgery in a network of 10 major clin- ics, 14 peripheral hospitals, and six regional hospitals. Finally, the third level is in charge of delivering complex health services and surgeries and highly specialized treatments in a network of three general hospi- tals and six specialized hospitals (children, women, geriatrics, psychiatry, and rehabilitation). Costa Rica: Good Practices in Expanding Health Care Coverage 205 Table 8.13 Costa Rica: Health System Inputs, 1995­2002 (per 1,000 inhabitants) Input 1995 2000 2002 Hospital beds 1.78 1.50 1.40a Physicians 0.85 1.32 1.75 Dentists 0.39 0.43 0.52 Pharmacists 0.35 0.45 0.55 Microbiologists 0.28 0.29 0.31 Nursesb 0.56 0.32 n.a. Total workforce 8.21 8.18 8.77c Sources: Information on hospital beds and physicians was taken from WDI 2006; information on dentists, pharmacists, microbiologists, nurses and total workforce comes from MIDEPLAN (http://www.mideplan.go.cr/sides/social/index.html). a. Last value corresponds to 2003. b. Figures refer to nurses employed by the CCSS only. c. Last figure corresponds to 2004. Supply of resources. The availability of health-related inputs has improved in the last decade (table 8.13). The number of physicians and pharmacists (per 1,000 habitants) increased more than 50 percent between 1995 and 2002; the number of dentists, by one third. Only nurses seemed to decline over time, but this situation is explained by three factors. First, the figure represents only nurses hired by the CCSS, an entity that has been representing a lower share of the total health work- force (though still the biggest). Second, in the same line, many nurses work inde- pendently or for the private sector. Third, there has been a brain drain of nurses migrating to other markets (especially to the United States) for better working opportunities. Despite this, the MOH-PAHO-CCSS 2004) estimated there were 1.1 nurses per 1,000 inhabitants in Costa Rica, a higher figure than in similar countries like Chile (1.0 nurses), Argentina (0.52 nurses), and Uruguay (0.7 nurses). Total staff working for the CCSS (still the most important health-related employer and the second biggest employer in Costa Rica) increased 27 percent (1995 to 2004) due to the significant growth among health professionals (3.2 per- cent per year). Policies to increase primary care coverage are a critical factor for understanding these accelerated dynamics. As part of the consolidation of the EBAIS, the CCSS hired many general practitioners and primary care technicians early in project implementation. Between 1995 and 1997, the CCSS created 1,180 positions to launch the first EBAIS, and in 2005 more than 3,000 employees worked for EBAISs. In contrast, hospital beds have been declining in the last years. Between 1995 and 2005, the number of beds (in CCSS hospitals) fell from 6,035 to 5,823 beds. This situation is in line with the reform objectives of improving efficiency by 206 Good Practice in Health Financing limiting the capacity of general and specialized hospitals, ultimately giving pri- mary care more predominance. Regulatory Framework The General Act on Health empowered the MOH to regulate the health sector through the Central Directorate of Health Services, the Directorate for Protection of the Human Environment, and the Directorate of Records and Controls. The MOH has the right to control benefits, set up tariffs, and define which providers can participate in the market. The MOH has organized a health regulation system in three subsystems: regu- lation of health services, regulation of health inputs, and regulation of the envi- ronment. The first subsystem regulates providers by establishing a minimum set of standards to be met by providers for licensure and accreditation. This regula- tion applies to both public and private providers and is oriented mainly to defin- ing standards in terms of infrastructure and health service quality. The MOH is also in charge of regulating pharmaceuticals (regulation of health inputs), and was therefore the institution that approved a drug-registration regulation for the distribution of new pharmaceuticals in the country. Jointly with the laboratories of the University of Costa Rica and the CCSS, the MOH controls the quality of medications. In terms of environmental regulation, serious conservation problems were observed in the last decade regarding water pollution (explained by inappropriate waste management in coffee factories), soil erosion, and progressive deterioration of the forests. The MOH with the Ministry of Environment implemented some specific measures to reduce environmental problems, and some improvements have been observed. For instance, the coffee-related pollution, once responsible for 21 percent of industrial pollution, has fallen to 5 percent in recent years. Despite significant progress in the field, regulation is still incomplete. First, the regulation of private providers is weak, unclear, and does not adequately protect patients. Minimum standards in terms of equipment, infrastructure, staff, health service quality, protocols, monopolistic competition, and market control are not sufficient to regulate private providers effectively. Similarly, the management agreement, the key CCSS tool for regulating its public and nonprofit providers, is a good instrument but incomplete due to the absence of sanctions for noncompli- ance with the negotiated terms. Other Components This section presents the main features of four elements that complement the Costa Rican health system: medical education, the supply and distribution of drugs, medical technology, and information systems. Medical education. Medical education has a long tradition in Costa Rica. The School of Medicine of the University of Costa Rica is one of the oldest faculties in Costa Rica: Good Practices in Expanding Health Care Coverage 207 the nation. Traditionally, a medical degree has conferred high status and is very attractive to high school students and to private universities6 as well. Beginning with two medical schools, the number started growing after 1985. By 2005, Costa Rica had 11 teaching faculties for medicine and other health-related disciplines (like physiotherapy). The number of graduates increased even faster. Between 1990 and 1995, the number of new doctors more than doubled, while the general population grew just 10 percent. The system has been unable to absorb the new doctors, especially because the CCSS (which used to hire more than 80 percent of the national doctors) now hires around 50 percent. The proliferation of new schools reflects the absence of a needs-based national plan for human resources and the paucity of regulations on the issue. Accreditation of new schools (in the hands of two national councils where the MOH has no participation) is done according to administrative criteria (e.g., fulfillment of faculty conditions, labora- tory standards) and not according to real needs. This situation also creates prob- lems in the market. The CENDEISS (the center in charge of planning human resource needs and training requirements within the CCSS) cannot provide slots for every new doctor. The graduates without a post usually start practicing on their own, pushing down the consultation costs. Pharmaceutical supply and distribution. Drug spending in Costa Rica exceeded US$200 million in 2005, and the CCSS is the most important single purchaser (38 percent of total drug expenditures). Both supply and distribution are concentrated in a few companies (oligopolistic markets), and the top seven CCSS suppliers fur- nish 50 percent of the institutional drug purchases. In the private market, compe- tition is greater, and the top five laboratories handle 28 percent of private drug purchases. Such high concentration has a negative impact on medicine costs. To improve access to some key medicines, the MOH has exerted partial control over some drugs. During 2000, the ministry made available a free phone service to ori- ent users about the price of certain drugs and to provide therapeutic information. Distribution is even more concentrated in a few companies. Although the mar- ket has more than 40 drug distributors, the top two firms account for 42 percent of drug sales, and the top four companies handle 61 percent of all sales. The Col- lege of Pharmacists lists about 850 pharmacies in the entire country. The CCSS has defined a list of official drugs (399 active principles classified in 54 therapeutic groups) that all the pharmacies must stock. In addition, every pharmacy must have a graduate pharmacist to provide free client services. Medical technology. Research and innovation practice is regulated by the MOH. Public health experts generally believe that the system devotes too few resources to improving research and innovation. Costa Rica has a wide range of institutions in charge of different health sector tasks. For example, public health research is developed by the Institute of Health Research and by the Clodomiro Picado Insti- tute (University of Costa Rica). There are three more public entities with very 208 Good Practice in Health Financing different profiles: a research institute on nutrition, a national drug center, and an institute on alcoholism and pharmacodependency. Information systems. Despite the existence of many heath information systems (HIS) in the CCSS, they play a minor role in policy making. Strategic planning, critically lacking before the reform, has been weakly incorporated into the daily tasks of the CCSS. The information collected by the HIS is used mostly for follow- ing up the contract performance but rarely for estimating population needs. The high rates of evasion and default force the CCSS to improve collection using mod- ern tools. In 2001, the CCSS installed the Sistema Centralizado de Recaudación (Centralized Collection System, SICERE) to ease management procedures for billing and contribution collection. Strengths and Weaknesses of the System Table 8.14 summarizes the major strengths and weaknesses of the Costa Rican health system. The list includes mainly the aspects assessed above and some other issues derived from the analysis. Table 8.14 Costa Rica: Major Strengths and Weaknesses of the Health Sector Strengths Weaknesses Solid institutions with more than 50 years of Falling rates of contributory coverage among operations; strong national health system economically active population Strong political support from all sectors Weak regulation in key areas like medical education Functional solidarity principle Increasing hospital costs Universal access to primary care centers and wide Problems with service quality, especially in access to public hospitals hospitals Positive incidence of benefits among poorer income Persistent gaps in equity and health outcome by groups, whose benefits exceed their financial gender and between regions contributions to the system High financial protection, especially for poorer Lack of adequate long-term strategic planning households; low out-of-pocket expenditures and no cost sharing mechanisms (formal) Extensive benefits package Slow transition to full performance-based system High levels of equity between rural and urban areas Weak internal organization of CCSS Stable public financing over the years High rates of evasion and default Strong external control from other public No freedom of choice institutions to avoid corruptive practices Adequate number of health professionals Lack of consolidated medical technology program Low dependence on external funds Excessive number of medical schools Cumulative experience transformed into concrete Not a patient-centered system outcomes Source: Author elaboration based on previous sections. Costa Rica: Good Practices in Expanding Health Care Coverage 209 Health Coverage Reforms Between 1943 and 1990, policies to increase health insurance coverage were ori- ented mainly to increase the breadth of coverage (that is, the number of individuals covered by health insurance). The health reform of 1994 shifted emphasis to depth of coverage (that is, the increasing of the share of the population with access to pri- mary care services). Policies Oriented to Increase Health Coverage Since the 1940s, with the creation of the CCSS, efforts were made to make social security coverage in Costa Rica universal and to maintain what had already been achieved in health. As first conceived, the health insurance scheme covered only public employees and manufacturing workers. Later, CCSS expanded the breadth of coverage to include farmers, independent workers, and unemployed/poor people. In 1961, the Congress approved Act 2,738 and established universal health insurance for workers and their families. This reform represents the turning point in health insurance coverage by including family dependents as insurance benefi- ciaries. This reform became the most important step toward universalization of health insurance in Costa Rica. In 1973, Act 5,349 transferred to the CCSS all public hospitals previously con- trolled by the Ministry of Health. The act defined the first bases of a social assis- tance program in which the government engaged to pay for health services for poor people. Although this effort was not a formal health insurance program for the poor, it at least guaranteed poorer families greater access to health services. Later, in 1975, the CCSS expanded health insurance coverage to farmers and independent workers. A year later, the Board of Directors (BOD) established the obligation of all inactive pensioners to be insured under the sickness and mater- nity scheme. In 1978, the CCSS created and regulated the voluntary health insur- ance scheme for independent workers. Then in 1984, the government of President Monge created the state-subsidized insurance plan for covering poorer families and unemployed persons. Unlike the program established in 1973, this was a per- manent insurance program with all the characteristics of a health insurance plan. The last two milestones in the long history of coverage-related policies occurred in 2000 and 2006. In 2000, the Act of Worker Protection declared all independent workers' obligation to join the health insurance program. More recently, the CCSS reformed the health insurance regulations for the first time in 10 years. The reform allows affiliated workers to cover handicapped brothers or sisters who cannot work. Also, the affiliated can insure brothers or sisters taking care of their parents if the parents are handicapped and over 65 years old. Finally, the reform expands health insurance protection to students younger than 25 and registered in non- university tertiary education centers (like university colleges and parauniversity institutes). Formerly, only students below 25 years of age and studying at universi- ties were protected by the legislation. 210 Good Practice in Health Financing The economic crisis at the end of the 1970s and the beginning of the 1980s caused a major upheaval in Costa Rica's public finances. Health sector invest- ments abruptly ceased, and service coverage could not be broadened. In addition, within the context of state reform at the beginning of the 1990s, a health sector assessment identified the most relevant problems to be solved to improve health system efficiency and effectiveness. These included: structural, organizational, and functional fragmentation; deficiencies in the regulatory framework; inefficiency and inequity in resource allocation; service-scheduling deficiencies arising from excessive centralization; nonexistence of demand self-regulation mechanism as a result of the privileged position given to tertiary care at the expense of integral individual care; lack of continuity in user care, caused by the nonexistence of a referral and counter-referral framework and by overspecialization in medical practice; dissatisfaction among service users and providers; and inadequate train- ing in human resource management. A critical problem in the sector before 1990 was low coverage for primary care services. before 1990, around 82 percent of the population was covered by health insurance. Nonetheless, the traditional PHC model was dilapidated, people did not have access to a nearby PHC facility (within 30 minutes of their residence). PHC coverage dropped from 46 percent in 1990 to 25 percent in 1995. Thus, the main challenge was not to increase breadth of coverage but depth of coverage. The Health Sector Reform: Policies and Evaluation To deal with these problems, the health sector reform was organized into three broad areas for action: the financing model, institutional reorganization, and modernizing management of the service provider network. Financing model. For almost 40 years, social security financing had experienced legal constraints on enforcing mandatory affiliation. The reform sought to reduce this problem by promoting two critical elements: (1) increasing contribution cov- erage by designing means of enforcing compulsory contributions, setting up sys- tems for tracking evasion, and extending contribution coverage to the independent workers; and (2) redesigning the financing model, redefining contribution rates for independent workers and pensioners, and modifying the system for calculating state contributions to the CCSS to cover the poor and the uninsured. Efforts to reform the financing model were not entirely successful, partly due to unforeseen events. The main results from the reform period may be summa- rized as follows: The gap between contributions and real expenses per insured began to narrow. As a percentage of GDP, health insurance contributions were stable at around 5 per- cent between 1990 and 2004. The differential between real income and real expense per direct insured has been falling. Between 1990 and 1999, the "surplus" per direct insured was 21 colones. After 2000, it declined to 10 colones, indicating that the subsidy to noncontributors had fallen to less than half. In 2003, the real Costa Rica: Good Practices in Expanding Health Care Coverage 211 contribution was insufficient to cover the real expense incurred in meeting the insured parties' demand for health services. The Costa Rica's ongoing demographic transition partly explains this spending behavior. Arias (2004) points out other factors affecting the trend toward finan- cial imbalance. These include evasion, late contributions, deficiencies in insurance administration, and inappropriate use of medical insurance by insured family members. Arias, for 2003, estimated evasion in the noncontributing, independent, and retired regimens at C 143,179 billion (US$359 million), equivalent to 51 per- cent of collections from the salaried worker regimen. Increases in the contributions have relieved the funding shortfall. Modifications in premiums are always a big change, because more than 90 percent of health insurance income comes from member contributions. Contributions began to increase, beginning in 2000, but other unsolved financing problems appeared: · State contributions for financing the regimens cover only 14 percent of the effective costs. · Contribution evasion is estimated at around 30 percent of the regulatory health insurance income. · High defaults on payments by employers and even the state (as an employer) also weaken the financial structure of the CCSS. In 2005, the state and employ- ers owed the CCSS C 50.2 billion (US$106 million), 86 percent of it owed by the state.7 The CCSS contribution coverage has been decreasing. Both salaried and non- salaried worker coverage has decreased in the last decade (figure 8.8). Between 2000 to 2004 the share of independent workers, contributing to the CCSS declined from above 80 percent to just under 40 percent--although a significant part of this decline is due to the introduction of a new method for estimating the size of the independent workers labor market. For salaried workers the declining trend implied a fall of the coverage rate from 75.3 percent in 1990 to 61.8 percent in 2004 in that group. In the last two years, the CCSS added 114,000 workers con- tributing to health insurance. Thus, for example, in September 2004, 1,076,028 workers were contributing to insurance, as compared with 961,266 in 2003, a 12 percent increase in coverage. Financial protection for the population, especially the poorer groups, is good, judg- ing by evidence from national surveys. According to estimates in the Survey on Household Revenues and Expenditures (SHRE 2006), Costa Rican households allocate 2.6 percent of their income to health spending, 4.1 percent of current, noncapital spending. Most household health expenditures go to specialist/hospital services and drugs, and such expenses affect urban households (5.1 percent of their income) more than rural families (3.6 percent). The share of health expenditures in the fifth quintile is approximately 6.6 percent of their income, 3.7 times higher than the share of health expenses in poorer households (first quintile). In other 212 Good Practice in Health Financing Figure 8.8 Costa Rica: Health Insurance Coverage, of Economically Active Population 1990­2004 90 82.51 80 75.34 70 66.45 61.80 70.6 60 50 coverage 40 % 39.14 30 20 10 0 1990 1992 1994 1996 1998 2000 2002 2004 salaried nonsalaried Source: Proyecto Estado de la Nación 2005. words, wealthier families are responsible for most of the out-of-pocket spending. Poorer households are not only less encumbered by health spending, but their bur- den also decreased between 1998 and 2004. In 1988, the SHRE reported that poor households allocated 2.1 percent of their income to cover health expenditures. By 2004, that share declined by 15 percent to 1.8 percent of their spending. Institutional reorganization. The institutional reorganization entailed strength- ening CCSS and MOH by establishing four strategic functions, separating man- agement and financing of health and pension benefits, and modernizing provider network management. The Law on Decentralization of 1998 was a key instrument in this reorganization. A key element of the sector-wide reforms was to clearly separate the stewardship functions, assumed fully by the MOH, from the financing, purchasing and provision functions which were managed by the CCSS. The MOH assumed the tasks related to oversight, coordination, and regulation of the system. The CCSS took over the insurer and health service provider functions at the three traditional levels, including health promotion and preventive services previously carried out by the MOH. This process went into operation beginning in 1998. One of the more important actions was the establishment of the Sector Council as a coordination body, under the supervision of the MOH, to be used to manage and lead sector policies. This was done despite the fact that its autonomy allowed the CCSS a fair amount of discretion in its own governance and policy formulation and planning. Costa Rica: Good Practices in Expanding Health Care Coverage 213 Four strategic functions were established within the MOH: political direction and leadership, health oversight, technological research and development, and health development regulation (MOH 2002). Some major progress in the ministry's oversight and regulatory function included the formulation of the National Health Policy 2002 to 2006, the publication of the joint health sector agenda for 2002 to 2006, implementation of the Institutional Evaluation and Development System (SEDI in Spanish), design and implementation of the national health account system, and evaluation of integral first-level care. The administrative separation of financing from purchasing and provision in health services, and separation of pensions from health, were the most prominent fea- tures of the CCSS institutional reforms in the past decade. Within the new health insurance structure, three basic functions were defined: health service financing, health service purchasing, and health service provision. Carrying out these new functions required certain changes, such as the creation in 1999 of the health service purchasing management area, responsible for service planning and pur- chasing activities and the approval of the Law on Decentralization of the CCSS Hospitals and Clinics in 1998. As a result of this process, overlapping subsidies between the two types of insurance (health and pensions) were eliminated, and administrative costs were efficiently allocated between the two insurers. The process, however, has run into some difficulties, and some challenges still lay ahead. They include: · Some interest groups within the CCSS are opposed to the reorganization. · The health service purchasing management area was assigned to the CCSS administrative division without any clear, direct link to the financial division. Therefore, the planning and purchasing processes lack an appropriate interre- lationship to facilitate financial and budget scheduling, resource allocation, and service provision. · The institutional reorganization needs to be redefined in terms of processes instead of structures. This means moving from the prevailing activity-oriented approach (enrollment, administration, statistics) to a more integrated approach (the health service purchasing cycle, for instance). This situation is equally required in both the MOH and the CCSS. In any case, the new institutional structure should be defined according to the new approach that favors increased decentralization, enhanced quality of care, improved health network manage- ment, and a performance-based orientation (Arce and Sáenz 2002). · Although the MOH has made significant advances in strengthening the stew- ardship function in the sector, there is a large unfinished agenda with regard to how the MOH interacts with the CCSS in terms of oversight of the financing and provision of care and in terms of the development of public health activi- ties in the sector. Furthermore, MOH activities in the environmental, food and drug safety and technology assessment have lagged considerably behind. 214 Good Practice in Health Financing · The regionalization system used needs to be brought back to the table because the CCSS and the MOH use two different systems. Staff, financial resources, and materials are unnecessarily duplicated, and efficiency suffers when attempting to coordinate sector policies. Law 7852 on Decentralization of 1998 contributed to the regulation process in three essential aspects of the health reform process. First, it promoted administrative decentralization in the CCSS. Second, it created Health Boards as community groups that promote the involvement of community leaders, democratically elected, to supervise the delivery of services from the perspective of responsive- ness to the population's needs and expectations. Finally, the law introduced the possibility of eliminating the life-time tenure enjoyed by hospital directors and promoted additional measures to improve incentives for hospital management. Service provider network management modernization. In provider network management, the modernization reform was intended to replace the traditional financing plan, based on the historic budget, with a more prospective arrange- ment, in which service production and compliance with quality standards were indicators that would define budget allocation. A first attempt was made at the end of the 1980s with the introduction of cooperatives and single payments that predefined a coverage region. Later, in 1997, management agreements were intro- duced, which generated major changes. Primary care has become more prominent in total expenses. Between 1997 and 2004, the proportion or resources allocated to primary care increased from 19 to 22 percent. This increase was financed by reducing the rate of growth of hospital expenditures, thereby reducing the overall allocation to hospital services in favor of increasing funds for primary care. The real expenses incurred over this period illustrate the impact that resource allocation has had on primary care. For ambu- latory services, actual expenses increased 78 percent, versus a 55 percent increase in hospital expenses. The targeting of health reform to the poorest areas, with the greater socio- economic needs, produced significant gains in access to care and health outcomes. Rosero-Bixby (2004) has shown that key indicators, such as infant mortality, decreased more rapidly in these areas because they were targeted in the first phase of the reforms.9 Primary care coverage results. Readjusting the health care model favored an inte- gral and social health promotion approach. The pivotal point of the readjustment was the creation of the health areas (103 areas in 2005) based on geographic- population criteria and criteria pertaining to accessibility, population distribu- tion, and the country's political-administrative divisions. Each health area was subdivided into segments that were assigned at least one EBAIS. The number of primary care establishments with an EBAIS team has grown constantly since Costa Rica: Good Practices in Expanding Health Care Coverage 215 Figure 8.9 Costa Rica: Primary Health Care Program Coverage, 1990­2003 100 90 87.00 86.00 80 81.00 PHC by 70 69.0069.00 61.00 60 covered 50 47.00 45.00 46.00 44.00 46.00 40 42.00 new PHC model 36.00 population 30 of 20 25.00 % traditional PHC model 10 0 1990 1992 1994 1996 1998 2000 2002 2004 Source: CCSS. 1995. Thus, for example, 232 establishments were founded in 1995, and by 2004 there were 855 EBAIS, (i.e., 3.7 times more that at the beginning of the program). Priorities were determined for each region, based on the population's epidemio- logical profile, which was a major step in defining needs-based plans. The new model that accompanied the health reform in 1994 strengthened the role of pri- mary care through the EBAIS. In terms of coverage, the rate jumped from 36 per- cent of the population in 1995 to 69 percent in 2000, and to 86 percent in 2003 (figure 8.9). According to the presidency of the Republic, practically everyone in Costa Rica was covered by the primary health care (PHC) program in 2005. As a result of the Reform, the percentage of people without equitable access to primary health services dropped by 15% between 1994 and 2000 in areas where health sec- tor reform was implemented in 1995­1996, whereas areas that had not yet initi- ated health sector reform in 2000 experienced only a 3% reduction. In addition to the reorganization of the network by health areas and the cre- ation of the EBAIS, another strategy developed by the CCSS to increase PHC delivery entailed the introduction of special partnership models between the CCSS and nonprofit institutions. The first attempts to resolve shortcomings in primary care coverage started in 1988 when the CCSS broke the traditional public-based delivery model and began purchasing health services from health cooperatives. Because the early experiences were successful in terms of quality and cost, the CCSS decided to expand the program during the reform period. By 2004, the CCSS had established management agreements with four cooperatives and one foundation (belonging to the University of Costa Rica), covering more than 10 percent of the Costa Rican population. 216 Good Practice in Health Financing Key Conditions for Success In both breadth and depth of coverage, Costa Rica is a high performer but not problem-free. In terms of breadth, coverage is close to universal, with almost 90 percent of the population affiliated to the CCSS. By 2006, 3.9 million people (89.2% of the population) were affiliated to CCSS in any of the health insurance programs that exist for that purpose (direct, self-insured, State, pensioners). Important internal problems regarding financing and administration persist, but over the years the CCSS has been transformed into one of the most solid institu- tions in Costa Rica. In terms of depth of coverage, the reform has also been success- ful. In less than 12 years, the system moved from a 25 percent coverage rate (primary care services) to 99 percent in 2005. Again, the EBAIS model improved access and the role of primary care in the system, although some problems persist in service quality and immunization coverage. The reform results suggest a net positive balance overall. Among the main fac- tors explaining the success of the reform are: Political commitment: the two most important political parties agreed change was needed, the new government in 1994 followed through, and stakeholders were involved in the process. Although the Partido Liberación Nacional and the Partido Unidad Social Cristiana did not agree on every detail, they did agree on the need for health care reform and strongly backed the process, allowing the Reform to continue despite alternation of the presidency in 1994. The rapid approval of the reform loans in the Congress was a signal of such political support. The Ministry of Planning and the CCSS itself were heavily involved in the design and imple- mentation of the reform. Opposition to the reform existed, but it was more the result of sporadic groups than the collective opposition of key stakeholders. Transparency and continuous communication were critical to facilitate the implementation. Mixed commissions were created to involve unions and hospitals in the discussion and the understand- ing of the process. This support took unusual dimensions. The pilot test of the new resource allocation model was, for instance, publicly supported by the hospitals included in the sample. In short, one of the main advantages of the Costa Rican case was the relatively general consensus among the main stakeholders: govern- ment/CCSS, opposition, unions, and providers. Origin of the reform: the idea for health reform came from within the health sec- tor, not from outside. This special feature has allowed the reform process to unfold smoothly without confrontations between key institutions like the CCSS and the MOH. Availability of funds: political support was expressed in economic support. The reform program was financed with external loans and local contributions. The first loan amounted to US$22 million from the World Bank, and it was part of a US$32 million project for financing the core component of the reform agenda. A second loan was approved by the Inter-American Development Bank, for US$42 million. The total cost of the second project amounted to US$60 million. Costa Rica: Good Practices in Expanding Health Care Coverage 217 B O X 8.1 Costa Rica: Cooperatives as Health Care Providers The introduction of market-like mechanisms by an administrative council, elected every began in 1988 with the introduction of the two years by the general assembly of all first health care cooperatives. Each coopera- cooperative members. Legally, all workers in tive was founded by the employees of a the cooperatives became shareholders and primary health care clinic. They formed periodically received earnings generated autonomous, legal entities that assumed by the cooperative. The cooperatives responsibility for managing the facility. The operated under private law, without public facilities were leased from the CCSS to the encumbrances related to contracting, firing, cooperative for a yearly fee of US$1, and all and management of personnel and resources. equipment and infrastructure were The cooperatives enjoyed income tax exemp- transferred to the cooperative. From this point tions and considerable political support. They on, the cooperative assumed full responsibility had full autonomy to manage the capitated for maintaining the transferred equipment payments from the CCS, but were obliged to and buying new equipment (unless coopera- present annual financial statements to the tives could prove that damage or deteriora- CCSS for external audit. All profits were either tion of the transferred equipment were reinvested in new equipment or infrastructure due to faulty construction or quality). The or distributed to cooperative members. In cooperatives bought drugs, medical supplies, return, the cooperatives were required to fol- and other inputs from CCSS centers at cost low the guidelines and policy objectives set plus 15 percent for administrative and ship- by the CCSS and the Ministry of Health.The ping expenses, or directly from the market. contracts signed with the CCSS obliged coop- The same services as in public clinics were eratives to provide the following services: provided, and the catchment population con- general and specialized medicine, emergency tinued to be the residents of the geographic care, minor surgery, dental care, pharmaceuti- area served by the cooperative. The coopera- cals, laboratory and radiology services, biop- tives received a yearly capitation fee based on sies, laboratory smear tests, social work the estimated number of members in the services, services related to rights and benefits geographic area. No additional services were verification, transportation for patients, and provided or fees charged by the cooperatives, support services. because the new model was supposed to Gauri, Cercone, and Briceno (2004) showed maintain an identical package of services with that cooperatives conducted an average of 9.7 only a change in management responsibility. to 33.8 percent more general visits, 27.9 to The cooperatives were, however, allowed to 56.6 percent more dental visits, and 28.9 to charge uninsured individuals seeking care 100 percent fewer specialist visits than CCSS on equal terms with the CCSS clinics. clinics. The number of nonmedical, Several groups had strongly opposed the emergency, and first-time visits per capita introduction of cooperatives because were no different from the traditional public they suspected that it was effectively a privati- clinics. These results suggest that the coopera- zation. This, they viewed as a renunciation by tives substituted generalist for specialist ser- the state of its responsibility to provide free or vices and offered additional dental services, low-cost health services. but did not turn away new patients, refuse In 1988, a self-managing cooperative, emergency cases, or substitute nurses for doc- COOPESALUD R.L., was named service admin- tors as care providers. istrator and put in charge of the CCSS Pavas Clinic. Cooperative management was selected Source: Gauri, Cercone, and Briceno 2004. 218 Good Practice in Health Financing Most of the funds were allocated for modernization of the primary care network. In total, US$17 million from Project 1 and US$19 million from Project 2 were assigned to primary care transformation. Wide participation of main stakeholders: CCSS workers (especially medical staff), MOH staff, medical doctors, and communities all participated in the reform process. The information given the stakeholders was transparent. Although several groups opposed the reforms, as was to be expected, continuous negotiation and consen- sus building allowed the main part of the reform to go forward. Nonetheless, the original reform agenda has still not been fully implemented, and some elements are missing. For instance, the issue of allowing patients the right to choose their provider has not yet been considered. Priority to rural and poor areas: the first EBAIS were launched in rural areas, par- ticularly in poor regions. Almost all of the first 229 EBAIS were opened in poor, rural areas, following social equity principles. This has allowed the system to gain quick benefits and to create a solid basis of support for further changes. This is a distinctive characteristic of the health coverage reform. Strong technical team: all members of the team were highly qualified. The group of professionals that designed and implemented the reform and the "political team" (the BOD and the executive president of the CCSS) had a solid background in similar tasks, an understanding of the national problems, and experience in negotiating key changes. Clear reform objectives: The CCSS was clear about its objectives. Although there were many problems at the time of the assessment, technical and political efforts had to be concentrated on just some of them. Putting too many objectives and "components" in the reform agenda may dilute the strength of the reform. Transformation of the CCSS for a sustained reform process. Changes in the model meant that some internal reorganization had to be made in the way the CCSS functions. Some measures were implemented, like the creation of a Purchasing Unit and the establishment of the CENDEISS, the department in charge of train- ing CCSS staff. Continuous monitoring and evaluation. The reform was constantly monitored to verify that different measures had been adopted and observe the outcomes. Focus on access to primary care. The health sector reform greatly benefited from expanding PHC services rather than concentrating on hospital care. The new strategy moved to a PHC-centered delivery model that increased coverage in rural areas and improved physical access to basic health services throughout the coun- try. This strategy yielded concrete benefits during the first two years of implemen- tation. In about a decade, the new PHC model achieved nearly full coverage. The outcomes of the new strategy were the result of multiple concurrent factors: insti- tutional transformations, community participation in decision making, increased budgets, enhanced participation by the for-profit and nonprofit sectors, adequate planning, and incorporation of quality standards. Costa Rica: Good Practices in Expanding Health Care Coverage 219 The Outlook for Further Reform Perhaps corruption is the most important enemy of the reform process at this moment. A critical case of corruption, known as "the Finnish scandal," hit the CCSS in 2004 when the general prosecutor of the republic brought charges for misappropriation of funds against the CEO of the institution, together with a for- mer president of the republic, members of the BOD and staff of Corporación Fis- chel.10 According to the brief, the CCSS had been granted a US$39 million loan from the government of Finland for modernizing the hospital network. At least half of the loan was tied to purchases of Finnish products. The contract was awarded to the Finnish consortium Instrumentarium Corporation Medko Med- ical, and a commission of US$8.8 million (20 percent of the value of the loan) was paid to Corporación Fischel (the consortium's Costa Rican representative). Part of this commission was paid to bank accounts of CCSS board members and former president. The scandal severely damaged citizens' confidence in the efficiency of the institution. The common perception of the effects of corruption scandals in the CCSS is that they greatly affect the health services. The 2004 scandals severely curtailed any effort to implement further changes in the organizational, financing, and delivery model. Many consulted experts believe that any attempt to change the status quo is now associated with corrup- tion and is under strict public scrutiny. As a result, reforms have slowed down and are more difficult to put into action. Despite substantial progress in the overall health model, the reform still lags in several areas. Some of those topics will definitely be "points of disagreement" between the CCSS and other stakeholders or within the CCSS. Among the most important issues are · The monopoly of the CCSS in the administration of health insurance schemes in Costa Rica. Should the market open to new insurers? Privatization has no room in Costa Rica at this moment (the CCSS is the public institution with the greatest support from the citizenship), but market openness should be dis- cussed in more detail. · Free choice of providers, especially of hospital services. · Implementation of new financing mechanisms beyond the current tripartite model. · The definition of a modern institutional organization. This was discussed in the early reform years but never appear in the final agenda. Moreover, the 2004 scandal was facilitated by the "restructuring process" implemented by the new CCSS authorities, which gave managers more power and decision-making autonomy. Thus, any plan to further change the administrative structure usu- ally arouses suspicion and encounters obstacles to its full implementation. 220 Good Practice in Health Financing Lessons for Policy Makers The enabling factors (financial, institutional, others) that create adequate condi- tions for success in the field of health coverage are analyzed in this last section. Some elements of the Costa Rican political economy are evaluated, and some lessons are derived for possible use by policy makers in low- and middle-income countries. Enabling Technical and Political Factors From the Costa Rican experience, the key enabling factors for the success of a coverage-oriented reform are The political actors must compromise to reach a consensus and follow through. In Costa Rica, the reform was designed by one government but implemented by the opposition in the next presidential term. Achieving commitment implies constant dialog, negotiation, and consensus building with the main stakeholders. Designing a health reform is a complex exercise that demands technically sound strategies. It requires adequate diagnosis of the sector, an economic and political strategy for implementation, and incorporation of country-specific conditions. When designing a reform, it is important to single out some key aspects on which most of the effort will be concentrated, instead of trying to solve every problem at the same time. Appropriate phases of the reform should also be discussed. Sustainability of the reform depends on the financial support throughout imple- mentation. This implies two things. First, there must be a "short-term" budget to launch the program, such as the two loans Costa Rica had. Second, it is critical to commit to investing resources for the long haul. The example of Costa Rica shows that, if sustained over time, health investments pay off. Since 1950, Costa Rica has allocated ever-increasing amounts of money to the health sector. As a result, life expectancy increased from 45 years in the mid-1940s to almost 80 years in the mid-2000s. Active participation by local communities in decision making is an important part of reform design. Final users who help shape the reform will understand the scope of the changes. Opposition (usually heated in health-related issues) will cool off. Ask which measures will yield positive benefits in the short term. It is hard for peo- ple to accept change if they cannot quickly see some benefits. In Costa Rica, the first steps were oriented to increasing coverage in rural and poor communities. Private participation is important. Anything but public service delivery invites opposition in many countries. However, in countries where the network of Costa Rica: Good Practices in Expanding Health Care Coverage 221 providers is thin and the needs urgent, the contribution of the private sector may be critical to avoid bureaucratic regulations that can delay infrastructure projects, for instance. Institutional Financing Arrangements Parallel to changes in the health sector, some internal administrative and organi- zational transformation should be implemented: Creation of the Purchasing Unit was the most important organizational change introduced in the CCSS during the reform period. Separating purchasing from provision functions has allowed a single department to concentrate fully on plan- ning, negotiating, monitoring, and evaluating the performance of health care providers. At the same time, this separation has permitted the Purchasing Unit to focus on coverage issues. Both clinical and nonclinical staff were trained on the new orientations of the model. During reform implementation, the CENDEISS did the training CCSS staff. Giving a training program permanent status, with its own budget and clear long-term objectives, might be a good idea. Adequate information goes hand in hand with the sectoral transformation. Good information systems allow an institution to improve its capacity to track contribu- tions, evasions, consultancies, costs, and reporting and to improve the planning cycle. The SICERE system, for example, enables the CCSS financial management to track revenue flows daily instead of only 30 days after the fact, as before. Restructuring payment mechanisms could be a powerful tool to increase coverage. In Costa Rica, the management agreements between the CCSS and primary care providers explicitly define coverage (e.g., immunization, smear) as a target to which payments will be linked. Clear incentives to promote administrative staff efficiency should be set up. Lessons for Low- and Middle-Income Countries The reform experience of Costa Rica suggests eight sets of lessons for low- and middle-income countries. Reforms can be applied even if the context is far from ideal. Almost everyone, including workers, opposed the creation of mandatory health insurance in Costa Rica in the early 1940s. Nonetheless, the president insisted that the people needed a health insurance system for financial protection. After intensive negotiations, the project was approved in the Congress with the support of the Roman Catholic 222 Good Practice in Health Financing Church, the governing party, and the Communist Party. Completely opposite forces, like the Church and the Communist Party, strongly supported the initiative. Health reforms take time for implementation--and results. Changing a health service delivery model takes time. A new model demands not only new staff and equipment, but also changes in perceptions and behavior of the professionals and their prospective patients. In Costa Rica, one change was the switch in orientation from cure to prevention. Achieving full primary care coverage with services took Costa Rica less than 15 years; fully universal health insurance coverage is still a goal, after more than a half-century. Experiences will vary from country to coun- try and from topic to topic. Expand primary care to reach universal coverage. The main focus of the health sector reform in Costa Rica was the expansion of primary care services as the main channel to provide universal access to the whole population. The new strategic ori- entation yielded important health outcomes. The success of the PHC-oriented strategy resulted from the convergence of multiple factors. They included · The top CCSS authorities knew a new model was needed. · Budget increases concentrated on PHC activities. · Cooperatives and the private sector were intelligently incorporated in health service delivery and regulated by management agreements that include specific coverage and quality targets. · Adequate planning of the new model favored division of the country into small catchment areas, each with at least one EBAIS. · Costa Rica is a small country with easy access to rural areas. · Heavy investments were made in infrastructure in the first stages of reform. · PHC coverage prereform was low, and all stakeholders recognized the need for a new model and came to support it. Money counts. Without enough money, good results cannot be sustained. Bud- gets must be adequate to trigger the reforms and to consolidate a base for further transformations. In Costa Rica, strengthening primary care was one of the pillars of the strategy to expand coverage. The share of PHC in the total health budget of Costa Rica increased 22 percent in 10 years. With the additional money, the CCSS to set up more than 855 EBAIS, each taking care of 1,000 households (4,000 per- sons) throughout the country by 2006. The public and the private sectors, working together, can do more to improve health conditions than either sector can achieve on its own. Costa Rica, breaking with the traditional public-centered delivery model, incorporated new provision models in which the private sector was given an important role. Both international experience and the Costa Rican case show that the private sector can be an extraordinary Costa Rica: Good Practices in Expanding Health Care Coverage 223 means of increasing coverage, especially service coverage. Participation of nonpub- lic entities should be limited to private firms, and to delivery, not financing. In Costa Rica, the spectrum of potential providers ranges from cooperatives to univer- sities to nonprofit associations and NGOs. The incorporation of cooperatives and private foundations complemented CCSS efforts to extend primary care through- out the country. By incorporating the management agreements and creating an adequate regulatory framework to normalize relations between the CCSS and pri- vate providers, the model was able to launch the EBAIS in poor districts. Think about the needs of the disadvantaged first. The decision to site the first EBAIS in rural and poor regions yielded multiple benefits to the reform process (see endnote 9). First, it increased equity and access in traditionally underserved communities. Second, starting with poor regions is likely to generate quick bene- fits. Quick benefits win recognition from different social groups, creating a solid base for further reforms. Be open and above board. Transparency and accountability mechanisms should be activated at the outset to enhance communication with all interested parties and build their trust. Pay attention to the financial health of the system. Costa Rica is good example of financial sustainability--but not cost containment. Contributions per insured person have been increasing, but so have costs. Real expenditure per direct con- tributor increased 5 percent during the 1990s, while the real contributions per direct contributor increased 11 percent (World Bank 2003). The difference has been enough to finance the deficit generated by other groups like pensioners, vol- untary affiliates, and state beneficiaries. Several issues explain the dynamic behav- ior of contributions vis-à-vis the decelerated growth in expenditures: · The CCSS implemented specific plans to reduce evasion and defaults. · Waiting lists were used as cost-containment mechanisms. · Channeling funds to PHC activities resulted in highly cost-effective interven- tions that allowed the system to achieve good results and helped in part to con- trol expenditures. · The existence of the Office of the Controller General (outside the organiza- tional framework of the CCSS) is important to forestall irrational spending by health authorities. The Costa Rican reform experience has shown that complex reforms in the health sector are possible if technical and political factors are considered and clear objectives are established and agreed by all stakeholders. While there is clearly an unfinished agenda to meet the challenges of quality, efficiency and user access, extensive analysis and consensus building will need to take place to ensure that the reforms continue. 224 Good Practice in Health Financing Endnotes 1. Mostly pre- and postnatal care and screening activities. 2. The colon (C) is the Costa Rican currency. The average reference exchange rate for 2003 by the BCCR US$1 = C 398.6. 3. The Lorenz Curve is a graphic representation generally used to show inequalities in the distribution of a determined variable (e.g., income, social spending, public subsidies, assets). On the x-axis, the graph shows the cumulative share of households/families. The y-axis presents the cumulative proportion of the assessed variable. Perfect equality is repre- sented by a 45° line. If the final curve lies above 45°, the variable presents a progressive dis- tribution. If the curve is below the 45° line, the variable is regressive. 4. Diverse estimates exist, each with a specific methodology offering different results. For example, Kleysen (1992) estimates that the relative weight of the private sector is 20 percent, while Sáenz and León (1992) calculated it at 23 percent, and Durán and Herrero (2001) at 31 percent. 5. Each EBAIS has a basic structure of one physician, one nurse, and one primary care technician. 6. In Costa Rica, there are four public universities and more than 40 private universities. 7. According to data published by La Nación newspaper (June 27, 2005), current debt could cover almost all the health needs for one year, including, among other items, buying medicine for all members, building new facilities, and social benefits for specific groups such as senior citizens. 8. The big reduction between 2000 (75 percent coverage) and 2001 (43 percent cover- age) is partly explained by a purging of the CCSS databases. The Law for Worker Protection forced the CCSS to extend, for a five-year term, compulsory health insurance to indepen- dent workers, so only a fraction of the original group remained as voluntary affiliates. 9. The proportion of the population with poor access to health services went from 30 percent to 22 percent in the first four geographic areas (Buenos Aires, Pérez Zeledón, Golfito, Turrubares) where reform began in 1995 to 1996 (Rosero 2004). 10. The case is still pending in October 2007. References Amenábar, Ana Victoria C. 2005. "Industria Farmacéutica en Costa Rica: El negocio de los medicamentos." Revista Actualidad Económica 18 (December 2004­January 2005): 308­9. Arce, Claudio. 2001. "El desempeño hospitalario: entre luces y sombras." San José: CCSS (Unidad Coordinadora Proyecto de Reforma, Gerencia de Modernización y Desarrollo). Arce, Claudio, and Carlos Muñoz. 2001. La Reforma Pendiente: Introducción de Libre Elec- ción en el Seguro de Salud. San José: CCSS. Arce, Claudio, and Luis Sáenz. 2002. "Hallazgos preliminares, logros y desafíos de la reforma en el sector salud de Costa Rica durante los años noventa." Report prepared for Proyecto Estado de la Nación [State of the Nation Project], San José. Arias, Rodrigo. 2004. Evasión, costos y financiamiento del seguro de salud. San José: Departa- mento Actuarial, CCSS. Gauri, Varun, James Cercone, and Rodrigo Briceno. 2004."Separating Financing from Pro- vision: Evidence from 10 Years of Partnership with Health Cooperatives in Costa Rica." Health Policy and Planning 19(5): 292­301. Costa Rica: Good Practices in Expanding Health Care Coverage 225 CCSS (Caja Costarricense del Seguro Social). 2005. Anuario Estadístico 2004. www.ccss.sa.cr. ------. 2007. Políticas prácticas en materia de cobertura contributiva en CCSS. San José: CCSS. Castro, Carlos, and Luis Sáenz. 1998. "La Reforma del Sistema Nacional de Salud." San José: Colección Tiempos de Cambio, MIDEPLAN. CCP (Centro Centroamericano de Población). Información Censal de los Cantones de Costa Rica (INFOCENSOS). San Jose: Universidad de Costa Rica. http://infocensos. ccp.ucr.ac.cr. Duran, Fabio. 2005. "Hacia dónde va la Seguridad Social de Costa Rica?" In Ensayos en Honor a Víctor Hugo Céspedes Solano, ed. Grettel Lopez and Reinaldo Herrera. San Jose: Academia de Centroamérica. Duran, Fabio, and Fernando Herrero. 2001. El Sector Privado en el sistema de salud de Costa Rica. In Serie Financiamiento del Desarrollo, Proyecto CEPAL/GTZ"Reformas de los sis- temas de salud en América Latina," Economic Commission for Latin American and the Caribbean, Unit of Special Studies of the Executive Secretary, Santiago. García G., Rossana. 2004. Curso de Gestión Local de Salud para Técnicos del Primer Nivel. El Sistema Nacional de Salud de Costa Rica: Generalidades. San José: Centro de Desarrollo Estratégico e Información en Salud y Seguridad Social (CENDEISSS), CCSS. Gauri, Varum, James Cercone, and Rodrigo Briceno. 2004. "Separating Financing from Provision: Evidence from 10 Years of Partnership with Health Cooperatives in Costa Rica." Health Policy and Planning 19(5): 292­301. INEC (Instituto Nacional de Estadística y Censos). 2006. Survey on Revenues and Expendi- tures. San José: INEC. ------. 2005. Encuesta de Hogares de Propósitos Múltiples. San José: INEC. ------. 2001. Encuesta de Hogares de Propósitos Múltiples. San José: INEC. Kleysen, Brenda. 1992. Los Gastos Privados en Salud en Costa Rica. Working Paper No. 154, Instituto de Investigaciones Económicas, San José. MIDEPLAN (Ministerio de Planificación y Política Económica). "Sistema de Indicadores sobre Desarrollo Sostenido, Variables e Indicadores Sociales." San José: MIDEPLAN. http://www.mideplan.go.cr/sides/social/index.html. MOH-PAHO (Ministry of Health­Pan American Health Organization). 2003a. Gasto y Financiamiento de la Salud en Costa Rica: Situación Actual, Tendencias y Retos. San José: PAHO. ------. 2003b. Las desigualdades de salud en Costa Rica: una aproximación geográfico-- poblacional. San José: PAHO. MOH-PAHO-CCSS (Ministry of Health­Pan American Health Organization­Caja Costar- ricense del Seguro Social. 2004. Perfil del Sistema de Servicios de Salud en Costa Rica. http://www.cor.ops-oms.org. OPS (Organización Panamerican de la Salud). Iniciativa Regional de Datos Básicos en Salud. Sistema Generador de Tablas. http://www.paho.org/Spanish/SHA/coredata/tabulator/ newTabulator.htm. Proyecto Estado de la Nación [State of the Nation Project]. 2005. 11 Informe del Estado de la Nación, Anexo Estadístico. San José: UNDP. http://www.estadonacion.or.cr/Compendio/ ind_compendio.html. 226 Good Practice in Health Financing Rosero-Bixby, Luis 2004. "Supply and Access to Health Services in Costa Rica 2000: A GIS-Based Study." Social Science and Medicine 58: 1271­84. Rosero-Bixby, Luis. Assessing the impact of health sector reform in Costa Rica through a quasi-experimental study. Pan American Journal of Public Health, Volume 15, Number 2, February 2004, pp. 94­103 (10). Sáenz, Alberto. 2005. "The Role of Equity and Solidarity within Social Health Insurance: Chances and Risks of the Costa Rica Way towards Universal Coverage." Paper presented at the International Conference on Social Health Insurance in Developing Countries. http://www.shi-conference.de/contribut.html. Sáenz, Luis B., and Miriam León. 1992. "Gastos de los hogares en servicios de salud priva- dos en Costa Rica: 1987­1988." Acta Médica 35 (1) 1992: 30­38. Superintendencia de Pensiones de Costa Rica (SUPEN). 2006. Informe al Comité de Vigilan- cia del Réginen de Invalidez, Vejez y Muerte. San José: SUPEN. Trejos, Juan Diego. 2004. "La Equidad de la Inversión Social en el 2000." Paper presented at the Octavo Informe sobre el Estado de la Nación en Desarrollo Humano Sostenible, UNDP, San José. http://www.estadonacion.or.cr/info2002/nacion8/ ponencias.html. United Nations Population Division. 2004. World Population Prospects: The 2004 Revision Population Database. New York: United Nations. http://esa.un.org/unpp/index.asp. WHO (World Health Organization) 2004 Global Burden of Disease Estimates. Geneva: WHO, Department of Measurement and Health Information. http://www.who.int/ healthinfo/bodestimates/en/index.html. World Bank. 2006. World Development Indicators 2006. Washington, DC: World Bank. ------. 2003. Costa Rica: El Gasto Social y la Pobreza. Washington, DC: Unidad Sectorial de Desarrollo Humano Departamento de América Central, Latin America and Caribbean Región, World Bank. Relevant Legislation Act 17, Constitutive Law of the CCSS (1943) Decentralization Act No. 7852 (1998) Act 2,738 (1961), which established universal health insurance for workers and their families Act 5,349 (1973), which transferred to the CCSS all public hospitals previously controlled by the Ministry of Health Act 7,374 (1993), which split responsibilities between the MOH and the CCSS Law 7,852 (1998) on Decentralization Act of Worker Protection (2000), which declared all independent workers' obligation to join the health insurance program Executive Decree No. 14,313 SPPS-PLAN, which formally created the health sector in Costa Rica and regulated its structure and organization Executive Decree No. 19,276-S (1989), which created the National Health System and establishes the General System Bylaws, in which the Ministry of Health became the gov- erning institution 9 Estonia: "Good Practice" in Expanding Health Care Coverage Triin Habicht and Jarno Habicht Background Estonia is a small (45,227 sq. km) upper-middle-income country in the Europe and Central Asia (ECA) Region of the World Bank, with a population of 1.4 million and a GDP per capita of US$5,328 in 2005.1 Compared with similarly sized European coun- tries, Estonia's population density is low, 30 inhabitants per square kilometer (km2). Until regaining political independence in August 1991, Estonia's health system was that of the Soviet Union. Health care was centralized, universal, and provided nominally free of charge to everyone. The first Health Insurance Act was approved by the Parliament even before Estonia regained independence. The main rationale for health reform was to introduce a system that would ensure secure, sustainable financ- ing for the health sector. In the late 1980s and early 1990s, all Estonians were open to change and eager to move away from Soviet system to a market economy. However, it was understood that privatization and other market economy reforms would have to be balanced by a strong, sustainable social sector that included health insurance and pension systems. The success of the health financing reform has been facilitated by strong economic growth, which has enabled the continuous expansion of health insurance revenues and has supported the restructuring and scaling up of the health system. These factors have guaranteed stability and continuity for the longer period, which Estonia has used to build up an institutionally sound health insurance system. The health insurance system is mandatory. Contributions are related to being active in the workforce. Noncontributing individuals (children, pensioners) represent almost half (49 percent) of the insured population, and their expenses are implicitly subsidized by the others. State contributions cover only about 4 percent of the insured population. By the end of 2003, 94 percent of the population was covered. The uninsured 6 percent, 227 228 Good Practice in Health Financing working-age individuals outside the formal labor market, are ineligible for social insurance. Introduction Estonia began radical economic reforms in the late 1980s to restore a market economy modeled on the planned economy under Soviet Union. Economic Situation and Main Social Trends Monetary reform was a first step toward economic modernization. A national cur- rency (Estonian kroon, EEK) was introduced with a fixed exchange rate and a currency board arrangement, liberal external policies, and, most important, the immediate removal of most restrictions on capital account transactions. The main characteristics of Estonian reforms have been radical and rapid introduction of market-oriented institutions. Privatization of the economy has been completed, and only a few strategic companies (e.g., Estonian Energy Company) remain under the central government. In the early years of independence (1991­94), Estonia's GDP contracted, but the recession ended by 1995. The economy recovered quickly, and GDP has grown every year since then. In 2004, the first year of full membership in the European Union (EU), the Estonian economy grew by 7.8 percent--5.6 percentage points higher than the EU average. Rapid growth assists the convergence process, and Estonia is witnessing a remarkable convergence in real terms to EU levels. Pur- chasing power parity per capita GDP increased to 51 percent of the EU level in 2004 from 35 percent in 1995. GDP per capita in 2004 rose to 14,555 international dollars from 6,289 in 1995. Economic growth is expected to continue at around 6 percent, about three times faster than the EU average. However, vulnerability to external shocks is high because Estonia is small open economy, with a large, persistent current account deficit (US$1.4 billon in 2004) and rapidly expanding gross external debt (US$10 billion in 2004). The purchasing power of the population increased by 30 percent between 1992 and 2002, but this growth was not evenly distributed among different socioeco- nomic groups. Income inequality, as measured by the Gini coefficient, widened substantially in the 1990s (to 0.38), but began a steady decline, reaching 0.36 in 2004. Nonetheless, large differences persist in average monthly income per house- hold member among income groups: the income of the richest 20 percent of the population is 6.1 times larger than that of the poorest 20 percent. However, the proportion of people living below the national poverty line decreased tremen- dously from 36.1 percent 1997 to 17 percent in 2003. The groups at highest risk of poverty are the unemployed (especially the long-term unemployed), large fami- lies, and single-parent families. The poverty headcount ratio at $1 a day and $2 (international dollars) were respectively 2 and 8 percent of the population in 2003. Estonia: Good Practice in Expanding Health Care Coverage 229 Estonia follows a conservative fiscal policy; a yearly balanced budget policy guarantees a favorable and stable environment for economic development. The government is committed to long-term fiscal sustainability and considers its aging population a macroeconomic policy priority. This conservative approach is also pursued in the health insurance budget, so far kept in balance. The Estonian taxation system is simple and transparent, with few exceptions and differentiations. The Estonian flat-rate personal income tax system is one of the most liberal tax regimes in the world. The goal of taxation policy is to motivate entrepreneurship and initiative and to tax consumption rather than earnings. This has guaranteed economic growth, but some fundamental changes are planned because Estonian labor is heavily taxed compared with neighboring countries.2 The high tax burden on labor stems mainly from the social insurance tax paid by employers on behalf of employees, 33 percent of each employee's salary. Of this revenue, 13 percent is earmarked for health insurance, and 20 percent funds pen- sions for current retirees. A reduction in the personal income tax rate is scheduled, from 26 percent in 2004 to 18 percent by 2011 with annual decrease by one percent point. This reform supports the government policy of low taxation of earnings. Demographic and Epidemiological Situation Similarly to countries across Europe, the Estonian population is aging. Around 20 percent of the population was over 60 years of age in 2002, a group expected to grow to more than 25 percent by 2025 (figure 9.1). Among other similarities with Europe, aging urbanization has also increased. By 2005, 69.3 percent of the popu- lation lived in the cities. Since 1991, Estonia's population has been declining, due mainly to negative nat- ural growth and emigration, particularly to Russia. In 2005, Estonia had a crude birth rate of 10.7, a crude death rate of 12.9, and a fertility rate of 1.37, below the population maintenance rate. Hence, the overall population natural growth rate is negative, ­3.8. The dependency ratio (the proportion of population aged 0 to 14 or over 65 years) is 48 percent, below than average EU rate. However, this is likely to increase in the future as the population continues to age. These demographic trends will have serious consequences for Estonia's health sector because of the reduced tax base. Health expenditures are financed mainly through labor-related taxes (65 percent), and children and the elderly are insured, but no contributions are made on behalf of these groups. The epidemiological transition, as the population ages, also puts pressure on the health budget with the growing burden of chronic illnesses that are costly to treat. The average age of mothers giving birth, including the average age at first birth, has continuously risen since 1995 from 23 years to 24.8 years in 2003.3 Birth activ- ity among women under 25 has also continuously decreased. Evidence of positive changes in family planning habits comes from the steady and continuing decrease in the absolute number of induced abortions and the number of abortions per 230 Good Practice in Health Financing Figure 9.1 Estonia: Population Pyramids, 2004 and 2025 a. 2004 b. 2025 80+ 80+ 75­79 75­79 70­74 male female 70­74 male female 65­69 65­69 60­64 60­64 55­59 55­59 50­54 50­54 45­49 45­49 age 40­44 age 40­44 35­39 35­39 30­34 30­34 25­29 25­29 20­24 20­24 15­19 15­19 10­14 10­14 5­9 5­9 0­4 0­4 6 4 2 0 2 4 6 6 4 2 0 2 4 6 % of population % of population Source: Eurostat 2006 (http://europa.eu.int/comm/eurostat/). 1,000 women between the ages of 15 and 49, from 50.0 in 1995 to 31.0 in 2003.4 In 2000, the number of induced abortions fell below the number of live births for the first time. The number of deaths has been stable, only slightly crossing the 18,000 mar- gins. More than half the deaths are caused by circulatory system diseases; cancer and accidents, poisonings, and acute injuries are the next most important causes. The causes of death differ by gender. In 2003, 63 percent of women's deaths were caused by diseases of the circulatory system (for men 47 percent), 17 percent by neoplasm (for men 20 percent), and less than 5 percent by accidents, poisonings, and acute injuries (for men 16 percent). The mortality and morbidity patterns are similar to those of East European countries. According to the latest burden of disease study in Estonia (Ministry of Social Affairs and University of Tartu 2004) the main causes are cardiovascular diseases, neoplasm, and external causes. Still, due to the burden from disability or decrease in health status other disease groups rank high: joint and muscle dis- eases, pulmonary diseases, and mental health disorders (figure 9.2). More than half the burden of disease is concentrated in the working age population, among men 58 percent. Less of the burden is among over 65-year-olds (39 percent) and young people aged 0 to 19 years (8 percent). There is clear difference among men and women: only 2 of the top 10 causes are same, and among men external causes are much more prevalent, 11 percent of the total disease burden. Estonia: Good Practice in Expanding Health Care Coverage 231 Figure 9.2 Estonia: Burden of Disease cardiovascular diseases 6.9 26.2 neoplasms 8.0 11.7 external causes 1.6 10.3 others 5.0 2.5 joint and muscle diseases 6.3 0.3 pulmonary diseases 3.7 2.1 psychiatric diseases 3.6 0.5 neurological diseases 2.8 0.8 0.8 diseases of digestive system 2.6 gentio-urinary diseases 1.2 0.4 0.9 malformations 0.6 0.5 infectious diseases 0.6 0 5 10 15 20 25 30 35 total burden of disease, % years lived with disability years lof life lost (mortality) Source: Ministry of Social Affairs 2004. Average life expectancy decreased in the first half of 1990s to a low turning point in 1994, 61 for men and 73 for women. This was the result of radical changes following independence, expressed in increases in mortality from external causes. Since 1995, life expectancy has slowly increased, to 66 for men and 77 for women in 2003, but remains among the lowest among EU countries (figure 9.3). Over the years the infant mortality rate has fallen from 14.8 in 1995 to 5.7 in 2002. The maternal mortality rate of 23 per 100,000 live births (7.7 in 2002) is very low. The five main risk factors creating the disease burden are smoking (8.3 per- cent), low physical activity (7.4 percent), high alcohol consumption (6.6 percent), overweight (5.1 percent), and low fruit and vegetable intake (3.8 percent). The smoking and alcohol consumption risk factors are concentrated among men, together causing more than 12 percent of the disease burden. Smoking alone causes 40 percent of cardiovascular diseases and 40 percent of cancers in Estonia. Political Environment Estonia is a parliamentary republic. The Parliament (Riigikogu) consists of one chamber with 101 members elected to four-year terms. Since independent Esto- nia's first elections, in 1992, all governments have been coalitions of two or three 232 Good Practice in Health Financing Figure 9.3 Estonia: Average Life Expectancy at Birth Compared with EU Countries, 2003 85 82.0 80 78.4 77.2 76.1 75 age 70.1 70 66.2 65 60 Estonia EU members EU members since before May 2004 May 2004 female male Source: Health for All Database 2006 (http://www.euro.who.int). political parties. Although none of the coalitions have governed for a full term, they have been stable enough to launch and implement economic and social reforms. Estonian political parties are at the center or to the right of the political spec- trum. To date, governments have been on the right, although social democratic values and ideology have become more visible in recent years. In the early 1990s, health and social sector policies were driven by social values. The coalition that was in government between 2004 and 2007 has agreed to lower the proportional rate of personal income tax and to make a one-shot increase in state pensions. In health policy, the coalition agreement enables wide interpretation and is not clearly formulated. However, the priorities are prevention of HIV5 and drug dependency, tackling the risk factors related to noncommunicable diseases (NCDs), remuneration policies (health services and pharmaceuticals), and improv- ing access to care. The second political layer in Estonia consists of 227 municipalities, which have budgetary autonomy and local tax-raising powers. Administratively, however, Estonia is divided into 15 counties, each run by a governor and administered by "the county government." Both the governor and the county government staff members are civil servants under the central administration. Many state agencies, including those engaged in health care administration and finance, operate not on a county basis, but through regional departments covering two to four counties. The role of the counties and municipalities in the health sector has varied. During the decentralization era in the early 1990s, a number of responsibilities were dis- Estonia: Good Practice in Expanding Health Care Coverage 233 tributed to counties, as was the fragmented health insurance system. Counties, with specially recruited staff members, were responsible for planning and over- sight of health care services. In the late 1990s, during centralization of planning and development, the role of planning primary health care and liaising with municipalities on behalf of government remained with the counties. The munici- palities also have some role in the health sector, running local public health activ- ities and supporting the health care sector (as well as owning some of the hospitals), but the practice varies across Estonia. Overview of Health Financing and Coverage This section describes the Estonian health financing system: the collection of funds, pooling, benefits package, cost sharing, and purchasing of health care ser- vices. Finally, the performance of the current system is described as well as its impact on efficiency and equity. Collection of Funds In 2004, total health expenditures were 5.1 percent of the Estonian gross domestic product (GDP).Among EU countries, this share ranges from similar to over 10 per- cent and averages 6.4 percent. The level of health expenditures in Estonia has been stable over time with small variations due to changes in the economic environment. In recent years, the health expenditure share of GDP has increased slightly because a good economic environment ensured stable tax collections and favored increases in out-of-pocket payments. The annual increase in total health expenditures in absolute terms was around 4 percent in the late 1990s and about 14 percent6 more recently. In 2004, total health expenditures amounted to US$618 million. The share of public expenditures in total health expenditures is relatively high, 76 percent, and 4.2 percent of GDP in 2004. In the total government budget, public health expenditures have been around 11 percent. Most public funding comes through the health insurance system--87 percent of health expenditure in the public sector and 66 percent of total health expenditure in 2004 (table 9.1). Table 9.1 Estonia: Total Health Expenditure, by Source, 2004 Expenditure Source (US$ millions) Share (%) Central government 52 8.5 Municipalities 8 1.3 Social insurance 406 65.7 Out of pocket 132 21.3 Other private 17 2.7 External sources 3 0.4 Total 618 100.0 Source: Ministry of Social Affairs 2006 (www.sm.ee). 234 Good Practice in Health Financing Figure 9.4 Estonia: Distribution of Funding Sources, 1999 and 2004 3.5 0.5 5.6 2.7 14.0 21.3 (%) funding of 66.0 65.7 composition 2.2 1.3 8.7 8.5 1999 2004 central government municipalities health insurance fund private household out-of-pocket other private sector external sources Source: Ministry of Social Affairs 2006. When the health insurance system was introduced in 1992, municipal and central government financing was expected to grow in line with health insurance fund- ing. Instead, their share has been decreasing, and the role of financing through health insurance is increasing (figure 9.4). The main source of health insurance revenues (98 to 99 percent) is a social health insurance contribution paid by salaried workers and self-employed people. This is also known as a"social tax," which covers health and pension contributions, respectively 13 percent and 20 percent of employee wages and of self-employed individuals' earnings. Health insurance contributions, related to the active workforce, cover about half the people. The health expenses of noncontributing individuals (49 percent of the insured population) are implicitly subsidized by the other categories, which shows great solidarity within the system. These noncontributing groups, includ- ing children, pensioners, disability pensioners, and students (during their nomi- nal study time) are eligible for the same benefits package as everyone else in the insurance pool without any contribution from either themselves or the state. The state officially makes contributions for only a small proportion of the covered population (4 percent), including individuals on parental leave with children under three years of age, individuals registered as unemployed (eligible for up to nine months' coverage), and caregivers of disabled people. The state's contribu- tion for these groups is defined annually when the state budget is approved and depends upon the number of eligible persons. However, the contribution rate is Estonia: Good Practice in Expanding Health Care Coverage 235 low, 15 times lower than the average contribution made by workers and the self- employed. In 2005, an increase the per person contribution rate is planned to make it equal to the health insurance tax paid on minimum salary. This cuts the difference in contribution rates from 15 to 2.5. Under Estonia's system, solidarity is high; half of the insured population contributes, and every insured person is entitled to the same benefits (cross-subsidization within the pool). In 2004, central government funding accounted for 8.5 percent of total expen- ditures, and local municipalities accounted for 1.3 percent of total health expendi- tures. The central government finances ambulance services, emergency health care services7 for uninsured people, and public health programs (e.g., HIV/AIDS; tuberculosis prevention; cardiovascular disease prevention). Local municipalities have no clear responsibility for covering health care expenditures and therefore financing practices vary widely.8 Municipalities mainly cover health expenditures of the uninsured (some extra care in addition to central government­covered emergency care), and support for travel to health care facilities, for selected public health programs, and for out-of-pocket payments by people facing high health expenditures. Other municipalities cover some costs of family physician (FP) ser- vices in their region, but in-kind contributions for FPs are also common (e.g., office space to set up practices). Municipalities that own hospitals pay for some of their capital costs. Out-of-pocket payments comprise statutory cost-sharing for Estonian Health Insurance Fund (EHIF) benefits, direct payments to providers for services outside the EHIF benefits package or from non-EHIF-contracted providers and informal payments. Out-of-pocket payments constituted about 21 percent of total health expenditures in 2004. Most out-of-pocket payments go for pharmaceuticals (53 percent) and dental care (23 percent). According to the National Health Accounts, households paid out of pocket for 45 percent of total pharmaceutical expenditure and 61 percent of dental care in 2004. Both shares have been increasing due to introduction of new drug reimbursement rules (introduction of reference prices) and exclusion of adult dental care from the EHIF benefits package. The share of out-of-pocket payments for total health expenditures has been increasing in Esto- nia (figure 9.4) and is expected to increase further. Still, the share of out-of-pocket payments is relatively low (21.3 percent) compared with similar countries in East- ern Europe (e.g., 45.9 percent in Latvia and 24.2 percent in Lithuania), but higher than in most of EU member countries, but about the same as for the EU as a whole after its May 2004 expansion (figure 9.5). Other private expenditure includes employer-paid health care travel insurance, employer-paid health checkups and pharmaceuticals (bought mainly by foreign visitors but also by corporations). However, the share of these expenditures is low, less than 3 percent of total health expenditures. Informal payments are not com- mon in Estonia and are not an important source of out-of-pocket payments, according to recent studies (Josing 2004; CIET 2002).This may be due to the intro- duction of formal copayments in 2002 or to the generally low level of corruption and informal payment practices. 236 Good Practice in Health Financing Figure 9.5 Estonia: Out-of-Pocket Payments Compared with the EU, 2004 25 22.8 21.3 20 15.5 payments expenditures) 15 health 10 total out-of-pocket of (% 5 0 Estonia EU members before EU members since May 2004 May 2004 Source: Health for All Database 2006. Private insurance funds plays little role in health care in Estonia. Until recently, the private insurance market consisted mostly of employer-paid health care travel insurance. Since 2002, one private insurer has offered health insurance, but it is not very popular (just a few hundred enrollees) because the benefits package is limited. In addition, the EHIF scheme enables people who otherwise would remain unin- sured to enroll in the public health insurance scheme. Eligibility for voluntary cov- erage by public scheme is restricted to Estonian residents who receive a pension from another country or people not currently eligible for membership, but who were members for at least 12 months in the two years before applying for voluntary membership. Voluntary members pay a contribution based on the previous year's average national salary and benefit from the same coverage as other insured people. These contributions are pooled with the funds covering the other insured people. Allocation of Funds and Purchasing Services The schematic overview of the organization of health financing arrangements in Estonia is given in figure 9.6. The core purchaser of health care services is EHIF, which purchases most of the care for insured persons (94 percent of total popula- tion), except ambulance service. Ambulance service is financed directly from the state budget and administered by the Health Care Board, an agency of the Min- istry of Social Affairs. Emergency care costs for the uninsured are covered from the state budget, but the administrative tasks are delegated to the EHIF, in the interests of efficiency, because the same payment methods and tariffs are used for uninsured and insured persons. Estonia: Good Practice in Expanding Health Care Coverage 237 Figure 9.6 Estonia: Overview of the Health Financing System, 2004 Ministry of Social Affairs 8.5% and its agencies 1.3% municipalities tax Estonian Health Insurance Fund (EHIF) taxes taxes social 65.7% EHIF regional branches general general earmarked 21.3% hospitalsa ambulatory specialist population careb employers primary carec copayments patients pharmacies public health programmes ambulance servicesd Source: Adapted by the authors from Jesse et al. 2004. Note: The bold line represents main sources of revenue. a. Fee-for-service plus daily rate plus some per case payments. Fifty percent of each case is reimbursed using DRG prices; contracts are close-ended case-volume contracts. b. Fee-for-service; close-ended case-volume contracts. c. Weighted capitation plus fee-for-service plus additional fixed payments. d. General budgets and fixed payment per provider unit. The EHIF funds are collected centrally to balance regional disparities in income. Pharmaceutical and temporary sick leave benefits (open-ended obliga- tions for the EHIF) are administered centrally, but most health care services funds are allocated to EHIF regional departments (four in 2006) by a crude capitation method, unadjusted for need. For primary health care, however, a more sophisti- cated formula is used.9 Overall, 98 percent of the EHIF health care services funds 238 Good Practice in Health Financing Figure 9.7 Estonia: The EHIF Contracting Process pooling capitation- needs selection of annual capped contract in the based assessment partners and cost and volume performance EHIF allocations in each negotiations contracts and for monitoring to EHIF region about contract strategic hospitals and utilization regions volumes a five-year review (regional basis) framework contract standard contract conditions negotiated and adjustments agreed between EHIF and of contract Estonian Hospital Union or volume according Estonian Society of to utilization Family Doctors Source: Couffinhal and Habicht 2005. are allocated to the regional branches. The rest remain centrally managed for a few expensive or infrequent procedures for which regional allocation would not be feasible. These include bone marrow transplants, peritoneal dialysis, some areas of oncology, and hematological treatment. The allocation of funds is further refined during the contracting process. The EHIF contracts with providers (not individual doctors) for all types of care: pri- mary care, specialist outpatient and inpatient care (except ambulance care and emergency care for the uninsured), and all other services in the benefits package. The EHIF contracting process is shown in figure 9.7. Since 2001, the EHIF has gradually introduced the need assessment as an input for purchasing decisions. Currently, these decisions are based mainly on historical data analysis about health care services utilization and existing queues or waiting times, but there are further plans to use more sophisticated methods. In 2005, the EHIF started to negotiate with medical specialist associations10 about their assess- ment of population needs in terms of their specialty to obtain detailed input from specialists and involve them in the planning process. The results of these negotia- tions are taken into account in EHIF budgeting and contracting processes and benefits package updates. The aim is to broaden the role of the medical profession in long-term direction setting for purchasing decisions. The next step in the contracting process is selection of prospective providers. Contracts are made only with providers licensed to work in Estonia by the Health Care Board. The EHIF practices selective contracting (choosing not to contract with certain providers). Selective contracting is intended to motivate service quality improvement and to introduce mild market competition into health care provision, but also to buy services for delivery in areas providers perceive as less attractive.11 Estonia: Good Practice in Expanding Health Care Coverage 239 Nevertheless, the EHIF is required to contract with all Hospital Master Plan hospi- tals (fewer than 20 strategic acute care hospitals), which have a historically deter- mined guaranteed contract volume. The main exception is dental care, which does not systematically contract with the EHIF and provides services privately. Also, 20 percent of outpatient care is purchased using selective contracting, where the EHIF announces public procurements and all providers can submit proposals. The EHIF negotiates the standard contract conditions yearly with providers representing associations such as the Society of Family Physicians and the Hospi- tal Association. This ensures that once the EHIF and the provider associations agree on the overall contract terms, they are universal, and all provider have to accept them. After selection of the providers and agreement on the contract, negotiations continue with selected providers of specialist care to determine the service volume (number of cases) and average price per case by specialty. These negotiations do not determine the actual payment method but constitute a planning element aimed at containing costs for each case. In terms of coverage, the agreement on the number of cases is more important. This supports the implementation of the EHIF objective of keeping access to care at least at the same level as in previous year (measured as overall cases per provider). As a result of these negotiations, contract volumes are agreed with each provider. Until 2005, a strict system was applied: if a provider exceeded the agreed volume during the contract period, the EHIF would not pay. This principle was followed to ensure EHIF solvency. Since 2006, additional cases can be covered, but then EHIF pays at reduced rates. In addition if additional revenues materialize or the projected number of cases has to be changed during the year, the contracts can be renegotiated. To keep abreast of real access, the EHIF monitors contract implementation by providers monthly via the Management Information System. That process applies only for specialist care because, for primary care, the contract volume is not negotiable but is determined in accordance with the payment rules. As mentioned, the actual payment methods, service prices, and benefits pack- age are not determined during the contract negation process (in Estonia they are viewed together in one "health services list," which determines all three). All pay- ment methods used by the EHIF are regulated by government-approved health care services list. Payment method revisions follow the same process as price revi- sions. Inpatient and outpatient providers of specialized care are paid using a range of payment methods that depend on the type of services provided: fee-for-service, visit fees, per diem, DRG-based, and case-based complex prices. During the late 1990s, a decision was made to move away from detailed fee-for-service payments and their perverse incentives and to increase case-based payments. This was moti- vated by the fact that the average length of stay was well above the proclaimed objective of 4.5 days for acute care cases. In addition, the bed occupancy rate in acute care hospitals decreased from 80 percent to 65 percent during the 1990s, which indicated that hospital capacity was being used inefficiently. Gradual 240 Good Practice in Health Financing Figure 9.8 Estonia: Payment Methods for Inpatient and Outpatient Specialist Care, 2005, actual share from budget 90 83% 80 70 60 50 expenditures 40 35% 31% total 30 24% of % 20 10% 10% 10 6% 0% 0 fee-for-service case-based per diem DRG (old) outpatient inpatient Source: EHIF, personal communication, July 2006. implementation of the diagnosis-related group (DRG) system began in April 2004. The proportion of DRG payment for each case was initially set at a low 10 percent to minimize possible risks under the new system. In 2005, the share was raised to 50 percent. However, several types of payment methods are used at the same time, with the intention of balancing different incentives (figure 9.8). FPs are paid through a combination of age-adjusted capitation and other types of payment (fee-for-service, lump sum to cover fixed costs) that make up the prac- tice budget (figure 9.9). For capitation payments, FP practices receive monthly pre- payments, recalculated twice a year to reflect changes in the patient list. An additional payment instrument was introduced in 2006, a quality bonus system in which targets are set at practice level to improve overall health care quality12 and promote good performers. Health care service prices are identical for all providers, and there are no adjustments for regional or hospital characteristics such as complexity of services provided or teaching status. In principle, prices cover all costs related to deliver- ing services except scientific and teaching activities, which are funded separately. All prices approved are maximum prices, and providers and the EHIF can agree on lower prices for the contracts with individual providers. Revision of service prices and payment methods can be initiated by provider or specialist associa- tions or by the EHIF. Each service should be evaluated in terms of four criteria: medical efficacy (evaluated by the relevant medical specialist association), cost- effectiveness (evaluated by health economists), appropriateness and compliance with the national health policy (evaluated by the Ministry of Social Affairs), and Estonia: Good Practice in Expanding Health Care Coverage 241 Figure 9.9 Estonia: Payment Methods for Family Physicians, 2005, as share of total budget 24/7 GP line other lump sum 0% 2% 11% fee-for-service 14% capitation (age-adjusted) 73% Source: EHIF 2006 (http://www.haigekassa.ee). the availability of financial resources (evaluated by the EHIF). The service list is usually updated annually. Benefits Package and Patient Copayments In Estonia health insurance coverage is mandatory with no opt-out possibility and intended for the whole population. Entitlement to health insurance coverage is based on residence in Estonia and membership in main categories defined by law. In 2004, mandatory health insurance covered 94 percent of the population in four main categories: persons who pay their own contributions as employees or self- employed persons eligible for coverage without contributing, in particular chil- dren and pensioners; persons covered by contributions from the state; and persons covered by voluntary agreements. Some minor alterations in entitlement rules have not caused changes in overall coverage. The EHIF provides two kinds of benefits for the insured: cash benefits (about 20 percent of total EHIF expenditures) and benefits in kind (rest of the budget, health care services paid by the EHIF). The cash benefits (reimbursement for ser- vices) include the costs of dental care for adults and some reimbursement for high pharmaceutical expenditures.13 In addition, the EHIF pays for temporary health- related incapacity for work in different forms as sickness benefits, maternity ben- efits, adoption allowance, and care allowance.14 Since the end of 2002, voluntary EHIF coverage has been extended to persons who might otherwise remain uninsured. Voluntary members pay a contribution of 13 percent of the previous year's national average salary as published by the Sta- tistical Office, and they are entitled to the same benefits as compulsory members. However, the number of voluntary members is still low. 242 Good Practice in Health Financing The uninsured (6 percent of the total population) are working-age individuals who are not employed in the formal labor market and are not eligible under other criteria to register as unemployed or disabled. The EHIF does not reimburse the uninsured for any kind of health care cost. The uninsured have the right to emer- gency care, which is reimbursed through the state budget but administered by the EHIF, which has the lowest transaction costs. Local municipalities used to cover emergency care costs of the uninsured, but this led a situation in which access to treatment was different by municipality. Discussions are ongoing about whether local municipalities should start to cover primary care and essential pharmaceuti- cal costs of the uninsured. Another option on the table is to allow municipalities to buy EHIF coverage for selected vulnerable groups in their constituencies. The EHIF covers a broad range of health care benefits (in-kind benefits for patients) such as family physician services, in- and outpatient specialist care, long-term care, rehabilitation, dental care for children, and prescription drugs. This feature is in part inherited from the old system in which the state funded and provided universal, comprehensive health care. Some health care services are excluded from the benefits package (for example, cosmetic surgery, alternative therapies, and opticians' services). An attempt has been made to introduce explicit rules for adding new services to the benefits package (see above) and for establish- ing the appropriate level of user charges. In addition to health care services, the EHIF finances health prevention and promotion programs. The EHIF administers its special health promotion fund, from which public health institutions, nongovernmental organizations (NGOs), and local activists can apply for funds for one-year projects on preset priority areas. At the same time targeted public procurements are organized in key areas (such as information sharing, guidelines for population). Disease prevention pro- grams funded include school health, reproductive health, screenings (breast can- cer, phenylketonuria, and hypothyreosis, cervical cancer, hearing in newborns), specific immunizations, and other activities. This has broadened the scope of ser- vices supported by the EHIF from purely curative to preventive services. Since 2007, the health promotion and disease prevention activities are coordinated and advised by merged commission. Different dental care benefits for children and adults are included in the bene- fits package. Since 2002, the EHIF has guaranteed free dental care for children and adolescents until the age of 19, including preventive and curative services. Adults must pay for dental care out of pocket but may receive partial reimbursement by the EHIF (US$11.51 per year for most people). The reimbursement rate is higher for some population groups with greater needs, such as pregnant women, moth- ers in the first year after childbirth, and people suffering from certain diseases that affect their teeth). The pharmaceuticals covered by the EHIF are defined on a positive list15 of medicines requiring a prescription from a physician. Guidelines for adding new pharmaceuticals to the positive list have been clarified, and reference prices have Estonia: Good Practice in Expanding Health Care Coverage 243 been introduced. In addition, efforts have been made to put more generics on the market and promote their utilization. This has been done not only to contain insurance costs, but also to promote cheaper but equally effective medicines for patients, still with a mixed outcome. Medical devices for certain diseases are also included as in-kind benefits and are subject to a coinsurance rate of 90 percent up to a yearly ceiling of US$1,534. Prescription drugs are generally subject to a US$3.82 deductible, and a further coinsurance percentage usually applies (50 per- cent up to US$15.41). Beyond this ceiling, the user covers all costs. Additional measures attempt to limit the burden for some categories of patients: a positive list of drugs for chronic conditions16 with 75 percent and 100 percent coinsurance rates and a lower deductible of US$1.59. Additional exemptions apply for chil- dren and retired people. Flat copayments are charged for some types of health care services--primary care physician home visits, outpatient care visits, and hospital bed-days. The gen- eral principle has been to move to free primary care without any patient cost shar- ing. Since 2002, the only copayment that FPs are allowed to charge is a US$3.76 home visit fee. Earlier, a small visit fee (US$0.61) was applied with some exemp- tions such as children and retired or disabled persons. For outpatient specialist care, the consultation fee is US$3.82, and this is applied to office visits as well. These fees are defined as maximum, but the provider can decide the actual amount between zero and US$3.82. Providers are not allowed to charge a visit fee for children younger than two years or for pregnant women. Inpatient care providers can charge patients a per diem rate for up to 10 days with a limit of US$2.21 per day. Care for children, pregnancy or delivery-related condi- tions, and emergency care are exempt. These rules apply to care for which the EHIF reimburses contracted providers. If a patient covers the full cost (as do uninsured patients), providers are allowed to set different user charges to cover treatment costs as long as these charges are "reasonable." Before 2002, there were no such explicit rules, and different providers applied different user charges even when the EHIF was reimbursing the costs. The health care service list approved in government also sets a maximum ceil- ing for coinsurance by service type, which the Health Insurance Act says can be up to 50 percent. Most services have no coinsurance except for in-vitro fertilization (30 percent coinsurance), abortion without medical indication (30 percent), and rehabilitation per diem for some illnesses (20 percent). Efficiency and Equity Performance of the Health Financing System In 2004, Estonia spent a relatively moderate 5.1 percent of GDP on health, yet the public share of total health expenditures is high, 76 percent. This allows good financial protection for most of the population. However, out-of pocket payments rose from 13.7 percent in 1998 to 21.3 percent in 2004, and financial protection is increasingly worrisome. In 1995, 0.3 percent of households had health expendi- tures higher than 40 percent of their budget after buying their food. In 2002, 244 Good Practice in Health Financing 1.6 percent of households faced health expenditures of that magnitude. (Habicht et al. 2006). The main objective of health financing system is to improve the people's health. The present health financing organization has been in place for 15 years its long- term objectives are ambitious. Therefore, intermediate objectives such as improv- ing service quality and access to care are more informative indicators of health financing performance. Access to care in terms of waiting times and number of treatment cases has been the most explicit and important objective of the EHIF. The EHIF is legally obliged to balance yearly revenues and expenditures and so far has done so except in 1999 due to economic crisis.17 This good performance is partly due to freedom to cut provider contract volumes to operate within dimin- ished revenues. But a more important reason has been the rapidly increasing health insurance tax revenues (allowing increases in contract volumes and service prices). The Health Insurance Fund Act also requires the EHIF to maintain adequate reserves to minimize potential financial risks. The EHIF has to have two types of long-term reserves: legal reserves of 6 percent of its budget and risk reserves of 2 percent. The accumulation of reserves began in 2002, and the reserve requirements were fulfilled in 2003. In terms of income inequalities, Estonia is a country where income is very unequally distributed. In 2004, the average income in the lowest decile was one- tenth of that in the highest decile. Health expenditures follow same pattern, and the low out-of-pocket spending by the poorest individuals probably reflects finan- cial barriers in access to care. For instance, the usual copayment for a specialist visit (EEK 50, about US$4), represents more than 6 percent of the monthly income in the lowest decile.18 This assumption is supported by empirical research correlating health care utilization with socioeconomic status (Habicht and Kunst 2005). Controlling for health status, lower-income individuals utilize fewer FP and specialist services, as well as dental care. It is expected that pharmaceutical use follows the same pattern. Estonian health insurance coverage is mandatory and targeted to the whole population. Still, about 6 percent of population remains uninsured. Because Estonian health insurance is closely related to labor market, the share of insured is dependent on the employment level. The share of uninsured decreased from 6.6 percent in 2000 to 5.5 percent in 2004, a reflection of increased labor market par- ticipation. Over the same time period, the unemployment rate decreased from 13.8 percent to 9.9 percent. The uninsured are among the working age population because children and pensioners are all covered. According to the Household Bud- get Survey 2003 (Statistical Office of Estonia 2004), the uninsured are mainly men (6 percent, 3 percent among women) in the 35 to 54-year-old age group. By socio- economic background, the uninsured are predominantly the less educated. They live in rural areas and poorer municipalities, where the average income is below the country's average. Estonia: Good Practice in Expanding Health Care Coverage 245 Overview of the Health Care Delivery System The health care delivery system can be divided into primary care and specialist care. The natural parts of the primary care are a well-developed family medicine system, ambulance services, and prescription medicines. Specialist outpatient and inpatient care, mostly at hospitals, is described below. Nursing homes, rehabilita- tion services, and long-term care are in the early stages of development in Estonia and are not discussed here. Primary Care--Family Medicine The introduction of primary care started in 1991 with the redesign of the Soviet- type polyclinic model for delivering primary care. Polyclinics were staffed by working medical school graduates with no specialized training and by pediatri- cians and other specialists such as cardiologists. The first step in primary care reform was to introduce family medicine as a medical specialty and to start post- graduate training for it in the early 1990s. Now there are enough family physicians to cover the whole population (figure 9.10). Since 1997, all citizens have been required to register with a family physician patient list, and family physicians have been entitled to become independent service providers. For doctors (as practice leaders), a motivation was the capitation fee, defined by the size of the registered list. Any patient who did not register with an FP was assigned to one. In 1996, only 8 percent of the population used a family doctor (EMOR 1996); by 2003, everyone did. The regulatory framework of 2002 stipulates primary care as the point of Figure 9.10 Estonia: Number of Family Physicians, 1993­2004 996 1,000 907 900 864 800 700 676 628 600 500 462 number 404 400 290 300 200 146 162 100 63 63 0 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Source: Ministry of Social Affairs 2006. 246 Good Practice in Health Financing entry to the health system for most conditions.19 Direct access is maintained for the list of outpatient specialists, for example, gynecologists. All family physicians are required to work with at least one family nurse in their practice to be contracted by the EHIF. However, there is a shortage of family nurses (fewer than 700 in 2004), who have better employment prospects abroad or in other industries (Atun 2004). This has put some limitations on further development of family medicine in Estonia, although special training programs for family nurses have been launched. Family medicine includes a range of ser- vices such as diagnosis, laboratory tests, investigations, treatment, follow-up, and also prevention and promotion activities. A further development in 2004 was the introduction of a family medicine call center funded by the EHIF to provide 24/7 access to advice from a family doctor. This line is free of extra charges for the whole population; and only the regular phone call fee is applied. In the first quar- ter of 2006, 24/7 took more than 36,000 calls, and use of this family medicine direct line service has been stable. Family physicians can operate as sole proprietors or as companies (general part- nerships or limited partnerships) with a list of registered patients. Despite the pri- vate practice option, the majority of family physicians have a contract with the EHIF.20 As mentioned, each family physician has a list of registered patients, which cannot contain fewer than 1,200 or more than 2,000 patients.21 The average prac- tice size was around 1,600 persons in 2005. The practices are organized around one family doctor with a nurse, but group practices have been springing up for effi- ciency gains in resource use (both equipment and human). The patient is free to choose and change a family physician. In 2005, only 13 percent of the population changed doctors and the main reason was a change in their place of residence. The family physician can refuse to register a person only if the practice list has reached the maximum size or if the applicant does not perma- nently reside in the practitioner's service area. Family physicians are required to have at least 20 visiting hours a week, and practices are to be open at least 8 hours a day.22 Patients should be able to see their family physician within one day for acute problems and within three days for care of chronic conditions. In actuality, around half the patients requesting an appointment can see the family doctor the same day (regardless of the condition), and the majority make the visit within three days. Good access and the services available at the family medicine level has increased patient satisfaction with the family medicine system. In 2005, about 90 percent of the population was satisfied with the family medicine system, as com- pared with 80 percent in 1999.23 The system for reimbursing family physicians is designed to provide incentives for giving preventive care, as well as for taking more responsibility for diagnostic services and treatment. Therefore, a mix of payment methods is used (capitation, fee for service, lump sum) to balance the financial risk associated with different incentives. Management of key chronic conditions at primary care level has improved in recent years (Atun et al. 2006). Estonia: Good Practice in Expanding Health Care Coverage 247 Primary Care--Ambulance Services The ambulance services organization in Estonia was inherited from the previous system with strong emphasis on prehospital care, as in many East European coun- tries. This is the mixed model, where the ambulance has dual roles: to diagnose and provide onsite treatment and to transport the patient to the hospital. Forty percent of the ambulance teams are led by a doctor and other 60 percent, by a spe- cialized nurse. Thus, the service can often be provided on site, and transportation to the hospital is not needed. As provider organizations, 80 percent of ambulance service providers belong to other institutions (e.g., hospitals), and 20 percent spe- cialize in emergency care. The ambulance services have gone through several stages in terms of financing schemes and services. From the early 1990s through 1996, ambulance services were financed by regional sickness funds with some additional transfers from the state budget. Insurance coverage defined entitlement to the services, and the provider billed for them. In 1997, financing was centralized in the Central Sickness Fund, but coverage was still governed by insurance status. However, centralization removed regional differences in service financing practices. In addition, the incen- tive to provide services changed because the payment for each visit was linked to the preparedness to act of the emergency care teams, headed by either a doctor or a specialized nurse. The general budget is used, and its size depends on the number of teams and their composition. Since 1998, ambulance services have been financed from the state budget (man- aged by the Ministry of Social Affairs). This ensured that everyone in Estonia (citi- zens as well as temporary residents) are entitled to the services. Previously, access to ambulances was related to insurance status. Although the number of teams (90 in 2005) and cars (126 in 2005) has decreased slightly, service availability has not. In the late 1990s, 230,000 calls for ambulances were answered, 240,000 in 2005. In 2002, responsibility for purchasing and monitoring ambulance services was shifted from the ministry to the newly created Health Care Board under ministry supervi- sion. Ambulance services and family medicine thus, to some extent, balance and fill in for each other to provide comprehensive access to services. Outpatient and Hospital Specialist Care Since the beginning of the 1990s, Estonian specialist health care has undergone broad reforms including centralization of highly specialized services and decen- tralization of outpatient specialist services. Currently, outpatient specialist care is provided by health centers, hospital outpatient departments, and specialists prac- ticing independently. Both public and private specialists can be contracted by the EHIF. Estonia inherited an excess of hospital beds from the 50-year Soviet era. In the early 1990s when health insurance system was established, a clear purchaser- provider split was introduced. The hospital sector was restructured, and lower-level 248 Good Practice in Health Financing Figure 9.11 Estonia: Number of Hospitals and Acute Care Admissions, 1985­2003 140 25 120 20 100 15 people hospitals of 80 100 hospital of per 10 number 60 number 5 admissions 40 20 0 1985 1988 1991 1994 1997 2000 2003 number of hospitals (left scale) acute care hospital admissions (right scale) number of acute care hospitals (left scale) Source: Ministry of Social Affairs 2006 (www.sm.ee). or lower-quality facilities were closed (figure 9.11). Quality standards were set and a licensing system introduced. Still, because many aspects remained unsolved in the changing regulatory framework, hospital sector reform was reinitiated in late 1990s, when the Ministry of Social Affairs commissioned the Hospital Master Plan 201524 to make projections about future hospital capacity. The plan pro- jected a three-fourths decrease (from 68 to 15) in the number of hospitals through mergers and other types of restructuring, a two-thirds reduction in the number of acute care inpatient beds, and concentration of acute inpatient care in 15 large hospitals by 2015. Complementing the reduction in the number of acute care beds, a system for rehabilitation and long-term care was to be developed. Criteria used for planning hospital capacity included sufficiently large population pools to support minimum service volume for quality and efficiency, development of medical technology, demographic and epidemiological projections, and accessi- bility (not more than 70 kilometers away, an hour's drive by car). According to a recent study using data from the late 1990s, access barriers were lower for 3.4 percent of the population living farther than 30 minutes from acute care hospitals (Rooväli and Kiivet 2006). Despite some variations, hospital care is accessible for different population groups in rural and urban areas (figure 9.12). Further reforms centralizing high-technology services would affect a quarter of the population. Access to general services will decrease for 10 percent of the pop- Estonia: Good Practice in Expanding Health Care Coverage 249 Figure 9.12 Estonia: Access to Medical Care, by Residence telephone consultation with doctor visit to family doctor visit to specialist visit to dentist hospitalization 0 5 10 15 20 25 30 35 40 45 50 respondents aged 25­74 Tallinn other urban area rural area Source: Kunst et al. 2002. Note: The proportions of respondents (age 25 to 74) in 1999 who had a telephone consultation with a doctor; visited a doctor, specialist, or a dentist in the previous six months; or was hospitalized in the previous 12 months. ulation, but the development of outpatient care and strengthening of the primary care already available can compensate for this decrease. This Hospital Master Plan was also an important input to the development of the Health Service Organization Act of 2001,25 which clarified the legal envi- ronment for service provision (hospitals but also all service providers in the health care sphere). All hospitals (public and private) have to operate under private law as foundations or joint-stock companies. This means that even hospitals owned by the central or local government must be run as private companies, with full managerial rights over assets, residual claimant status, and access to financial mar- kets. Hospitals are governed by supervisory boards, composed of members named by the owners or founders who are mainly from the public sector. All hospitals must be licensed by the Health Care Board. Licenses for meeting minimum stan- dards are issued for five years. Estonia has succeeded in reorganizing the hospital sector to improve efficiency and systemic responsiveness. In 1990, Estonia had about 120 hospitals and 14,000 acute care beds. The number of acute care beds had fallen to 8,600 by 1998 and to about 6,000 by 2003. The contraction is related to the establishment of the hospi- tal licensing system under which small hospitals providing predominantly long- term care lost their acute care status and were turned into nursing homes. Other 250 Good Practice in Health Financing Figure 9.13 Estonia: Number of Doctors and Nurses per 100,000 Inhabitants, 1998­2004 650 600 550 500 inhabitants 450 100,000 400 per 350 number 300 250 200 1998 1999 2000 2001 2002 2003 2004 number of doctors number of nurses Source: Ministry of Social Affairs 2006 (http://www.sm.ee). hospitals have been turned into primary care centers for outpatient care. In terms of hospital closures, access to care as measured by acute care hospital admissions has remained same (figure 9.12). The reduction in the number of acute beds has been due to hospital mergers. The average length of stay was been reduced from 14 days in 1990 to 6.4 days in 2003, but the bed occupancy rate is still low--68 percent in 2003 compared with 77 percent in 1994. One of the most critical aspects of the Estonian health system is human resource planning, and the number and quality of health care professionals. Since the early 1990s, the entire health workforce has been independent and outside the public service. This means that doctors and nurses are free to contract with providers such as hospitals for their services. This freedom has provided motiva- tion for professionals but has also made long-term planning difficult. At the beginning of health care reforms in the early 1990s, it was assumed that there was an oversupply of doctors, particularly in certain specialties. At the same time, there was and still is a shortage of nurses (figure 9.13), and specialist services are unevenly distributed around the country. After independence, underinvest- ment in health facilities and human resources was a major source of cost savings, resulting in relatively low salaries and poor morale among doctors and nurses. More recently, salary increases have improved staff motivation. The prospect of free movement of medical professionals within the EU has put further pressure on human resources in the Estonian health system. This has been reflected in the high Estonia: Good Practice in Expanding Health Care Coverage 251 number of documents processed by the Health Care Board for health care profes- sionals wanting to work outside Estonia (more than 300 doctors and around 150 nurses in 2004). In 2004­05, 4 percent of Estonia's doctors and 1 percent of its nurses left. However, in 2006, the outflow slowed, and professionals now leaving are seeking short-term work aboard rather than stays of several years. The current challenge, in addition to the migration of professionals, is the aging of profession- als, where the medium-term impact would be higher. Pharmaceuticals The pharmaceutical sector was reformed during the 1990s with the aims of estab- lishing a drug regulatory authority, creating a modern legislative framework, introducing a system for reimbursing drugs and privatizing pharmaceutical ser- vices (e.g., pharmacies). During the 1980s, the range of pharmaceuticals available was 20 years behind the times, and only a few, privileged patients had access to contemporary medication. The new reimbursement scheme for medicines was introduced in 1993. Since then the reimbursement category is determined by dis- ease severity, medication efficacy, and the patient's ability to pay. Whereas the lack of effective medication was the main issue until 1992, the increased of cost of medicines has become a major driving force for reforms (figure 9.14). The cost of medicines in Estonia has gone up for three main reasons. First, medication use has Figure 9.14 Estonia: Cumulative Increase in EHIF Pharmaceutical and Health Care Services Expenditures, 1993­2006 1,900 1,700 in 1,500 100) = 1,300 increase (1993 1,100 900 cumulative 700 of expenditures 500 index 300 100 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005e 2006e pharmaceuticals health care services Source: EHIF 2006 (http://www.haigekassa.ee). a. Forecast for 2005 and 2006. 252 Good Practice in Health Financing Figure 9.15 Estonia: Use of Cardiovascular Medicines, by Groups, 1994­2005 250 200 day per 150 1,000 per 100 DDDs 50 0 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 antihypertensives diuretics beta-blocking agents calcium channel blockers ACE inhibitors angiotensin II antagonists HMG CoA inhibitors fibrates Source: State Agency of Medicines, Annual Statistics on Medicines 2006. increased--triple the 1990s' budgets. Second, older drugs are being replaced by more effective, but more expensive, medications. And third, new pharmaceuticals, previously unused or unavailable in Estonia, have been introduced (Kiivet and Harro 2002). The last two qualitative changes have expanded drug coverage and allow the prescribing doctor to use more effective and safe medicines. In volume terms, the amount of prescription pharmaceuticals used per capita doubled during the 1990s: from 314 defined daily doses (DDDs) per 1,000 per day in 1994, to 574 in 1999 and 754 in 2004. In monetary terms, the pharmaceutical market increased 2.5 times between 1999 and 2005. Examples of important changes in medicine use patterns in Estonia include treatments for ulcers, hypertension, depression, and inflammatory joint diseases, all targeting more chronic conditions needing con- tinuous medication. Looking only at the treatment of hypertension (figure 9.15), a move toward evidence-based prescription can be observed. ACE inhibitors are used more widely, almost as much as the European average, and use of other evidence-based pharmaceuticals has at least doubled. The changes in general were driven by open- ing the market to new medicines, by the targeted insurance reimbursement sys- tem embodied in the positive list, and by information sharing on evidence-based practice with service providers and public. Estonia: Good Practice in Expanding Health Care Coverage 253 Thus, pharmaceuticals of known quality and proven efficacy have been made available, and the well-functioning reimbursement system supports public access to them. The State Agency of Medicines under the Ministry of Social Affairs is fully responsible for the control of all pharmaceutical activities--registration, quality control, and other activities involving drugs for veterinary as well as human consumption. Until 2003, EHIF led the pharmaceuticals initiative. Since then, the Ministry of Social Affairs, with its special policy unit for medicines, has been responsible for all pharmaceuticals reimbursement policies. For prescription pharmaceuticals, reimbursement is either 90 percent or, for drugs used to treat more serious dis- eases, 100 percent. Over-the-counter pharmaceuticals, vitamins, mineral supple- ments, herbal remedies, and the like are not reimbursed. All pharmaceuticals used in hospitals are free for inpatients, similar to other inpatient services for insured persons. Since the development of the first essential drug list for Estonia in 1992, the list has been used as a guide for medicine usage and reimbursement. Later, hospitals used it to develop their own formularies. In hospital, all medicines are free for patients covered by the EHIF. Health Coverage Reforms Since regaining independence in 1991, Estonia has been going through extensive health care reforms to expand services and insurance coverage. These reforms can be evaluated against two main goals: to ensure access to high-quality services and to provide financial protection (through insurance coverage with a broad benefits package and limitations on out-of-pocket payments). Chronologically, Estonian health care reforms can be divided into four phases: the early 1990s, the mid-1990s, the late 1990s and early 21st century, and the cur- rent system. In this section, the last 15 years of health care reforms and their impact on health coverage are described chronologically in terms of the two main goals. The reform milestones are depicted in figure 9.16. Figure 9.16 Estonia: Milestones in Health Sector Reform, 1992­2003 Health Insurance whole country establishment Health Care Hospital Fund Act, Health covered by of health Organization Act, beginning of Master Service Health effective insurance licensing of primary care Plan Organization Insurance primary care system providers reform 2015 Act Act system 1992 1994 1997 2000 2001 2002 2003 Source: Authors. 254 Good Practice in Health Financing The Early 1990s: Establishing the Social Health Insurance System to Ensure Sustainable Financing During the 1980s, when Estonia was part of the Soviet Union, health care coverage was virtually universal in terms of breadth. In reality, the depth of coverage varied among population groups. Services were well developed in some specialties such as maternal and child health, but in other areas, the use of modern technology or clinical methods for treatment lagged practice in West European countries. The first health care reform, before Estonia became officially independent, began with financing. The objective was to build a health insurance system to secure sustainable financing for the health care sector. From the centralized inte- grated state model of Semashko, the system would be shifted to a decentralized social health insurance model, protecting health funds through earmarked taxes and enhancing systemic efficiency and responsiveness. Preparations began in the late 1980s when changes were foreseen and need to establish a functioning health system emerged. The first Health Insurance Act was approved by Parliament, also before Estonia's official independence. The Estonian Medical Association (representing the majority of doctors) played a significant role in successful implementation of health insurance reform. It saw in the insurance system a possibility to ensure sustainable funding for medical care in a new economic environment. This support was important in streamlining the reform. Another important success factor was the consensus and commitment of political parties to the system between 1993 and 1995 after independence and during a period of political stability. The health insurance system was designed around regions, where 22 noncom- peting sickness funds were established. This was one part of the health care decen- tralization plan. The sickness funds were organized by county or city governments authorized to approve the sickness fund statutes and rules for calculating health care benefits. The Ministry of Social Affairs exercised a supervisory role over the health insurance system. To enhance national coordination, the government set up an Association of Sickness Fund. One reason for establishing a Bismarckian type of health insurance system was to ensure a sound revenue base for the health care system. Another important rationale was to relate health insurance closely to the labor market to give people incentives to participate in the formal labor market. Thus, the health insurance tax was introduced with a contribution, set at 13 percent of employees' salaries, paid entirely by employers. Other contribution rates were set for different kind of entrepreneurs. The 13 percent rate on salaries still applies, but due to changes in the legal environment, other types of entities (e.g., self-employed farmers) have been reorganized but pay the same 13 percent rate. At first, the health insurance tax was a separate tax. In 1994, it was incorporated in the social tax as a share ear- marked for the health care system. Estonia: Good Practice in Expanding Health Care Coverage 255 In the early stages, the regional sickness funds collected the health insurance tax, and there was no central pooling or any risk adjustment of funds. Therefore, more deprived areas had lower revenues, which curtailed their access to care. Recognition of this shortcoming of regional organization impelled administrative changes. The established health insurance system was mandatory with no opt-out exceptions, but people were allowed to buy supplementary private insurance for uncovered expenditures. From the beginning, insurance coverage was almost uni- versal because as most population groups were eligible. Out-of-pocket payments for health care services were almost nonexistent dur- ing the Soviet era. Formal out-of-pocket payments were not introduced at first. In 1993, Estonia introduced the prescription pharmaceuticals reimbursement sys- tem, based on the positive list principle26 and limited cost-sharing. The introduction of a purchaser-provider split was seen as one tool for facili- tating downsizing of the provider network. Sickness funds were contracting with service providers to ensure necessary care for insured persons in their region. Providers were paid according to the price list approved by Ministry of Health (later the Ministry of Social Affairs). The price list was based on the German price list, where fees for different services were adjusted for local prices. Before 1995 copayments were US$0.38 per prescription for medicines on the positive list with a coinsurance rate of 100 percent (defined by disease groups and pharmaceutical names) and for population groups such as disabled persons, chil- dren under three years old, and people over 70 years of age. Similar copayments were applied for medicines with a coinsurance rate of 90 percent, and slightly higher copayments for prescription medicines with a coinsurance rate of 50 per- cent. The policy resulted in a small share for kept the share of private expenditures low in the early 1990s and access to medicines good. This first wave of reform laid the base for a functional health system by estab- lishing a sustainable and fully operational health financing system, an inevitable prerequisite for the later reforms. The health insurance system was crucial to ensure breadth and depth of coverage after the collapse of the Soviet system while Estonia started to build up a social security system. An important change while expanding depth of coverage was the introduction of new reimbursement system for prescription medicines, which made medicines available with limited out-of- pocket payments. Another important change was the introduction of a purchaser- provider split to obtain transparent contractual arrangements for achieving efficient use of resources. Mid-1990s: Decentralizing the Provider Network to Ensure Access to Modern, High-Quality Health Care Reorganization of the provider network marked the second phase of reforms. After the introduction of the purchaser-provider split, health care organizations 256 Good Practice in Health Financing were gaining more autonomy than they had had under the Semashko type of sys- tem. The new freedom opened space for incentives for restructuring the provider network to improve systemic responsiveness and efficiency. When the Health Ser- vice Organization Act came into force in 1994, the health service planning func- tion was delegated largely to the municipalities to dilute the central government's role. In 1991, the provider licensing system had been enhanced, an important pre- condition for decreasing hospital network capacity and ensuring quality. In 1994, after a detailed review of all providers, substandard providers were closed. How- ever, legislation enacted up to that point did not provide for supervision and accountability. In line with hospital sector reorganization, primary health care reform was ini- tiated. The first step in primary health care reform was to introduce the family medicine as a separate medical specialty (in 1993) and to start postgraduate train- ing at the University of Tartu, the only university providing training for future doctors. Respecialization training had been introduced in 1991. The basic frame- work for family medicine, set up in the Health Service Organization Act, was amplified in the Family Practice Act during 1994, but the act remained a draft and reform slowed down, because there were no appropriate incentives. In 1997, the primary care reform plan was reinitiated by ministerial decree. Administrative changes were also going on in the health insurance system. The first years of experience showed that a fully decentralized, uncoordinated system of sickness funds was too fragmented. Revenue collection was also fragmented, lacked the vital central pooling and risk-adjustment arrangements, and con- tributed to widening inequalities between regions. To strengthen central functions such as planning, redistribution of revenues between regions, and control of financial resources, the Central Sickness Fund was established in 1994 and the regional sickness funds were subordinated to it. The revenue collected was pooled centrally and reallocated to the regions on capitation basis. The regional sickness funds were cut to 17 in an attempt to improve resource use efficiency. Jointly with introduction of Central Sickness Fund, the State Health Insurance Council (SHIC) was established. The SHIC consisted of 15 members,27 nomi- nated by the organizations represented on the council and mandated to serve for three years. The main SHIC responsibilities were to approve the state health insur- ance budget and to develop the price list for health care services. The role of the SHIC was mainly advisory. Concurrently with establishment of the SHIC, Regional Health Insurance Councils were also established to advise each of the regional sickness funds. Although their role was advisory, they had a remarkable influence on contracting processes. Changes were also made in health insurance eligibility criteria by defining more precisely the groups eligible for health insurance without contributions. As a result, some segments of the working-age population remained uninsured.28 The motivation behind this change was to give clear incentives to participate in the legal labor market and to decrease the share of informal payments, a serious con- Estonia: Good Practice in Expanding Health Care Coverage 257 cern due to the ongoing privatization and other radical changes in the economic environment. Another important amendment of the Health Insurance Act was to allow regional sickness funds financial discretion only to the extent resources were avail- able. In the early years when resources were plentiful, this provision was not so important. The providers' capacity to deliver enough services was the challenge at that time, but it was feared that, as provider capacity expanded, the health insur- ance system might be unable to cover all expenses. Regional fund managers have strictly followed this principle of balanced revenues and expenditures, Only once, in 1999, did operating expenditures exceed revenues, and the deficit was covered from central sickness fund reserves. The balanced budget principle was followed mainly because of the funds' desire to demonstrate independence from the state budget and ability to function autonomously in the health sector. Since the early 1990s, Estonia's general fiscal policy has always supported balanced budgets in every sector. This commitment is bound by strict regulation and generally conser- vative fiscal policy, which disapproves of public sector deficits. In 1995, patient copayments for primary care and specialist visits were intro- duced. The copayments were low (EEK 5, less than US$0.50) and many popula- tion groups were exempt. The copayments were introduced mainly to decrease informal payments, which were common during the Soviet era, and to rationalize health care service utilization by eliminating unnecessary visits. However, this decision was politically sensitive because people were used to free medical care and their willingness, but also their ability, to pay for medical care was limited. After 1995 cost sharing for pharmaceuticals was also increased by raising copay- ments and putting ceilings on EHIF-covered benefits. Copayments for pharma- ceuticals continued to increase modestly in the mid-1990s but more rapidly in the late 1990s and in the current century. The second wave of reforms ensured further development of service provision. The primary health care system was strengthened through family medicine reform, and ambulance services were separated from the health insurance system, ensuring access to ambulance care for the whole population. To ensure the quality of the provider network, a clear licensing system was created that resulted in clo- sure of substandard providers. As a result, the number of hospitals decreased, enabling more efficient use of resources and securing service quality. Late 1990s and Early 21st Century: Recentralizing to Enhance Depth of Coverage After Estonia regained independence, the health care system was widely decentral- ized to municipalities and counties. The weakness of implemented reforms lay in the lack of preparation in terms of staff training, accountability procedures, and guidelines for policy sustainability. In the health insurance system, some recentral- ization was already going on in 1994, with the establishment of the Central Sick- ness Fund. But it was evident that the provider network restructuring needed 258 Good Practice in Health Financing reinforcement. Municipality-level planning of provider-related functions did not work. Protecting the interests of local providers often took precedence over system- level efficiency and accountability, and the municipalities, as administrative units, lacked the necessary revenue base and competences. Some functions clearly had to be recentralized and the legal status of providers established. The Ministry of Social Affairs had to take the stewardship role in planning the provider network. The important milestone in the third reform phase was the Hospital Master Plan 2015. This was a necessary accelerator for continuing hospital sector reform with clear targets. Until 1998, primary care was provided mostly in polyclinics and ambulatories, owned by the municipalities, and by a few private providers. In 1998, some pri- mary care planning functions were recentralized from municipality to county. To foster the primary care reforms, the Ministry of Social Affairs introduced the pri- mary care reform plan in 1997 with goal of providing the whole population with family physician services by 2003. Ensuring access to primary care has been seen as important precondition for centralizing specialist care and downsizing hospital network capacity. All Estonian inhabitants, insured and uninsured, must be regis- tered with a family physician.A new mix of payment methods for family physicians was introduced. To encourage physicians to retrain and become certified as family medicine specialists, they were offered an incentive of about US$80 per month. Overall coverage, including the ambulance services, has been expanding. Enti- tlement was broadened to the whole population, and service quality was improved, even with the smaller number of teams available. These accomplishments con- tributed to the overall development of the health care sector with centralized hos- pital services, new opportunities for high-quality services, and access to doctors and nurses at the primary care level. In the meantime some administrative changes were made in health insurance system. In 1999, the social tax collection function was assigned to the Tax Agency, which transfers the revenue earmarked for health care to the health insurance fund. This allows the health insurance fund to concentrate on the purchasing function. Tax revenue collections have also increased. Over the decade, as Estonia's own legal system evolved, the legal status of sick- ness funds grew fuzzy, with some features of public independent legal person and some of a state agency related to the Ministry of Social Affairs. In 2001, the estab- lishment of the EHIF as a public, independent legal body with seven regional departments enabled clarification of the roles of central and regional departments. The legal status of service providers was also defined more precisely, clarifying relations between purchaser and providers. In 2003, seven regional departments were merged into four departments as a natural step in centralization, each cover- ing 200,000 to 500,000 insured persons. In 2002, the Health Insurance Act clarified the regulation of all aspects of the health insurance system, including benefits, reimbursement lists and levels for health services and drugs, maximum levels of Estonia: Good Practice in Expanding Health Care Coverage 259 cost sharing for insured people, and contractual relations between the EHIF and providers. The most important feature of the third wave of health coverage reforms has been the tidying up the overall regulatory framework for the health financing sys- tem and provider network. In terms of breadth of coverage, changes were made in insurance eligibility criteria, and some population groups were excluded. However, the share of uninsured population has not increased, because employment has strengthened. The third phase has also been marked by growing out-of-pocket expenditures for pharmaceuticals and for dental care, which was excluded from the benefits package for adults and replaced with a low yearly monetary compensa- tion. In addition, maximum ceilings for patient copayments were set for specialist visits and hospital stays, keeping FP visits free of charge for the insured population, but leaving the uninsured to cover all costs but emergencies out of pocket. Further Directions and Challenges The fundamental changes to build a wide-coverage health care system in Estonia were made in the early stages of reform, the early 1990s. The further incremental arrangements were made to support the primary health care and hospital sector reforms and to strengthen the EHIF purchasing function. Now some fine tuning of the system is needed to improve performance. One continuing subject of debate concerns the low level of health expenditures in relation to GDP. Additional funding could be one answer. In the long term, however, financial sustainability is a big concern because the care of an aging pop- ulation will absorb more and more health resources. In terms of public financing, the broadening of the health insurance revenue base (taxing other incomes in addition to salaries) is one option. A second option is to persuade local munici- palities to increase their financing by expanding their responsibilities, for exam- ple, by providing health care services for the uninsured. Because both alternatives demand strong political commitment, it is doubtful that radical changes will be made any time soon. The only decision made in 2006 was to increase the state's contribution rate, which will increase the budget by less than 2 percent and affects only 4 percent of the insured. Another topic under discussion is an expansion of private financing by foster- ing more favorable conditions for private insurance or by increasing out-of- pocket payments. Out-of-pocket payments have been creeping up in an attempt to activate macro-level cost-containment in public funding by rationalizing use of health care services and pharmaceuticals. Therefore, the impact of rising out- of-pocket payments on different social groups should be evaluated, considering that current evidence shows that access to medicines may already be difficult. Introducing targeted exemptions could be considered, especially for people with chronic conditions. In addition, supply-side measures such as prescription bud- gets, active feedback to doctors, and rational prescribing training programs have 260 Good Practice in Health Financing also been discussed. Nonetheless, in the absence of a strong commitment to con- tain costs, funding has mounted every year. The uninsured are the most vulnerable population group in terms of high out- of-pocket health expenditures. Currently the insurance eligibility policy for the working-age population is deliberately tied to labor policy, and health insurance is seen as one incentive for landing a job in the formal market, subject to payroll taxes. This link is unlikely to be cut in the near future, thus barring the way to fully universal coverage. In terms of service delivery, fine tuning presents the biggest challenges. The strengths of the current delivery system are strong family medicine centered on primary health care. This system covers for a wide range of first-tier services with- out copayments and with minimal waits. It is complemented with the ambulance services for care outside normal working hours and on weekends, which now counts as a fourth of all health care visits. This slight overuse of this service can be considered a weakness in the delivery system, because many of those "visits" turn out to be cases treatable by family physicians or just calls for transportation (National Audit Office 2004). That was one of the reasons for opening the increas- ingly popular 24/7 primary care call line, in late 2005. The challenge lies in making the delivery system more patient centered and in coordinating care at the primary level with development of additional nursing and rehabilitation services. This would enable more efficient use of resources and ensure better access and quality of care. Continuity of care presents another chal- lenge: having outpatient care and high-quality hospital services close to patients without long waits. All this hinges on the availability of human resources, particu- larly nurses, who for several years have been migrating to neighboring countries or leaving nursing for other jobs. Health Coverage Reforms: Lessons for Other Countries Health care coverage has improved in Estonia over the last 15 years, and health financing reforms have played a key role. The main enabling economic, political, and institutional factors are summarized in this section, together with some lessons learned in the reform process that may be useful to other countries. Successes and the Reasons Estonian health financing reform has succeeded in establishing a health insurance system that ensures nearly universal coverage, The society's commitment to radi- cal changes, the medical profession's support, and political consensus have been the most important enabling factors of health financing reform. In the late 1980s and early 1990s, the entire population was open to change and willing to move away from the Soviet system to a market economy. Politicians as well the general public were open to radical changes, and there was political consensus for estab- lishing a social health insurance system. Another vital factor for the development Estonia: Good Practice in Expanding Health Care Coverage 261 of the health insurance system has been robust economic growth,29 allowing con- tinuous expansion of health insurance revenues and restructuring and scaling up of the health system. These factors have guaranteed stability and continuity long enough to build an institutionally sound social health insurance system. Support from the medical profession early in the reform was crucial. Doctors, nurses, and other health workers saw in the new health insurance system, with yearly revenues from the earmarked tax, a sustainable way of financing health care, contrary to budget allocations in Soviet times. The health insurance contri- bution rate (13 percent of salary) was decided upon after looking at examples from other social health insurance systems, but calculated in a way that would bring in more revenues than projected to be needed in the early 1990s. This cush- ion ensured enough resources in the early years of change--to the medical profes- sion's satisfaction. Some Unresolved Issues Because Estonia was emerging from a Semashko-type centralized health care sys- tem, the incentive to decentralize was strong. Thus, in the early years of reform, both the financing system and the provider network were decentralized. However, the regions' varying capacity to conduct all necessary functions soon led to recen- tralization. In 1994, the Central Sickness Fund was established to supervise the regional sickness funds. The service delivery planning function was also decen- tralized, and the role of Ministry of Social Affairs decreased. However, the regions needed central oversight to organize and plan service delivery, and the ministry had to enhance its stewardship role. The correct balance has not yet been struck between the stewardship role and direct control of functions in an environment in which neither service providers nor the medical profession are under the direct control of government. Some Lessons for Other Reformers For low- and middle-income countries, nine lessons can be derived from Estonia's experience with health coverage reforms. 1. Health system revenue collection should be in line with a country's general fiscal policy and take into account labor market policies and trends. In Estonia, the mainly salary-based health insurance system is closely related to the labor market. At the same time, solidarity makes the system cohesive and comprehensive; some noncontributing groups are covered: children, pensioners, and some special groups (e.g., disabled persons, mothers on job leave). For the working-age population, the eligibility criteria for health insurance are set in such a way as to give strong incen- tives to participate in the formal labor market. Therefore, a part of the working-age population (the informally employed) remains uninsured. The underlying ratio- nale has been to support overall labor market policies, focused in the 1990s, on 262 Good Practice in Health Financing decreasing the size of the informal labor market, but now emphasizing increased labor market participation, particularly among the long-term unemployed. The revenue collection arrangement, payroll taxes, has supported privati- zation, labor market development, and decreased unemployment. It has also ensured a stable resource base for the health insurance system and has allowed the Estonian Health Insurance Fund to provide a broad benefits package. In the long run, however, relying solely on salary-related contributions could jeopardize financial sustainability as the economy develops and the population ages. There- fore, to broaden the revenue base, diversification options should be considered when the system is designed. Although the system has ensured sustainable fund- ing in Estonia so far, its relation to fiscal policies and labor market patterns should be continuously evaluated. 2. Formalization of the labor market and the economy as part of efforts to decrease corruption also has positive impacts on the health sector. Informal payments (money for services) were common in Estonia until the early 1990s. When health care reforms began, these informal payments were a serious concern. To formalize payments so both patients and doctors know where they stand, copayments for primary and specialist care were introduced by regulation in the mid-1990s. Lately, however, informal payments have resurfaced as an issue, especially in some spe- cialties, but patients are not complaining publicly. These considerations prompt the conclusion that economic formalization and clear regulation are not enough to clear away invisible barriers to care posed by informal payments. Patients have to be informed about their rights, and processes have to be set up for them to make complaints about violations. 3. Out-of-pocket payments play a large and growing role in health financing but should not be allowed to become a barrier to care for the less well-off. The Eston- ian health financing system relies mainly on public resources, which enables it to protect most of the public from illness-related financial hardships, but the growth of out-of-pocket expenditures poses a risk. To some extent, this growth is related to increased incomes and changing consumption patterns. Out-of-pocket pay- ments are used to ration health care and are an important source of revenue for many countries' health systems. For some people, however, out-of-pocket expen- ditures are a serious access barrier to necessary care, especially for prescription medicines. Their impact on access for different population groups should there- fore be evaluated and monitored. If necessary, income- or health-related targeted exemptions should be introduced. 4. All levels of government need to be involved in designing the health financing system so that rational, transparent allocations can be made among competing uses. When the health insurance system was introduced in Estonia, the central government and local municipalities were expected to continue financing health care at their historic levels. In practice, they gave other development sectors Estonia: Good Practice in Expanding Health Care Coverage 263 greater priority and less funding to health care. Now most of the public funds for health care come through the health insurance system. Therefore, countries plan- ning to introduce a mandatory health insurance system should be aware that this may send a signal to the central and local governments that they can reduce their financial support to the health sector, with adverse effects on long-term sustain- ability of the health finance. Once again, building the finance system on diverse revenue sources can ensure better sustainability in transition economies and in countries where demographic changes are taking place. 5. In a changing environment, roles and functions throughout the health care sys- tem have to be constantly watched and adjusted to improve performance. To keep up with changing times, Estonia has already had to make some adjustments in its new health care system. The mandatory social health insurance system, as con- ceived, was decentralized. Noncompeting, regional sickness funds collected the insurance tax and bought services, each for its own enrollees. After only a year, it was obvious that an organization with a pool of less than 1.5 million is administra- tively ineffectual and that the lack of risk-adjustment mechanisms would widen disparities between regions because of differences in regional wealth. Therefore, the Central Sickness Fund was established. Now, funds collected regionally are pooled centrally and allocated to each region based on capitation. Centralization enabled improvements in administrative efficiency and central pooling of funds. Centralization of other functions has continued. Client services and relations with insured persons have been developed further at the regional level and moved into Internet-based services. Now, for example, an insured person can check on the validity of its insurance, using Internet-based banking services. Planning and overall monitoring are done centrally. Purchasing and service-quality monitoring by trustee doctors is performed by regional branches of the insurance fund. Func- tions of this kind need to be analyzed in each country and tailored to its particu- lar health system. The ability to make necessary structural changes to improve systemic performance should be built into the design. 6. Centralization has enabled strengthening of the purchasing function and effi- ciency improvements in resource use. One strong purchaser wields much more market power over providers than many disconnected local sickness funds. Con- solidation of purchasing power was also an important enabling factor for restruc- turing the provider network, because the hospital network also had to be merged and restructured. The EHIF purchasing process facilitated that restructuring. Without it, the process would have been complicated by dealings with the all the independent regional sickness funds and their individual local interests. 7. Delineation of clear, transparent, contract obligations was the basis for achiev- ing cost containment, service quality, and access to care. The health insurance fund has devoted considerable effort to enhancing the purchasing function. Besides planning, needs assessment, and monitoring, contracts have been an 264 Good Practice in Health Financing important component of that effort and the foundation for the purchasing func- tion. In Estonia, unlike some societies, contracts have always been considered seri- ous agreements between parties. This tradition of respect for contracts was the basis for accountability in honing the purchasing function. Standard contract conditions are now negotiated with provider associations (not single providers), which streamlines the contracting process and makes it transparent. Estonia's experience shows the importance of insurer-provider contracts that specify the responsibilities on both sides for improving performance in terms of access and quality of care. The contracts have also been a vital cost-containment mechanism, because they specify the minimum number of treatment cases and the upper limit of volume. 8. For equitable service coverage, a comprehensive provider network should be put in place for every level of care. Estonia did this by developing family medicine centered on primary care, restructuring ambulance services, and reorganizing the hospital sector by introducing high-technology services and providing access to modern medicine. These changes were sequenced. Health financing reforms sup- ported the development of the provider network by developing appropriate finan- cial mechanisms (as incentives for primary care). With the development of the provider network, a new balanced system was achieved, providing coverage for a wide range of services. New issues have arisen related to continuity of care and integrating disease pre- vention and health promotion services into the current system. The ongoing chal- lenge has been to improve coverage, step by step (e.g., by providing new curriculums, staff training, equipment and facilities, and financial incentives) at different levels of care. 9. Information technology has played an important role in helping providers and health financing institutions extend and improve their service. Estonians are receptive to IT technologies, not least in the health sector. In the mid-1990s, the providers used to send the health insurance fund bills on paper. Now all health care providers have personal computers, and electronic communications are used between the EHIF and its partners (patients, employees, providers) to increase data processing efficiency and speed up outcomes. Today all transactions between health care providers and pharmacies are also done electronically. Electronic data trans- mission, besides making transactions more efficient, enables collection of high- quality data for assessing provider performance and service use by the insured. Endnotes 1. World Bank. World Development Indicators. 2006. 2. The tax wedge for low-income workers was 37.2 percent in Estonia in 2004, com- pared with the European Union average of 36.4 percent (Eurostat 2006, http://europa.eu. int/comm/eurostat/). 3. Ministry of Social Affairs 2006, http://www.sm.ee. Estonia: Good Practice in Expanding Health Care Coverage 265 4. Ministry of Social Affairs 2006, http://www.sm.ee. 5. According to latest data available (October 2007), there were almost 6,200 HIV- positive individuals (Health Protection Inspectorate 2007, http://www.tervisekaitse.ee). 6. The inflation rate was 3.0 percent in 2004 and 4.1 percent in 2005 (Ministry of Finance 2006, http://www.fin.ee). 7. According to Health Care Services Organization Act (2001), in Estonia emergency care means health services provided by health care professionals in situations where post- ponement of care or failure to provide care may cause death or permanent damage to the health of the person requiring care. 8. In 2005, the share of health expenditures from total budget varied from 8 percent to 0 percent by municipalities. 9. Age-adjusted capitation plus additional allocations through fee-for-service and lump-sum payments. 10. In 2005, pulmonology and oncology were priority areas. 11. The criteria for provider selection are: the proximity of service provision to patients; the share of provision of services in day care; the experience in the last contracting period (rejected claims for reimbursement, complaints by patients); and the price of ser- vices (lower price offer winning EHIF preference). 12. The expected outcome is improvements in the quality and effectiveness of preven- tive services and monitoring of chronic diseases. Specific outcome indicators have been agreed and are being monitored. The first "round" of quality bonuses was paid in 2007 on the basis of results for 2006. 13. Insured people who spend more than a given amount out of pocket on listed drugs in a year can claim partial reimbursement. Since 2003, additional financial protection has been provided to those who face high pharmaceutical expenditures: the EHIF reimburses 50 percent of a yearly cost between US$460 and US$767, and 75 percent beyond, up to a limit of US$1,534. Any additional cost is not covered. 14. For details, see the Health Insurance Act, http://www.legaltext.ee. 15. Developed by Ministry of Social Affairs and updated as needed. 16. The government-approved list of eligible conditions is available in Estonian at https://www.riigiteataja.ee/ert/act.jsp?id=1008084. 17. The deficit was then covered by fund's own reserves, and no extra allocations from state budget were made. 18. According to Income and Living Conditions Survey 2003, 7 percent of the adult population did not get needed FP care for economic reasons (mostly uninsured, who face out of pocket payments), long waiting times, and long travel distances. For specialist and dental care, these shares were, respectively, 8 percent and 15 percent. Access barriers are big- ger for lower socioeconomic groups and rural dwellers. 19. If a specialist is contacted directly, the patient has to bear the full cost of services (except in some narrow specialties such as selected infectious diseases, fertility and gyne- cology related cases, and traumas). 20. If the FP has no contract with the EHIF, the patient bears the full cost of service. 21. Exemptions are allowed in some regions if a county governor and the EHIF agree. 22. This is monitored by trustee doctors under a system developed by the EHIF for quality control through, for example, visiting practices, reviewing clinical work, auditing, and checking on access. 23. TNS EMOR (1999­2003), Faktum (2004­05). 266 Good Practice in Health Financing 24. The HMP 2015, done by Swedish consultants in 1999, was published in April 2000 (SC Scandinavian Care Consultants AB. 2000). The plan, used as a basic planning exercise, was developed further in 2002 and 2006. 25. The regulation covers all health care service provision in Estonia (previous legisla- tion from mid-1990s with updates) from primary care to hospital services. In addition, patient-provider relations are regulated by the Obligations Act, and provider-purchaser relations by insurance regulations and contracts. 26. The positive list included a list of pharmaceuticals reimbursed at different levels of copayment (with a small deducible). 27. One representative each from the Ministry of Social Affairs and the State Social Insurance Board, one county doctor, one municipal doctor, one provider representative, five representatives of employer unions, and five representatives of insureds' unions. 28. Data on coverage from this period are not available. 29. GDP growth in real terms was 7.1 percent in 2003, 8.1 percent in 2004, and 10.5 percent in 2005. References Atun, R. A. 2004. "Advisory Support to Primary Health Care Evaluation Model: Estonia PHC Evaluation Project." Final Report, WHO Regional Office for Europe, Copenhagen. Atun, R. A., N. Menabde, K. Saluvere, M. Jesse, and J. Habicht. 2006. "Introducing a Com- plex Health Innovation--Primary Health Care Reforms in Estonia (Multimethods Evaluation)." Health Policy 79: 79­91 CIET International (Community Information, Empowerment and Transparency Interna- tional). 2002. "Curbing System Leakages: The Health Sector and Licensing in Estonia." CIET International, Tallinn. Couffinhal, A., and T. Habicht. 2005. "Health System Financing in Estonia: Situation and Challenges in 2005." HSF Working Document, Health Systems Financing Programme, WHO Regional Office for Europe, Copenhagen. Faktum. 2004. Elanike rahulolu arstiabiga 2004 [Population satisfaction with health care]. Survey. Tallinn: Faktum. Faktum. 2005. Elanike rahulolu arstiabiga 2005 [Population satisfaction with health care]. Survey. Tallinn: Faktum. Habicht, J., K. Xu, A. Couffinhal, and J. Kutzin J. 2006."Detecting Changes in Financial Pro- tection: Creating Evidence for Policy in Estonia." Health Policy and Planning 21: 421 ­31. Habicht, J., and A. Kunst. 2005."Social Inequalities in Health Care Services Utilization after Eight Years of Health Care Reforms: A Cross-Sectional Study of Estonia, 1999." Social Science and Medicine 60: 777­87. Jesse, M., J. Habicht, A. Aaviksoo, A. Koppel, A. Irs, and S. Thomson. 2004. "Health Care Systems in Transition: Estonia." WHO Regional Office for Europe for the European Observatory on Health Systems and Policies, Copenhagen. Josin, M. 2004. Korruptsiooni ja varimajanduse levik Eestis [Corruption and the black mar- ket in Estonia]. Tallinn: Estonian Institute of Market Research. Kiivet, R., and J. Harro, eds. 2002. Health in Estonia 1991­2000. Tartu, Estonia: University of Tartu. Estonia: Good Practice in Expanding Health Care Coverage 267 Kunst, A. E., M. Leinsalu, A. Kasmel, and J. Habicht. 2002. Social Inequalities in Health in Estonia. Main Report. Tallinn: Estonian Ministry of Social Affairs. Rooväli, L., and R. A. Kiivet. 2006. "Geographical Variations in Hospital Use in Estonia." Health and Place 12(2): 195­202. Ministry of Social Affairs, Estonia. 2006. "Years Lost Due to Burden of Disease: Links with Risk Factors and Cost Effectiveness to Reduce Risks." Report [in Estonian], Tallinn. SC Scandinavian Care Consultants AB. 2000. "Estonian Hospital Master Plan 2015." Sum- mary Report. Tallinn. Statistical Office of Estonia. 2004."Household Living Niveau 2004." Special data query pro- vided by Statistical Office, Talinn. TNS EMOR. 2003. Elanike rahulolu arstiabiga 2003 [Population satisfaction with health care]. Survey. Tallinn: EMOR. ------. 2002. Arstiabi kasutamine elanike poolt 2002 [Utilization of health care services by the population]. Survey. Tallinn: EMOR. ------. 2001. Arstiabi kasutamine elanike poolt 2001 [Utilization of health care services by the population]. Survey. Tallinn: EMOR. vices by the population]. Survey. Tallinn: EMOR. ------. 1999. Arstiabi kasutamine elanike poolt 1999 [Utilization of health care services by the population]. Survey. Tallinn: EMOR. ------. 1996. Arstiabi kasutamine elanike poolt 1996 [Utilization of health care services by the population]. Survey. Tallinn: EMOR. National Audit Office of Estonia. 2004. "First-Level Emergency Care Organization" [in Estonian]. Tallinn. 10 The Kyrgyz Republic: Good Practices in Expanding Health Care Coverage, 1991­2006 Melitta Jakab and Elina Manjieva Background The Kyrgyz Republic is a small (191,300 sq. km), landlocked, mountainous, low- income Central Asian country. Its population (5.2 million) is predominantly rural (CIA 2006). In 2004, per capita GDP was US$431 (Recent Economic Development, Jan­Dec 2004, WB). The Kyrgyz Republic devotes 5.6 percent of GDP to health care, about average for its income level. Some 41 percent of health spending is public, but out-of-pocket payments (formal and informal) account for 50 percent of health expenditures, a higher proportion than in comparable countries. In terms of health outcomes, it does slightly better, with a life expectancy of 68. In 1991, at the time of the breakup of the former Soviet Union (FSU), the Kyrgyz Republic had a norm-driven, centrally planned, general revenue­financed health sys- tem in which free health care was every citizen's right. Between 1991 and 1996, the Kyrgyz economy collapsed, and GDP contracted by more than half with equivalently large reductions in government funding for the health system. As a result, the Kyrgyz Republic found itself saddled with an inefficient and unaffordable health system and gradual erosion of health benefits coverage. Coverage erosion manifested itself in increased out-of-pocket payments both formal and informal. Reforms starting in the mid-1990s diversified health sector financing, centralized the flow of public funds into a national pool, clarified entitlements to health benefits, introduced provider payment reforms, strengthened primary care, rationalized the delivery of hospital services, updated treatment protocols, and broadened consumer choice. The Kyrgyz reforms were successful at halting the process of coverage erosion and at reversing the earlier trends in some areas by 2004. At the heart of the reforms was the recognition that efficiency gains had to be made before coverage and equity issues 269 270 Good Practice in Health Financing could be directly addressed due to tight resources and excess physical capacity. Although changing the provider payment mechanisms and restructuring service delivery are usually recognized as instruments for improving efficiency, they became critical preconditions for improving equity in the Kyrgyz context. Recognition of the interlinked nature of efficiency and equity meant that the reform instruments were realistic, appropriately sequenced and relied on internal resources in the system rather than on injection of additional government or donor funds. Several factors explain why coverage reform featured strongly in the Kyrgyz health system agenda: availability of data and information on the erosion of coverage and use of this information through advocacy; linking health reforms to the wider poverty reduction agenda, stressing to the connections between erosion of coverage and poverty; support of the president in the early phases of the reforms; support of the international community; and the equity orientation of the health reform architects and their recognition of the reality of the limited fiscal space. Introduction This section provides the economic, demographic, social, and political context for the health sector reforms discussed in this chapter. The extent of economic down- turn and its impact on government spending for social services, including health care, are highlighted This trend provides a vital contextual factor for understand- ing why efficiency and equity were seen as inseparable goals of the reforms in the Kyrgyz context where equity could be achieved only through increased efficiency and not through injection of external resources. Economic Environment The Kyrgyz Republic, one of the poorest republics in the Soviet Union, depended heavily on subsidies from Moscow. 1 The loss of subsidies and the economic con- sequences of disintegration led to a painful transition process. Between 1991 and 1995, GDP declined by more than 50 percent. Strong economic growth in 1996 and 1997 was not enough to noticeably improve most peoples' standard of living. GDP per capita has still not recovered to the preindependence level (figure 10.1). The growth spurt of 1996 and 1997 was followed by another downturn in 1998, attributed largely to the spillover effects of the Russian financial crisis. Since then, the GDP growth rate has stabilized at around 5 percent a year. Extreme poverty2 continued to increase during the early years of transition, reaching a rate of 23.3 percent in 1999. The absolute poverty rate3 was also high and rising, from 52 percent in 1996 to 64 percent in 1999 (World Bank 2001). However, in 2000 both extreme and absolute poverty started to decline: extreme poverty fell by half, and absolute poverty fell from 63 to 49 percent in 2000 and 2003 (World Bank 2005a). Measured against the international poverty line of US$2.15 per day, however, poverty in the Kyrgyz Republic is significantly higher: in 2003, it was estimated at 70 percent. Unfortunately, the lack of consistent interna- The Kyrgyz Republic: Good Practices in Expanding Health Care Coverage 271 Figure 10.1 The Kyrgyz Republic: Key Economic Indicators, 1990­2004 a. GDP per capita, 1990­2004 2,500 $) 2,000 1,500 interntional 1,000 (current PPP 500 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 b. GDP growth, 1990­2004 15 10 5 0 change ­5 annual %­10 ­15 ­20 ­25 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Source: WDI Dataset, 2006. tional poverty line estimates before 2000 makes longitudinal comparisons difficult, but recent health and education indicators seem to reflect overall improvement in the well-being of the population. The Kyrgyz Republic became the first state in Central Asia to be welcomed into the World Trade Organization (WTO), in December 1998. The growth recovery was led by agriculture, gold mining, and trade. In 2004, gold exports represented 40 percent of total exports. Apart from gold, the Kyrgyz Republic relies on a few other exports such as cotton, electricity (bartered for natural gas and coal), and tobacco, making it vulnerable to changes in commodity prices (World Bank 272 Good Practice in Health Financing 2005b). More than half of the population is employed in the agricultural sector, which has a low productivity4 and, consequently, low wages. Political Economy The Kyrgyz Republic is a unitary state with a strong central government. It is divided into eight regions (oblasts), including the capital city, which are sub- divided into 43 districts (rayons) (Oxford Policy Management 2006). The presi- dent appoints the regional governors subject to the approval of local councils. The Law on the Basic Principles of the Budget (No. 78, June 11, 1998) establishes revenue assignments (including a tax-sharing system and local revenue options) and defines expenditure responsibilities between the national and subnational governments.5 However, there is a mismatch between subnational public service responsibilities and available revenue-raising instruments. The system of inter- governmental fiscal transfers is such that subnational and subregional govern- ments have limited control over revenue sources and few incentives to raise their own revenues. This, coupled with large differences in regional fiscal bases, results in chronic underfunding of services for which local governments are responsible, except in the capital city Bishkek and a few other relatively rich districts. Until 2006, health and education were considered local government responsibilities, which contributed to geographical disparities in the quality of those services. In 2006, financing of health services was transferred to the central budget, with the exception of the capital city. Demographic, Epidemiological, and Social Environment Although two-thirds of Krygyzstan's population is still rural, there is a high rate of rural to urban migration and emigration that is not officially registered. Life expectancy at birth fell in the early transition years from 68.3 years in 1990 to 65.7 years by the mid-1990s but began to recover and by 2004, at 68.2 years, was close to its pretransition level (WDI Dataset, 2006). The adult literacy rate is over 90 percent. Gross enrollment ratio in secondary schools is also improving after the record low of 85 percent in the mid-1990s (NSC Dataset 2005). The Kyrgyz Republic has a fairly young population: 31 percent of the total population is under 15 years of age. However, as shown in figure 10.2, the popula- tion is gradually getting older: the total fertility rate and the child dependency ratio have both declined in the last 15 years (from 3.69 to 2 births per woman, and from 65 to 50 children per 100 working age adults). Similarly to other countries in Central Asia and the Caucasus, the Kyrgyz Republic is confronted with a dual challenge of controlling high rates of both com- municable and noncommunicable diseases at the same time. In addition to high childhood mortality due to infectious diseases, countries in this region are strug- gling with staggering adult mortality due largely to cardiovascular diseases, cancer, and injuries (World Bank 2004a). Control of TB and HIV present a further chal- lenge to governments in this region. The Kyrgyz Republic: Good Practices in Expanding Health Care Coverage 273 Figure 10.2 The Kyrgyz Republic: Population Pyramids, 1990 and 2005 a. 1990 b. 2005 90­94 90­94 75­79 male female 75­79 male female 60­64 60­64 45­49 45­49 age age 30­34 30­34 15­19 15­19 0­4 0­4 20 15 10 5 0 5 10 15 20 15 10 5 0 5 10 15 % of population % of population Source: United Nations Population Division 2004. The leading cause of mortality and morbidity among both men and women in the Kyrgyz Republic is cardiovascular disease (table 10.1), and the largest gain in life expectancy in the Kyrgyz Republic, 4.52 years, will come from reducing that toll (World Bank 2004a). In terms of years of life lost due to premature mortality and disability, as measured by disability-adjusted life years (DALYs), neuropsychiatric Table 10.1 The Kyrgyz Republic: Leading Causes of Mortality and Disability, 2002 Total Total deaths Share DALYs Share Disease (thousands) (%) Disease (thousands) (%) Cardiovascular 21 47 Cardiovascular 199 17 diseases diseases Malignant 4 9 Neuropsychiatric 164 14 neoplasms conditions Respiratory 3 7 Unintentional 121 11 diseases injuries Infectious and 3 7 Infectious and 98 9 parasitic diseases parasitic diseases Unintentional 3 6 Perinatal 93 8 injuries conditions Other 11 24 Other 465 41 Total (all causes) 45 100 Total (all causes) 1,140 100 Source: WHO Global Burden of Disease Dataset 2004 (http://www.who.int/healthinfo/bodestimates/en/index.html). 274 Good Practice in Health Financing Figure 10.3 The Kyrgyz Republic: Leading Causes of Infant Mortality, 2004 respiratory diseases 19% other 6% causes arising during the perinatal period 58% congenital anomalies 12% infectious and parasitic diseases 5% Source: RMIC 2004. conditions are the second most important cause of disability after cardiovascular diseases; however, cancer is the second most important cause of death. Child mortality indices in the Kyrgyz Republic have been steadily improving. The infant mortality rate (IMR) declined from 28.2 per 1,000 live births in 1997 to 20.9 in 2003.6 Under-five (U-5) mortality during the same time period improved from 42.1 to 27.6 per 1,000 live births. However, these improvements hide large intra-country disparities with some regions showing almost no improvement. According to Ministry of Health (MOH) data for 2004, the high/low IMR ratio is 2.1 (RMIC 2004). One of the largest contributors to infant mortality in the Kyrgyz Republic is respiratory disease. Regardless of recent declines, deaths from respira- tory diseases still account for almost one-fifth of all infant deaths (figure 10.3). Infectious and parasitic diseases accounted for 16 percent of infant mortality in 1999. However, their share has been steadily declining, reaching 5.3 percent in 2004. As infectious and parasitic causes of infant deaths have declined in impor- tance, the share of causes arising during the perinatal period (including asphyxia, infection, and premature deliveries) has increased, from 33 percent in 2000 to 58 percent in 2004 (RMIC 2004). This trend highlights the need for continued atten- tion to prenatal care programs in primary health care (PHC) and the quality of care for management of deliveries and newborns.7 Although the infant mortality rate due to respiratory infections decreased sig- nificantly in all regions between 1997 and 2004 (table 10.2), the pace of improve- ment has been varied. The Issyk-Kul and Osh regions had similar indicators in 1997, but by 2004, they showed stark differences: 25.3 deaths per 10,000 births in Issyk-Kul as compared with 81.3 deaths per 10,000 births in Osh. The Kyrgyz Republic: Good Practices in Expanding Health Care Coverage 275 Table 10.2 The Kyrgyz Republic: IMR from Respiratory Infections, 1997­2004 (per 10,000 births) Percent change, Country/locality 1997 2001 2003 2004 1997­2004 Kyrgyz Republic 122.6 67.9 61.6 48.0 ­60.9 Bishkek 31.3 25.2 26.0 14.4 ­54.1 Batken oblast 150.4 105.7 106.3 77.8 ­48.3 Jalal-Abad oblast 145.7 78.7 65.1 4202 ­71.1 Issyk-Kul oblast 131.6 34.2 30.1 25.3 ­80.8 Naryn oblast 118.3 68.1 51.7 61.1 ­48.4 Osh oblast 136.1 84.6 90.4 81.3 ­40.3 Talas oblast 150.5 71.6 57.0 51.4 ­65.9 Chui oblast 60.5 38.0 39.9 25.2 ­58.4 Osh City -- -- 38.2 40.0 -- Source: RMIC, Health of the Population and Performance of Health Facilities in the KR. 1997­2004. Note: -- = not available. Political Context The Kyrgyz Republic, in the early years of independence, was viewed by the interna- tional community as one of the few success stories of economic and political reform in Central Asia and the country became known as an "island of democracy."8 In contrast to the other Central Asian states, independent media, nongovernmental organizations (NGOs), political parties, and civil society organizations were allowed to develop. However, starting in the mid-1990s, President Akaev began to tighten his grip on power, reducing the powers of parliament and curbing the activ- ities of some political parties, mass media, and NGOs. Growing opposition to the president's powers brought the country close to a crisis on several occasions (Insti- tute for War and Peace 2003). By 2005, the Akaev had lost support among citizens and alienated some of his former allies and key national and regional elites. Parliamentary elections in February to March 2005 were widely perceived as unfair and caused the initial wave of protests conducted mostly by supporters of individual candidates. Over the fol- lowing two weeks, the wider opposition joined in, and the agenda broadened to national issues. As the protests reached the capital, the president fled the country with his family and eventually resigned from the post he had held for 15 years. The new government, formed in 2005, has not yet met expectations of fur- ther democratization, increased transparency, and accelerated economic and social development, leading to widespread popular disillusionment. Economic and political reforms are slowly progressing, but the overall political uncertainty creates an unpredictable policy environment not conducive to long-term struc- tural reforms. 276 Good Practice in Health Financing Overview of Health Financing and Coverage This section describes the development of the health financing system under the Manas Health Sector Reform Program. Health financing trends, the evolution of coverage, and how financing and payment reforms aimed to increase coverage of health benefits are discussed. The discussion of the financing and payment reforms in two parts corresponds to two phases of the reforms. The first phase (1997­2000) was a preparatory phase with piloting of elements of the future health financing system and intensive capacity building. The second phase (from 2001 and continuing to this date) focused on the nationwide implementation of major structural reforms. At the end of the section, several indicators of equity and efficiency the reforms aimed to improve are reviewed. Trends in Health Care Financing Public spending on health declined from 4.0 percent of GDP in 1995 to a low of ­1.9 percent of GDP in 2001, improving to 2.2 percent in 2003 (World Bank 2004b). At the same time, out-of-pocket expenditures rose steadily (table 10.3). There are no reliable estimates of out-of-pocket expenditures prior to 2000, because most of these were informal and thus not officially acknowledged. According to most assessments, health expenditures declined both as a share of GDP and as a share of total government expenditures, showing that the decline in public spending resulted not only from the decline in public expenditures in gen- eral, but also from changes in government spending priorities. The share of health expenditures in both republican (central) and local budgets has declining since 1996 (figure 10.4). Declining public expenditures translated into budget shortfalls for covering the costs of existing structures and previous activity levels. As a result, patients were asked to contribute toward the cost of their care for basic items such as medicine, syringes, IV tubes, bandages, notebooks, light bulbs, linen, and food. These expenses were in addition to informal payments to health care personnel. In contrast to the "free" health care services provided during Soviet times, health care services had become expensive by 2000 for the average Kyrgyz household. By 2000, the mean out-of-pocket spending was KGS 337 (US$8). Conditional on reporting any contact with the health system (visit, hospitalization), the mean out-of-pocket payment was KGS 1,846 (US$46), the equivalent of five times monthly per capita consumption. Total out-of-pocket expenditures were com- posed of expenditures for outpatient visits (14 percent), outpatient drug pur- chases (58 percent), and payments associated with hospitalization (28 percent). Depth of Benefits Packages In Soviet times, the coverage of health services was both wide and deep. Publicly funded and provided services were available to all citizens. There was no explicitly defined benefits package, and there was no private sector (Sargaldakova et al. The Kyrgyz Republic: Good Practices in Expanding Health Care Coverage 277 Table 10.3 The Kyrgyz Republic: Key Health Financing Indicators, 2000­2007 2000 2001 2002 2003 2004 2005 2006 2007 In nominal terms Total health expenditures (million som) Budget 1,248 1,335 1,478 1,528 1,809 2,148 2,421 3,706 MHIF 105 120 142 197 338 255 467 581 OOP 1,521 1,885 2,254 2,628 3,091 3,491 3,922 4,406 SWAp 253 457 Total 2,875 3,340 3,874 4,354 5,238 5,893 7,062 9,150 Per capita health expenditures (in som) Budget 255 270 297 305 353 419 473 707 MHIF 21.5 24 29 39 66 50 91 111 OOP 310.8 382 453 524 604 682 766 841 SWAp 49 87 Total 587 677 779 869 1,023 1,151 1,379 1,745 As share of total health expenditures Budget 43.4% 40.0% 38.2% 35.1% 34.5% 36.4% 34.3% 40.5% MHIF 3.7% 3.6% 3.7% 4.5% 6.5% 4.3% 6.6% 6.3% OOP 52.9% 56.5% 58.2% 60.4% 59.0% 59.2% 55.5% 48.2% SWAp 3.6% 5.0% Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% As share of GDP Budget 1.9% 1.8% 2.0% 1.8% 1.9% 2.1% 2.2% 2.7% MHIF 0.2% 0.2% 0.2% 0.2% 0.4% 0.3% 0.4% 0.4% OOP 2.3% 2.6% 3.0% 3.1% 3.3% 3.5% 3.5% 3.2% SWAp 0.2% 0.3% Total 4.4% 4.6% 5.2% 5.3% 5.8% 6.1% 6.3% 6.6% In real terms Total health expenditures Budget 1,248 1,244 1,449 1,469 1,721 2,005 2,217 2,990 MHIF 105 112 139 190 322 238 428 468 OOP 1,521 1,757 2,210 2,527 2,941 3,259 3,592 3,555 SWAp 231 369 Total 2,875 3,113 3,798 4,186 4,984 5,502 6,236 7,382 Per capita health expenditures Budget 255 252 291 293 336 392 427 570 MHIF 22 23 28 38 63 47 82 89 OOP 311 356 444 504 574 637 692 678 SWAp 45 70 Total 587 631 764 835 973 1,076 1,246 1,408 Source: Adyljan Temirov et al., Policy Research Paper #48, CHSD, Bishkek, April 2008; FMR for 2007, MOH, March 2008. Note: GDP deflator 2000=100 (in som). Private expenditures for 2007 are projected based on the growth rate between 2005 and 2006. OOP = out-of-pocket. 278 Good Practice in Health Financing Figure 10.4 The Kyrgyz Republic: Health Expenditures as Share of the State Budget, 1995­2003 30 28.4 28.0 27.7 27.6 25.5 25.8 25 23.7 23.4 22.1 20 budget 15 13.5 12.3 health 11.6 11.7 10.8 of 10.1 9.9 10 9.0 % 8.5 5.8 5.4 5.3 4.7 4.8 4.5 4.3 5 3.7 3.4 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 state budget republican budget local budget Sources: Treasury data; WHO staff calculations. 2000). Health services, including inpatient drugs, were largely free of charge at point of service. The only exception was outpatient medicines and occasional gifts to providers. The Soviet health system provided high levels of financial protection. The provider network was extensive at both primary and secondary care levels, and geographic access for the population was good. The only official rationing of services was through waiting lists where several privileged groups of patients (e.g., war veterans, families of military servicemen or employees of the Ministry of Interior, pregnant women, disabled individuals, residents of high-altitude areas) received preferential treatment. As a legacy of the Soviet health system, all Kyrgyz citizens were entitled to the same benefits when the transition began. However, these benefits increasingly went unfunded after the drastic loss of budgetary revenues during the 1990s. According to laws existing at the time, health facilities could not, by charging offi- cial fees, cover the growing funding gap between their budgets and the cost of inherited benefits. They could charge only for services that were not part of the "guaranteed volume of services" and provided only by highly qualified specialists (Meimanaliev 2003). At the same time, the definition of what constituted "guaran- teed volume of health services" was vague and undefined. Thus, declining budgets and vague laws spawned informal charging practices. To continue delivering ser- vices, physicians began to ask patients to contribute to the cost of their care by buying medicines and supplies and paying medical personnel. These informally charged payments, unrecorded in accounting books, grew throughout the 1990s at The Kyrgyz Republic: Good Practices in Expanding Health Care Coverage 279 the discretion of medical personnel, unregulated by the state or even by hospital management. Although entitlements did not change during the transition period, drastic budget cuts and the emergence of informal payments led to a de facto reduction in the depth of coverage. From 1996 on, the Kyrgyz government became seriously engaged in health sec- tor reform. From a health financing perspective, the reform consisted of two phases. The first phase lasted from 1997 to 2001 with the introduction of the Mandatory Health Insurance Fund (MHIF) and a small complimentary payroll tax. The second phase was launched in 2001 with a complete reform of the fund- flow and purchasing mechanisms, benefits packages, and service delivery struc- ture. Both phases are described in greater detail below. Phase 1 began in 1997 with the introduction of the MHIF. A small payroll tax contribution was collected from the economically active (employed, self- employed, and agricultural workers). Pensioners and the registered unemployed were funded out of the pension and unemployment funds, respectively. Collection of these contributions was entrusted to the Social Fund, which was to make appropriate transfers to the MHIF (Law on Insurance Contributions for the State Social Insurance 1998; Law on Medical Insurance 1999). In the first three years after the introduction of the mandatory health insurance scheme, coverage reached 30 percent (Meimanaliev 2003). The main impediments to increasing enrollment were lack of trust in the insurance system, limited employment in the formal sector, and low tax compliance. The rate of coverage by the medical insur- ance scheme increased as new financing sources were added. Thus, in 2000, cover- age had reached almost 70 percent when the republican budget started funding insurance contributions for children under 16 years of age (including full-time students under 21) and social welfare recipients (Kutzin et al. 2001). By 2001, 83 percent of the population was covered by the mandatory health insurance. The remaining 17 percent were primarily unemployed individuals who failed to register as such. A large part of this group consisted of housewives, because insurance provided to formal sector employees did not cover their fami- lies unless they were covered as a member of some other group; for example, as recipients of welfare benefits paid to women with more than five children. The insurance status of students (18 to 21 years old) has been left unclear. According to legislation on the books, they are insured by the state, but the annual budget law has yet to provide funding for this group. In terms of depth of coverage, the introduction of the MHIF did not signifi- cantly change entitlements. The benefit offered by the MHIF to the insured relative to the uninsured was access to free drugs during inpatient care when the drugs were subject to user fees (table 10.4). As described above, however, because most out-of-pocket payments in hospitals were informal, this benefit was notional. At the same time, the additional public financing channeled to hospitals was meant to reduce formal or informal payments by patients. There is no direct evaluation of whether this occurred, but the effect was likely small because the share of the 280 Good Practice in Health Financing Table 10.4 The Kyrgyz Republic: Coverage in Phase-1 (1997­2001): Population Groups, Sources of Financing, and Benefits Population group Year Source of financing Services provided/Depth of coverage Formal sector 1997 2% from the wage bill administered Drugs at hospital level, salary bonuses employees, by the Social Fund (SF) 1998 Emergency care at the primary levela except civil servants 2000 Additional/outpatient drug package Civil servants 1998 2% from the wage bill administered Drugs at hospital level, salary bonuses by the SF Emergency care at primary level 2000 Additional/outpatient drug package Pensioners 1997 Value of 1.5 x minimum salary adminis- Drugs at hospital level, salary bonuses tered by the SF (Pension Fund) 1998 Emergency care at primary level 2000 Additional/outpatient drug package 2003 Republican budget Coverage unchanged Registered 1997 Value of 1.5 x minimum salary adminis- Drugs at hospital level, salary bonuses unemployed tered by the SF (Unemployment Fund) 1998 Republican budget Emergency care at primary level 2000 Additional/outpatient drug package 2002 Effectively not covered: no funding provided this group by Republican budget Children and 2000 Value of 1.5 × minimum salary Drugs at hospital level, salary bonuses students Republican budget under 21 Emergency care at the primary level Additional/outpatient drug package Welfare 2000 Value of 1.5 x minimum salary Drugs at hospital level, salary bonuses benefits Republican budget recipients Emergency care at the primary level Additional/outpatient drug package Farmers 1997 2% of the land tax administered by Drugs at hospital level, salary bonuses SF (5% in 2000, 6% in 2003) Emergency care at primary level Additional/outpatient drug package 2002 Health insurance policies (400 KGS = Coverage unchanged US$10/year) administered by MHIF Self-employed 1998 2% of the value of 3 x minimum salary Drugs at hospital level, salary bonuses Emergency care at primary level 2000 Additional/outpatient drug package 2002 Health insurance policies (400 KGS = Coverage unchanged US$10 / year) administered by the MHIF Sources: Law on Insurance Contributions for the State Social Insurance for 1998, 1999, 2000, 2003; Law on Medical Insurance of the Citizens of the Kyrgyz Republic, October 18, 1999, with amendments from April 21, 2003, and July 15, 2003. Note: The reimbursement principles: primary level, per capita for emergency care (1998); hospital care, per treated case for drugs and salary bonuses (1997). a. By law emergency care at the primary level is to be provided to the entire population but per capita reimbursement is provided only for the insured. The Kyrgyz Republic: Good Practices in Expanding Health Care Coverage 281 health sector budget allocated through the MHIF was less than 5 percent of total public expenditures on health in 1998 (Kutzin et al. 2001). The real benefit from the introduction of the MHIF in this first phase was institutional, allowing step- by-step introduction of population and output-based purchasing mechanisms in a previously input-based environment. Phase 2, launched in 2001, marked the introduction of significant health financing and service delivery reforms after a long preparatory and piloting period (5 to 6 years). A key instrument of these reforms was explicit definition of entitlements to coverage through the State Guaranteed Benefits Package (SGBP). The primary role of the SGBP was to enumerate the rights and obligations of patients and the state with regard to provision of health services and clarify the entitlements of different population groups (Government Decree No. 98, Febru- ary 25, 2002). The SGBP specifies the following: · Primary care is provided free of charge for the entire population with certain lab and diagnostic tests against copayment. · Hospital care is provided against formal copayment. Copayment is a flat fee payable upon admission. · Copayments vary with insurance status, exemption status, case type (delivery, surgery, medicine), and referral status (required written referral from primary health care physician. · Exemption categories were designed based on categorical targeting and disease types to protect populations with high expected health care use. Providers receive a higher payment for treating exempt patients to prevent selection. · An additional outpatient drug benefit was also introduced to subsidize the price of medicines for primary care sensitive conditions in order to reduce unnecessary hospitalizations (e.g., anemia, ulcers, pneumonia, hypertension). The use of copayment is regulated. Twenty percent can be used to complement the salary of personnel, and 80 percent must be used for key inputs such as medi- cines, medical supplies, laboratory supplies, and food. Copayment collections are reported monthly to the MHIF (Oblast Departments); use of copayment rev- enues, quarterly. The SGBP and the associated copayment system were not aimed specifically at any income group or geographic area. Copayment rates vary with insurance sta- tus, exemption status, case type (delivery, surgery, medicine), and referral status (see, for example, tables 10.5 and 10.6). Two exemption categories were designed, based on categorical targeting and disease types to protect populations with high expected health care use: · Targeting based on social categories. Assistance was intended to reach economi- cally vulnerable groups, but they were defined largely in terms of social and demographic characteristics, whereas, for example, war veterans, people over 282 Good Practice in Health Financing Table 10.5 The Kyrgyz Republic: Copayment Rates for Surgery, 2004 (KGS) Locality Exempt Insured Uninsured Without referral Bishkek, without Republican HF 260 900 1,602 2,330 Bishkek, Republican HF 260 1,080 1,440 2,730 Chui 260 780 1,130 2,120 Issyk-Kul 260 780 1,130 2,070 Naryn 260 690 1,040 2,060 Talas 260 690 1,040 1,900 Osh 260 560 920 1,550 Osh City 260 650 1,020 1,740 Jalal-Abad 260 650 1,020 1,720 Batken 260 560 920 1,640 Source: Law on State Guaranteed Benefits Package for 2004. Table 10.6 The Kyrgyz Republic: Copayments for Treatment without Surgery, Diagnosis, Minor Surgery, 2004 (KGS) Locality Exempt Insured Uninsured Without referral Bishkek, not Republican HF 200 690 970 1,790 Bishkek, Republican HF 200 830 1,110 2,100 Chui 200 600 870 1,630 Issyk-Kul 200 600 870 1,590 Naryn 200 530 800 1,580 Talas 200 530 800 1,460 Osh 200 430 710 1,190 Osh City 200 500 780 1,340 Jalal-Abad 200 500 780 1,320 Batken 200 430 710 1,260 Source: Law on State Guaranteed Benefits Package for 2004. 75 years of age, and the disabled regardless of their income were fully exempt from any fees. · Targeting based on medical condition/disease type. Interventions to prevent and cure diseases with important public health consequences and externalities were also exempt from charges (TB, AIDS, syphilis, polio, diphtheria). Hospitals set up reserve funds to subsidize copayments for the uninsured very poor that did not fall under any of the exemption categories. These funds were to The Kyrgyz Republic: Good Practices in Expanding Health Care Coverage 283 be financed by setting aside 10 percent of copayments. About 9.5 percent of patients in Chui and Issyk-Kul received free treatment, 43 percent of them funded from hospital reserve funds from March to December 2001 (Ibraimova 2002). However, there is no systematic monitoring of the functioning and impact of these reserve funds. The Outpatient Drug Benefit Program, also called the Additional Drug Package (ADP), was introduced in 2000 as a small pilot test in seven pharmacies and grad- ually rolled out to the entire country by 2003. The ADP allows insured patients to buy prescribed drugs at discount prices at pharmacies under contract to the MHIF. The subsidy amount, based on a reference price, is about 50 percent of the refer- ence price. The fund reimburses the difference between retail price and the patient pays the pharmacy. By 2004, the number of participating pharmacies and drug sales outlets more than doubled, to 800, up from 357 in 2002 (MHIF Dataset, 2006). However, the geographical disparity in distribution noted earlier is also pre- sent here: the high/low ratio between oblasts is 8.8 (excluding the capital). For example, in 2004, in Jalal-Abad oblast there were 35 pharmacies as compared with only 4 in Talass oblast (MHIF Dataset, 2006). The list of ADP-reimbursed drugs was carefully selected from the essential drug list to include those necessary for managing primary care­responsive conditions (Meimanaliev 2003). A constant effort is being made to improve cost-effectiveness and to include as many generic drugs as possible. The share of generic drugs pre- scribed by doctors has increased to more than 90 percent, and the package includes 74 generics (MHIF Dataset, 2006). According to a recent evaluation of the outpa- tient drug benefits program in the Kyrgyz Republic (Ibraimova, Kadyrova, Jakab, draft), the introduction of the ADP has led to reductions in overall prices and in price variations across the geographical areas for covered medicines (e.g., atenolol for management of hypertension). In itself, the SGBP did not change the de facto coverage gap that emerged because of lack of public funding, but it did clarify entitlements. It made it clear that public funds would cover about 50 percent of hospitalization costs and patients were responsible for paying for the rest. Out of necessity, this was the pre- vailing practice, but physicians were free to charge whatever they liked. The intro- duction of the SGBP aimed to end this practice and match entitlements to available public funding and eliminate unfunded mandates. The success of these reforms is examined later. Financing and Payment Revenue collection, pooling, and purchasing fall into three distinct periods: pre- 1997, 1997 to 2000, and 2001 to 2006. Revenue collection, pooling, and purchasing prior to 1997. Prior to 1997, health services were funded from general tax revenues and out-of-pocket payments. Public funding was fragmented along administrative structures: republican-level 284 Good Practice in Health Financing (central-level) providers were funded from national taxes; oblast (region) facili- ties, from oblast taxes; city facilities, from city taxes; and rayon (municipal) facili- ties, from rayon taxes. There was no funding and decision making across these administrative boundaries. This fragmented system of resource allocation led to overlapping population coverage in most urban centers and undercoverage in rural areas. For example, residents of Bishkek had access to both the republican and city health systems, while rural dwellers had direct access only to rayon health facilities unless they traveled to Bishkek or an oblast center. In addition, the capi- tal and oblast centers had more resources than the other areas to fund their health systems. The need to eliminate this duplication of services became acute urgency with the sharp decline in the state health budget when 75 percent of government spending went for staff and infrastructure, leaving little for direct medical expen- ditures. The resulting savings could be allocated to underserved areas and variable costs, but first resource allocation had to be changed. According to the input-based budgeting then current, annual budget alloca- tions for health facilities depended on the number of beds and staff positions no matter what the output. Eighteen input categories were used for budgeting (e.g., personnel, drugs, utilities), and managers could not reallocate across line- item categories if need or opportunity arose without a long procedure of obtaining permission from the Ministries of Health and Finance. Moreover, there was no mechanism for reinvesting savings. For instance, if a health facility managed to decrease its utilities costs, it could not reallocate those savings to increase staff salaries or buy medicine. The input-based budgets, allowing no managerial author- ity to move funds across different line items, had to be replaced to give managers incentives to increase throughput with the minimal amount of resources. Most important, however, this could increase the depth of coverage because resources freed up through increased efficiency would allow health managers to acquire more drugs and medical supplies that patients would otherwise have to cover. Revenue collection, pooling and purchasing 1997­2000. The revenue collection and pooling arrangements changed slightly with the introduction of the MHIF in 1997. Most revenues remained as described above, and the resources managed by the MHIF were added to the revenue mix as a small complementary public resource allocation mechanism. In 1998, the general tax­funded share of the health sector amounted to 95 percent of public expenditures on health, while rev- enues pooled through the MHIF made up the remaining 5 percent (Kutzin et al. 2001). During this phase, there was no change through the budget in the purchas- ing mechanisms, which remained based on historical line-item budgets. The real significance of this reform period was to create the foundation for the second phase in the new MHIF institutional structure and allow it to learn pur- chasing and use of prospective provider payment mechanisms using small amounts of funds. The MHIF began to contract with a small number of providers and in two years had contracts with all central district (rayon) hospitals. For these contracted facilities, the MHIF introduced new reimbursement procedures: case- The Kyrgyz Republic: Good Practices in Expanding Health Care Coverage 285 based payment to hospitals and capitation-based payment to primary-level health facilities. These will be described further in the next section, because they gained in importance as the MHIF-managed resources grew. Revenue collection, pooling, and purchasing 2001­06. The introduction of signif- icant health financing reforms began in 2001 in two oblasts. The reforms were sys- tematically rolled out to the rest of the country by including two more oblasts every year, gradually changing the health financing system by 2004. In the Kyrgyz Republic, these reforms are referred to as the "Single-Payer Reforms," capturing well their main intent: to end fragmentation of the pooling mechanism by creat- ing regional purchasing pools. Oblast purchasing pools were established as departments of the MHIF, pooling all oblast, rayon, and city tax revenues. Oblast finance departments transferred all tax revenue allocated for health into the oblast departments of the MHIF. These tax revenues were complemented by the payroll tax revenues collected by the MHIF and transferred to providers using the same payment mechanism. This change created an opportunity to reallocate resources across city-rayon bound- aries within oblasts.9 The role of the MHIF changed dramatically with the introduction of single- payer reforms because it was made responsible for pooling state budget funds as well as health insurance premiums at the regional (oblast) level. Thus, it became the sole purchasing agency for health care services in every region. It also gained additional roles in quality assurance and the development of health information systems because these were considered part of the contract-monitoring functions. Once oblast purchasing pools had been created, the old line-item provider pay- ment mechanisms had to be replaced. Once pooled at the oblast level, public funds were transferred to capitation-based outpatient care and to hospitals on a per-case basis. Payment rates per case were defined prospectively, and payment to hospitals was made retrospectively (monthly). The system for grouping cases is modeled on the U.S. diagnosis-related-groups (DRGs) but was created from Kyrgyz utilization and cost data. The case-based payment system introduced the concept of output- oriented payment to the Kyrgyz health system. According to Kutzin and others (2002: 12), "in doing so, the MHIF challenged one of the fundamental weaknesses of the former system: low productivity." This change gave providers incentives to downsize within their facility. The capitation rate for PHC facilities is based on the available budget and the estimated number of insured persons that will be enrolled in family group prac- tices (FGPs) for the coming year. Previously, citizens were assigned to a district physician according to their place of residence. The new regulation allows them to choose the FGP in which they want to enroll and to switch at each annual registra- tion period. So, in theory, practices that attract more patients receive more capita- tion funds, an incentive to provide high-quality and user-friendly services to attract users (Atun 2005). In practice, this works only in large urban settings because rural areas and small towns have relatively few providers. Allocation of primary care 286 Good Practice in Health Financing resources based on a capitation formula adjusted by coefficients for rural, small towns, and mountainous areas allows equalization of resources across the country. Beyond the financing mechanism, these reforms have important institutional dimensions. The MHIF was established as a parastatal social insurance fund under the Ministry of Health. It was important to create the oblast purchasing pools out- side the core public bureaucracy and establish them as part of this parastatal entity. Had centralization of pooling taken place within the confines of regular oblast budgets, prospective payment could not have been implemented. Because public administration and budgeting was (and still is) governed by the same rigid input- based, line-item budgeting mechanisms as during Soviet times, the incentives for downsizing and savings would not have materialized. Pooling alone without changing provider payment mechanisms would not have changed resource flows and allocation patterns. From 2006, the oblast purchasing pools were further centralized into one national purchasing pool. This shift offers further scope for cross-subsidization across regions and should lead to further improvement in geographical equity. It also expands the risk pool to the national level, a positive development for a coun- try with a small population. The changes made as a result of this new pooling arrangement do not have direct implications on any other elements of the system, such as source of funds, collection, or purchasing methods (Kutzin, O'Dougherty, Jakab 2005). This is important because it means no significant additional training is required for administration of the system--which could expose the key stake- holders to "reform" exhaustion. Equity The drop in public funding and the growth of out-of-pocket payments affected demand for health services. By 2000, significant inequalities in utilization emerged, leading to a disproportionate capture of public expenditures by the rich and conse- quently poor targeting. The utilization ratio of the PHC and outpatient specialist services among the poorest quintile as compared with the richest quintile was about 2:5 (figure 10.5). The difference in utilization by income was slightly smaller in case of hospitalization: the ratio of the poorest quintile to the richest quintile was about 3:5. Limited data are available on how these trends affected the distribution of health outcomes. The 1997 DHS Survey is the only data point on the distribution of health outcomes in the period prior to reforms, and it does show the expected poor-rich gradient. According to the survey, infant mortality was 1.8 times higher in the poorest quintile than in the wealthiest, child mortality was almost twice as high, and the stunting rate was 2.4 times higher among children from the poorest quintile than among the richest. Considering the pronounced growth in poverty during the 1990s, it is difficult to disentangle causality and attribute the emerging health inequalities to deterioration of general household welfare or to health sys- tem factors such as erosion of coverage. The Kyrgyz Republic: Good Practices in Expanding Health Care Coverage 287 Figure 10.5 The Kyrgyz Republic: Access to Health Services, by Income Level, 2000 20 one 15.62 15 least at 10.29 hospitalization 10 reporting or visit 6.51 5.78 population 5 doctor of % 0 PHC hospitals poorest richest Sources: NSC HBS 2001, and HSC HIS 2004; WHO staff calculations. Efficiency Limited resources and excess capacity were general features of the health sector in all FSU countries, and the Kyrgyz Republic was no exception. As shown in figure 10.6, the Kyrgyz Republic had a large hospital infrastructure even compared with wealthier countries. The number of hospital beds per 100,000 in the Kyrgyz Republic exceeded the number of hospital beds in the European Union (EU) by more than 25 percent (863.6 against 687.9). Figures for the average length of stay and hospital admission rate were also higher for the Kyrgyz Republic than in the EU member countries, both old and new, but lower than the CIS average. The increase in inequality demonstrated earlier and the erosion of coverage were inexorably linked to health system inefficiencies. First, two-thirds of the health sec- tor budget was spent on expenditures related to utilities and personnel. This was due to the combination of the ponderous infrastructure inherited from Soviet times, which was sucking away a bigger and bigger piece of a shrinking health care budget. As a result, fewer resources were left for other inputs, such as medicines and supplies, also needed for good quality medical care. So until downsizing could be finished, the inefficiencies manifested in the large hospital infrastructure meant that coverage would remain shallow and patient contributions high. Second, in addition to poor technical efficiency outlined above, there were problems with allocative efficiency in distributing health expenditures between primary and hospital care. Resources were allocated disproportionately to hospital- 288 Good Practice in Health Financing Figure 10.6 The Kyrgyz Republic: Comparative Indicators of Hospital Efficiency, 1995 a. Hospital beds per 100,000 people 1,200 1,129.5 1,000 863.6 800 760.1 887.9 600 number 400 200 0 Kyrgyz Commonweath of European European Union Republic Independent States Union members since May 2004 b. Average length of stay 18 16.42 16 14.70 14 12 11.18 11.54 10 days 8 6 4 2 0 Kyrgyz Commonweath of European European Union Republic Independent States Union members since May 2004 based curative care while primary health care was neglected. Primary care did not fulfill the gatekeeper function, and there was a strong tradition of self-referral by patients to specialists and hospitals. In 2001, when the major reforms in health care financing were launched, only 15 percent of total health expenditures went to pri- mary care facilities. Hospitalizations were frequent for conditions that could be treated in primary care (e.g., hypertension, ulcers, anemia, and pneumonia). This inefficiency also added to the financial burden on patients and contributed to increasing inequality in use patterns presented above. The Kyrgyz Republic: Good Practices in Expanding Health Care Coverage 289 Figure 10.6 The Kyrgyz Republic: Comparative Indicators of Hospital Efficiency, 1995 (Continued) c. Inpatient care admissions per 100 people 25 20.0 20 17.0 17.8 16.6 15 number 10 5 0 Kyrgyz Commonweath of European Union European Union Republic Independent States members since May 2004 d. Health expenditures, by category utility costs direct medical 13% expenditures 25% personal expenditures 62% Sources: WHO Health for All Database; Central Treasury data quoted in G. Purvis et al. 2005. It was widely recognized that tackling inefficiencies by reducing hospital capac- ity and shifting treatment for primary care­responsive conditions out of hospitals was a precondition to improving coverage and equity in the health system. In light of postindependence economic developments, the inherited infrastructure and care patterns were not sustainable. Efficiency gains in the health system were the only possible way to increase coverage in this fiscal context. 290 Good Practice in Health Financing Overview of Health Delivery System To achieve efficiency gains, service delivery had to be restructured and reoriented. This section focuses on the changes in the inherited Soviet service delivery system (Atun 2005; Kutzin 2002; and Meimanaliev 2003). Supply and Organization of Hospitals, Physicians, and Public Health Programs This section describes primary, secondary, and tertiary care service delivery sys- tems and reforms, including organization, physical infrastructure, and human resources. The section concludes by showing the impact of the changes on the effi- ciency of service delivery as seen in the allocation of public expenditures between the fixed and the variable costs. Primary health care. There are three types of primary health care providers in the Kyrgyz Republic: Feldsher-Obstetrical Ambulatory Points (FAPs), family group practices (FGPs), and family medicine centers (FMCs). FGPs are the main providers of PHC. An FGP usually consists of three to five doctors. There are three organizational forms of family group practices: freestanding and autonomous, a unit within a free-standing polyclinic (in urban areas), and a unit within a hospital-based polyclinic. FGPs have to meet licensing and accreditation criteria before they can be contracted by the MHIF (Atun 2005). Of the 701 FGPs in the country, 670 are part of FMCs (RMIC Dataset, 2006). FAPs, which are health care facilities in remote rural areas, offer the most basic services such as antenatal and postnatal care, immunizations, and health educa- tion. The 875 FAPs across the country each serve between 500 and 2,000 people (RMIC Dataset, 2006). FMCs are the largest outpatient health facilities, staffed by 10 to 20 medical professionals. They offer services ranging in scope from general care to specialized care and instrumental diagnostics, thus combining primary care services and secondary outpatient care. FMCs, renamed oblast- and rayon- level polyclinics, are now staffed by general practitioners and specialists, who also work in outpatient departments. There are now 85 FMCs in the country, which include 670 FGPs and 816 FAPs (RMIC Dataset, 2006). PHCs are generally accessible, even in rural areas. Primary health care facilities are located close to patients' homes, with a median distance of 1 to 2 kilometers (Atun 2005). For most patients (73 percent) the travel time to the nearest health facility is less than a half hour. Most patients walk, and only one in three incurs travel expenses when attending health facilities (26 percent in Bishkek to 43 per- cent in Batken) (Atun 2005). However, the physical presence of a FAP or an FMC does not always translate into access to health services because of shortages of basic medical supplies and, increasingly, health professionals willing to work in rural areas. Poor living conditions, low salaries, and a rising demand for health workers in more prosperous neighboring countries such as Kazakhstan and Rus- The Kyrgyz Republic: Good Practices in Expanding Health Care Coverage 291 sia are the main factors behind the growing problem of unfilled vacancies in many rural health facilities (MOH 2006). Soviet medicine put heavy emphasis on hospital care, and primary care was neglected. Primary care facilities did not have independent legal and administra- tive status and functioned mostly as hospital subdivisions. The result was low- quality and underfinanced primary care providing a narrow range of services and limited gatekeeper function. To address this, early health system reforms in the mid-1990s focused on replacing the system of district doctors with family practi- tioners along the following main lines: · Primary care was organizationally and financially separated from hospital care. Independent FGPs were set up. The population was required to enroll with an FGP of their choice. · Human resources training was overhauled in an attempt to institutionalize family medicine training at three levels: undergraduate training; postgraduate training (two-year clinical internship for graduates of higher medical edu- cation institutions); continuous training (training trainers, such as doctors and nurses, in family medicine; retraining doctors and nurses in family medicine; and qualifications upgrading of practicing specialists and trainers in family medicine). Most family doctors and nurses (between 60 and 70 percent) par- ticipated in retraining in the field of family- and evidence-based medicine under different programs supported by ZdravPlus, Scientific Technology and Language Institute (STLI), the Swiss Cooperation Office, and the World Bank. · On the basis of evidence-based medicine, 162 clinical protocols on the most prevalent diseases were developed and introduced. · The State Guaranteed Benefits Package (SGBP) was introduced, specifying people's health care benefits and payment obligations. Within the framework of the SGBP, all citizens, irrespective of their insurance and enrollment status, are entitled to free primary health care services. · An ADP was introduced, entitling insured individuals to drugs for primary care­responsive conditions at reduced prices. · Primary health care began to receive additional funding from payroll taxes through the MHIF. · New provider payment methods were successfully introduced for FGPs based on per capita financing, including partial fund-holding for medicines in the ADP. Capitation payments are a more equitable way of distributing resources at PHC level than line-item financing based on historical performance. · Outreach activities have been initiated to improve relations with the public and involve local communities in the decision-making process relating to PHC services. 292 Good Practice in Health Financing Secondary and tertiary care. Secondary care is provided by oblast merged hospi- tals at oblast level and territorial hospitals at the level of rayons and cities. The structure of all in-patient care facilities at oblast and rayon levels is similar and consists of an administrative-logistical unit (management, human resources, planning, economic, and accounting departments), outpatient diagnostic depart- ment (ODD) and treatment units. The main difference between the territorial and oblast hospitals is in the number of treatment units--that is, in the variety and complexity of conditions that can be treated on site. There are 51 territorial hos- pitals (former city- and rayon-level hospitals), including 1 children's hospital in Bishkek, and 7 oblast hospitals (one per region). Tertiary care is provided in eight central-level hospitals, including the National Center for Oncology, the National Cardiology and Therapy Center, and the National Center for Surgeries. These facilities provide highly specialized services and are meant to serve the entire country, although studies of referral patterns show that 90 percent of patients reside in the capital and neighboring Chui oblast. The medical personnel at these hospitals also are better educated and more expe- rienced than staff members working in secondary-level hospitals (Meimanaliev 2003). According to Article 4 of the Law on Health Care Organizations (August 13, 2004), there are two types of ownership of health care organizations: private and public. The private sector has a marginal role in service delivery, specializing mostly in optic and dental care services, gynecology, and urology. Privately owned health facilities are concentrated in the capital. The limited evidence available sug- gests a growing, though small, number of private sector providers. As explained above, hospital network downsizing was essential to improve technical and allocative efficiency. This was achieved through centralizing the pooling mechanism and changing provider incentives and administrative mechanisms. Introducing case-based payment created incentives for down- sizing, and pooling created the opportunity to rationalize across administrative boundaries. As most hospitals were built on a pavilion design, operating in 15 to 20 small buildings, within-facility downsizing had great potential for savings on fixed costs. The unnecessary buildings were demolished, rented out (e.g., to pharmacies), or transferred to other public uses (e.g., health promotion cabinets). In 2001­04, physical capacity in the hospital sector was reduced from 1,464 buildings to 784, and the square footage was reduced by 39.6 percent with corresponding changes in the total operational area, utility costs, and maintenance costs (Purvis et al. 2005). Simultaneously, personnel were cut by about 30 percent. At the same time, across- facility downsizing involved merging facilities serving overlapping populations through administrative mechanisms. This allowed reallocation of expenditures from fixed costs (personnel, infrastructure, and maintenance) to direct medical expenditures (drugs and supplies) (figure 10.7). The Kyrgyz Republic: Good Practices in Expanding Health Care Coverage 293 Figure 10.7 The Kyrgyz Republic: Reallocation of Public Expenditures in the Single-Payer System from Fixed Costs to Variable Costs 100 90 16.4% 36.6% 80 70 60 50 40 percentage 30 83.6% 63.4% 20 10 0 2000 2004 maintenance of infrastructure and personnel patients Source: MHIF Dataset in Purvis et al. 2005. Regulatory Framework The Law on Health Protection (January 9, 2005) established the general legal framework and articulated roles and responsibilities of state bodies involved in health care service provision. It is a key legal document that regulates the quality of health services, including licensing and accreditation processes for health pro- fessionals and organizations (Article 81). The Medical Accreditation Commission (MAC) was established in 1997 as an independent agency responsible for accredi- tation of health organizations. At its current capacity, the MAC can carry out 50 accreditation surveys a year. Accreditation for health facilities is voluntary. The validity periods vary according to the score received. The Medical Insurance Law (October 18, 1999) provided a basis for financing the health care system through compulsory and voluntary medical health insurance. According to the Law on Health Care Organizations (August 13, 2004), "state and municipal health organizations have financial autonomy within the bound- aries set out in their organizational charter," which is approved by the owner (Article 35). Organizational charters are developed on the basis of standard char- ters that are approved by a Ministry of Health order. These charters allow health facility management to make decisions on allocation of resources across different inputs. Moreover, management is allowed to maintain commercial bank accounts for copayments retained by the facility. Retention by the facility is particularly important: previously all officially collected user fees were turned over to the cen- 294 Good Practice in Health Financing tral budget where they could be reallocated to other needs. The lack of "own resources" contributed to the rise in informal payments and lack of transparency in financial management of health facilities. Other Components of the Delivery System Medical education, pharmaceuticals, medical technology, and health information systems are other vital parts of the delivery system. Medical education. Until the recent closures, eight higher educational facilities provided medical education in the Kyrgyz Republic. Recognizing the growing problem of poor quality of medical education, a joint commission led by the Min- istry of Education closed five institutions and one pharmaceutical education department--none of them accredited. This action was welcomed as an attempt to reduce surplus capacity in medical education and improve the quality of remaining training programs. Graduate and postgraduate training is provided by the Kyrgyz State Medical Academy (KSMA), and the Kyrgyz State Medical Insti- tute for Retraining and Continuing Medical Education (KSMIRCME). Family medicine principles have been introduced into undergraduate curricu- lums, but undergraduate programs have not been reoriented toward general prac- tice, and students still choose their specialized educational track the first day of medical school. Teaching remains largely theoretical and didactic, with limited clinical opportunities for medical students to supervise patient care, and no unified educational standards across educational institutions. KSMA and KSMIRCME are participating in regional efforts that have been initiated with donor assistance to begin to move toward the World Federation of Medical Education accreditation standards. As part of the National Health Reform Strategy for 2006­10, a working group has been set up to review the undergraduate medical curriculum and develop a strategy to improve both its content and the teaching/educational processes used to teach it. Strengthening family medicine programs and making them more attractive for students has a direct impact on ensuring access to health care services, particularly for the rural population. As noted in the discussion of health service delivery, the lack of family doctors, particularly in rural areas, is increasingly problematic. There is also a constant shortage of applicants to family medicine programs. In 2005, KSMA offered 70 clinical residencies but filled only 7 places; in 2006, it offered 50 residencies and filled only 14. There are 14 medical (nursing) colleges located throughout the Kyrgyz Repub- lic where nurses and feldshers10 are trained. In 2005, 2,365 people were graduated from medical colleges. The nursing curriculum has been revised, the training period extended to three years, and new disciplines introduced (e.g., introduction to law, reproductive health and family planning). New government educational standards have been developed for eight specialties for mid-level medical person- nel, including pharmacists. The Kyrgyz Republic: Good Practices in Expanding Health Care Coverage 295 Pharmaceutical sector. The pharmaceutical sector changed dramatically in 1997 with the adoption of a new drug policy that allowed privatization of pharmacies (Meimanaliev 2003). Simultaneously, the MOH established a Department of Phar- maceutical Provision and Supplies as the main regulatory body responsible for accreditation of pharmacies, regardless of the ownership, and licensing of pharma- cists and pharmaceutical organizations (Government Decree No 556, September 26, 1997). By 2000, there were 1,285 pharmacies and drug outlets in the Kyrgyz Republic (MOH Report 2004). However, they are unevenly distributed across the country, because running a drug outlet in remote rural areas is unprofitable. To address this issue, the MOH allows personnel with clinical backgrounds, after a month of training, to sell drugs in villages that do not have a private pharmacy. Two laboratories control the drug quality. Currently, 70 percent of imported drugs must be tested, and all of them are tested. The remaining 30 percent come from reliable manufacturers and are exempt from testing (GMP certified). Medical technology. The use of advanced medical technology is limited first and foremost, by the highly constrained resource environment. In 2002, recognizing that sustainable, transparent, and equitable mechanisms had to be instituted for providing expensive, high-technology medical care, the government created the High-Technology Fund (Government Decree No 287, May 7, 2002). The fund is financed from four sources: the republican (central) budget, external donors, pri- vate charitable organizations, and humanitarian assistance. Since its creation, however, it has been chronically underfinanced and has had little impact on patients' access to high-technology services. The shortage of financing, at least from external donors, is partly due to lack of clear operational guidelines for the fund. The decree is vague on basic rules of access to services. The government has yet to decide on the maximum amount allowed per service or procedure, rules for allocating the available budget across types of procedure, rules for determining eligibility (clinical and financial needs), and the extent of and any exemption from copayments. Health information system. The Kyrgyz Republic has a well-developed health information system for a country of its income level. The system has three main levels: national, the MHIF; oblast, the oblast departments of the MHIF; and pro- vider level. This system feeds information into five databases: the MHIF database on the insured population; the enrollment database with primary care facilities to enable capitation-based payment; the hospital admission database with case cod- ing to enable case-based payment managed by the Republican Medical Informa- tion Center; the outpatient care utilization registry also managed by the RMIC; and the Additional Drug Package. However, because the health management information system development was driven largely by the needs of health care financing reforms, it has largely ignored other important areas. Thus, information on health status and quality of care indicators is not always consistent and reliable. Also, financial resources 296 Good Practice in Health Financing needed for system maintenance are lacking. High telecommunication costs also pose a significant barrier to further HMIS development. The Health Coverage Reforms Depth of coverage was the problem that emerged in the Kyrgyz Republic during the transition period. At the start of transition, the entire population was entitled to free health services of every kind, with the exception of a few specialized ser- vices and outpatient drugs for which small fees were charged. During Soviet times, these entitlements were funded, and an extensive service delivery infra- structure reached most rural areas, at least with paramedical if not primary health care posts. Until 1991, coverage was high--both in terms of breadth and depth. The dramatic loss of budget revenues after independence from the Soviet Union made these entitlements unaffordable and increasingly unfunded. In the early transition years (1991­95), providers were left to their own devices to find a way to adjust to this dramatically reduced financing. At the time, the input-based budget formation process, which rewarded capacity, gave no incentives for effi- ciency. Thus, the adjustment took place by introducing informal charging prac- tices. Patients were required to pay for medical personnel, medicines, medical supplies, and even nonmedical supplies. Although entitlements did not change on paper during this time, the depth of coverage eroded, marked by increasing out-of-pocket payments. There was a mismatch between de jure and de facto enti- tlements, and a deep sense of disillusionment with the health system. In the 1990s, the Kyrgyz Republic emerged as the second poorest country in the FSU. Poverty rates hovered around 50 percent, and the potential for fast eco- nomic growth was crippled. It became clear that the fiscal context would remain limited, at least in the short run, and that the depth of coverage could not be restored to pretransition levels from the depleted public coffers. Reforms to create a transparent and sustainable system of coverage required two difficult steps: (1) eliminating the mismatch between the de jure and de facto entitlements by clarifying the respective roles of the government and individuals in funding health services within the prevailing fiscal space, and (2) downsizing the inherited service delivery infrastructure to achieve and rechannel efficiency gains from fixed costs to variable costs and lighten patients' financial burden. Because the coverage problems and their causes were entangled in a broad web of performance problems and inadequacies of the inherited Soviet health system, a systemic approach was necessary. The ensuing reforms had to encom- pass the entire spectrum of the health system; they could not address isolated reform objectives or use single instruments. The reforms thus reposed on four main pillars: · Clarification of entitlements through the introduction of the State Guaranteed Benefits (SGBP) package. The primary role of the SGBP was to define the rights The Kyrgyz Republic: Good Practices in Expanding Health Care Coverage 297 and obligations of patients and the state with regard to provision of health services and clarify the entitlements of different population groups. In itself, the SGBP did not change the de facto coverage gap that emerged due to lack of public funding, but it did clarify entitlements. It was based on the acknowledg- ment that public funds were sufficient to cover only about 50 percent of hospi- talization costs and that patients were responsible for paying for the other 50 percent. Out of necessity, this was the prevailing practice but charging practices were left entirely to the discretion of physicians. The introduction of the SGBP aimed to end this practice, match entitlements to available public funding, and eliminate unfunded mandates. · Centralization of previously fragmented pooling arrangements. Oblast purchas- ing pools were created in 2001 in two oblasts and gradually introduced nation- wide by 2004. The new system replaced the previously compartmentalized financing arrangements with city revenues funding city hospitals, rayon rev- enues funding rayon hospitals, and oblast revenues funding oblast hospitals. In this fragmented context, the service delivery system could not be restructured across these rigid administrative boundaries. Thus, pooling reform was a pre- condition for reconfiguring the service delivery structure and thus to achieve efficiency gains. In 2006, the pooling arrangements were further centralized at the national level, moving to one pool for purchasing most health services and promising great potential for equalizing funding across the country. · Replacement of historical line-item, capacity-linked budgeting with population- and output-based provider payment mechanisms. Pooling was a necessary but insufficient condition to initiate restructuring of the hospital sector. As long as the budgeting process was based on capacity and the payment mechanism was based on rigid line items, there were no incentives to downsize. The intro- duction of case-based payments was a milestone in this regard. The health sec- tor became unique in the public administration by severing the link between inputs and budgets. · Restructuring of service delivery with strengthened primary health care and more efficient and streamlined hospital care. Pooling and payment reforms created an enabling environment for downsizing hospital capacity. Downsizing was not left entirely to the discretion of hospital management. The MHIF, the newly formed Hospital Association, and a qualified consulting company worked closely with each hospital to demonstrate the benefits of downsizing and develop a restructuring plan. Outside the capital city, significant downsizing took place, as described above, with documented reductions in fixed costs. Simultaneously, arduous efforts were made to strengthen primary care and broaden the scope of services. To reduce unnecessary hospitalizations, particu- lar attention was devoted to primary care­responsive conditions such as ulcers, pneumonia, hypertension. 298 Good Practice in Health Financing In sum, the Kyrgyz coverage reforms were closely linked to an overall reform agenda trained on the creation of an efficient and sustainable health system. Chronology The Kyrgyz health system reforms described above became known as the Manas National Health Care Reform Program (1996­2005). The reforms were imple- mented in two main phases. Phase 1 (1996­2000) focused on refining the reform design; building capacity in the health sector for policy development, implemen- tation, and evaluation; launching primary care reform; establishing the MHIF; and piloting new financial mechanisms. Phase 2 (2001­05) involved major struc- tural changes in health care financing and service delivery (box 10.1). In 2006, a follow-on phase was introduced with the Manas Taalimi National Health Care Reform Program (2006­10). In this phase, the achievements of the Manas reforms will be taken forward, the remaining weaknesses addressed, and a new generation of reforms introduced. B O X 10.1 The Kyrgyz Republic: The Manas National Health Care Reform Program 1997­2000 2001­05 Establishment of the Mandatory Health Development and Implementation of Insurance Fund and piloting of reforms Single-Payer System 1. Approval of the Manas National Health 1. Government Decree on Introduction of a Care Reform Program (1996) New Health Care Financing Mechanism in 2. Family group practices enrollment Health Facilities (2001) campaign in Issyk-Kul oblast (1996) 2. Government Decree on Program of State 3. Law on Drugs (1997) Guarantees fo Issyk-Kul and Chui Oblasts 4. Licensing and accreditation process (2001) started (1997) 3. Piloting of the Outpatient Drug Benefits 5. Case-based payment to hospitals from Package in Bishkek (2001) insurance funds introduced, rolling-out of 4. Transfer to output-based financing in primary health care reforms to three addi- Issyk-Kul and Chui oblasts (2001) tional oblasts and the capital city Bishkek 5. Single payer system joined by Naryn and (1997­98) Talas, followed by Batken, Jalal-Abad, and 6. Case-based payment from budgetary funds Osh oblasts (2002­03) to selected hospitals introduced in Bishkek 6. Countrywide introduction of Outpatient (1999) Drug Benefits (2003) 7. Laws on Interpretation of the Law on Princi- 7. Countrywide coverage with State ples of the Budget and the Law on Local Guaranteed Benefits Package (2004) Self-Governments, laying the foundations 8. Law on Single-Payer System in Health for the single-payer system: Care Financing (2004) (a) Pooling of rayon/city funds to finance 9. Law on Health Care Organizations (2004) health care allowed at higher levels; 10. Law on Health Protection (2005) (b) "Copayment"introduced in the health system that is outside of the Treasury system and is tax exempt (2000). Source: Authors' compilation. The Kyrgyz Republic: Good Practices in Expanding Health Care Coverage 299 Evaluations The financial protection impact of the Kyrgyz reforms is still being evaluated, but already it is clear that they are associated with a declining financial burden of hos- pital care, especially for the poor, and increasing transparency. In addition, utiliza- tion of both outpatient and inpatient health services has become more equal. Nonetheless, large out-of-pocket payments for health care still put a sizeable financial burden on households nationwide. Interpretation of national trends has been hampered by the phased introduc- tion of the reforms. What drives nationwide results is unclear: is it the lack of reforms in late-reform areas or the reforms themselves in the early areas? The pre- liminary results of an ongoing evaluation to separate these effects shows evidence that the financial burden associated with hospital care has been lightened and informal payments are less prevalent in the early-reform oblasts than in the late reformers. Concomitantly, growing expenditure on outpatient medicines has emerged as a next generation coverage/financial protection problem. These find- ings are expanded below, but a definite assessment of reform benefits has to await analysis of the household survey of 2007. Depth of coverage. Analysis of the 2001 and 2004 household surveys show that nationwide, out-of-pocket expenditures continued to grow (figure 10.8). Out-of- pocket payments include payments in outpatient settings for outpatient drugs and Figure 10.8 The Kyrgyz Republic: Trends in Out-of-Pocket Payments, 2000­03 (KGS) 400 360.2 350 300 250 person 200 173.0 per 150 SOM 111.9 95.9 100 52.3 41.9 50 0 outpatient outpatient drugs inpatient 2000 2003 Sources: NSC 2001; Household Budget Survey, Health Module, NSC, 2004; Kyrgyz Integrated Household Survey, Health Module; WHO staff calculations. 300 Good Practice in Health Financing both formal and informal payments at the time of hospitalization. This increase is fueled by spending on outpatient medicines. People consume more drugs, and drugs cost more. The large increase is due to both a price and a quantity effect. Analysis of out-of-pocket patterns by consumption quintiles shows that the share of out-of-pocket payments in household resources grew significantly among the poorest quintile to 7 percent by 2003. The second, third, and fourth quintiles also spent more of their resources on health in 2003 than in 2000. In all four quin- tiles, spending on outpatient medicines drove the results. The richest quintile expe- rienced a slight reduction, mainly because the better-off are using less health care. Because the reforms were introduced in phases, they had been implemented in only four of the eight oblasts when the follow-up survey was done. Thus, these nationwide trends do not allow identification of cause and effect. An ongoing evaluation of the Kyrgyz reforms on financial protection uses a difference-in- difference approach to isolate the prereform effect. According to preliminary results, out-of-pocket payments for hospitalization were similar in reform and control oblasts prior to the introduction of the reforms, controlling for many other observable household characteristics. During the reform period, out-of- pocket payments increased significantly in control oblasts by nearly SOM 600 (US$15). In contrast, out-of-pocket payments increased by only SOM 200 (US$5) per individual in reform oblasts during the same period. This leads to a difference- in-difference estimate of the reform effect of minus KGS 393 (US$10), equivalent to 29 percent of the prereform out-of-pocket payments in reform oblasts. The results show that the reform effect has benefited mostly the lower-income groups. This suggests that the reforms have been successful at limiting the increase in out- of-pocket payments for hospitalization, and the effect size is quite large (Jakab 2007). In sum, whether the reforms have improved depth of coverage is not yet a clear. Nationwide data show that out-of-pocket payments still put a sizeable burden on households seeking treatment, but a more refined evaluation of the reform effect shows that the reforms have considerably improved financial protection, at least for hospital care. Equity in access to care. Despite an increase in the burden of out-of-pocket pay- ments, utilization of health care services across socioeconomic groups has become more equal over the 2000­03 time-period for both outpatient and inpatient care (figure 10.9). In 2000, the richest 20 percent of the population used outpatient services twice as frequently as the poorest 20 percent. By 2003, the visit rate dropped, but only among the richer half of the population. The poorest quintile experienced no change in their rate of visits, and the visit rate for the second income quintile increased significantly. Similar trends were observed for hospital- izations. The hospital utilization rate declined for all socioeconomic groups. For the poorest, it declined only marginally while for the richest it declined by 40 per- cent. As a result, the rich are using hospital care only slightly more frequently than The Kyrgyz Republic: Good Practices in Expanding Health Care Coverage 301 Figure 10.9 The Kyrgyz Republic: Access to Outpatient Care and Hospital Care, 2000 and 2003 a. Access to outpatient care b. Access to hospital care 18 12 16 10 14 at 12 8 reporting visit 10 reporting 6 one 8 hospitalization 1 least population 6 4 at population least of of % 4 % 2 2 0 0 1 2 3 4 5 Poorest 2nd 3rd 4th Richest quintile 2000 2003 Sources: NSC 2001; Household Budget Survey, NSC 2004; Kyrgyz Integrated Household Budget Survey; WHO staff calculations. Note: Outpatient care includes primary and specialist care. The population was grouped into five quintiles using consumption data and quintile data calculated by the National Statistical Committee. Each quintile group contains 20 percent of the population ranked from poorest to richest. the poor. On the whole, the reforms appear to play an equalizing role on utiliza- tion across the country and across socioeconomic lines. Transparency. The single-payer system has improved health system transparency for the population by creating a clear system of benefits and entitlements through the SGBP and the copayment policy. Previously, confusion about entitlements, coupled with great pressure on providers to make up for a drastic decline in public funding, gave rise to a widespread system of informal payments, in particular for hospital care. At the time of hospitalization, most patients had to make informal to payments to physicians for medicines and other supplies. However, there was no formula for regulating the contribution (in-kind and in-cash) that patients were required to contribute, and physicians could decide for themselves how much and whom to charge according to their own understanding of social justice. Various surveys indicate that patients found this system opaque, subjecting them to unpre- dictable financial demands at the time of illness. The cost of hospitalization was not known in advance, not even at the time of admission, and expenditures mounted steadily during the course of a hospital stay. 302 Good Practice in Health Financing Figure 10.10 The Kyrgyz Republic: Mean Payment by a Public Hospital Patient, 2000 and 2003 2,500 2,072 1,995 2,000 1,828 1,500 1,378 1,193 1,231 KGS 1,000 500 2000 2000 2000 2003 2000 2003 late reform early reform admission and lab medicines, supplies, personnel food Source: NSC 2001; Household Budget Survey, NSC 2004; Kyrgyz Integrated Household Budget Survey; WHO staff calculations. Note: "Early reform" includes Issyk-kul, Chui, Naryn and Talas; "late reform" includes Osh, Jalal-Abad, Batken. Medicines, supplies, and personnel covers unequivocally informal payments, because patients should receive these services free of charge against their copayment. In 2000, also largely informal admission and lab payments, because there was no charge at that time for admission. In 2003, admission and lab payments are in early reform oblasts is a formal copayment, but in Bishkek and late-reform oblasts it reflects a mix of copayment and informal payment due to the varied timing of these policies. The reforms clarified entitlements, but problems remain. Figure 10.10 shows a breakdown of out-of-pocket payments for hospitalization, by category. The pay- ment amount reflects the mean payment per episode of hospitalization among patients admitted to a public hospital at least once in the year preceding the survey. The dark solid area is formal and informal payments made upon admission and for lab tests; formal and informal payment that cannot be separated based on this survey. The checkered area represents payments for medicines, supplies, and per- sonnel, all of them informal payments. The white area is the amount paid for food. The figure shows that the overall out-of-pocket payments at the time of hospi- talization grew the most slowly in early-reform oblasts. This indicates that the reforms did slow the growth of out-of-pocket and the resultant burden on house- holds. In areas where single-payer reforms and copayment were introduced latest, informal payments increased, by 73 percent in the late-reform oblasts of Osh, Jalal- Abad, and Batken, and by 24 percent in Bishkek. In contrast, informal payments The Kyrgyz Republic: Good Practices in Expanding Health Care Coverage 303 decreased by 2 percent in the early-reform oblasts of Issyk-kul, Chui, Naryn, and Talas where the reforms had more time to take hold. As the evaluation above shows, depth of coverage, equity of access, and trans- parency have tangibly improved. However, some issues still need to be addressed before the benefits of the reforms reach the entire population. A key impediment to reaching this goal is lack of awareness of entitlements among vulnerable groups. Although the SGBP clearly regulates entitlements, many people do not know that primary care is free or that insured individuals have a right to the Outpatient Drug Benefits Package. At the hospital level, many patients continue to pay informally, despite publicly displayed copayment schedules. Lack of awareness of rights is a particular problem among internal migrants who do not know that their health insurance benefits are portable and that their enrollment with a family group prac- tice is not linked to a permanent place of residence (World Bank 2005c). Scaling-Up The Kyrgyz reforms were piloted and the introduced in phases. Although in many contexts, pilots do not get scaled up for nationwide implementation, the Kyrgyz experience is positive from this perspective. The reforms were introduced in two additional oblasts each year between 2001 and 2005, when the entire country was finally covered. In the interim, resistance grew and side effects began to show. To carry out these politically sensitive, unpopular reforms, strong political ownership was essential. Gradual implementation allowed learning-by-doing, which is criti- cal when introducing completely new financing and institutional mechanisms. In addition, phased implementation also allowed policy makers to devote sufficient attention to the oblasts and providers that were next in line. Lessons for Other Countries The successes of the Kyrgyz health reform are due in part to the focus on a com- prehensive approach rather than on isolated instruments or "magic bullets." The Kyrgyz reforms did not attempt to find a magic bullet to achieve every health sys- tem objectives such as the"right" benefits package or the"right" provider payment reform. The reforms involved a coordinated package of changes in pooling, pur- chasing, design of the benefits package, and patient copayments. A key element of this package was a carefully structured and well-aligned financial incentive system on both the provider and the patient sides. In addition, reforms of health care financing were complemented with harmonization of the organization and struc- ture of service delivery at both the macro- and microlevels. This included restruc- turing of facilities and extension of provider autonomy. Enabling Factors Complex reforms require careful sequencing of various reform steps. The Kyrgyz reforms stand out in this regard. First, it was implicitly acknowledged that, without 304 Good Practice in Health Financing efficiency improvements through downsizing, other health system objectives such as equity and quality cannot be improved, given the fiscal constraints. By combin- ing financing and service delivery reforms, the Manas reforms were extremely suc- cessful in achieving efficiency gains, which paved the way for focusing on other health system objectives in the next phase. Second, the health financing reform measures were also carefully sequenced. Changes in revenue collection and pool- ing were the initial steps to ensure optimal flow of public funds first. Then, new purchasing arrangements to provider incentives were introduced. These steps brought transparency to the public sector. The last step, introduction of a benefits package, created a transparent social contract between the people covered and the government providing coverage through taxation. Close attention to the institutional aspects of the reforms was important to ensure sustainable results. Progress on efficiency and quality would have been much more difficult had the MHIF not been formed as a parastatal agency in charge of purchasing. The single-payer system has successfully incorporated "strategic pur- chasing" in its day-to-day operations. This included output-based payment mech- anisms, cleverly structured incentives for referrals and exemptions, the ADP and regular monitoring of quality. All these tools provided explicit incentives for effi- cient and high-quality delivery of health care. Had the MHIF not been set up out- side the core public sector, it would not be able to purchase services "strategically" and would have had to rely on hierarchical line-item input control as other organi- zations in the core public bureaucracy. This type of input-based budgeting was a major source of inefficiency in the system inherited from the USSR, and it could it not have been overcome without changing purchasing methods. Phased implementation and successful piloting were an effective implementation approach that helped build capacity and stakeholder support and allowed learning by doing. The Issyk-kul Intensive Demonstration Site provided an opportunity for the basic health reform model to be developed. Many of the early adjustments and fine-tuning of the reforms had already taken place in Issy-Kul. Also, Issy-Kul became a visible symbol of reform, greatly assisting the expansion of these reforms to other parts of the country and the further expansion of the family group prac- tice model. Issyk-Kul continues to serve as a demonstration site for the country as the reforms here have been deepened over time. For example, the inpatient copay- ment policy was first implemented in Issy-Kul and Chui (another oblast where the reforms have been ongoing for at least five years). The finding that the greatest and most sustained reduction in informal payments occurred in Issyk-kul is likely to be due to this longer experience with the reforms, which allowed capacity building and commitment to reform principles. Strong MOH coordination and collaboration with the development partners facili- tated harmonized activities for reform design and support. Donor presence in the The Kyrgyz Republic: Good Practices in Expanding Health Care Coverage 305 health sector is strong in the Kyrgyz Republic. However, experience from other highly donor-dependent countries indicates that donor support is often frag- mented and overlapping. That means that the health sector often benefits less from donor support than it should and has fewer results to show at the end of a long period of donor support. This was not the case in the Kyrgyz Republic where donor collaboration has been strong. The activities of most major donors in the health sectors were well aligned with the government's reform plan and the donors made an effort to harmonize their procedures and positions. Having a well-articulated sector strategy facilitated this unified approach. The donors supporting the same health sector strategy became important political advocates for its continued implementation and success. By gradually moving to a "single voice" on many pol- icy issues, they provided powerful political support and contributed to the continu- ity of the Kyrgyz reforms over a decade. Due to the successful alignment and harmonization of donor support over a decade, the Kyrgyz health sector has moved to a Sector-Wide Approach (SWAp) in 2006 for the first time in the region. Key Financing Factors The decline in public expenditures during reform implementation limited the impact on financial protection and transparency. Oblast funding began to decline after successful restructuring. Oblast-level pooling and output-based payment have achieved their intent: the Kyrgyz health system underwent significant down- sizing and became slimmer between 1996 and 2003 with resultant reductions in utility and maintenance costs. The rest of the public sector has not taken the same giant steps as the health sector to move to output-based payment of public providers. Declining staff and beds triggered the well-instilled response of the Soviet budgeting system: if inputs decline, so too should public funds. This mech- anistic response has removed savings from the health sector, reducing provider incentives to embark on painful downsizing processes elsewhere. The introduc- tion of copayment also had a crowding-out effect on public spending. The period of accelerated decline in public funding, especially at oblast level, coincided with the introduction of copayment. This seemed to create a crowd-out effect in bud- get negotiations, and marginal resources were allocated to sectors that did not have a chance to collect additional revenues through copayments. The slow pace of reforms in the overall public finance system created a challenging operating environment for the health sector and limits its ability to achieve effi- ciency, quality, and equity gains. The health sector gradually but firmly moved away from input-based, line-item budgets and administrative control mechanisms and began to switch to performance management and output-based payment of public providers. These changes have vastly improved transparency and efficiency in public health sector management. At the same time, these ideas remained revolu- tionary within the overall public system, which still operates largely on the inherited 306 Good Practice in Health Financing principles of input-based budgeting and strict line items. This mismatch between the pace of public finance reform and health finance reform has created significant problems for the health sector in the annual budget negotiation process. Key Political Economy Factors Due to political pressures to eliminate copayment without a commensurate increase in public funding, unfunded mandates have returned to the health sector in recent times. In the 2006 revision of the SGBP, copayment was eliminated for children under five, deliveries, and pensioners over 75 years of age. This move, a result of political pressure from the new government, was not accompanied by any clear and transparent calculations of the cost and affordability of the policy. As a result, commensurate funding is not likely to be available to compensate providers for the loss of official copayments, and this will usher in the return of informal payments and deterioration of hard-won improvements in trans- parency. The impact of this policy is currently under evaluation. Replicability in Low- and Middle-Income Countries The Kyrgyz reform in its entirety is particularly applicable in transition economies confronted with excess capacity and drastically reduced fiscal space. The main logic of the Kyrgyz reforms was to squeeze efficiency gains out of the system and use the savings to improve the depth of coverage. This set and sequencing of reform instruments have less relevance in countries without significant excess capacity in infrastructure and staff. Endnotes 1. Unless otherwise noted, economic and poverty data in this section are taken from the World Development Indicators (WDI) Dataset, 2006. 2. The extreme poverty rate is defined as the proportion of the population with insuf- ficient expenditures (or consumption) to purchase a food basket of 2,100 calories. 3. The absolute poverty rate is defined as the proportion of the population too poor to obtain the proper amount of food and necessary nonfood consumables. 4. The share of agriculture in GDP is 34 percent, according to WDI. 5. Government functions are funded from the central budget and mixed functions, from the central and local budgets. Local functions are defined as expenditure obligations of local self-governments. 6. IMR increased to 25.7 in 2004, but it is not strictly comparable to earlier years because the government introduced new live birth criteria to conform with the interna- tional standards. 7. For more details, see Asymbekova, unpublished report. 8. This description of the political situation is based on two reports by the Interna- tional Crisis Group (2001; 2005). The Kyrgyz Republic: Good Practices in Expanding Health Care Coverage 307 9. The centralization of financing reforms was completed in 2006 with oblast pur- chasing merged into one purchasing pool for the entire country, further increasing the scope from cross-subsidization and transfers. 10. A feldscher is equivalent to a physician's assistant or nurse practitioner in the United States. The following definition of feldscher is the correct one: Feldsher or Obstetrician Posts are primary rural health facilities that provide basic predoctor services such as first aid, preventive care and antiepidemic work, sanitary and hygiene promotion, early detec- tion of and initial measures against infectious outbreaks. FAPs have only mid-level health personnel (up to three people), such as obstetricians, doctor's attendants and nurses. References Asymbekova, Gulnara. "Analysis of Maternal Mortality for 1999 in the Kyrgyz Republic." Unpublished report for the Ministry of Health. Bishkek: MOH. Atun, Rifat. 2005. Evaluating Manas Health Sector Reform (1996­2005): Focus on Primary Health Care. Policy Research Papers. Bishkek: WHO-DfID Manas Health Policy Analy- sis Project. CIA (Central Intelligence Agency). 2006. World Fact Book. Washington DC: CIA. https:// www.cia.gov/cia/publications/factbook/geos/kg.html. Ibraimova, Ainura. 2002. Report to the Government of the Kyrgyz Republic on the Activities of the MHIF for 2001. Bishkek: Mandatory Health Insurance Fund. Ibraimova, Ainura, Ninel Kadyrova, Melitta Jakab. Washington, DC: CIA. Draft. Contracting Mechanism to Improve Access to Essential Drugs--the Kyrgyz Outpatient Drug Benefit. Bishkek: WHO-DfID Manas Health Policy Analysis Project. International Crisis Group. 2005. Kyrgyzstan: A Faltering State.Asia Report. Bishkek/Brussels: International Crisis Group. ------. 2001. Kyrgyzstan at Ten: Trouble in the "Island of Democracy." Asia Report. Bishkek/ Brussels: International Crisis Group. Institute for War and Peace Reporting. 2003. "Bishkek Braced for Aksy Anniversary." News Report RCA No. 188, March 4. http://www.iwpr.net/?p=rca&s=f&o=177056&apc_ state=hruirca2003 Jakab, Melitta. 2007. "An Empirical Evaluation of the Kyrgyz Health Reform: Does It Work for the Poor?" PhD dissertation, Harvard University, Cambridge, MA. Kutzin, Joseph. 2000. "Toward Universal Health Care Coverage: A Goal-Oriented Frame- work for Policy Analysis." Working Paper, World Bank, Washington, DC. Kutzin, Joseph, Ainura Ibraimova, Ninel Kadyrova, Gulaim Isabekova, Yevgeniy Samyshkin, and Zainagul Kataganova. 2002. Innovations in Resource Allocation, Pool- ing and Purchasing in the Kyrgyz Health System. Policy Research Papers. Bishkek: WHO-DfID Health Policy Analysis Project. Kutzin, Joseph, Ainura Ibraimova, Ninel Kadyrova, Tilek Meimanaliev, and Tobias Schüth. 2001. Addressing the Informal Payments in Kyrgyz Hospitals. Policy Research Papers. Bishkek: WHO-DfID Health Policy Analysis Project. Kutzin, Joseph, Sheila O'Dougherty, and M. Jakab 2005. Fiscal Decentralization and Options for the Kyrgyz Health Financing System: Reflections on Three Options. Policy Research Papers. Bishkek: WHO­DfID Manas Health Policy Analysis Project. 308 Good Practice in Health Financing Meimanaliev, Tilek. 2003. Kyrgyz Model of Health Care System. Bishkek: Uchkun. Meimanaliev, Adilet-Sultan et al. 2005. Health Care Systems in Transition: Kyrgyzstan. Copenhagen: European Observatory on Health Care Systems. MHIF (Mandatory Health Insurance Fund). 2006. Dataset provided upon request. MOH (Ministry of Health), Kyrgyz Republic. 2006. Manas Taalimi Health Reform Strategy. Bishkek. NSC (National Statistical Committee). 2004. Kyrgyz Integrated Household Budget Survey. Bishkek. ------. 2001. Household Budget Survey. Bishkek. ______. 2005. Dataset provided upon request. Oxford Policy Management. 2006."Kyrgyz Republic: Public Financial Management Assess- ment." Final Report, January 18, 2006. Oxford Policy Management, Oxford, U.K. http://siteresources.worldbank.org/INTECAPUBEXPMAN/Resources/PFMkyrgyz ReportFinal090806.pdf. Purvis, George, et al. 2005. Evaluating Manas Health Sector Reforms (1996-2005): Focus on Restructuring. Policy Research Papers. Bishkek: WHO-DfID Manas Health Policy Analy- sis Project. RMIC (Republican Medical Information Center). 1997­2005. Health of the Population and Performance of Health Facilities in the Kyrgyz Republic. Bishkek: Ministry of Health of the Kyrgyz Republic. ______. 2006. Dataset provided upon request. Sargaldakova, Acelle, et al. 2000. Health Care Systems in Transition: Kyrgyzstan. Copen- hagen: European Observatory on Health Care Systems. United Nations Population Division. 2006. World Population Prospects: The 2004 Revision Population Database. http://esa.un.org/unpp/index.asp. World Bank. 2006. World Development Indicators. Washington, DC: World Bank. ------. 2005a. Kyrgyz Republic Poverty Update: Profile of Living Standards in 2003. Wash- ington, DC: World Bank. ------. 2005b. Kyrgyz Republic: Country Economic Memorandum. 2 vols. Washington, DC: World Bank. ------. 2005c. Operationalizing the Health and Education MDGs in Central Asia: Kyrgyz Republic Health Education Case Studies. 2 vols. Washington, DC: World Bank. ------. 2004a. Millennium Development Goals for Health in Europe and Central Asia: Rele- vance and Policy Implications. Washington, DC: World Bank. ------. World Bank. 2004b. Public Expenditure Review. 2 vols. Washington, DC: World Bank. ------. 2001. Poverty in the 1990s in the Kyrgyz Republic. Washington, DC: World Bank. ------. 2006. World Development Indicators (WDI) Database. World Bank, Washington, DC. WHO (World Health Organization). Various years. Health for All Database. http://data. euro.who.int/hfadb. ------. 2004. Global Burden of Disease Estimates, 2002. http://www.who.int/healthinfo/ bodestimates/en/index.html. The Kyrgyz Republic: Good Practices in Expanding Health Care Coverage 309 Legal Norms Government Decree No. 98 on Copayment for Drugs, Food and Certain Medical Services Provided by State Health Facilities, February 25, 2002. Government Decree No.556 on Establishment of the Department of Drug Provision and Medical Equipment under the Ministry of Health, September 26, 1997. Government Decree No. 287 on High-Technology and Expensive Medical Services, May 7, 2002. Law on the Basic Principles of the Budget, June 11, 1998. Law on Health Care Organizations, August 13, 2004. Law on Health Protection, January 9, 2005. Law on Insurance Contributions for the State Social Insurance for 1998, January 17, 1998. Law on Insurance Contributions for the State Social Insurance for 1999, December 26, 1998. Law on Insurance Contributions for the State Social Insurance for 2000, January 18, 2000. Law on Insurance Contributions for the State Social Insurance for 2003, February 15, 2003. Law on Medical Insurance of the Citizens of the Kyrgyz Republic, October 18, 1999, with amendments on April 21, 2003, and July 15, 2003. Law on State Guaranteed Benefits Package for 2004. 11 Sri Lanka: "Good Practice" in Expanding Health Care Coverage Ravi P. Rannan-Eliya and Lankani Sikurajapathy Background Sri Lanka, a rainy, 66,000-square-kilometer island in the Indian Ocean near the equator, is in the South Asia Region of the World Bank. Only 15 percent of Sri Lanka's 20 million people live in cities. It is a lower-middle-income country, with GDP per capita of US$965 in 2004. The expansion of health care coverage in Sri Lanka, with its focus on the poor, dates from the 1930s, and many of the initial motivations continue to be important influences. By far the most important one for health services has been democracy. In the 1920s, conditions in the island were much like those in most other British colonies. Government intervention in health was limited to providing health care to a small urban population that operated the colonial infrastructure and administration and an equally small workforce involved in export agriculture, and to a sanitary regime designed to control major epidemic threats such as cholera. Democracy based on uni- versal suffrage was introduced in 1931 expressly to empower the poorer groups in society and women and to put pressure on the elites to pay closer attention to social and health conditions. After 1931, the political economy of the island changed irrevocably as the political power base shifted from urban residents to the majority rural population. The impact of democracy on health was accentuated by the emergence of competitive politics along a left-right dimension with two-party competition well embedded by the late 1950s, a rural bias in the delimitation of electorates where each national legislator typically represented fewer than 10,000 voters in the 1930s, and a single-member constituency system that encouraged politicians to engage in parish-pump politics to maximize the government infrastructure built in their districts. The introduction of democratic politics forced successive governments to continuously expand free public 311 312 Good Practice in Health Financing health services into rural areas where voters wanted the same standards established earlier for the urban population. Once democracy had served to establish a widely dispersed government health infrastructure, accessible by all, it then acted to ensure its survival under often diffi- cult, fiscal conditions. Subsequently, successful market-oriented and reform-minded governments in Sri Lanka have generally understood that the cost of adequate public sector health services accessible to the poor was a small fiscal price to pay for the polit- ical support that they engender to enable other more important economic reforms. Introduction Sri Lanka's distinctive history, economy, people, politics, and health conditions have contributed to impressive achievements in health care over the past 50 years. Economic Environment Sri Lanka's economy was historically based around agriculture, primarily rice cul- tivation, but several centuries of active trade resulted in a society that was more open to outside influences and interactions than most Asian countries. Prior to British occupation of the island in the late-18th century, large-scale irriga- tion agriculture and later spice exports provided a base for government taxation and dictated key aspects of government organization, and a tight, state-led social organization. The British introduced coffee, tea, and rubber cultivation, and by the end of the 19th century a classic dualistic export economy emerged (Snodgrass 1966). Cash crop exports brought prosperity and a trade surplus, and their taxation gave the government a ready revenue source. After independence in 1948, Sri Lanka's econ- omy was highly trade dependent, although most of its people were involved in sub- sistence rice cultivation. Tea, rubber, and coconut made up more than 95 percent of exports, and living standards were the highest in South Asia. Relative prosperity continued until the Korean War commodity boom in the 1950s. Then, declining commodity prices and a failure to diversify exports led to economic stagnation, ever-tighter import controls, and inward-oriented import-substitution policies (Bruton 1992). Income stagnated and unemployment was high (more than 20 per- cent). Mounting social tensions contributed to two Maoist insurgencies and an ethnic-based separatist conflict after 1970, which have presented major challenges for Sri Lanka's economy. Under a new government in 1977, Sri Lanka became one of the first developing countries to embark on economic liberalization, pursued ever since. Trade was liberalized, export taxes on cash crops removed, and the economy opened up. In return, Sri Lanka benefited from substantial Western aid inflows for more than a decade. These policies led to substantial improvement in economic growth (table 11.1), averaging 3 to 4 percent real per capita income growth ever since, despite the series of debilitating internal conflicts that started in the early 1970s. Growth has been led by export-oriented manufacturing, initially concentrated in gar- Sri Lanka: "Good Practice" in Expanding Health Care Coverage 313 Table 11.1 Sri Lanka: Economic Indicators 1930­2005 GDP per GDP per External capita capita Revenue Expenditure Debt ODA Year (1990 US$) (1990 PPP$) (% GDP) (% GDP) (% GDP) (% GDP) 1930 180 945 ~10 ~10 ~0 ~0 1950 273 935 16 20 3 0 1970 316 1,130 20 27 18 1.7 1990 577 1,935 22 31 72 5.7 1995 704 2,636 20 31 67 4.5 2000 844 3,626 17 27 55 0.4 2005 962 4,390 16 24 48 3.4 Source: Central Bank of Sri Lanka (2006); Institute for Health Policy databases; estimates of pre-1950 GDP originally prepared by author for Rannan-Eliya and de Mel (1997). Note: ~ = approximate. ments but now diversifying. By the 1990s, more than 75 percent of Sri Lanka's exports were industrial products. Continuing economic growth in recent years has pushed unemployment to less than 7 percent of the workforce, raised income in 2005 to more than US$1,000 per capita, and modestly reduced the number of Sri Lankans living in poverty (table 11.2). More substantial reductions in poverty have not occurred, because recent economic growth has been associated with increasing income inequality, and living standards for the lowest income quintile have hardly changed. Although official development assistance (ODA) remains significant, private foreign direct investment (FDI) is now more important for growth, but not to the same extent as in other Southeast Asian economies. Table 11.2 Sri Lanka: Social Indicators 1930­2005 Poverty Poverty Infant Population head count head count Literacy mortality Life Year (millions) (>, France: Harmattan. MSP (Ministère de la Santé Publique). 1976­1996. "Série des bulletins épidémiologiques." MSP, Tunis. ------. 1986. "1956­1986 Trente ans au service de la santé: un engagement, une éthique." MSP, Tunis. NHIF (National Health Insurance Fund). 2006. Informal data. NHIF, Tunis. Omrane, A. R. 1983. "The Epidemiologic Transition, A Theory: Preliminary Update " Jour- nal of Tropical Pediatrics 9: 305­17. Tunisia: Good Practice in Expanding Health Care Coverage 437 UNDP (United Nations Development Program) 2004a. "World Population Prospects: The 2004 Revision." Population database. http://esa.un.org/unpp/index.asp. ------. 2004b."Stratégie du phénomène de la pauvreté en Tunisie." Etudes et rapports sur le développement, <>, July 2004. UNPD (United Nation of Population Division). World Mortality Report 2005. New York: UNPD, Department of Economic and Social Affairs. ------. 2001. Human Development Report. Brussels: Ed. De Boek. ------. 1999. "World Urbanization Prospects: The 1999 Revision." New York: UNPD. WBI (World Bank Institute). 2004. "Public Health in the Middle East and North Africa: Meeting the Challenges of the Twenty-First Century." Learning Resource Series Publica- tion 29163, WBI, Washington, DC. WHO (World Health Organization). 2006. "World Health Statistics." http://www.who.int/ whosis/whostat2006/en.index/html. ------. 2004. Department of Measurement and Health Information, December. http:// www.who.int/healthinfo/bodestimates/en/index. htm. ------. 2000. The World Health Report 2000--Health Systems: Improving Performance. Geneva: WHO. World Bank. 2006a. World Development Report: Equity and Development. Washington, DC. ------. 2006b. "Etude du secteur de la santé en Tunisie." Human Development Depart- ment, Middle East and North Africa Region, Washington, DC. ------. 2006c. "World Development Indicators." http://www.ddp-ext.worldbank.org. ------. 2005. "World Development Indicators." http://www.ddp-ext.worldbank.org. ------. 2004a. "Making Deeper Trade Integration Work for Growth and Jobs." Republic of Tunisia, Development Policy Review. Report 29847-TN, World Bank, Social and Eco- nomic Development Group, Middle East and North Africa Region, Washington, DC. ------. 2004b. "Project Performance Assessment Report: Republic of Tunisia." Report 31017. http://www-ds.worldbank.org/servlet/WDSContentServer. ------. 2004c. "World Development Indicators." http://www.ddp-ext.worldbank.org. ------. 2003a. "Vers une meilleure gouvernance au Moyen-Orient et en Afrique du Nord: Améliorer l'inclusivité et la responsabilisation." Report on the development in the Mid- dle East and North Africa. World Bank, Washington, DC. ------. 2003b. "Stratégie de développement touristique en Tunisie." World Bank, Wash- ington DC. ------. 2000. "Poursuivre l'intégration à l'économie mondiale et pérenniser le progrès économique et sociale de la Tunisie." Sociale and structural review, World Bank, Wash- ington, DC. 14 Vietnam: "Good Practice" in Expanding Health Care Coverage--Lessons from Reform in Low- and Middle-Income Countries Björn Ekman and Sarah Bales Background With a population of about 84 million and a GDP per capita of US$540, Vietnam is one of the larger low-income countries in the East Asia and Pacific Region of the World Bank. It is one of the world's most dynamic economies, with economic growth averaging between 6 and 8 percent a year over the past decade. Vietnam is predomi- nantly rural: only 25 percent of the population lives in urban areas, but the ratio of the rural to urban population is rapidly falling. In 2003, Vietnam spent about 5.4 percent of its GDP on health--approximately US$26 per capita, which is about the average for its income level. Its health outcomes are excellent: in 2004, life expectancy was 70.3 years and the infant mortality rate was only 17.4 per 1,000 live births. Vietnam's health care financing reforms have been motivated by three main con- siderations: its historic emphasis on good health, the important poverty-eradication component of health, and a desire to make health care affordable for the poor. Prior to the reforms, Vietnam's health system was geared toward provision of "health for all" via an extensive network of community health services and intercommunal polyclin- ics for primary care and government hospitals for higher levels of care. Coverage was fairly widespread, but the quality of care was low primarily due to a lack of resources. The introduction of user fees in the late 1980s, however, posed barriers to access, and resulted in dissatisfaction among poor and nonpoor. In response, the government introduced targeted health financing support, initially as part of other broader poverty eradication policies and subsequently in the shape of stand-alone health financing programs for the poor and other target groups, including students and children. The political-institutional framework of Vietnam--a one-party state--makes it difficult to assess the political economy of reforms, which take place within a less-open 439 440 Good Practice in Health Financing system. For example, the lack of formal political opposition to the governing party and the absence of a free press in which political debate can be observed, complicates the study of the extent to which affordable health care is a central political-economic issue. Although the political economy dimensions of the health financing reforms have been little documented, some issues can be noted. First, the political accountability of the Communist Party depends on an ability to provide the population with, among other services, affordable health care. Public dis- satisfaction with user fees became an "embarrassment" for the party and led to reforms. More recently, with the National Assembly's expanding role in the country's general political debate, the minister of health has had to respond to questions and criticisms from the representatives about what the government is doing to help the poor obtain health care. Making the government accountable for the implementation of the program on Health Care Funds for the Poor (HCFP), for example, has con- tributed to some recent program adjustments. Introduction In the decade following the end of the armed conflict with the United States and the reunification of the country in 1975, Vietnam pursued a socialist economic policy, with the government controlling the production of most goods and ser- vices. Since then, liberalizations have been introduced that have affected the health sector. Economic Environment In the immediate postwar period with an estimated gross domestic product per capita of US$1301 a year, Vietnam was one of the world's poorest countries. Poor economic performance, stagnant school enrollments, and widespread malnutri- tion were endemic. By the mid-1980s, the Communist Party realized that a change was necessary to sustain food security and improve the welfare of the Vietnamese people. Thus, the government adopted an economic reform program referred to as Doi Moi [renovation]. The program launched a series of policy changes continuing to this date. The most important changes involved extensive agricultural and trade liberalization. In the health sector, a far-reaching deregulation of the health care system was introduced in the early 1990s. Among other things, the changes meant that health professionals were allowed to operate private clinics and that private entrepreneurs could sell drugs in private outlets. User fees for health care were introduced to raise revenues. In addition, a social health insurance program was initiated and is now being expanded by further measures to broaden coverage and improve efficiency. Table 14.1 shows the evolution of a set of key economic indicators since 1990. At the outset, economic growth in Vietnam lagged that of other geographically or economically comparable countries. Since then, as a direct result of the economic reform program, overall GDP growth has been comparable to that of other fast- growing countries in the region and well above that of other low-income countries Vietnam: "Good Practice" in Expanding Health Care Coverage 441 Table 14.1 Vietnam: Economic Indicators, 1990­2004 Indicator 1990 1995 2000 2001 2002 2003 2004 GDP per capita (constant 2000 US$) 226 305 397 419 444 471 502 GDP per capita, PPP 989 1,419 2,012 2,170 2,336 2,518 2,745 (current international US$) GDP growth (annual %) 5.10 9.54 6.79 6.89 7.08 7.34 7.69 Poverty gap at US$1 a day (PPP) (%) -- 181 92 0.53 0.5 n.a. n.a. Poverty headcount ratio at -- 581 372 -- 28.90 -- <20 national poverty line (% of population) Revenue, excluding grants n.a. n.a. n.a. n.a. n.a. n.a. 22 (% of GDP) Tax revenue (% of GDP) -- -- -- -- -- -- 17 External debt (% of GDP) -- -- 38.60 37.90 34.90 34.10 34 Aid (% of GNI) 3.12 4.08 5.47 4.50 3.70 4.52 4 Unemployment, total -- -- 2.30 2.80 2.10 2.30 2 (% of total labor force) Inflation, consumer prices (annual %) -- -- ­1.71 ­0.43 3.83 3.10 7.80 Sources: WDI 2006; IMF Article IV Consultations, PRGF Reports; National Survey Data 1993, 1998, 2002. Note: -- = not available. and some lower middle-income countries. Growth rates have been particularly strong in per capita terms due to Vietnam's reduction in fertility rates and sub- sequently limited population growth rates. The most recent estimate of gross national income (GNI) per capita is current US$540 (WDI 2006), thus placing Vietnam among the low-income countries that are rapidly moving into the group of lower-middle-income countries. The changes in per capita income over this period suggest that the policies, in terms of economic stability, have provided Vietnam with a national income per capita of around US$2,700 a year in purchasing power terms. This is above the average for the low-income country group. With its rapid economic growth Vietnam's income poverty rates have fallen steeply.Although estimates and definitions vary and are surrounded by some uncer- tainty, income poverty fell from around 58 percent in 1993 to 29 percent in 2002 (World Bank 2003a: i). The most recent estimates suggest that poverty has fallen even further, to a little more than 15 percent of the population in 2006 (GSO 2006). Extreme poverty is relatively limited in Vietnam, and the remaining absolute poverty is concentrated mainly in some disfavored regions (flood-prone areas and mountainous regions) and among ethnic minorities (WHO 2003), who make up about 13 percent of the population. Income distribution remains an issue in Viet- nam with the most well-off 20 percent receiving around 45 percent of available income compared with only around 7 percent for the poorest fifth of the popula- tion (WDI 2006). 442 Good Practice in Health Financing In terms of government revenue, recent estimates suggest that Vietnam has a revenue-to-GDP ratio of around 22 percent, which is expected to increase slightly. The overall budget deficit is around 4.5 percent of GDP, suggesting that the over- all fiscal situation is sustainable. Furthermore, national accounts data show that Vietnam has a tax revenue-to-GDP ratio of around 17 percent, mainly from income and profit taxes (IMF 2005). Vietnam has managed to limit its external debt over this period. It is not a highly indebted poor country (HIPC) that can benefit from debt cancellation ini- tiatives. Recent estimates put its total external debt at 33 percent of GDP, and debt servicing is less than 10 percent of total exports of goods and services (ibid.). An important issue for its external relations was accession to the World Trade Organi- zation (WTO) on January 11, 2007, which is expected to affect the Vietnamese health sector in general and the pharmaceutical market in particular. Inasmuch as pharmaceuticals make up half of all health spending, drug prices are expected to fall significantly with possibly far-reaching consequences for both access to medi- cine and to health financing generally. Vietnam has also limited its dependence on overseas development assistance (ODA). In the early reform period, former Soviet-bloc countries and Sweden pro- vided assistance, a large part of it for the health sector. Since the late 1990s, when relations with the rest of the world were normalized, inflows of aid, including loans from development banks, have increased to 5 percent of GNI, much more than the regional average but substantially less than, for example, several coun- tries in Sub-Saharan Africa. Demographic, Epidemiologic, and Social Environment Vietnam covers 330,000 square kilometers (207,000 sq. miles) with a population of around 84 million resulting in some 255 inhabitants per square kilometer. The population consists of several different ethnic groups. The largest group, the Kinh, accounts for around 87 percent of the population. The remaining 13 percent con- sists of some 50 different ethnic minority groups with distinctive ethnolinguistic and cultural characteristics (WHO 2003). A quarter of the population lives in urban areas, although with migration from the countryside to the larger cities, the ratio of rural to urban population is falling. Vietnam has pursued a relatively strict family and population policy over the past two decades that has resulted in a fertility rate barely above replacement level. Projections suggest that the population growth rate will be further reduced to less than 1 percent after 2025 (WDI 2006). Table 14.2 presents selected population, demographic, and social indicators for various years since 1990. Life expectancy at birth was much higher in Vietnam than in other countries at similar income levels well before the mid-1980s when the economic reform program was initiated. One likely factor for Vietnam's good health outcomes is the comparatively high level of education, as reflected in the high literacy rates for both men and women. The Vietnamese literacy rates, for previous years not shown in the table, are also sig- nificantly higher than those of other low- and even some middle-income countries. Vietnam: "Good Practice" in Expanding Health Care Coverage 443 Table 14.2 Vietnam: Selected Population and Social Indicators, 1990­2004 Indicator 1990 1995 2000 2001 2002 2003 2004 Population, total (millions) 66.2 72.9 78.5 79.5 80.4 81.3 82.2 Population ages 0­14 (% of total) 38.92 37.02 33.46 32.67 31.87 31.08 30.30 Population ages 15­64 (% of total) 56.13 57.96 61.14 61.89 62.67 63.46 64.24 Population ages 65 and above 4.95 5.03 5.40 5.44 5.46 5.46 5.45 (% of total) Median population age 20.20 21.30 23.10 -- -- -- -- Rural population 79.74 77.79 75.68 75.20 74.72 74.24 73.76 (% of total population) Population growth (annual %) 2.18 1.80 1.29 1.23 1.16 1.10 1.04 Age dependency ratio 0.78 0.73 0.64 0.62 0.60 0.58 0.56 (dependents to working-age population) Life expectancy at birth, female years) 66.84 69.42 71.52 -- 72.20 72.50 72.80 Life expectancy at birth, male (years) 62.82 64.96 66.70 -- 67.30 67.60 67.90 Life expectancy at birth, total (years) 64.78 67.14 69.05 -- 69.69 69.99 70.29 Birth rate, crude (per 1,000 people) 28.78 23.76 19.43 -- 18.50 18.30 18.00 Death rate, crude (per 1,000 people) 7.26 6.68 5.80 -- 6.20 6.10 6.10 Fertility rate, total (births per woman) 3.62 2.67 1.90 -- 1.87 1.80 1.80 Mortality rate, infant 38.00 32.00 23.00 -- -- 19.00 17.40 (per 1,000 live births) Mortality rate, under-5 (per 1,000) 53.00 44.00 30.00 -- -- 23.00 23.20 Mortality rate, adult, female 153.15 -- -- -- 130.06 -- -- (per 1,000 female adults) Mortality rate, adult, male 215.02 -- -- -- 181.67 -- -- (per 1,000 male adults) Maternal mortality ratio -- -- 130.00 -- -- -- -- (modeled estimate, per 100,000 live births) Literacy rate, adult female -- -- -- -- -- -- 86.92 (% of females ages 15 and above) Literacy rate, adult male -- -- -- -- -- -- 93.92 (% of males ages 15 and above) Literacy rate, adult total -- -- -- -- -- -- 90.28 (% of people ages 15 and above) Source: WDI 2006. Note: -- = not available. Figures 14.1 shows population pyramids for the years 1990, 2005, and 2020 (estimated), respectively. The figures reflect the effect of the government's family planning and population policy, one prominent feature of which is the active encouragement of a maximum of two children per family. 444 Good Practice in Health Financing Figure 14.1 Vietnam: Population Pyramids, 1990, 2005, and 2020 a. 1990 b. 2005 90­94 90­94 75­79 male female 75­79 male female 60­64 60­64 age45­49 age45­49 30­34 30­34 15­19 15­19 0­4 0­4 20 15 10 5 0 5 10 15 20 15 10 5 0 5 10 15 % of population % of population c. 2020 90­94 75­79 male female 60­64 age45­49 30­34 15­19 0­4 15 10 5 0 5 10 15 % of population Sources: United Nations Population Division, World Population Prospects, the 2004 Revision Population Database (http://esa.un.org/unpp/index.asp). Although the projections indicate that the total population will grow to around 125 million by 2025, it is plausible that Vietnam will be able to sustain its popula- tion with social services. The demographic transition of an aging population will, however, require continuous adaptation in the social service delivery system to the needs of elderly people. Vietnam has made significant progress in its national health programs for tar- geting specific diseases. For example, it has dramatically reduced the incidence of and death from malaria. It has also reduced the incidence of polio, neonatal tetanus, and leprosy to the point where they are no longer regarded as threats to the health status of the population. Vietnam: "Good Practice" in Expanding Health Care Coverage 445 Table 14.3 Vietnam: Estimated Burden of Disease, 2002 Top five diseases, by cause Estimated total Rank GBD Cause deaths per 1,000 people 1 Cardiovascular diseases 160.0 2 Infectious and parasitic diseases 74.6 3 Malignant neoplasms 64.1 4 Respiratory diseases 51.3 5 Unintentional injuries 36.1 Burden of disease, by disease category Estimated total GBD rank / cause deaths per 1,000 Share (%) 1 Communicable, maternal, perinatal, 126.2 24.5 and nutritional conditions 2 Noncommunicable diseases 341.1 66.1 3 Injuries 48.5 9.4 Total 515.8 100.0 Top five diseases, by cause Estimated total Rank GBD cause DALYs, 1,000 1 Infectious and parasitic diseases 2,337 2 Neuropsychiatric conditions 2,195 3 Cardiovascular diseases 1,389 4 Unintentional injuries 1,389 5 Sense organ diseases 809 Burden of disease, by disease category Estimated total GBD rank/cause DALYs, 1,000 Share (%) 1 Communicable, maternal, perinatal, and nutritional conditions 4,305 32.2 2 Noncommunicable diseases 7,334 54.9 3 Injuries 1,721 12.9 Total 13,360 100.0 Source: World Health Organization, Department of Measurement and Health Information, December 2004 (http://www.who.int/healthinfo/bodestimates/en/index.html). 446 Good Practice in Health Financing The estimated global burden of disease (GBD) for the main disease categories is shown in table 14.3. With respect to morbidity patterns, Vietnam has under- gone an epidemiological transition along with the economic and social changes over the past two decades. In 1986, communicable diseases accounted for more than half of the entire estimated disease burden, with noncommunicable diseases making up almost the full balance (World Bank 2001). By the end of the 1990s, noncommunicable diseases accounted for 65 percent of the main causes of illness. Now communicable diseases make up around 27 percent, and accidents and injuries account for the rest. Accidents and injuries are expected to continue ris- ing, not least as a result of road traffic incidents. As table 14.3 indicates, Vietnam faces a dual burden of disease. Morbidity and mortality from noncommunicable diseases and injuries are rising, while the bur- den of disease from communicable diseases remains heavy. As to specific causes of morbidity and mortality, it is estimated that the top 10 causes of hospital admissions were: pneumonia, acute pharyngitis and acute ton- sillitis, acute bronchitis and acute bronchiolitis, influenza, transport accidents, diarrhea and gastroenteritis of presumed infectious origin, essential primary hypertension, gastritis and duodenitis, diseases of the appendix, and intracranial injuries (Ministry of Health [MOH] information system). Similarly, the main causes of hospital deaths were: intracranial injuries, HIV, pneumonia, intracere- bral hemorrhage, transport accidents, acute myocardial infarction, stroke (not specified as hemorrhage or infarction), heart failure, respiratory tuberculosis, and intentional self harm (suicide). Information on causes of sickness and death is, however, very thin--deaths from the causes mentioned above explain only 5 per- cent of all deaths. Political and Governance Environment Vietnam is a one-party socialist republic with a president as chief of state and a prime minister as head of government. The legislative branch, which elects the president from among its members, consists of a national assembly with represen- tatives from the 64 provinces (comprising 611 districts/provincial towns and 10,600 communes/wards). The cabinet is appointed by the president and ratified by the national assembly. The national assembly is gradually assuming a more prominent role in policy debates and decisions, including in the social sectors. Given the country's relatively high decentralization, this is an important factor in terms of implementation of policies and possible adjustments. Formally, Vietnam is governed by People's Committees at various levels of the administra- tion, including the commune. These committees have far-reaching authority over mobilization and allocation of resources at their respective level of administra- tion. Furthermore, while it is fair to say that Vietnam is a centrally planned soci- ety, focused on production targets, attempts at active commune and village level participation in decision-making processes have been ongoing for many years. In contrast with its administration, in terms of implementation Vietnam is highly decentralized, even in an international perspective (World Bank 2004, Vietnam: "Good Practice" in Expanding Health Care Coverage 447 chap. 8). For example, spending at the subnational level has increased steadily since the early 1990s and now constitutes around 45 percent of state budget expendi- tures. An important factor in the decentralization process is the extent of inter- provincial transfers, which are determined by the share of locally raised revenues that each province can retain and the grant received from the central government. Because resource availability varies considerably among provinces, some provinces are net contributors while most are net recipients of such grants. In general, the grants are directed toward a set of expenditure areas that include social and eco- nomic activities. A concern has been raised that the allocation formula is not suffi- ciently commensurate with local needs and resource mobilization abilities (World Bank 2003a). In particular, moving toward allocations based on development out- comes, including equity concerns, would most likely improve the incentive struc- ture of the provincial and local authorities. This is a particular concern because the health sector is often given low priority when shortages of funds force cuts. Another decentralization issue is the distribution of autonomy among admin- istrative and service delivery units, including health care providers such as hospi- tals and clinics. The main point of conferring authority over resources is to improve operational efficiency. However, concern has been raised that the absence of transparency, accountability, and sound monitoring might lead to inappropri- ate use of resources and rent-seeking behavior among staff and managers of spending units (World Bank 2003a). This is a risk when decisions about invest- ments in expensive equipment are made without any higher-level coordination and control. Continuity of care between different levels of the system is also threatened under the strong decentralization push underway. As noted above, Vietnam has relied largely on a planning process that looks almost exclusively at the attainment of certain predefined production targets. Recently, the government has initiated a process that would give greater considera- tion to development goals (MPI 2005). Although this entails a much more analyt- ically demanding planning process, it could be expected that the country would thereby manage to sustain or improve upon its recent record in economic develop- ment and poverty reduction. Overview of Health Financing This section provides an overview of the Vietnamese health financing system. It starts by looking at the recent time trends in health financing followed by a description of the various health financing programs currently in existence in Vietnam and briefly reviews the main reimbursement systems. Finally, the section looks at some equity issues related to financing and utilization of care. Trends in Health Financing Vietnam's health financing indicators show a continuing increase in per capita health expenditures but a relatively constant share of GDP spent on health (table 14.4). Private spending accounts for about three-fourths of all spending. Out-of- pocket spending accounts for the largest share, although it has seen a decline as 448 Good Practice in Health Financing Table 14.4 Vietnam: Per Capita Total Expenditure on Health, 1999­2003 Indicator 1999 2000 2001 2002 2003 Per capita total expenditure on health at 111 129 145 143 164 (US$ at international dollar rate) Total expenditure on health 8.6 10.1 11.5 11.5 13.3 (US$ billion at international rate) Total expenditure on health as share of GDP (%) 4.9 5.3 5.5 5.1 5.4 General government health expenditure as share 6.7 6.0 6.7 5.1 5.6 of total government expenditure (%) Out-of-pocket spending as share of total 58.2 62.7 59.2 58.1 53.6 health expenditures (%) Social insurance as share of general government 9.5 10.5 13.7 15.8 16.6 expenditure on health (%) Private health expenditures as share of total 67.3 72.0 70.8 71.9 72.2 health expenditures (%) Public health expenditures as share of total 32.7 28.0 29.2 28.1 27.8 health expenditures (%) External resources for health 3.4 2.7 2.8 3.5 2.6 (% of total expenditure on health) Source: WHO 2006. health insurance has increased its share. Starting in 2003 with the establishment of the Health Care Fund for the Poor (HCFP) and in 2005 with free health care for children under age six, the government has greatly increased its spending commit- ments on health care. These new entitlements to free health care constitute an impressive increase in government funding of health care not yet reflected in the figures in the following tables. Vietnam is not overly dependent on ODA for health care spending in general. Only 2.6 percent of total health spending comes from external sources, although the share of external funding for preventive care is somewhat higher (Vietnamese NHA). Table 14.5 shows the share of health expenditures by type of service, and table 14.6 indicates the sources of funds by type of service. Most health spending goes for self-treatment--over-the-counter purchases of prescription and nonprescrip- tion drugs--not as part of a medical consultation, which accounted for a rela- tively stable high share of around 42 percent of all health spending from 2000 to 2003 (Vietnamese NHA). These purchases, mainly at private pharmacies, are not covered by health insurance or exemption programs. Self-treatment is the pri- mary treatment choice of both poor and rich (Chang and Trivedi 2003). Spending on inpatient care accounts for about 27 percent of all spending on health in Vietnam, mostly at state-run hospitals funded by user fees, govern- ment subsidies and health insurance reimbursements. Outpatient care accounts for about 14 percent of total health spending half of which comes from the state budget. Primary health care (PHC), funded largely by employers and school health programs, accounts for 7.5 percent of all spending. Vietnam: "Good Practice" in Expanding Health Care Coverage 449 Table 14.5 Vietnam: Health Spending, by Type of Service (percentage) Health care type 1999 2000 2001 2002 2003 Inpatient 30.1 29.5 26.7 26.4 27.0 Outpatient 14.2 12.0 16.4 13.7 13.5 Self-medication 38.1 43.5 41.7 42.0 42.1 PHC and school health 4.8 4.9 6.3 7.7 7.5 Preventive and public health 10.5 7.9 6.6 7.7 7.4 Other 2.3 2.2 2.3 2.4 2.5 Total 100.0 100.0 100.0 100.0 100.0 Source: NHA 1998­2003; MOH, Statistical Publishing House 2006. Table 14.6 Vietnam: Sources of Funding, by Type of Service, 2003 (percentage) State Health Health care type budget insurance User fees ODA Other Total Inpatient 27.2 15.2 37.5 1.4 18.8 100 Outpatient 49.6 19.8 10.1 1.3 19.3 100 Self-medication 0.0 0.0 100.0 0.0 0.0 100 PHC and school health 18.3 0.0 18.2 1.0 62.5 100 Preventive and public health 73.4 0.0 2.2 24.4 0.0 100 Other 67.3 10.8 0.0 6.1 15.8 100 Total 22.5 7.0 55.1 2.6 12.8 100 Source: MOH 2006. Preventive health programs have declined in importance to 7.4 percent of total spending, three-fourths of it from the state budget and most the balance from external assistance. Government administration and regulation of the health sec- tor, research, and training, account for a very small share of the total. Health Financing Programs As Vietnam undergoes its transition to a market economy, the government con- tinues to stress preventive care, maintaining state budgetary and ODA funding of national target programs and other preventive health measures. The government also continues to invest in the grassroots health network to implement these pro- grams and to deliver primary and maternal and child health care. However, in recent years, financing of curative health care in Vietnam has increasingly relied on mobilizing resources from society (under the term xa hoi hoa, socialization), primarily through user fees at government facilities, expansion of the private sec- tor, and a contributory social health insurance scheme. The share of expenditures of state-run health facilities covered by the government budget declined, from 450 Good Practice in Health Financing Table 14.7 Vietnam: Health Insurance Coverage, by Program Type, 1998­2004 Health insurance holders 1998 1999 2000 2001 2002 2003 2004 Percentage of population covered 12.7 13.4 13.4 15.8 16.5 20.3 27.7 by health insurance Of which: --poor (%) -- 0.6 1.1 1.9 2.1 4.6 4.8 --nonpoor (%) 12.7 12.8 12.3 13.9 14.4 15.7 22.9 Sources: 1998­2002 Health Insurance Statistical Yearbooks. 2003­04 provisional figures from Vietnam Social Insurance Agency as presented in Lieu et al. 2005. Note: -- = not available. 62.7 percent in 1998 to 49.6 percent in 2003 (Lieu et al. 2005). User fees and, to a lesser extent, health insurance have taken up an increasing share, ensuring fund- ing to maintain and improve the quality of health care facilities. Social health insurance was introduced in 1992 as a solution to help mobilize resources and create a more appropriate mechanism for payment of user fees (table 14.7). The Vietnamese Communist Party and government have decided that health insurance should be the main health care financing mechanism and have set the challenging goal of reaching universal health insurance coverage in the near future. At the same time, social health insurance has gradually become the main mechanism for subsidizing users in an effort to increase government health spending to promote health equity goals and overcome the negative conse- quences of user fees. Figures from the Vietnam Social Insurance Agency (VSIA) indicate growing coverage of the population by health insurance, especially among the poor who receive government assistance to buy health insurance. Cov- erage by a prepayment mechanism is even higher if other government health financing programs are included. The next section briefly describes the public and private health financing mechanisms for curative care currently available in Vietnam. Compulsory social health insurance. As mentioned above, compulsory social health insurance was begun in 1992. There are two groups of beneficiaries, those who have made contributions to health insurance and those whose contributions were paid by governmental agencies. The contributory compulsory social health insurance covers all Vietnamese employees with a contract of three or more months as well as pensioners and peo- ple on disability. Despite the compulsory nature of social health insurance for these groups, compliance is low for people in the private sector. Even among peo- ple making contributions for health insurance, there is some evidence of income underreporting so as to pay lower amounts into social insurance. Many groups in the population are eligible for compulsory social health insur- ance paid through the government budget not covered through their employ- Vietnam: "Good Practice" in Expanding Health Care Coverage 451 ment. The first group is individuals who have served the government or provided meritorious service to the country, including elected officials, political beneficia- ries who served in the revolution or wars protecting the nation, and family mem- bers of current military or police officers. The second group, the "vulnerable," for whom the government finances social insurance contributions, includes the elderly aged 90 and older, social assistance beneficiaries (such as the disabled, elderly without family to care for them, orphans), and foreign students on schol- arships from the government of Vietnam. In 2003, the government put in place the HCFP to ensure financing of health care services for the poor, ethnic minorities in disadvantaged provinces, and resi- dents of remote communes (under Program 135). Provinces were allowed to choose whether to implement health care for the poor by purchasing health insur- ance or by reimbursing facilities directly for services used by the poor. When Decree 63 on health insurance was issued in 2005, the direct reimbursement option was discontinued, and the provinces were required to buy health insurance for these target groups with funding from the central budget. Voluntary health insurance. Voluntary social health insurance, by law, covers indi- viduals who need health insurance but are not covered under the compulsory scheme, such as the self-employed or the nonworking population, or individuals who have compulsory health insurance but want to contribute to voluntary health insurance to supplement their compulsory coverage. The voluntary part of the social health insurance program is intended for groups of people rather than indi- viduals, so as to create a risk pool, reducing the risk of adverse selection. Most of the beneficiaries covered under voluntary health insurance are school pupils and stu- dents, a relatively easy group to enroll to meet the minimum coverage rate under the insurance regulations. Other options for risk pools include communes or vari- ous mass organizations and associations, but it has been more difficult to enroll enough members. Even then, the joiners tend to already have health problems and therefore pose a higher risk than healthy individuals to the insurance scheme. In addition to social health insurance, several private and state insurance com- panies offer health insurance, usually as an add-on to life insurance products including policies for pupils and students as well as older working people to cover hospitalization, surgery, and emergency transport. In 2003, according to the latest National Health Accounts (NHA), private health insurance reimbursements to facilities constituted about one-third of all health insurance reimbursements (MOH 2006). Other government-sponsored programs. There are two other government pro- grams of note. One, by Decision 139 of October 2002 on the HCFP, covers health care costs for the poor and other target groups. It also provides--in theory-- assistance to the nonpoor faced with unexpected and catastrophic health expendi- tures, but in practice implementation has been weak. 452 Good Practice in Health Financing The other, under the Law on Protection and Care of Children, began in 2005. All Vietnamese children under 72 months of age are eligible for free health care at all government health facilities, with facility services reimbursed directly from the state budget. In 2006, the prime minister asked the MOH and Vietnam Social Security (VSS) to study how to convert this scheme into health insurance coverage for children under age six. Benefits Packages of Health Financing Programs Compulsory and voluntary health insurance pay for outpatient and inpatient diagnosis and treatment at government and private facilities that have signed a contract with the social insurance agency agreeing to accept reimbursement at the rates set for public facilities. The package is generous, covering the costs of consul- tations, diagnosis, treatment, and rehabilitation during the time of treatment at the health facility; lab tests, diagnostic imaging, and other diagnostic techniques; medicines on the list drawn up by the MOH; blood and transfusions; medical procedures and surgery; use of materials, medical equipment, and treatment bed; and prenatal exams and assistance at delivery. Transportation costs for referrals from district- to higher-level hospitals are also covered by health insurance for the poor, political beneficiaries, and people living or working in mountainous and remote areas. There are also some exclusions. For example, health insurance does not reim- burse fees for certain health problems or treatments because of moral hazard issues, the elective nature of the intervention, or because those health problems are covered by other government programs.2 Patients may also have to make copayments for any additional cost incurred beyond what basic services would have cost at an appropriate level of care in government facilities.3 For children under age six, the benefits include consultations, diagnosis, and treatment at all primary care public facilities (including commune health stations, district hospitals), and, if formally referred, treatment at provincial and central- level facilities. In an emergency, a family can seek treatment for the child at any level of facility without a referral. If the family of a child requests use of certain high-tech equipment, the family must pay additional fees. When this policy is converted to health insurance coverage for children under age six, the benefits will likely be the same as for other insured groups. Financing and Payment This subsection discusses the main health financing and payment mechanism currently used in the Vietnamese health financing system, including contribution rates, sources of financing, and provider payment procedures. An important issue of health financing is the financial sustainability of the programs; this issue is dis- cussed toward the end of the subsection. Contribution rates and financing sources. Compulsory insurance contribution rates are set at a modest 3 percent of contractual salary and basic allowances, pen- Vietnam: "Good Practice" in Expanding Health Care Coverage 453 sion, social insurance payments, scholarship, or minimum wage, depending on the insured individual's entitlement group. Formally employed workers pay 1 per- cent of salary and their employers pay 2 percent. For retirees and people receiving social insurance benefits, the contributions are paid by the VSS. For the other groups, the government budget covers contributions. For the poor and the elderly aged 90 and older, the contribution is a fixed amount of US$3.10 per person per year with the contribution paid from the state budget. Voluntary social health insurance contributions range between US$1.90 and US$10.00, depending on the locality and the type of risk pool (e.g., school, association, commune) and are paid by individuals and their households. For children under age six, the government has allocated US$4.70 per child per year, but if the total amount required exceeds this amount, funds from the reserve budget line of the national budget would be used to pay the excess. The HCFP requires provinces to allocate US$4.40 per eligible beneficiary, with 75 percent paid in from the state budget and most of that used to buy health insurance. The remaining amount, if it can be mobilized, can be used to pay for health care of ineligible groups confronted with catastrophic health care costs. In the past, because of strict reimbursement ceilings imposed on hospitals, and despite low contributions, the health insurance fund has run a surplus. Since pas- sage of Decree 63 on revised health insurance regulations in 2005, there has been some concern about the solvency of the social insurance fund because contribu- tions remained low while the benefits package was expanded to include more than 200 additional medicines and many expensive high-tech services. In 2005, VSS estimated that health insurance reimbursements exceeded total contributions to health insurance by US$8.6 million, an amount equivalent to about 4.3 percent of total health insurance contributions. Currently this shortfall is being financed from accumulated reserves, but VSS has warned that these reserves may become exhausted in the coming years. Now, finding a solution for this issue is a top prior- ity, by either increasing contribution rates or tightening cost controls. To gauge their adequacy, contributions can be compared with per capita curative care spending. In 2003, according to NHA estimates, per capita annual spending amounted to US$7.10, counting only health spending at government hospitals and curative care facilities. Adding-in private curative care including self-treatment, annual per capita spending was US$20.30 (MOH 2006). There is some concern that with the generous package, coverage of a higher and higher share of people for whom low contributions have been paid (poor, elderly and soon children under age six), and with adverse selection among the voluntarily insured, major problems may arise in balancing the insurance fund. In addi- tion, the cost of health care is fast increasing with the development of new tech- nology and drugs, and contribution amounts will have to be raised periodically to keep up. Provider reimbursement procedures. Reimbursement of providers from health insurance and direct reimbursement of services for children under six and the 454 Good Practice in Health Financing poor not yet covered by health insurance is on a fee-for-service basis. The amount reimbursed to facilities is based on the partial schedule of user fees charged to uninsured patients, although in some cases the health insurance reimbursement is less than what facilities charge the uninsured. Drugs and disposables are reimbursed at the facility's purchase price. The "major drugs list" of the MOH lists the drugs reimbursed by health insurance. It was updated in 2005 and now includes 646 modern and 91 traditional medicines. Regulations on certain drugs require committee approval before use, and the drug list also specifies which level of hospital (district, provincial, central) is allowed to dispense each type of medicine. A list of procedures with prices reimbursed by health insurance was developed in 1995, and a price list covering new services was appended to it in 2006. Private facilities that sign contracts with the health insurance agency are reimbursed at the same rates as government facilities. Inherent in the fee-for-service payment mech- anism are incentives to overprescribe certain health services and medications. Sepehri, Chernomas, and Akram-Lodhi (2005) found evidence that the inpatient admission rates and the average length of stay were higher for better-off individu- als and for the insured. To overcome this provider moral hazard problem (or provider-induced demand), Decree 63 called for alternative payment mechanisms, including capita- tion and case-mix options (such as diagnosis-related group, DRG). In 2004, WHO negotiated an experiment at selected districts in Hai Phong province. Initial results of the experiment were favorable both in terms of maintaining service quality, health insurance fund solvency, and provider and patient satisfaction. However, in 2005 when new health insurance regulations came into effect cover- ing a wider range of services, especially high-tech services, the health insurance agency would not allow an increase in the capitation amount to cover the higher cost. Therefore, providers did not want to continue with the experiment. The MOH Department of Planning and Finance is doing a costing study as part of the process of developing a case-mix payment system to be pilot-tested in 2008. It is hoped that the case-mix system will improve the cost-effectiveness of health spending by reducing use of unnecessary services and drugs. This will contribute greatly to ensuring solvency of the health insurance fund. Measures to ensure solvency of the insurance fund. At present, the health insur- ance scheme reimburses facilities only the partial user fees charged to patients pay- ing out of pocket. The government continues to subsidize health facilities by paying salaries, buying equipment, and paying part of operating costs. If the government allows facilities to charge for the full cost of health care, the cost to patients and the insurance fund will increase substantially. Lieu, Long, and Bales (2007), studying the costs of treating five different medical conditions at provin- cial hospitals, found that the amount charged covers between 39 percent of full costs for pneumonia in children and 71 percent of full costs for intracranial injury. Vietnam: "Good Practice" in Expanding Health Care Coverage 455 Currently the MOH intends to continue subsidizing health care but would like to shift from direct payment to providers to payment through users, but it has not yet identified the appropriate mechanism for this purpose. In the hospital costing study by Lieu, Long, and Bales (2007), drugs accounted for between 30 percent and 70 percent of the full cost of services. The cost of diag- nostic procedures (depreciation, labor, and materials) was not found to be a high share of the total cost of care, but their share may increase with the development of high-tech services. The major drug list and list of user fees reimbursable by health insurance could theoretically be used to control costs. However, in establishing the lists of drugs and services, little attention was paid to the cost-effectiveness crite- rion. Thus, the benefits package is extremely generous--and ineffective as a tool for controlling costs. Imposing ceilings on benefits is another way of ensuring solvency of the health insurance fund. Individual medical services costing less than VND 7 million are reimbursed by insurance, but any amount over VND 7 million is reimbursed in different amounts for different beneficiaries. Copayments are often instituted to discourage overuse of health services (moral hazard), but all copayments were eliminated in 2005 (Decree 63) in an effort to attract prospective health insurance subscribers. Neither ceilings on individuals nor copayments, however, gives practitioners any incentive to deliver cost-effective services. In health care, practitioners, not consumers, usually decide which services to provide and what to charge. To avoid overspending the health insurance fund under the fee-for-service reimbursement mechanism, the VSS has capped the amount each first-contact facility will be reimbursed. In principle, the insurance fund available for reim- bursements at each hospital is based on the number of people insured and the average contribution amount in the province. In practice, however, each facility subtracts an amount from the insurance fund as payment for referrals. The amount is calculated as 110 percent of the amount actually reimbursed to referral facilities in the previous year. Because no caps are imposed on services provided by referral hospitals, there are no incentives for cost-effective services at this level. Inherent in this payment mechanism is an incentive for first-contact hospitals to refer patients. Although some expenditures in excess of the cap may be reimbursed by health insurance, there is a risk that they will not be or that reimbursement will be inor- dinately delayed. To avoid overspending, hospitals create their own controls on services to the insured, which often means that insured patients receive different services from those offered to patients paying user fees. Sometimes hospitals ask insured patients to "voluntarily" agree to pay the differential between the health insurance reimbursement and the amount the hospital would collect from fee- paying patients. This practice, though forbidden by the health insurance regula- tions, is not stopped because of sympathy with the difficulties of staying within the health insurance fund budget. 456 Good Practice in Health Financing Equity: Health Indicators and Outcomes, and Their Distribution Several studies have reported inequitable utilization of health care services across living-standards groups, with the poor less likely to seek any care and, when they do, going to lower-level facilities. These studies have also shown differentials in the burden of health care expenditures with the heavier burden on the poor (Lieu, An, and Long 2005; Knowles et al. 2003; Minh, Khang, and Dung 2004; Chien 2005; Dong et al. 2002). Earlier efforts by the Vietnamese government to improve health equity and access to health care for children under age six involved unfunded exemptions at government hospitals and provision of health insurance to a small share of the poor who were hard to target. The Vietnam National Health Survey (VNHS) of 2001/02 covered the period just before the approval of the HCFP policy (Decision 139). Low insurance coverage was found among the poor, ethnic minorities, and inhabitants of mountainous regions. In addition, fee waivers or health insurance reimbursements were obtained for only a third of inpatient visits by people eligi- ble for Decision 139 benefits, compared with two-thirds of children under age six and nearly three-fourths of the people covered by health insurance (MOH 2003). The negative consequences and inequities in access to curative health care and financial burdens for the poor have been documented in numerous studies (Knowles et al. 2003; Minh, Khang, and Dung 2004; Chien 2005; Dong et al. 2002; Wagstaff and van Doorslaer 2003). Wagstaff and van Doorslaer found that out-of- pocket expenditures for care contributed to increasing the number of absolute poor by 3.4 percentage points at the end of the 1990s. The HCFP policy was developed because of the documented inequities in health outcomes, access to health care, and differential financial burden on the poor. Evi- dence of the strong impact of health insurance on access to health care helped to orient the policy toward subsidizing users through health insurance, although orig- inally the policy also allowed direct government reimbursement to facilities for ser- vices used by the poor. Implementation of the HCFP policy began in 2003, and the results of the first impact evaluation only recently became available. Implementa- tion of the policy on free health care for children under age six only began in 2005. During the implementation process, administrative reports from the provinces were compiled to identify shortcomings and implementation difficulties. Targeting specific populations. The government's objective in targeting assistance to cover user fees is not based simply on need (that is, providing health care for the poor and the very old). It is also politically motivated (for example, covering peo- ple who have performed meritorious service to the country, nonpoor ethnic minorities) and anchored in concern about the health of future generations (health care for children younger than six). One of the early complaints from the provinces about Decision 139 was prompted by perceived unfairness of providing assistance to some better-off peo- Vietnam: "Good Practice" in Expanding Health Care Coverage 457 ple who happened to be part of the target group such as certain residents of disad- vantaged communes or better-off ethnic minorities in mountainous provinces. In addition, it was felt that the many near-poor ethnic minorities residing outside of the 12 provinces targeted as disadvantaged also deserved assistance. In the first revision of the HCFP policy, in 2006, one of the main changes was to reduce leak- ages to better-off groups and undercoverage among near-poor ethnic minorities. Its implementation has been hindered by a lack of clear-cut criteria for select- ing the individuals who should receive assistance with catastrophic health expen- ditures from among the group of people who are ineligible for free health care cards or health insurance. The lack of well-defined criteria also makes it difficult to evaluate whether provinces are using these funds appropriately. According to administrative records of implementation, only about half the provinces have actually implemented this policy provision. Overview of the Health Delivery System in Vietnam This section provides an overview of the Vietnamese health service delivery sys- tem, including the supply and organization of health care and the regulatory framework. Among other things, the data show that Vietnam is comparable with most other low-income countries in terms of service availability, albeit with large variation in quality across various parts of the country. Supply and Organization of the Health Care System The health care system in Vietnam is organized in four tiers: the MOH, the Provin- cial health bureaus, the district health centers (DHCs), and the commune health stations (CHSs) (World Bank 2001). The MOH, together with the people's com- mittees at the various administrative levels, is responsible for formulating and exe- cuting the country's health policy and programs. In addition to the MOH, several separate units and institutes function outside the ministry in areas such as train- ing, research, and policy development in especially important areas, including family planning. In addition, 31 central-level hospitals provide mainly preventive care services together with training, research, and supervision of some targeted preventive care programs. The second tier, the provincial health bureaus, is fairly influential, considering Vietnam's extensive decentralization. The provincial health bureaus, part of the provincial governments under the provincial people's committees, are charged with the overall planning of provincial health services and programs. Immediately below the provincial health bureaus are the District Health Centers (DHCs). In 2006, as part of the decentralization process, the DHCs were split into three entities with distinct responsibilities: district health offices, district hospitals, and district preventive health centers. District hospitals and preventive health cen- ters, for the most part, still supervised and managed by the provincial health bureaus, ensure curative and preventive care for district residents. The district 458 Good Practice in Health Financing health office, under the district people's committee, is responsible for government administration regarding overall management of the commune health stations, food safety and inspections, and disease prevention. Finally, at the lowest level of the health care system are the CHSs, charged with delivering primary health care and executing the national health target pro- grams. In addition to the CHSs, a system of village health workers (VHWs) pro- vides residents of remote villages with basic health care, information-education communication (IEC), and first aid. In terms of infrastructure, Vietnam has around 1101 hospitals with a total of 138,675 hospital beds. Fifty-five percent of the hospitals are district hospitals, 24 percent are provincial hospitals, and around 3 percent are central hospitals under the responsibility of the MOH. In addition, there are some 77 non-MOH hospi- tals that belong to other ministries and government units. In 2006, there were 62 private hospitals, a number that is expected to increase rapidly as a result of fur- ther deregulation and the government's intentions to further diversify health ser- vice provision. As to the organization of health care and programs, most primary health care is delivered at the CHSs, although much is also provided higher up in the system in district, and even provincial, hospitals. The primary health care network in Vietnam has a long history. Until the mid-1980s, most care was provided by the local agricultural work brigades. This grassroots system formed the backbone of the service delivery system, but under the economic reforms at that time, these primary health care providers were largely dismantled. Recognizing the ensuing difficulties, in 1994 the government assumed the responsibility for paying the salaries of all health workers. The national health programs, implemented from the top down, are an impor- tant feature of health care delivery in Vietnam. They are free or available at very low fees to the general public. The government has approved a revised list of national target programs in health to include: HIV/AIDS, community mental health (mainly schizophrenia and epilepsy), TB, leprosy, malaria, cancer, malnutri- tion, reproductive health, military-civilian medical collaboration, and an ex- panded program of immunization (EPI). In the CHS a large range of maternal and child health services are also provided, including prenatal care, deliveries, and health care for small children. Coverage of these primary health care and preven- tive programs is nationwide although evidence from a recent national health sur- vey (MOH 2003 [VNHS 2001/02]) indicates that the programs are not being implemented uniformly across provinces and districts. In addition to the public facilities, Vietnam's growing private health sector pro- vides outpatient care and traditional medicine services and sells pharmaceuticals. As people's ability to pay for health care evolves and current policy reforms take hold, the size and scope of the private health care sector is expected to grow. Infor- mation is scant, however, about the size and scope of the private health care sec- tor because private facilities come under the jurisdiction of the provincial health Vietnam: "Good Practice" in Expanding Health Care Coverage 459 bureaus, and many private providers are not licensed. Distinguishing public from private facilities is also difficult, because many public doctors also operate private practices and many public hospitals are setting up semiprivate wings providing more"hotel-like"services to increase their revenues under the policy for promoting hospital autonomization. The most recent data on health care utilization suggest that private delivery of care may be as high as 30 percent of all care, with large vari- ations across provinces and with a concentration in the urban areas (MOH 2003). Statistical Overview of the Health System Using recent international health system statistics, this section presents a set of key health system indicators for Vietnam on human resources, service coverage, and some related health outcomes. In terms of nonfinancial health system inputs, table 14.8 shows human resources devoted to health care. At around 1.3 government health workers per 1,000 inhabitants, Vietnam is on a par with other low-income countries. As professional training improves, the qualifications of health system personnel are also expected to improve. In addition to these formal health care providers, the many alternative (traditional) health care providers constitute a large and well-recognized part of the delivery system. Viet- nam's ratio of 23 hospital beds to 10,000 inhabitants in 2003 compared with 39 for the region. The average numbers hide many details but seem to show that Vietnam has developed a relatively well-staffed delivery system with decent physical infra- structure, including hospitals and lower-level clinics. Vietnam compares favorably in terms of per capita availability of health care facilities, but service quality, access, and utilization vary across regions. For exam- ple, although people in the most remote and mountainous areas have access to primary health care facilities, use is limited by poor quality, restrictions in physical access, financial obstacles, and possibly cultural factors. Furthermore, CHSs are perceived as implementing only target health programs, although their mandate extends to providing curative services as part of primary health care. There is some concern about low CHS utilization, estimated at an average of 12 consulta- tions a day (World Bank 2001). Reasons for the low utilization rates at primary care level include the availability of alternative providers, staffing irregularities, and perceptions of low quality. Table 14.8 Vietnam: Human Resources for Health, Selected Indicators Physicians Nurses Midwives Pharmacists Total Density/ Density/ Density/ Density/ Density/ 1,000 Number 1,000 Number 1,000 Number 1,000 Number 1,000 (2003) 42,327 0.53 44,539 0.56 14,662 0.19 5,977 0.08 1.3 Sources: WDI 2006; WHS 2006. 460 Good Practice in Health Financing Table 14.9 Vietnam: Selected Health Care Coverage and Health Outcome Indicators, 1990­2004 Indicator 1990 1995 2000 2001 2002 2003 2004 Births attended by skilled health staff -- -- 69.60 -- 85.00 -- 90.00 (% of total) Immunization, DPT (% of children 88.00 93.00 96.00 96.00 75.00 99.00 96.00 ages 12­23 months) Improved water source 72.00 -- -- -- 73.00 -- 80.00 (% of population with access) Improved water source, rural % of rural population with access) 67.00 -- -- -- 67.00 -- -- Improved water source, urban (% of urban population with access) 93.00 -- -- -- 93.00 -- -- Improved sanitation facilities (% of population with access) 22.00 -- -- -- 41.00 -- 32.00 Malnutrition prevalence, height -- -- 36.50 -- -- -- -- for age (% of children under 5) Malnutrition prevalence, weight -- -- 33.80 -- -- 28.40 -- for age (% of children under 5) Diarrhea treatment (% of children -- -- -- -- 39.00 -- -- under 5 receiving oral rehydra- tion and continued feeding) Prevalence of HIV, total -- -- -- 0.30 -- 0.40 -- (% of population ages 15­49) Malaria prevention, use of -- -- 15.80 -- -- -- -- insecticide-treated bed nets (% of under-5 population) Incidence of TB (per 100,000 202.17 -- -- -- -- -- 176.48 people) Tuberculosis cases detected -- 29.61 82.06 83.40 86.87 85.38 88.81 under DOTS (%) ARI prevalence (% of children -- -- 9.30 -- -- -- -- under 5) ARI treatment (% of children -- -- 60.40 -- 71.00 -- -- under 5 taken to health provider) Contraceptive prevalence -- -- -- -- 78.50 -- -- (% of women ages 15­49) Source: WDI 2006. Note: -- = not available. Table 14.9 presents a set of selected health service coverage indicators for Viet- nam and some health outcome variables considered to be influenced by the health care system. Among other things, the table indicates that immunization coverage in Viet- nam is high. Since the mid-1980s, coverage has increased from less than 55 per- cent to more than 95 percent (World Bank 2001). Although immunization rates Vietnam: "Good Practice" in Expanding Health Care Coverage 461 may be particularly high, other indicators of health service coverage also show similar increases. These improvements have contributed to the achievements in bringing down morbidity and mortality rates in Vietnam. Regulatory Framework The MOH is the key regulator in the health sector. The ministry is responsible for developing national policies and monitoring and supervising the national public health programs. An important policy document is the Health Sector Master Plan to the Year 2010 and Vision to the Year 2020, recently developed by the MOH Department of Planning and Finance in collaboration with other sector depart- ments. The government envisions a continued emphasis on care at grassroots while simultaneously developing more advanced medical technologies. The ministry is also responsible for developing and drafting central government health sector deci- sions and sending them to the provinces with specific implementation guidelines. The private sector is regulated by the Ordinance on Private Practice of Medi- cine and Pharmacy, enacted in 1993 and updated in 2003, and by the decree detailing the implementation provisions. The ordinance stipulates the procedures and professional requirements for obtaining a private license to practice. The reg- ulations specify 10 types of private health facilities, four types of private tradi- tional medicine facilities, and four types of private pharmaceutical facilities. Licenses for private providers, including pharmacies, are issued by the provincial health bureaus after testing and screening the applicants. The health insurance regulatory environment is changing fast. Currently, health insurance is regulated by Decree 63 of 2005, and VSS has overall responsi- bility for its implementation, including collecting contributions, managing the fund, and reimbursing providers. With technical assistance from the WHO, a law on health insurance is being developed (ILO 2006), and it is expected to be passed by the national assembly in 2008. As a reflection of the growing importance of health insurance as a financing instrument, a department for health insurance has been formed within the MOH. The new department is formally charged with developing legislation and strategies for attaining universal health insurance cov- erage and is expected to take an active part in the supervision and monitoring of current and new programs. Additional Components of the Health Delivery System Together with the Ministry of Education, the MOH is responsible for educating and training all health care personnel at national and regional health training facilities. Vietnam has several medical universities that prepare health personnel and specialists in most medical areas. In line with its overall responsibility for the health sector, the MOH is also charged with generating most health statistics and information. The VSS is responsible for collecting and compiling health insurance data, a task expected to become more important as insurance coverage expands. 462 Good Practice in Health Financing The government of Vietnam is planning technical advancements for the health sector so that it can also provide services and products for export purposes, including medical treatments for foreign visitors and pharmaceutical exportation. Central-level hospitals are becoming more technically sophisticated, and many demanding medical procedures can already be performed in Vietnam. Although improvements in medical technology will most likely play an important role in enhancing the quality of medical care, how much this process will help Vietnam achieve its major public health goals is not clear. For example, the main pediatrics hospital in the largest city can already deliver advanced neonatal care to save early preterm newborns. Such services are likely to be out of reach, however, for the many poor households on the fringes of society and the still large number of mothers giving birth at home. The pharmaceutical sector has a prominent place in the Vietnamese health care system. Medicines, for example, absorbed half of all health spending in 2003 (MOH 2006). They were dispensed by a variety of providers only partly controlled and supervised systematically by the authorities. Despite existing laws and institu- tions to regulate the sale of pharmaceuticals, Vietnam is experiencing serious problems with antibiotic resistance, partly as a result of irrational use of drugs due to poor supervision and monitoring. Formally, responsibility for inspection and control of the pharmaceutical sector lies with the Drug Administration of Viet- nam (DAV) and the MOH Department of Therapy. In practice, however, enforc- ing a policy for rational use of drugs is difficult, as indicated by the fact that 73 percent of all reported illnesses are treated by self-medication with over-the- counter drugs bought from public or private drug vendors (MOH 2003). The Health Coverage Reforms in Vietnam The health system in 1986, prior to the Doi Moi series of socioeconomic reforms, was described in detail in a review of Vietnam's health sector (MOH 2001), on which this section strongly draws. Rationale for Reforms in Vietnam The guiding principle of the Vietnamese health system prior to reform was "health for all." At grassroots, an extensive network of CHSs and intercommunal polyclin- ics provided the population with primary health care. Government hospitals treated people needing more intensive care. Despite the wide coverage, the quality of health care suffered from a lack of resources. From a severe shortage of trained health workers at independence and very short-term basic training, the medical training of Vietnamese health workers improved as the system developed and medical schools expanded. Whereas in 1945 there were 180,000 persons per doctor, by 1985 the ratio had fallen to 3,137. Nevertheless, poor access to developments in medical science for train- ing medical workers and low salaries were important causes of low-quality care. Materials and medicines, especially western drugs, were distributed through the Vietnam: "Good Practice" in Expanding Health Care Coverage 463 health service but were generally insufficient, especially in CHSs. The health sector relied heavily on traditional medicine for treating common ailments, and distrib- ution of the limited supply of modern drugs was tightly controlled. Supervision of health care services through inspections was infrequent due to limited availability of vehicles, fuel shortages, and absence of telephones in addition to limitations in clinical competence. During the pre-reform period, only 3 to 4 percent of the state budget was allo- cated to health care, insufficient to cover the recurrent costs of government health facilities. Few funds were available to upgrade health facilities or provide medical equipment. Only some 8 percent of the perceived need for medical equipment was met. The basic operating costs of facilities above the commune level were cov- ered solely through the government budget. The financing of CHSs depended on revenue from the agricultural cooperatives. Foreign assistance from the former Soviet Union and other socialist countries dried up abruptly at the end of the 1980s. Because of the high dependency on for- eign aid and the substantial government budget deficit at the time, health financ- ing in Vietnam was severely constrained. Many health facilities did not even have enough money to maintain routine activities or pay salaries. Much of the medical equipment was obsolete or broken. The drug supply, already inadequate, wors- ened at every level. No longer could the government subsidize current expendi- ture for all health facilities, and alternative sources of finance had to be mobilized urgently to keep the health system afloat. Chronology of Reforms The major economic and social reforms began with the Sixth Communist Party Congress, in 1986. The economy has gradually made the transition from a cen- trally planned to a socialist-oriented market economy. International trade and exchanges of information have increased. Dramatic economic growth and rapid improvements in living standards have been the result. However, gaps between rich and poor and between better-off and disadvantaged regions have widened in terms of both economic development and access to quality health care. A series of health system financing reforms have been put in place at first to overcome resource constraints of the health system and more recently to narrow the gaps between rich and poor and advantaged and disadvantaged regions to ensure health for all. This section describes the main health financing reforms since Doi Moi began in 1986. A general overview of the chronology of the main reforms in Vietnam is provided in table 14.10. The replacement of cooperative agriculture with family-based production in the Doi Moi reforms meant that the commune people's committees lost revenue from which to finance social services. The real income of CHS workers declined-- their salary payments could be delayed by months--and their morale plummeted. Starting in 1987 as part of the market-oriented reforms, private health and pharmaceutical sectors were officially sanctioned. As a result, Vietnam saw a rapid 464 Good Practice in Health Financing Table 14.10 Vietnam: Chronology of Health Reforms, 1989­2006 Year Reform measure 1989 Fee for service Private practice Pharmaceutical sales 1992 Health insurance Legal document on private practice 1994 Upgrading of the CHS system Salary for the CHS staff Health Policy Unit at MOH 1996 National drug policy Policy for social mobilization (Xa hoi hoa) of health care National target programs in health 1997 Provincial health system model 1998 Hunger Eradication, Poverty Reduction Program: Health Cards for the Poor 1999 Redefine goals of health sector reform 2003 Health Care Funds for the Poor (HCFP) 2004 Implement free health care for children under age 6 2005 New health insurance regulations deepening coverage 2006 Drafting revision of HCFP and health care for the near-poor Drafting health insurance law Source: Adapted from Adams 2005. increase in private medical practice and retail pharmacies, but primarily in better- off areas and cities. Most private practices were small, and many of them were run by moonlighting government health workers. Greater freedom given to state pharmaceutical production and import companies and an increasing number of private pharmaceutical producers and distributors resulted in a rapid rise in the volume and diversity of pharmaceuticals on the market. As service at CHSs dete- riorated, attendance at these facilities fell. People, including rural residents, resorted increasingly to more convenient and friendly over-the-counter drug pur- chases or private consultations for basic medical care despite its greater expense. Also in 1987, in response to the dire state of many CHSs, the government decided that the provincial budget should subsidize salaries for a set number of CHS staff members per commune (typically three). However, by the early 1990s, only a third of CHSs were being subsidized in this way, and in 1994 the prime minister signed Decision 58/TTg, designating CHS staff members as government workers and allocating funds for their salaries to be paid from provincial budgets. This salary support has greatly improved income and morale among CHS work- ers, and it contributed to the strengthening of the Vietnam's public primary health care system. Partial service fees and charges for drugs and diagnostics were officially intro- duced in public hospitals and other health facilities in 1989. The revenue was Vietnam: "Good Practice" in Expanding Health Care Coverage 465 retained by the collecting facilities and was to be used to improve services. Certain population groups (such as children under age six, the poor, residents of remote or mountainous areas) were to be granted partial or total exemption from charges by the individual health facility, but this policy was inadequately funded and proved difficult to implement effectively or fairly. As a result, user charges have deterred people, especially the rural poor, from seeking public health care, and especially care at more expensive, higher-level hospitals. Although charging service fees has not been official policy for CHSs, many of them do levy a small fee. They may also earn substantial revenues from pharmaceutical sales. Health insurance was introduced in Vietnam in 1992 to lighten the cost burden on the government health budget and increase people's access to health care. There was an initial surge in the number of people with compulsory, employment- related coverage and voluntary student insurance. Official documents have stated universal health insurance with equity of coverage and increased cost sharing of health service financing as a goal for 2010. However, attempts to extend voluntary health insurance to the general population have not met expectations and have led to serious problems of adverse selection. In 1998, and again in 2005, health insur- ance regulations were revised, gradually eliminating differences in the benefits packages between types of insurance and increasing the package of services and drugs reimbursable through health insurance. Cost-control measures put in place by the social security administration in charge of health insurance have forced some hospitals to underprovide health services to ensure that their costs do not exceed their annual allocation of health insurance funds. Public health spending increased rapidly in real terms in the 1990s. In 1994, the MOH initiated a program to upgrade medical equipment in public hospitals. By 2000, central and provincial hospitals had most of the necessary equipment. Of district hospitals, 90 percent had X-ray machines; 70 percent, ultrasound machines; and 92 percent, ambulances. The Communist Party and the govern- ment issued guidelines for consolidating and strengthening the basic rural health network. External assistance helped finance the rehabilitation of CHSs and DHCs, focusing on disadvantaged areas. By 2002, 98 percent of the communes had a functioning CHS, and the goal is for every CHS to have a doctor by 2010. Although the government is still increasing its contributions to the health sec- tor, it is strongly pushing a policy of social mobilization (cost sharing) to mobilize additional funds for the health sector. An important component of this policy is the policy on semiautonomous public hospitals laid out in Decree 10 in 2002 and revised in 2006 by Decree 43. These facilities would move from state budget fund- ing to a self-financing status with decentralized management powers. The purpose is to create incentives and conditions for public hospitals to operate more effi- ciently. There is some concern, however, that this policy will encourage commer- cialization of the Vietnamese hospital service, with perverse incentives to overuse investigations, drugs, and services to raise revenues and to concentrate on money- making activities to the detriment of other essential, but less lucrative, aspects of health care. 466 Good Practice in Health Financing Household health care costs increased rapidly after the reforms and put a par- ticularly heavy burden on poor rural households. Equitable access to, and utiliza- tion of, health care has always been a high priority of both the public health service and the Communist Party in Vietnam. Therefore in the mid-1990s, a pol- icy of providing government-subsidized health insurance cards for the poor was put into effect to overcome the financial barriers to health care that resulted from the user-fee policy. Insufficient funds meant that not all the poor were covered, however, and the contribution per card was inadequate to cover the cost of ser- vices. To overcome these weaknesses, Decision 139/2002/QD-TTg provided for all provinces and cities to establish an HCFP with a major subsidy from the central state budget. The target beneficiaries include all poor people, as well as residents of disadvantaged communes and ethnic minorities in disadvantaged provinces. The amount allocated to the fund per poor person was increased substantially, and the provinces were allowed to use it either to purchase health insurance or to reimburse hospitals directly for health care services used by targeted beneficiaries. In 2005, the Law on Protection and Care of Children came into effect, mandat- ing free health care for children under six years of age. State budget funds were allocated, and the policy is being implemented as a direct reimbursement scheme from the state budget to state facilities providing health care to children under age six. Work is beginning on a health care policy for the elderly. Evaluation of Health Coverage Reforms Vietnam's impressive achievements in reducing poverty and improving health and other outcomes among the population have attracted considerable interest from the international community, including health policy researchers and ana- lysts. Quite a few studies have been published on several key outcomes, but only a handful of studies have applied rigorous quantitative analysis to gauge the causal effects of the policy interventions. The most important contributions to the impact are reviewed below. Health insurance in Vietnam has been shown to have important impacts on health outcomes, access to health care, and out-of-pocket expenditures, especially for low-income individuals. Four recent studies have looked at the impact of the compulsory social health insurance program. Wagstaff and Pradhan (2005) applied quantitative techniques, including double- difference and matching, to evaluate the impact of the government's social health insurance program on four outcomes: health, utilization, out-of-pocket spending, and poverty (nonmedical spending). Using panel data from the 1990s, they found that Vietnam's social health insurance has had a positive impact on height for age and weight for age of young school children and body mass index (BMI) of adults. This study also found that, for young children, health insurance increased use of primary care facilities and reduced self-medication. Among older children and adults, health insurance led to an increase in the use of hospital inpatient and out- patient care. Finally, the authors found that health insurance led to a reduction in Vietnam: "Good Practice" in Expanding Health Care Coverage 467 annual out-of-pocket spending on health and an increase in nonmedical house- hold consumption. Using econometric modeling on the same data as the previous study, Sepehri, Sarma, and Simpson (2006) also found a positive effect of the compulsory social health insurance program on out-of-pocket spending. Controlling for unobserved heterogeneity, social health insurance reduced out-of-pocket expenditures between 16 and 18 percent with a more pronounced reduction for low-income individuals. Wagstaff (2005) analyzed the effect of a negative health shock on income, med- ical spending, and household consumption. The author found that this type of health insurance does provide financial protection against the adverse effects of illness. Finally, Chang and Trivedi (2003) also found a strong negative insurance effect on use of self-medication because the insured were much more likely than the uninsured to seek a medical consultation in case of illness. The application of these evaluation techniques to household survey data, including panel data, pro- vides strong evidence of a positive and causal impact of this type of insurance on key policy outcomes. With regard to the voluntary component of the Vietnamese health insurance program, Jowett, Contoyannis, and Vinh (2003) used a special survey to collect information from individuals on health care use and spending in three Viet- namese provinces in the late 1990s. Among other findings, the authors reported a 200 percent reduction in the average out-of-pocket spending for outpatient care. As to the effect of voluntary health insurance on utilization, Jowett, Deolalikar, and Martinsson (2004) found that, using the same data source and controlling for potential selection bias, voluntary health insurance in Vietnam led to higher use of outpatient facilities and public providers and less reliance on self-treatment and private providers, particularly strong effects at lower income levels. Although these studies are not national in scope, the finding that voluntary health insurance has had a positive impact is important for the government's policy development in the health sector. Nguyen (2003) evaluated the impact of the health card initiatives introduced in the late 1990s as part of special programs to reach target groups. The author applies first-difference techniques coupled with propensity score matching on national household survey data from 2002 to analyze the impact on health care utilization and out-of-pocket spending. The results indicate only limited or no impact of the health care program on these outcomes. A recent report discusses several reasons for the apparent absence of an impact from these efforts to target the poor and other socially excluded groups, including incomplete implementa- tion of the program, low levels of coverage, and finally, the need to improve the operation of the scheme so as to provide an effective protection mechanism (World Bank 2003a). The policy of free health care to children under six has not yet been formally evaluated. However, a recent report discusses this policy, based on a case study of 468 Good Practice in Health Financing one central-level children's hospital in Hanoi (UNDP 2005). The report notes the hospital's increased patient load as a result of the policy. While it is too early to know for sure, the study discusses the possible effect on quality of care that the increase in utilization may have. In addition, future evaluations of this policy might usefully look at the benefit incidence of utilization of child care that may change to the advantage of the urban rich. The impact of the program on Health Care Funds for the Poor has recently been evaluated using nonexperimental techniques, including propensity score matching (PSM) with national survey data. Among other things, the evaluation shows that the program has led to an increase in service utilization by the program beneficiaries (Axelson 2007). For example, those covered by the HCFP use services around 6 percent to 10 percent more than do the comparison persons. The pro- gram reduces out-of-pocket expenditure around 18 percent. The impact evalua- tion shows that catastrophic out-of-pocket payments are reduced by around 17 percent (personal communication). In addition, small studies designed to evaluate Decision 139 in particular provinces or regions have also shown promising results. Households interviewed in a WHO study in Bac Giang and Hai Duong provinces reported increased utilization of health services after implementation of the HCFP, especially for inpatient care (Axelson et al. 2005). The study also found a signifi- cant increase in health care seeking at the CHS as the first contact among HCFP beneficiaries. Although these preliminary findings are encouraging, it is safe to say the imple- mentation of this important program is not problem-free. The quality of primary health care, especially in remote areas, is still poor. Health insurance contributions are too low to cover the cost of the expanded package of services promised the insured, and dangerous deficits threaten the solvency of the health insurance fund. Incentives for both public and private health care facilities to overprescribe medicines or expensive diagnostic services have led to a rapid escalation in health expenditures and waste. The lack of adequate supervisory and quality control sys- tems have exacerbated the overprovision of drugs and services and delayed fur- ther improvements in quality of care. Main Conditions for Success During the implementation process of developing a policy for health financing in Vietnam, certain conditions for success have become evident. Providers need ade- quate funding to continue to provide health services. Although the state remains a primary financing source for government health services, user fees have helped fill the gap left by inadequate state financing so that provision, and the attempt to improve quality, can continue. Inadequate reimbursement of services by health insurance has led to underprovision of services or unofficial copayments by the insured. Finding out what full costs are is a precondition to ensuring their full coverage by existing financing sources. Only now, however, are studies being done to find out the full cost of services provided at state facilities. Vietnam: "Good Practice" in Expanding Health Care Coverage 469 Implementation of assistance programs for the poor or other target groups requires both clear targeting and adequate financing. The initial user-fee policy called for exemptions or reductions for the poor and children under age six. This policy failed because the facilities were not adequately reimbursed for the services provided, and targeting was unclear. A new policy to ensure health care for the poor promised a small amount of funding to buy health insurance for the very poorest. However, it did not provide clear criteria for targeting these people, and the contributions did not cover service costs. Since then, the policy for the poor and for children under age six has evolved so that more funding is available, and the targeting criteria are clear. Monitoring and evaluating the impact of health financing policy has been important in orienting revisions or developing corrective policies for negative impacts. Findings from evaluations of the impact of user fees on access to care and the heavy burden of health expenditures on the poor was instrumental in con- vincing the government of the importance of a policy to assist the poor. Adminis- trative monitoring of this policy, although it was not comprehensive and did not rely on a rigorous research process, was still adequate to show that earlier policies did not provide adequate coverage. This evidence impelled the development of the HCFP to ensure coverage for all poor and several near-poor groups. Assessment of Remaining Health System Strengths and Weaknesses Compared with many other countries in a similar economic situation, Vietnam's health system is performing well, particularly at the primary health care level, despite some regional variations. Historically, the strong emphasis on the grass- roots level with a wide organizational reach has likely contributed to the good health outcomes. The government continues to envision strong development of the health sector with an emphasis on developing medical technology in terms of both products and services. Moreover, its limited financial inputs compared with other countries and its favorable health outcomes suggest that Vietnam's health sector is relatively efficient. Nonetheless, many difficulties persist in the performance of the health service delivery system. One problem that has been observed is high reliance on hospital care. Compared with other countries with similar national incomes, Vietnam ranks among the highest in terms of hospital beds per population, lower only than some of the countries of the former Soviet Union with its strong emphasis on hospital services (WHO 2006). Related to this is the inadequacy of the referral sys- tem, which leads to low utilization at lower levels and high use at upper levels of care, even for simple conditions. Moreover, the rational use of care is hampered by some weaknesses in the monitoring system where health information data are not systematically analyzed for use in planning and management. A third issue is the observed differences in access to and utilization of health ser- vices in Vietnam. These differences have likely increased in the recent past and may continue to do so for some time. As discussed elsewhere in this chapter, however, 470 Good Practice in Health Financing the government has introduced measures to address these inequalities. The most important of these measures are Health Care Funds for the Poor and the policy of free health care for children under six years of age. Finally, the problem of irrational use of drugs and antibiotic resistance is an area in which the government will have to step up counteracting measures. Medical staff training, effective supervision, and the purposeful enforcement of existing regulations are important factors in this respect. Health Coverage Reforms: Enabling Factors and Lessons from Vietnam The main health coverage reforms in Vietnam over the past two decades have been described in this chapter. During this period, the health sector has undergone many substantial transformations, including the introduction of cofinancing for health care, the private provision of services, and the gradual introduction of var- ious types of prepayment mechanisms, including social health insurance and free health care for certain target groups. The discussion has also shown that coverage rates have increased steadily in terms of both services and health insurance. For example, today more than 90 percent of all children receive vaccinations com- pared with only around 55 percent in the mid-1980s. Over the same period, cov- erage by some type of health insurance has increased to almost 60 percent of the total population. Despite these successes, many problems remain, as illustrated by widespread child malnutrition, substantial inequalities in access to and utilization of health services, and the existence of catastrophic out-of-pocket payments for health care. Enabling Factors The relative success of the Vietnamese government in extending health coverage to an increasingly larger share of the population, in terms of both services and health insurance, can be attributed to many different factors both inside and out- side of the health sector. Among the most important such enabling factors outside of the health sector, three stand out. The first of these external factors is the rapid and sustained economic growth of the past 15 to 20 years that has enabled the government to embark on a series of ambitious health financing reforms and set the goal of universal health insurance by 2010. Although health spending as a share of GDP has remained largely constant over this period, per capita spending on health has increased from less than US$15 to over US$25. Moreover, the growth of national income has enabled the government to significantly increase its share of health spending. This has meant that substantial additional resources are now being channeled to the new entitlements of free health care for the poor and for children. The importance of economic growth for sustainable financial support to gov- ernment provision of health insurance is all the more evident, considering that Vietnam has received limited amounts of development assistance for health dur- Vietnam: "Good Practice" in Expanding Health Care Coverage 471 ing this period. On average, less than 2 percent of total health spending has come from foreign assistance, compared with the much larger contributions seen in other low-income countries. Without economic growth of 6 to 7 percent a year, these reforms would still be wishes. Moreover, economic growth in Vietnam has been inclusive in that widespread poverty has been reduced. In light of the well- established importance of poverty reduction for improved health (and vice versa), this is a critical component of the Vietnamese experience. One final result of sustained economic growth and limited dependence on development assistance for health is the fact that Vietnam has been able to develop and pursue policies that have been genuinely domestic in their scope and aims. This relative independence in policy making is not unique to the health sector. It can be seen in the country's overall development approach, which has meant that ownership of programs has been high and the incentive and ability to correct faulty implementation have been present. The second enabling factor external to the health sector is the relatively well- structured and organized Vietnamese society. The political structure of the coun- try, with at least nominally effective channels of communication from the central level down to the community level, has enabled the government to implement national policies fairly uniformly and in a timely fashion. As described above, the health care delivery system has a clear structure from village level up to the ter- tiary hospitals at the central level. This, combined with political stability in the postwar period, has most likely contributed to the government's ability to imple- ment many of its policy programs. Moreover, the relatively structured organization of Vietnamese society will probably be an important asset when the country computerizes its public man- agement system, including the health information system, enabling the govern- ment to strengthen its monitoring and evaluation system. An effective information system will enable the authorities to implement even more effectively the special antipoverty programs targeting certain groups in difficult socioeco- nomic circumstances. Although recent reports have emphasized the limited leak- age under these programs, the scope for further improvements should be evident from the broadened data bases after computerization of the health and other information systems. One final enabling factor that can be seen as external to the health sector is the high esteem with which the Vietnamese people and the government have held health. For example, around a third of the population does some kind of regular exercise to stay healthy (MOH 2003). And the leadership has always emphasized public health as an important instrument for reducing widespread poverty. These sociocultural factors, combined with nutritional and healthy food consumption habits most likely also contribute to the good health outcomes in Vietnam com- pared with most other comparable countries. It can also be reasonably assumed that these factors have helped facilitate the uptake of the health coverage reforms, because the Vietnamese understand the importance of health for well-being. 472 Good Practice in Health Financing In addition to these external factors, at least three elements within the health sector can help explain Vietnam's good performance in providing its people with health service and financial coverage over the past 20 years. The first of these is the inheritance of a wide network of primary health care services. The grassroots level of the system has long been seen as the most important part of an effective and equitable service delivery system. Over the course of reform, the government has built strategically on these past investments to further develop their efficiency in providing the population with needed services and interventions. Moreover, a clear policy goal of the government has long been able to provide the necessary interven- tions for disease prevention and eradication. The strategic approach of providing effective preventive care for specific infectious diseases has led to a decline in the burden of disease from those conditions and a relative increase in the disease bur- den due to conditions that cannot be easily prevented by traditional means. The second important internal factor has been the timing, scope, and sequenc- ing of the identified reform measures. In terms of service provision, the govern- ment identified a set of priority health problems and diseases on which to concentrate its attempts to diminish the burden of disease. Although other prob- lems may have existed, setting priorities was seen as a necessary means for putting scarce resources to the best use. Furthermore, the priorities have been allowed to adapt to the gradually changing epidemiological pattern, for example, by the inclusion of HIV/AIDS prevention in the national disease priority program and the issuing of policies on tobacco control and prevention of accidents and injuries. Finally, besides skilled health professionals and clinicians, effective and equi- table health care requires extensive planning and management capacities, and the Vietnamese health sector is blessed with such skills. Within every level of the sys- tem, a formal structure delineates the management responsibilities of each type of provider. One result is the production of clear operational plans for the coming activity period. The weakness of the system is patchy monitoring. Related to the health planning and management skills of the health system is the capacity to develop useful health policies. A strategic effort on behalf of the central health authorities is to develop skills and capacities in health policy making, including analysis of the efficiency and distributional impact of policies and programs. Lessons from Health Coverage Reforms in Vietnam Many of the health coverage reforms that Vietnam has introduced can be imple- mented in other countries, but some of the enabling factors might not be easy to replicate. For example, special funds to finance the health care needs of certain groups can theoretically be introduced in other contexts. The long-term sustain- ability of such programs, however, presumably requires relatively high economic growth for extended periods of time and rapid declines in the size of beneficiary groups. With regard to political economy factors, an important dimension of the Viet- namese experience has been economic and political stability. The one-party struc- Vietnam: "Good Practice" in Expanding Health Care Coverage 473 ture of the Vietnamese governance system has enabled the country to develop policies that have been allowed to be implemented over a long period of time without interference from varying and sometimes competing political factions. The policy lesson from this factor points to the need to reach agreements on health policy reform measures that span the political divide in multiparty democ- racies. Otherwise, health reform measures may not be allowed to run their course over a sufficiently long time to take effect. In terms of good value for money, the importance Vietnam has put on primary health care can serve as an important lesson to other low- and middle-income countries now introducing health coverage reforms. As noted, the grassroots level has been the backbone of the health care system. It has enabled the health sector to gradually introduce ever more advanced organizational and medical procedures, such as provider payment reforms and liberalizations of the health care produc- tion markets. Low-income countries need to build a national capacity to develop, imple- ment, and evaluate health coverage programs and other policies. Moreover, coun- tries should attempt to use such capacities most effectively by pooling analytical resources, thereby creating a critical mass for effective health policy analysis. Lessons from Vietnam in this respect also include the strategic use of development assistance without allowing it to become too dominating. Generally, Vietnam has used its development assistance in health to develop health policy and other core competences and to implement the national disease priority programs. In terms of health insurance coverage, the experiences of Vietnam with design and sequencing of implementation may be of interest to other countries. While universal coverage may be the long-term goal, countries should not attempt to reach it too quickly or by means of only one or two alternatives. Vietnam initiated its health financing coverage attempts by introducing social health insurance in the early 1990s. The eligible groups were well defined, and the benefits package of ser- vices was identified, although in the case of Vietnam it may be too broad. Gradu- ally, more people were covered by prepayment as other programs were introduced, including voluntary health insurance and free health cards for the poor. Importantly, the most recently implemented insurance benefits package was defined to dovetail with the initial program, thereby creating a certain level of transparency in the health insurance system. This pattern was repeated with intro- duction of the Health Care Funds for the Poor in 2003. Creating a uniform bene- fits package in this way limits the scope for overlap within the health insurance system, which reduces transaction costs. Should the government choose to liberal- ize the health insurance market by allowing private for-profit agents to enter, this last point may become even more important because private actors would then compete with existing insurance programs. From the client's perspective, a rela- tively uniform and generous benefits package would facilitate the choice between alternative programs and prevent private firms from skimming the market. 474 Good Practice in Health Financing Inasmuch as every country attempting to achieve universal health coverage faces public resource limitations, one lesson from Vietnam concerns the impor- tance of targeting those scarce resources to the people most in need. Recent analy- sis has shown that, compared with many other countries, Vietnam has been able to develop effective targeting mechanisms with limited leakage and undercoverage (World Bank 2003a). Because it has also been shown that these achievements are a result of a pragmatic approach by local authorities (Vietnam Dev Report 2004), another lesson from Vietnam would be that the successful implementation of tar- get programs requires some operational decentralization. For administrative cost- effectiveness, it was also important that the health sector could latch on to the targeting being done by the Hunger Eradication and Poverty Reduction program, instead of having to fund the heavy administrative work of doing its own targeting. Finally, the case of Vietnam also highlights the importance of introducing health coverage reforms gradually and incrementally. Vietnam introduced social health insurance for the formally employed first then made a concerted effort to cover the poor and the socially excluded. This approach gave the authorities important opportunities for learning by doing, lessons that will be valuable when tackling the persistent and most difficult challenges of increasing coverage rates for the rural poor. The importance cannot be sufficiently emphasized of having good information and knowledge about the health care­seeking behaviors of dif- ferent target groups, such as children, the rural poor, and women. Endnotes 1. Unless otherwise noted, the October 24, 2006, exchange rate is used throughout this chapter: US$1 = Vietnamese Dong (VND) 16,080. 2. These exceptions are treatment of leprosy; medicines for treatment of TB, malaria, schizophrenia, epilepsy, and other diseases already covered by government-funded pro- grams; diagnosis and treatment of HIV/AIDS, except when HIV tests are part of protocols for treatment of other diseases or if the individual was infected with HIV through work; syphilis and gonorrhea; immunizations, nursing care, early diagnosis of pregnancy, health checkups, family planning services, and infertility treatment; plastic surgery, prosthetics, false teeth, eyeglasses, and hearing aids; occupational diseases, labor accidents, and war- related accidents; treatment costs in suicide attempts, self-inflicted harm, drug addiction or health problems associated with illegal activities or behavior; health assessments for legal reasons; and consultations, treatments, rehabilitation, or home deliveries. 3. For example: special requests for services such as selecting their own practitioner, type of bed, or specific facility; bypassing a primary care facility without a referral; seeking treatment at a facility that does not have a contract with health insurance; seeking care at a private facility; or seeking treatment overseas. References Adams, S. 2005."Vietnam's Health Care System: A Macroeconomic Perspective." Paper pre- pared for the International Symposium on Health Care Systems in Asia at Hitotsubashi University, Tokyo, January 21­22. Vietnam: "Good Practice" in Expanding Health Care Coverage 475 Akal, A. 2004. "A Comprehensive Review of Health Insurance and Financing Regulations: FAQs." Background paper, WHO, Hanoi, September. Axelson, H. 2007. "Impact of Decision 139 on Health Care Utilization: A Study Using Propensity Score Matching." Background Paper on Workshop on Impact Evaluation of Decision 139, Hanoi. Axelson, H., D. V. Cuong, N. K. Phuong, T. T. M.Oanh, D. H. Luong, and K. A. Tuan. 2005. "The Impact of the Health Care Fund for the Poor on Poor Households in Two Provinces in Vietnam." Paper presented at Forum 9, Global Forum for Health Research, Mumbai, September 12­16. Health Strategy and Policy Institute and World Health Organization, Hanoi. Baulch, B., Truong Thi Kim Chuyen, Dominique Hougton, and Jonathan Hougton. 2002. "Ethnic Minority Development in Vietnam: A Socioeconomic Perspective." WPS 2836, World Bank, Washington, DC. Chang, F. R., and P. K. Trivedi. 2003. "Economics of Self-Medication: Theory and Evi- dence." Health Economics 12(9): 721­39. Chien T. T. T., ed. 2005. Health Care for the Poor in Vietnam. Hanoi: Medical Publishing House. ------. 2003. Vietnam Health Report 2002, Hanoi: Medical Publishing House. Dieleman, M., P. V. Cuong, L. V. Anh, and T. Martineau. 2003. "Identifying Factors for Job Motivation of Rural Health Workers in North Vietnam." Human Resources for Health: 10. http://www.human-resources-health.com/content/1/1/10. Dong, Tham Tat, Pham Trong Thanh, Dam Viet Cuong, Duong Huy Lieu, and Nguyen Hoang Long, eds. 2002. User Fee, Health Insurance and Utilization of Health Services. Hanoi: Vietnam­Sweden Health Cooperation Programme. Glewwe, P., Nisha Agrawal, and David Dollar, eds. 2004. Economic Growth, Poverty, and Household Welfare in Vietnam. Washington, DC: World Bank. Government of Vietnam. 2003."The Comprehensive Poverty Reduction and Growth Strat- egy (CPRGS)/Poverty Reduction Strategy Paper (PRSP)," Government of Vietnam/ International Monetary Fund/World Bank, Hanoi. Government of Vietnam and World Bank. 2005. Vietnam--Management of Public Finances for Growth and Poverty Reduction: Public Expenditure Review. Vols. 1 and 2. Hanoi: Finance Publishing House. GSO (General Statistical Office). Various years. "Vietnam Household Living Standards Survey." GSO, Hanoi. Gwatkin, D., Adam Wagstaff, and Abdo Yazbeck. 2005. Reaching the Poor with Health, Nutrition, and Population Services: What Works, What Doesn't, and Why. Washington, DC: World Bank. Health Strategy and Policy Institute and World Health Organization. 2004. "Assessment of Decision 139 Implementation." Informal paper, Hanoi. ILO (International Labour Office). 2006. "Enhancing Social Protection in Vietnam- Developing the New Social Security Law." Informal paper, Hanoi. IMF (International Monetary Fund). 2005. "Vietnam: 2004 Article IV." Consultation­Staff Report, IMF, Washington, DC. 476 Good Practice in Health Financing Jowett, Matthew, P. Contoyannis, and N. D. Vinh 2003. "The Impact of Public Voluntary Health Insurance on Private Health Expenditures in Vietnam." Social Science and Medi- cine 56: 333­42. Jowett, M., A. Deolalikar, and Peter Martinsson. 2004. "Health Insurance and Treatment- Seeking Behaviour: Evidence from a Low-Income Country." Health Economics 13: 845­57. Knowles, James C., Nguyen Thi Hong Ha, Dang Boi Huong, Nguyen Khang, Tran Thi Mai Oanh, Nguyen Khanh Phuong, Vu Ngoc Uyen. 2003. Making Health Care More Afford- able for the Poor: Health Financing in Vietnam. Vol. 1: Main Report, MOH/ADB TA- 3877-VIE. Hanoi. ------. 2003. Making Health Care More Affordable for the Poor: Health Financing in Viet- nam. Vol. 2, Annexes, MOH/ADB TA-3877-VIE, Hanoi. Lieu, D. H., N. H. Long, D. B. Huong, S. Bales, N. K. Phuong, C. Jiaying, M. Segall, and H. Lucas. 2005."Alternative Approaches to Ensure Health Care for the Rural Poor: Lessons for Vietnam from the Chinese Experience." Draft Report, Ministry of Health, Hanoi. Lieu, D. H., Nguyen H. Long, and Sarah Bales. 2007. Study on Treatment Costs for Selected Disease Groups in Provincial General Hospitals. Hanoi: Medical Publishing House. Lieu, D. H., N. Q. An, and Nguyen H. Long. 2007. Study on Health Care Financing from a Household and User Perspective." Hanoi: Medical Publishing House. Minh, H. V., N. Khang, and N. V. Dung. 2004. "Health Care Financing for the Poor in Viet- nam," Working Paper No. 2004-8, World Bank, Washington, DC. Ministry of Finance, Vietnam. 2005. "Medium-Term Expenditure Framework," Hanoi. MOH (Ministry of Health, Vietnam). 2006."National Health Accounts 1998­2003." Hanoi: Statistical Publishing House. ------. 2005. "Treatment Costs for Selected Disease Groups at Provincial General Hospi- tals."VSHC/HPC, Hanoi. ------. 2004a. "Assessment of 1-Year Implementation of Decision 139." MOH, Hanoi. ------. 2004b. National Health Accounts: The Process of Implementation and Development in Vietnam. Hanoi: Statistical Publishing House. ------. 2003. "Vietnam National Health Survey 2001/02." Main report, MOH, Hanoi. ------. 2001. 55 Years of Developing Vietnam's Revolutionary Health System [55 nam phat trien su nghiep y te cach mang Vietnam] Hanoi: Medical Publishing House. MPI (Ministry of Planning and Investment). 2005. The Five-Year Socioeconomic Develop- ment Plan 2006­2010. Hanoi: MPI. Nguyen,V.C. 2003."Assessing the Coverage and Impact of Vietnam's Programs for Targeted Transfers to the Poor Using the Vietnam Household Living Standard Survey 2002," World Bank, Hanoi. Sepehri A., S. Sarma, and W. Simpson. 2006. "Does Nonprofit Health Insurance Reduce Financial Burden?--Evidence from the Vietnam Living Standards Survey Panel." Health Economics 15: 603­16. Sepehri A., R. Chernomas, and H. Akram-Lodhi. 2005."Penalizing Patients and Rewarding Providers: User Charges and Health Care Utilization in Vietnam." Health Policy and Planning 20(2): 90­99. Vietnam: "Good Practice" in Expanding Health Care Coverage 477 UNDP (United Nations Development Programme). 2005."Preliminary Assessment of Free Health Care for Under-Six-Year-Old Children Policy at Central Pediatric Hospital." Report to UNDP workshop on Financial Policy Analysis, Hanoi. Van de Walle, D. 2004. "Testing Vietnam's Public Safety Net." Journal of Comparative Eco- nomics 32: 661­79. Van de Walle, D., and D. Gunewardena. 2000. "Sources of Ethnic Inequality in Vietnam." Policy Research Working Paper 2297, World Bank, Washington, DC. Wagstaff, Adam. 2005."The Economic Consequences of Health Shocks."World Bank Policy Research Paper 3655, Washington, DC. Wagstaff, Adam, and Eddy van Doorslaer. 2003. "Catastrophe and Impoverishment in Pay- ing for Health Care: With Applications to Vietnam 1993­1998." Health Economics 12: 921­34. Wagstaff, Adam, and M. Pradhan 2005. "Health Insurance Impacts on Health and Non- medical Consumption in a Developing Country," WPS 3563, World Bank, Washington, DC. World Bank. 2006. World Development Indicators (WDI). World Bank. Washington, DC. ------. 2005. "Vietnam: Managing Public Expenditure for Poverty Reduction and Growth--PER and IFA." World Bank, Hanoi. ------. 2004. "Vietnam Development Report 2005: Governance." World Bank, Hanoi. ------. 2003a. "Vietnam Development Report 2004: Poverty." World Bank, Hanoi. ------. 2003b. "Ninh Thuan PPA, Poverty Task Force." World Bank, Hanoi. ------. 2001. "Growing Healthy: A Review of Vietnam's Health Sector." World Bank, Hanoi. WHO (World Health Organization). 2006. World Health Report 2006: Working Together for Health. Geneva: WHO. ------. 2003. "Health and Ethnic Minorities in Vietnam." Technical Series No. 1, WHO, Hanoi. A APPENDIX About the Coeditors and Contributors THE COEDITORS Pablo Gottret is the lead economist of Health, Nutrition, and Population in the Human Development Network of the World Bank. Before joining the World Bank in 2002, he was viceminister of budgeting in his native Bolivia (1987­90) and chief executive of the Regulatory Body for Private Pensions, Private Insur- ance, and Securities (1998­2002). Gottret led the technical teams that developed Bolivia's Capital Markets Law, Insurance Law, and reforms to the Pension System Law approved by the Bolivian Congress between 1997 and 1998. He has worked in several countries in Latin America, Africa, and Eastern Europe. Gottret has a PhD in economics from Texas A&M University with specializations in econometrics, natural resource economics, and finance. He has published several books and arti- cles in refereed journals. George J. Schieber is a public finance economist with more than 30 years of domestic and international experience in health policy and economic issues, working in more than 40 countries. He has been an assistant professor of econom- ics at the University of Pittsburgh, a staff member at the Urban Institute, and an economist at the OECD, Paris. He has held various positions in the U.S. govern- ment, including the director of the Office of Research and Demonstrations in the U.S. Health Care Financing Administration, and at the World Bank, including health and social protection manager in the Middle East and North Africa Region. He is currently a consultant to the World Bank and various governments. He has published more than 60 books and articles. Schieber holds a PhD in economics from Syracuse University. Hugh R. Waters is a health economist and associate professor in the Department of Health Policy and Management at the Johns Hopkins Bloomberg School of Public Health. His areas of expertise are (1) health insurance and health financing reforms; (2) evaluation of the effects of health financing mechanisms on access, 479 480 Good Practice in Health Financing equity, and quality; and (3) economic evaluation of health care interventions. He has 19 years' experience working with public health programs and has worked extensively as a consultant with the World Bank, World Health Organization, and other international organizations. He currently teaches a course on comparative health financing systems at the Bloomberg School of Public Health. He holds a PhD in public health economics from the same institution and an MS in interna- tional economics from Georgetown University. THE AUTHORS Hédi Achouri, M.D., graduated from the Faculty of Medicine of Tunis (1980). He practiced general medicine and was regional director for public health until 1986. Since 1987, he has been director of public hospitals in the Tunisian Ministry of Public Health headquarters. He was a founding member and president of the Maghreb Economics and Health Systems Network, a nongovernmental organiza- tion working in the fields of health economics and health systems management in the Maghreb Region (Algeria, Morocco, and Tunisia). He is currently chairman of the Board of Habib Thameur Hospital (Tunis), a member of the Board of the National Health Insurance Fund, and general director of health public facilities in the Tunisian Ministry of Public Health headquarters. Among his recent publica- tions is Advances in Implementing Social Security: Lessons from Tunisia (GTZ, Ger- many, 2007). Chokri Arfa was a health economist at the National Institute of Public Health (1995-2002). He is currently assistant professor and head of the Research Depart- ment at the National Institute of Labor and Social Studies, University of Carthage. He is the coauthor of several health economics studies and a book on "Economie et Santé: Evaluation et stratégies de mise en oeuvre". His research interests are national health accounts, health performance, and health surveys. He has worked as a short-term consultant for several national and international studies. His most recent expertise is Tunisian's health accounts and monography of health insur- ance. In 2007, he was awarded a prize at the European Workshop on Efficiency and Productivity Analysis X in Lille, France, for work on how to expand the use of production capacity in Tunisian's local hospitals. Sarah Bales, a statistician, works as a health policy adviser to the Ministry of Health in Vietnam. She has extensive experience in health economic and policy analysis and has contributed statistical expertise to several national health and household surveys in Vietnam. Most recently, she was the principal statistical expert for the impact evaluation of the Vietnamese health financing program for the poor. Ricardo Bitrán D., an industrial engineer from the University of Chile, with an MBA in finance and PhD in economics from University of Boston, is a senior About the Coeditors and Contributors 481 health economist with 25 years of experience in Chile, the United States, and more than 40 developing countries. At the University of Chile he is professor of public health in the School of Public Health, professor of health economics in the Mas- ter's in Public Policy Program, and adjunct assistant professor. His areas of exper- tise include economic evaluation of health investment and reform projects; study of health care markets; analysis of equity, efficiency, and financial sustainability; and delivery and financing of primary health care services. During his 10 years with Abt Associates Inc. of Cambridge, Massachusetts, first as a health economist and then as a senior scientist, he was also research director for the HFS Project, a US$18-million initiative funded by the U.S. Agency for International Develop- ment. Since 1994, he has been president and senior partner of Bitrán & Asociados. His 20 years of academic experience have encompassed teaching posts or assign- ments in Chile, China, France, Lebanon, Thailand, the United States, Zaire, and many other countries. He has been senior adviser to the GAVI Alliance and World Bank on health immunizations issues and academic director of the World Bank Institute Flagship Program in Health Sector Reform and Sustainable Financing in Latin American and the Caribbean Region. Bitrán is author of academic articles, books, and scientific papers. James Cercone is an economist and president of Sanigest Internacional, a health care management and consulting company. He developed his interests in health economics and health reform through engagements with the World Bank (1990­94) and the government of Costa Rica (1994­97). He has more than 16 years' experience in the health sector, with particular emphasis on Latin America and transition countries. In the past five years, Cercone has been involved in the startup of several venture capital funds focusing on health sector investments, as well as provided investment and investment advice to clients around the world. His experience covers more than 40 countries, with emphasis on health policy reforms, private sector investment, impact of corruption in the health sector, hospital restructuring, health insurance reform and purchasing strategies, devel- opment of provider payment mechanisms, performance improvement, eco- nomic analysis of health projects, monitoring and evaluation, development of HIV/AIDS strategies, and management of health care services. Cercone has worked extensively on the design, implementation, and evaluation of projects for the World Bank, Inter-American Development Bank, ILO, and other develop- ment organizations. Björn Ekman is an economist from Lund University in southern Sweden. He is currently assistant professor in the Health Economics Program (HEP) at Lund University. His research interests are health financing and health systems development in low- and middle-income countries and the economics of mater- nal and child health. He has worked as a consultant for WHO, Sida, and ADB in Asia and Africa. 482 Good Practice in Health Financing Lisa Fleisher is a senior analyst with the Health Systems 20/20 Project, funded by USAID and implemented by Abt Associates Inc. She has worked in international health and development for the past six years. Her areas of expertise include health financing, health system strengthening, health policy analysis, and global public health. Currently, Fleisher's work is focused primarily on providing technical sup- port to the HIV/AIDS and National Health Accounts components of Health Sys- tems 20/20. Before that, Fleisher was a public health specialist at the World Bank, where she worked on health financing and health systems, issues related to the global health aid architecture, and the World Bank's new strategy for health, nutri- tion, and population. Other experience includes the implementation and adminis- tration of USAID-funded activities for youth reproductive health and HIV/AIDS prevention in Africa and Latin America, as well as grant making in health services research focused on domestic health care financing and organization. Fleisher received her master of public health in international health from the Johns Hop- kins Bloomberg School of Public Health and her BA in human biology from Stan- ford University. Jarno Habicht, M.D., is head of the World Health Organization Country Office in Estonia, in the WHO Regional Office for Europe. Before joining WHO, he worked for the Ministry of Social Affairs in Estonia, the World Bank, and the Estonian Health Insurance Fund. He has previous experience in health services research and management of research programs. He is currently a PhD student in public health at the University of Tartu, Estonia. Triin Habicht heads the Health Economics Department in the Estonian Health Insurance Fund. She has also been working for the Estonian Ministry of Social Affairs in the public health area. Habicht is currently doing her PhD in economics at the University of Tartu, Estonia. Melitta Jakab is a WHO policy adviser in the Kyrgyz Republic, working mostly on health care financing issues and equity analysis. Previously, she worked at the World Bank in Washington, D.C., and in Hungary. Jakab has a PhD in health pol- icy with concentration in economics from Harvard University. Adam Leive, currently an economist with the Fiscal Affairs Department of the International Monetary Fund, worked in the Health, Nutrition, and Population Department of the World Bank during research for this volume. He has a master's degree in health economics from the University of York, United Kingdom, and a BA from Princeton University in the Woodrow Wilson School of Public and Inter- national Affairs. Elina Manjieva, a WHO consultant, works in the Center for Health System Development under the Ministry of Health of the Kyrgyz Republic. Previously, About the Coeditors and Contributors 483 she was a junior professional associate in the Human Development Unit of the World Bank Country Office in the Kyrgyz Republic. Elina has an MA in develop- ment economics and southwest Asian studies from the Fletcher School, Tufts University. José Pacheco Jiminez holds an MA in development economics from the Institute of Social Studies in The Hague, the Netherlands. Between 2001 and 2003, he was a staff member of the Ministry of Finance of Costa Rica and worked as a consultant for the Congress, the National Bank, and the Telecommunications Company of Costa Rica. Since 2003, he has worked for Sanigest Internacional, where he is direc- tor of the Department of Social Economics and coordinator of R&D. Pacheco has participated in more than 25 projects in 17 countries around the world, particu- larly in Latin America, the Caribbean, Eastern Europe, and Central Asia, where he has focused on payment mechanisms, contracting, health insurance, financing, and poverty analysis. Diana Margarita Pinto Masís received her MD and master in health administra- tion from the Pontificia Universidad Javeriana, Bogotá, and her doctor of science from the International Health Policy and Economics program at the Harvard School of Public Health. She worked as a technical adviser at the Colombian Min- istry of Social Protection, developing and executing projects to evaluate the design of the benefits packages and the financial sustainability of the health system. She is currently associate professor in health economics at the Department of Clinical Epidemiology of Pontificia Universidad Javeriana School of Medicine and an associate researcher at Fedesarrollo. Her research interests include impact evalua- tion of health policy and health interventions, technology assessment, and health system performance assessment. Ravi P. Rannan-Eliya is a physician and economist who was graduated from the University of Cambridge with degrees in political science and medicine. After his internship in the United Kingdom, he specialized in international public health, earning a master's degree in public health and then his doctoral degree in health economics from Harvard University. From 1993 to 1997, he was a member of the research faculty at Harvard University, where he worked on a series of research projects in Latin America, Africa, and Asia. From 1997, he established and devel- oped the leading health economics research group in Sri Lanka, at the Institute of Policy Studies, before transforming it in 2005 into a full-fledged, independent research center, the Institute for Health Policy, which he now heads. He has under- taken research and consulted in more than 30 countries, working with the World Bank, WHO, ADB, and other agencies and governments. His current research focuses on issues of health systems financing, aging, equity, and noncommunica- ble disease. 484 Good Practice in Health Financing Lankani Sikurajapathy was a research assistant at the Institute for Health Policy at the time of writing and now works for UNFPA in Sri Lanka. She was graduated with a degree in sociology with social policy from University of Warwick, before completing a master's degree in comparative social policy at the University of Oxford. While at IHP, she worked on a number of projects related to health sys- tems financing and development of district health performance indicators. Ajay Tandon is a senior economist with the Health, Nutrition, Population hub at the World Bank in Washington, D.C. He has a PhD in economics from Virginia Tech. He has worked on a variety of issues related to health economics, including the measurement and evaluation of health system metrics. He worked previously with the Asian Development Bank in Manila, Philippines, as well as the World Health Organization in Geneva, Switzerland. He also held research appointments at the University of Oxford and Harvard University. Suriwan Thaiprayoon is a new health policy analyst. She earned her undergradu- ate degree in Nursing Science before she obtained her master's degree in interna- tional public health (Honors) from the University of Sydney, Australia. Her main interests are universal health care coverage, mechanisms of international coopera- tion and healthy public policy. Thaiprayoon served as a technical assistant to the Thai public health minister from October 2006 to February 2008. She currently works at the Bureau of International Health, Ministry of Public Health, Thailand. Gonzalo Urcullo C., an economist from the Catholic University of Bolivia, holds a master's in management and public policy from the University of Chile. For more than a decade, he has worked in key posts in Bolivia's government, assessing the Council of Ministries in the areas of industry, tourism, telecommunications, min- ing, environment, and transport. He has worked in the fields of economic regula- tion, promotion of competition, public credit, and decentralization of planning. Urcullo also has experience in social areas. He has a diploma in labor economics from San Simon University in Cochabamba, Bolivia, and has worked in the Min- istry of Education. He has also worked formulating public development strategies to reach the Millennium Development Goals. Urcullo has published in the areas of pension systems, regulation, economic development, and health financing. Since January 2006, he been an associate at Bitrán & Asociados, working on health pro- jects in Bangladesh, Bolivia, Chile, El Salvador, Guatemala, and Peru. Suwit Wibulpolprasert, M.D., is a general practitioner, a public health specialist, an administrator, and a policy advocate. He began his career as a director and a practitioner in four rural district hospitals in Thailand (1977­85). Later he was the director of the Northeastern Public Health College, the director of the Techni- cal Division of the FDA, the director of the Bureau of Health Policy and Plan, and assistant permanent secretary and deputy permanent secretary of the Ministry of About the Coeditors and Contributors 485 Public Health. His main interests are in health policy and planning, and interna- tional health. He has been extensively involved in research and development in the areas of human resources for health, health economics and health care financing, international trade and health, health promotion, health information, and phar- maceuticals. He has published more than 100 papers, reports and books locally and internationally. In Thailand, Dr. Suwit is the editor of a local journal for paramedical personnel and has produced radio and television programs on health and social issues for more than 15 years. He is currently president of the Folk Doctor Foundation; eval- uation board member of the Thailand Research Fund; and board member of the Health Systems Research Institute, the National Health Security Board, the Thai Medical Research Council, the National Nanotechnology Centre, the Mahidol University Council, and the National Science and Technology Board. As part of his international involvements, he represents Thailand in many international health forums and the World Health Assembly. He also represented Thailand and the Southeast Asia Region as a member and vice chair of the govern- ing board of the Global Fund to Fight AIDS, TB, and Malaria from mid-2001 to March 2004. He was also president of the Intergovernmental Forum on Chemical Safety from November 2003 to September 2006 and a member and vice chair of the WHO executive board in 2004­07. At present, Dr. Suwit is a board member of the Health Metrics Network and the Mekong Basin Disease Surveillance Network. He chairs the steering committee of Asian Partnership on Avian Influenza Research and the steering committee of the Asia­Pacific Action Alliance on HRH. He is also a member and represents the chair of the Program Coordinating Board of the UNAIDS, and he is a member and chair of the Program and Policy Com- mittee of the Interim Board of the Global Health Workforce Alliance. He served as a deputy permanent secretary at the Ministry of Public Health, Thailand, in 2000­03. Currently, he serves as a senior adviser in disease control and is also responsible for health policy and international health works of the ministry. Index ACE inhibitors, 252 economic context, 89­92 ADP. See Kyrgyz Republic, Additional Drug economic growth, 90f, 93 Package elderly programs, need for, 128 aging population, 185f elections, 99 AMG. See Tunisia, medical assistance Explicit Health Guarantees (GES), 104, programs 125f, 129 ARP. See Colombia, Administradoras de Family Health Plan, 134 Riesgos profesionales Fondo Nacional de Salud (FONASA), ARS. See Colombia, under National Social 28, 29, 67, 100, 101f, 106, 128 Health Insurance beneficiaries, 111f, 113f benefits and contributions, 112, 112t births, attended by skilled personnel, by coverage, 106t, 107f country, 21f equity, 111 Bismarckian coverage model, 39, 254 funding sources, 109 brain drain, 360 GES and supply deficit, 125 hospital admissions, 108 case studies. See country studies maximum contribution, 134n catastrophic payment headcount, 12b membership by income level, 110 CCSS. See Costa Rica, Social Security Fund regulation of, 102 CENDEISS, 207 service rationing, 114 Chile. See also country studies services provided and average basic benefits package, 104 spending per beneficiary, 108t Beneficiaries Bill of Rights, 125 type of care provided, 106 birth and mortality rates, 94 future reform priorities, 127­128 budget surpluses due to copper, 90 GDP growth of and per capita, 91f catastrophe insurance, 125 government subsidies, 111, 127, 133 Collective Capitalization Fund, 134n government system, 100 consumption of services, 113 health care coverage, 106, 106t contributions need to be adjusted, 128 of health problems, 101 copayments, 110f reform highlights, 125 country summary, 27­30 health care sector, 123, 124 Curative Medicine for Employee law, product variety, 117­118 121 health care spending, 28, 102t, 103 debt, 90, 92f, 93f health care system, 115­118 disease burden, 94­95, 96t asymmetrical regulation, 128 487 488 Good Practice in Health Financing decentralization of, 123 insurance reform evaluation, 121­125 efficiency, 113­114 life expectancy, 27, 96, 97f, 99f, 124t equity, 109­113 mandatory health insurance (MHI), factors limiting development of, 100, 101f, 124 123­124 split of financing and delivery, 102 lack of coordination and moral medicines registered, 116 hazard, 129 minimum worker contributions, 114 lessons for other countries, 129­134 moral hazard, 113 potential problems, 128­129 mortality rate, 94, 124t reform under military government, National Health Service (SNS), 27­28, 122 119­120 regulatory issues, 116­118 centralized planning and flaws, health indicators, 96t, 124t 121­122 under military regime, 122­123 out-of-pocket spending (OOPS), 102t, health insurance reforms, 118­120, 119f 103, 103f, 134n Health Law, 120, 122 physicians, 27 health outcomes, 27, 89 political background, 98­100, 99b health problems covered, 104, 105t population, 92, 95f health professionals, number of, 116 aging, 93, 94, 124 health services investment increase, 134n poverty comparison, 98f hospital admissions, 108t primary and emergency care, and hospital management improvement, 127 targeted diseases, 104 hospitals and beds, 27, 115­116, 115f private health spending, 103 infant mortality, 96, 97f, 99f programs, 116 and maternal mortality, 124t public health insurance providers, 106 Instituciones de Salud Provisional public health spending, 65­66, 110f (ISAPREs), 28, 29, 100, 101f, 106, as share of GDP, 103t 122, 134n share of total spending, 103 age discrimination among reforms, 27­30 beneficiaries, 112 regulation and ISAPREs, 63 and decentralization, 120 reproductive rate, 93 beneficiaries, 108, 111f, 113f risk selection, adverse selection, and benefits proportional to cream skinning, 114 contributions, 112­113 salary scheme for health professionals, coverage, 107f, 1069t 114 funding sources, 109 SEREMIS, 123 market concentration, 117 Single Health Care Service (SUS), 121 market transparency, behavior, and social context, 98 barriers to, 118 social security system coverage, 106, maximum contribution, 134n 106t, 107f, 108­109 membership by income level, 110 beneficiaries by income and insurance open and closed, 134n type, 111f regulation of, 102 summary of country study, 27­30 risk selection, 128­129 Superintendent of Health, 28 services provided, 106, 108t tax revenue, 92t variety of plans, 114 unemployment, 93t Index 489 unidades de fomento (UF), 134n health care facilities' services, 160, 160t wealth distribution, 98 health care facilities, number and type, welfare program spending, 104t 159 CHS. See Vietnam, commune health health care plans, 32­33, 178n stations health care provider market, 159­160 CNAM. See Tunisia, National Health health care reform motivation, Insurance Fund 162­163 CNRPS. See Tunisia, National Pension and health care reforms, 160, 162 Social Protection Fund health care spending, 137, 167 CNSS. See Tunisia, Caisse Nationale de distribution and trends, 168f Sécurité Sociale indicators, 153t Colombia. See also country studies health care system, 31, 67, 174 Administradoras de Riesgos health insurance coverage strategies, 163 Profesionales (ARP), 152 health outcomes, 30 background, 137 health services, reasons for Beneficiaries Identification System nonutilization of, 153, 154f (SISBEN), 32, 70, 144, 163 hospital beds, 30 budget constraints' effect on universal hospital services, 159 coverage, 169 hospitalization use, 157, 157f closing of rural and urban disparities in hospitalizations and consultations, insurance coverage, 165­166, 167f source of payments, 154f compulsory health plans, 32­33 households and catastrophic payments, conflict impacts on health, mortality, 155, 155t and disease, 142 informal sector enrollment, 175­176 copayments, 146, 151t information systems, 70, 175 country study summary, 30­34 institutional arrangements, 177 country summary, 30­34 Law 100, 32 debt, 138t Law 715, 171­172, 179n decentralization, 162, 172 legal framework, 177 demographic profile, 139t life expectancy, 30, 137, 138, 139t disease, 141t, 142 managed competition model, 137, 174, economic and political support, 178 175 economic indicators, 138t Mandatory Insurance for Motor Vehicle economy, 59, 138 Accidents (SOAT), 152, 153t family benefits funds, 179 maternal mortality rate, 139t fertility rate, 138, 139t, 140f medical care, types of, 149t financial protection indicators, 153 Ministry of Social Protection, 175 GDP per capita, 138t mortality, 138, 139t, 141t government and political environment, National Health Superintendency, 159 139, 142 National Social Health Insurance health care access improvement, 155, 156f (NSHI), 32, 33, 142 health care coverage, 32, 66, 163 and coverage expansions, 163­165 interventions for disease burden, 153t ARS, 157, 158, 173­174 supplementation, 152t benefit packages, 148, 151­152 health care delivery organization and comparison of CR, SR, and PS administration, 173­175 benefits, 150t 490 Good Practice in Health Financing compulsory health plan (POS), 148, payments to insurer, 146 153t, 179 payments to providers, 147 compulsory health plan (POS-S), 148, per capita payment unit (UPC) value, 151, 153t 146, 146t contributory regime (CR), 137, 143, physicians, number of, 30, 159 151t, 164f population, 138, 138t, 139t, 140f, 145t and universal coverage, 163 preventive services, use of, 155, 156f enrollment growth, 164 private health expenditures, 167 enrollment, contributors, private health plans, 158 beneficiaries, 165f private insurance, 152 funds, 166 provider standards, 160 market development, 173 reforms, 31­34, 157 reducing tax evasion, 176 enabling factors, 177­178 coverage, 151, 152, 152t sandwich population, 175­176 covered interventions, 148 Social Security and Guarantees Fund Empresas Solidarias de Salud (ESS), (FOSYGA), 143, 172t 158, 173 balance determinants, 171t evasion of contributions, 169 balance of compensation fund, 169f future challenges, 175­177 economic impact on revenues and Health Promoting Companies (EPS), expenditures, 168 143, 157­158 impact of contribution evasion, 169 institutional roles, 174­175 subfund, 166 legal framework, 142 social security expenditures, 167 monitoring and oversight, 159 solidarity and guarantee funds, 144f Partial Subsidies (PS), 147, 151, 153t subsidies and sin taxes, 147 replicability in other countries, 178 subsidy transition from supply to SISBEN, 145t, 163 demand, 170 subsidized regime (SR), 144, 179n tax evasion, 33 contributors, 165f top health conditions by disease or enrollment growth, 164 cause, 141t financing changes, 166­167, 176 unemployment, 138t, 170f funding, 145, 145f universal health insurance coverage goal, funding sources, 172t 176­177 resource growth, 170­171 universal health insurance coverage resources and expenditures, 171f, milestones, 161t 172f commercialization of the health sector, use of public providers, 158­159 54, 465 summary of schemes (SR, CR, PS, competition among insurers, 131 subsidies, and special), 148t conditional cash transfers (CCTs), 11 total enrollees by plan, 158, 158f contracting, 263­264 National Social Health Insurance System contributions, 36­37 (NSHIS) COOPESALUD R.L., 217b update of benefit package, 139 copayments, 67, 70. See also country other insurance schemes, 147 copper, 93t out-of-pocket spending (OOPS), 154f, cost barriers, 31 167 cost containment, 29, 34 Index 491 Costa Rica. See also country studies Sector Council, 212 accountability, 37 unions, 202 background, 183 health care spending, 188­189 Basic Teams for Integral Health Care and contributions, 199f (EBAIS), 204, 215, 224n and coverage, 36 budget allocation per capita, 194, 196 and equity and sustainability, 198, 200 CENDEISS, 207 distribution, 190f child and under-5 mortality rates, 194 households, 196 country summary, 34­38 indicators, 189t debt, 184t, 185 health care system, 37, 201 demographic profile, 185­186, 187f health indicators, 187t, 193, 193t dentists, 205t health insurance coverage by income, disease burden, 186, 187t 195t economic profile, 184­185, 184t health insurance financial balance, 201t environmental regulation, 206 health outcomes, 34, 183 fertility rates, 186 equity of, 193­194, 196 financial protection, 211 health services access of the poor, 224n fiscal problems, 184 health services use, 194 health care cooperatives as providers, 217b hospital beds, 34, 186, 205, 205t health care coverage, 34­35, 36, 37 hospital performance contracts, 192 reforms, 209 immunization, 186, 193t health care sector income and social spending distribution, constitution, 204 198f factors jeopardizing financial infant mortality, 186, 193t sustainability, 200 and child mortality, 196t financing model reform efforts, insurance regulation, 201 210­211 Law 7852 on Decentralization, 202, 214 impact of economic crisis, 210 Law for Worker Protection, 224n information systems, 221 life expectancy, 34, 186, 193t institutional reorganization, 212­214 maternal mortality, 186, 193t institutions involved in, 202­203 medical education, 206­207 primary care, 217b medical technology, 207­208 coverage, 210, 214­215, 215f medicine supply and distribution, 207 provider network management, 214 Ministry of Health, 202, 206, 213 provider network organization, mortality and morbidity profile, 186, 203­204 187t reform National Development Plan (NDP), 188 Health Boards, 214 nurses, 205, 205t issues remaining, 219 Office of the Controller General, 201 lessons for other countries, out-of-pocket spending (OOPS), 189 221­223 outpatient consultations by income, 195f policies and evaluation, 210 payroll fees, 191t political and financial support, physicians, 34, 205, 205t 216 political and government profile, 188 stakeholder participation, 218 population, 185­186, 186t, 200 targeting the poor, 214 poverty, 184 492 Good Practice in Health Financing private health care sector, 204 strengths and weaknesses of, 208, 208t private health care spending, 189 training, 221 pro-poor redistribution of resources, transfer of all public hospitals, 209 189 workforce, 205 provider network, 204 stakeholder participation, 37­38, 216 public health research, 207 taxation as source of financing, 191 public health spending, 66, 196­197, transparency, 37, 63 197t unemployment, 184t by income, 199f universal health coverage, 209 public resources consumed by income, University of Costa Rica, 203 198 costs, escalation of, 131 purchasing, 38 Country Policy Institutional Assessment reforms, 35­38 Index (CPIA), 62­63 lessons learned from, 37­38 country studies, 6­7, 57, 63 rural and poor priority, 218 GDP per capita, 60f Social Security Fund (CCSS), 34­38, 62, political freedom trends, 64f 183, 192, 199f political stability and commitment, benefits package and ability to modify, 61­62 190 qualifying as good performers, 17 conditions for reform success, 216, 218 reform characteristics shared, 58­59 contribution decrease, 211 selection of good practice performers, cost-sharing, 202 71­72 coverage, 194 summaries (See countries) decentralization, 202 CSMBS. See Thailand, Civil Servant efficiency of, 196­198, 200 Medical Benefit Scheme equity gaps, 194 CUP. See Thailand, primary care evasion and default, 200 contracting unit financing sources, 191, 197, 197t CUS. See Thailand, secondary care Finnish scandal, 219 contracting unit information systems, 208 CUT. See Thailand, tertiary care management agreement, 192 contracting unit payment mechanisms, 192 CWC. See Sri Lanka, Ceylon Workers pension regime, 203 Congress primary care focus, 218 pro-poor resource distribution, 189 data availability, 15, 72, 73 providers, no choice of, 202 decentralization, 68 purchasing from cooperatives, 215 decision making, evidence-based, 69­70 purchasing unit creation, 221 demand-side incentives, 11 regulatory framework, 206 democracy, 45, 56n. See also under Sri reorganization of, 218 Lanka resource supply, 205 dependency index, 134n separation of financing and DHC. See Vietnam, District Health Centers purchasing, 213 diagnosis-related groups (DRGs), 50 services included, 192 disability-adjusted life year (DALY), 134n solidarity and equitable access, 198 discrimination, avoiding, 131 spending by source and use, 196 disease burden, 446. See also country Index 493 economic growth, 33, 59, 61 health care spending, 233, 259 effectiveness, proven, 4 by source, 233t efficiency gains, 70 municipalities' responsibilities, 235 efficiency types, 329­330 health care system, 39, 263 EHIF. See Estonia Health Insurance Fund financial sustainability, 259 EPS. See Colombia, Health Promoting revenue collection and labor market, Companies 261­262 ESS. See Colombia, Empresas Solidarias de solidarity, 235 Salud state contributions for Estonia. See also country studies noncontributing groups, 234 allocation of funds and purchasing health financing system overview, 237f services, 236­240 Health Insurance Act, 257, 258­259 ambulance service overuse, 260 health insurance eligibility criteria, 256 background, 227 Health Insurance Fund Act, 244 coinsurance, 255 health insurance system, 233, 254, 256 contracting, 264 mandatory, 227 contribution rates, 235, 254 revenues from social tax, 234 copayments, 255, 257, 262 health outcomes, 39 country summary, 38­41 Health Service Organization Act, 249, 256 debt, 228 HIV, 265n demographic and epidemiological hospital access, 248­249 situation, 229­231 hospital beds, 29, 231 disease burden, 230, 231 Hospital Master Plan 2015, 248, 258, 266n economy, 59, 228 hospital sector restructuring, 247­248, emergency care, 265n 249 family medicine as a medical specialty, hospitals and acute care admissions, 248f 256 human resource planning, 250 family medicine call center, 246 hypertension treatment and ACE family planning, 229­230 inhibitors, 252 family practice requirements and income inequalities, 228, 244 reimbursement, 246 infant and maternal mortality, 231 financial protection, 243 information technology, 69, 264 fiscal policy and taxation system, 229 life expectancy, 39, 231, 232f health care access barriers, 265n mandatory health insurance (MHI), health care coverage reform, 253 67­68 lessons for other countries, 260­264 Ministry of Social Affairs, 258 regulatory framework, 259 mortality and causes of, 230 unresolved issues, 261 municipalities' responsibilities, 242, 259 health care delivery system, 245­253, 260 national health service (NHS) to MHI health care sector transition, 40 informal payments, 262 nurse shortage, 246 reform, 253f, 257­259 out-of-pocket spending (OOPS), 235, provider network decentralization, 236f, 243, 255, 259 255­257 and MHI, 40, 41 Health Care Services Organization Act barrier to access, 262 (2001), 265n outpatient and specialist care, 247­251 494 Good Practice in Health Financing pharmaceutical sector, 253, 257, 265n hospital contracts, 239 reform, 251­253 increase in drug and services pharmaceutical use, 252, 252f expenditures, 251f physicians, 39, 245f payment methods, 239, 240, 240f, 241f physicians and nurses per inhabitants, programs, medicines, dental benefits, 250f 242­243 political environment and government provider selection, 238, 265n structure, 231­233 purchasing decisions, 238 polyclinics, 245 revenue and expenditure balance population, 229, 230f requirement, 244 poverty, 228 service prices, 240 primary care, 247, 257 specialist care, 243 and family medicine, 245­246 Estonian Medical Association, 254 reform, 256, 258 evaluation of reforms, 4­5, 74 private health financing expansion, 259 evidence base, enhancing, 71, 73 private health insurance market, 236 evidence dissemination, 74­75 private health spending, 235 evidence, lack of, 3, 4 provider network, 264 public health spending, 66 failures, evaluation of, 74 purchaser-provider split, 255 family medicine as a medical specialty, reforms, lessons learned from, 40­41 245, 256 revenue collection, 40, 41 FAPs. See Kyrgyz Republic, Feldsher- service providers' legal status, 258 Obstetrical Ambulatory Points sickness funds, 254, 255, 257 FGPs. See Kyrgyz Republic, family group Central Sickness Fund, 256 practice social health insurance model, financial protection, 17, 22, 72 decentralized, 254 indicators, 12­14 State Agency of Medicines, 253 measuring, 10 State Health Insurance Council (SHIC), financing mandates, 67­68 256 FMCs. See Kyrgyz Republic, family taxation, 39, 229, 258, 264n medicine centers transition to market economy, 227 FONASA. See Chile, Fondo Nacional de unemployment, 244 Salud uninsured population, 244 Former Soviet Union, 42 workforce participation incentive, 260 FOSYGA. See Colombia, Social Security Estonia Health Insurance Fund (EHIF), 40, Guarantees Fund 41, 235 FP. See Estonia, family practice benefits and copayments, 241­243 funding, use of various sources, 132 coinsurance, 243 contracting process, 238­239, 238f G-8, 75n core purchaser of health care services, 236 GES. See Chile, Explicit Health Guarantees coverage categories, 241­242 governance, 47, 50, 133 efficiency and equity, 243­244 government fiscal commitment to coverage family practices (FPs), 240, 241f expansion, 67 funds collected centrally and allocated gross domestic product (GDP) per capita, regionally, 237 18t, 22f, 60f Index 495 HCFP. See Vietnam, Health Care Funds for health insurance schemes, categorization the Poor of, 10 health care access, 31, 36­37. See also health insurance system, Bismarckian, 39, country 254 health care coverage, 9, 16. See also country; Health Metrics Network (HMN), 4 universal health care coverage health outcomes, 4, 15, 16, 17 and financing mandates, 67­68 correlations between income, spending, and spending increases, 66 and literacy, 26t health care coverage reform, factors for indicators, 19t success, 220 Health, Nutrition, and Population outcome health care delivery indicators, 19t determinants, 5f health care financing, 33­34 hospital bed supply, 24, 25f. See also health care scheme, reduced fees or country charges, 395­396 hospital production unit (UPH), 192 health care sector HSRI. See Thailand, Health Systems quality assurance and control, 415 Research Institute reform, 221­223 hypertension treatment, 252 workforce by occupation, 205t health care service delivery reform, 69 IHPP. See Thailand, International Health health care service indicators, 21f, 24 Policy Program health care services, effect of supply on immunization, by country, 19t social security system, 131 implementation factors, 68­75 health care spending, 12­14, 20, 26t infant mortality rate (IMR), 19, 19t, 20f. and health indicators, 20f See also country and income, by country, 18t, 21f informal sector, 33, 34, 58 as share of GDP per capita, 22f information dissemination, 74­75 government share vs. income, 23f information systems and evidence-based insured consume more, 65 decision making, 69­70 public expenditures, 65­66 information technology, 264 relating to outcomes, 16 institutional requirements, 130­131, 221 sustaining, 58 institutions, factor in health financing health care system, 19, 57, 263 reforms, 59, 61­63, 65 health care, percentage of people not International Monetary Fund (IMF) receiving, 19t Structural Adjustment Program, health financing, 10. See also private 400 financing demand-side incentives, 11 Kyrgyz Republic. See also country studies efficiency gains and copayments, 70 Additional Drug Package (ADP), 283 enabling environments for reform, 57 background, 269­270 for expansion, 33­34 benefits transition out of Soviet health sources, 394 system, 278­279 stakeholder participation in design, case-based payment system, 285 262­263 child mortality, 274, 286 health indicators, 15­16, 17, 20f copayments, 281­283, 282t health insurance design and elimination of, 306 implementation, 473 country summary, 41­44 496 Good Practice in Health Financing demographic, epidemiological, and hospital beds, 41, 274, 274f, 275t, 286, social environment, 272­274 287, 288f disease burden, 272, 274 hospital downsizing, 292, 297 donor collaboration, 304­305 hospital efficiency, 288f, 289, 289f economic environment, 270­272 hospital payment by patient, 302f family group practices (FGPs), 285, 290 hospital reserve funds, 282­283 family medicine, 291, 294 hospitalization, 288, 300­301, 302 family medicine centers (FMCs), 290 human resources training, 291 Feldsher-Obstetrical Ambulatory Points infant mortality, 306n (FAPs), 290 and child mortality, 286 fertility rate and child dependency, 272 causes, 274, 274f, 275t financial management, 44 informal charging practices, 278­279 financial protection, 299 informal payments, 302 fiscal control, 272 Issyk-kul Intensive Demonstration Site, GDP per capita and growth, 271f 304 health care coverage, 43, 270 lack of public awareness of health care benefits packages, 276, 278­279, entitlements, 303 281­283 Law on Health Care Organizations, 292, depth issues during transition years, 293 296 Law on Health Protection, 293 reform life expectancy, 41, 272 chronology, 298 literacy rate, 272 evaluation of, 299­303 Manas National Health Care Reform main pillars, 296­297 Program, 298, 298b pooling and case-based payments, Manas Taalimi National Health Care 297 Reform Program, 298 health care delivery system overview, 290 Mandatory Health Insurance Fund health care facilities, organizational (MHIF), 43, 284, 298b charters and accreditation, 293 and single-payer reforms, 285 health care financing indicators, 277t benefits of insured vs. uninsured, health care financing reform, 279, 280t, 279­280 281, 307 Medical Accreditation Commission health care financing trends, 276 (MAC), 293 health care organization ownership medical education, 294 types, 292 medical technology, 295 health care sector, 293 mortality and morbidity and DALYs, allocative efficiency, 287­288 273­274, 273t reforms, 303­305 National Health Reform Strategy, 294 health care system, 42, 67, 301 out-of-pocket spending (OOPs), 42, 43, efficiency, 287­289 276 health information system, 295 trends, 299f, 300 health outcomes, 41, 286 payroll tax, 43 health service delivery, restructuring, 297 pharmaceutical sector, 295 health services access, 286, 287f, 300­301 physician, 41 High Technology Fund, 295 political context, 272, 306 hospital and outpatient care access, 301f "island of democracy," 275 Index 497 population, 272, 273f maternal mortality rate, 20f. See also poverty rate, 270, 306n country poverty reduction, 42 mean positive gap, 12b President Akaev, 275 measles immunization, 19t, 21f primary care, 288, 290, 291 measuring outcomes, 4 primary care provider types, 290 MHIF. See Kyrgyz Republic, Mandatory primary facility capitation rates, 285 Health Insurance Fund provider payment methods, 284, 291 Millennium Development Goals public finance reform effects, slow pace (MDGs), 3 of, 305­306 MOOHs. See Sri Lanka, Medical Officers of public health spending, 277t, 278f Health units decline, 276, 305 MSP. See Tunisia, Ministère de la Santé low per capita, 66 Publique reallocation in single-payer system, MWS. See Thailand, Medical Welfare 293f Scheme purchasing, 285, 286, 304 reform replicability, 306 National Health Accounts (NHAs), 5 reforms, 42­44 national health services (NHS), 10, 58 scaling-up, 303 NDP. See Costa Rica, National resource allocation and duplication of Development Plan services, 284 NHSB. See Thailand, National Health revenue collection, pooling, and Security Board purchasing, 283­285 NHSRC. See Thailand, National Health secondary and tertiary care, 292 System Reform Committee single-payer system, 285, 298b, 301 NSHI. See Colombia, national social health State Guaranteed Benefits Package insurance (SGBP), 281, 283, 291, 306 clarification of entitlements, 296­297 out-of-pocket spending (OOPS), 12b, 18t, taxation, 43 22, 24f. See also country transition from Soviet Union, 270 overshoot, 12b, 14t uninsured, 279 Pavas Clinic, 217b labor market formality, 131 payment reforms, 69 labor market formalization, 262 payment schemes, 49­50 life expectancy, 19, 19t. See also country payments, 13f. See also country; out-of- literacy, 24, 25f, 26, 26t, 65. See also country pocket payments Lorenz Curve, 198, 224n PCT. See Tunisian Medical Store performance, 11, 15­16, 26n MAC. See Kyrgyz Republic, Medical defining "good," 58, 71­72 Accreditation Commission global, 17, 19­20, 22, 24, 26 managed competition model, 137, 174­175 pharmaceutical supply and regulation, mandatory health insurance (MHI), 10, 58. 336­337 See also country pharmaceuticals supply and regulation. MASSTE. See Tunisia, Ministry of Social See also country Affairs, Solidarity, and Tunisians physician supply, 24, 25f, 47. See also Abroad country 498 Good Practice in Health Financing policy, 62­63 SISBEN. See Colombia, Beneficiaries commitment and continuity, 65­68 Identification System political context, 30 SNS. See Chile, National Health Service political freedom trends, by country, 64f SOAT. See Colombia, Mandatory Insurance political stability and commitment, 61­62 for Motor Vehicle Accidents poor, targeting of pro-poor policies, 67 social insurance, 36, 55, 254 poverty comparison of Latin American social security systems, 130, 131­132 countries, 98f solidarity, 36, 66­67, 132­133 primary care expansion, 38 Sri Lanka, 336. See also country studies primary care focus, 68 background, 311­312 primary care strategy success factors, 222 Buddhism, 339 primary care, universal coverage of, 222 catastrophic protection, 321, 345, 349 private health financing, 29, 66. See also Ceylon Malaria Epidemic, 341 contributions; copayments Ceylon Workers Congress (CWC), 329 private health sector funding, and child birth attendance, 327f sustainability, 132 consumer quality differentials to "Proven Success for Global Health," 4 discourage the rich, 320 provider network, 264 contraception use by quintile, 328f public-private health financing, 22 country summary, 44­47 purchaser-provider split, 255 democracy, 311, 317­318, 325, 329 purchasing, 11, 38, 263­264, 304 and accountability, 347, 348 purchasing unit, creation of, 221 transition to, 339­340 demographic indicators, 313t reform. See also country disease burden, 315 commonalities across countries, 58­59 diversified society, 314 enabling environments for, 57 economic growth, 61 evaluation of, 74 economic history, 312­313 flexibility and mid-course corrections, economic indicators, 313t 70­71 family planning care, 322 length of time and sustainability, 38 fertility rate, 315, 315t motivations for, 58 fiscal deficit, 314 political and economic factors for government organization, 316 success, 132­133 health care access, 46, 35, 321­322 sequencing of, 69 equity of, 348 regulation, 29­30, 130­131 key conditions for expanding, 347 regulatory capacity, 53, 63 coverage expansion, 345t results, 3, 4, 5. See also evidence growth of, 335 revenue collection, 9, 22, 23f, 261 use and spending, 331t risk pooling, 9, 68 health care benefits package, 321 health care coverage expansion, 311 sandwich population, 33 health care facilities used, public vs. SGBP. See Kyrgyz Republic, State private, 334t Guaranteed Benefits Package health care financing, 320, 321t, 324 SHI for low-income countries, 58 health care sector, 341 SHIC. See Estonia, State Health Insurance efficiency, 347, 349 Council reform Index 499 evaluation of, 344­347 medical technology, 337 historical background, 338­339 Ministry of Health, 343 lessons for other countries, Ministry of Health facilities, 334 348­350 morbidity and mortality, 315, 317t politics and sustainability, 347 official development assistance (ODA), preconditions, 340 313, 313t scaling up and targeting, 338, 338t, out-of-pocket spending (OOPS), 323 346, 347 physicians, 44 resource allocation and targeted and nurses, beds, admissions, public subsidies, 350 vs. private, 334t technical efficiency gains, 346 compulsory posting and dual health care spending, 46 practice, 344 benefit-incidence of, 319­320 plantation workers, 328­329, 339 by provider, 352t population, 313t, 315, 315t, 316f compared to other countries, 331t poverty, 313t, 322t hospital coverage, 318­319 private health care sector, 334, 346 trends, 319t private health services, financed through health care system OOPS and employer, 323 efficiency, 330­332, 337 productivity improvements, 331­332, allocative, 332­333 343­344 quality vs. access, 343 public health care provision, 335f reform, 341­344 public health spending, 320, 324f health outcomes, 44, 314, 326, 345­346 by source, 351t inequality, 324­326 financed through general tax revenue, health services network, 333­334, 342 323 health services, rationing of, 322­323 low per capita, 66 hospitals, 319t, 331 poor vs. rich, 320t beds, 44, 313t, 314, 335 reforms, 45­47, 55 emphasis on, 333, 342­343 revenue collapse due to liberalization, Ministry of Health expenditures, 333t 314 technical efficiency comparison, 332t Sri Jayewardenapura General Hospital, household health spending and poverty, 322 322t targeting and equity impacts, 326, 328 infant mortality, 313t, 314, 315t transition from British colony, 343 by asset quintile, 327f unemployment, 313 by district, 326t user fee elimination, 342 by social group, 325t western medicine exposure, 339 life expectancy, 313t, 314, 315t SSS. See Thailand, Social Security Scheme literacy, 313t stakeholder support, 70 LTTE terrorist group, 328 success, defining, 71­72 maternal mortality, 314, 315t supervision, lack of, 29­30 Medical Council, 336 SUS. See Chile, Single Health Care Service medical education and regulation, 335­336 targeting the poor, 474 Medical Officers of Health (MOOHs) TDRI. See Thailand Development Research units, 333­334 Institute 500 Good Practice in Health Financing Thailand. See also country studies National Health System Reform "30 baht treats all diseases," 379 Committee (NHSRC), 378 background, 355 out-of-pocket spending (OOPS), 49 brain drain, 359 physicians, number that resigned, 372f catastrophic illnesses and expenses political system, 356 decrease, 374­375, 375t population, 357f Civil Servant Medical Benefit Scheme poverty reduction, 374­375 (CSMBS), 48, 363, 364t primary care, contracting unit (CUP), country summary, 47­50 369, 370b demographic transition, 356 private health insurance, 364t, 365 disease burden, 356­357, 358t private health sector facilities, 359 economic context, 355­356 provider choice, 368 economic crisis, 356 provider satisfaction, 370, 371f economic growth, 61 public health spending, low per capita, Health Card Scheme, 361, 363 66 health care financing reform, 362 public support, 49, 377, 379 health care financing sources, 362, 363t reforms, 47­50 health care system, 359­362 reimbursement for providers, 49­50 budget allocation, 360, 360f, 368, 369t revenue collection, 50 hospitals, 358 Rural Doctor Society, 360, 365­366 health care worker distribution, 371, 372 secondary care, contracting unit (CUS), Health for All, 360 370 health insurance coverage, 362t, 375t services covered, 355 health insurance model, 365f Social Security Scheme (SSS), 48, 361, health outcomes, 47, 357 363, 364t Health Systems Research Institute stakeholder participation, 50 (HSRI), 376 technical capacity, 50 health workers workload and tertiary care, contracting unit (CUT), dissatisfaction, 370 370 health workforce and training, 358 Thai-Rak-Thai party, 362, 366 hospital beds, 47 Triangle That Moves the Mountain infant mortality rate, 47, 357 strategy, 377­378, 378f institutional support for UCS, 63, 65 uninsured population decrease, 374 International Health Policy Program Universal Coverage Scheme (UCS or the (IHPP), 376 30 baht scheme), 48­49, 63, 65, life expectancy, 47, 357 363, 364t, 365 literacy, 47 adverse effects of allocation reform, maternal mortality, 357 371­372 medical tourism, 358 agenda, 365­366 Medical Welfare Scheme (MWS), 47­48, and reform component, 367­370 50, 361, 363 benefit package classifications, 367 National Health Act, 378 capitation payments, 368, 371, 373, National Health Security Act, 362 373f National Health Security Bill, 356, 366 consumer satisfaction and service National Health Security Board (NHSB), quality, 375­376, 376t 372, 377 copayments, 380 Index 501 evidence of success, 374­377 health care coverage, 51­52, 402t, 417 formulation process, 365 by scheme, 395f inclusion and exclusion list, 367t expansion highlights, 432­433, 434 lessons learned, 377, 379­380 primary care, 418t primary care system, 369­370 rates vary, 397 public hospitals in financial trouble, health care delivery, 409­413, 425 373, 373f health care financing, 401, 402t, 403, resulting difficulties, 370 403f, 425 service coverage expansion, 376 health care network formation, 432 technical support, 376 health care provision, 414­415, 433 use of by in- and outpatients, 371t by facility type, 409t Voluntary Health Card Scheme (VHCS), health care scheme, free, 395 48 health care sector Thailand Development Research Institute demand regulation, 415 (TDRI), 376 development, 433­434 Tunisia. See also country studies employment structure, 410­411 administrative capacity, 53 human resources development, background, 385 412­413 billing, 405 infrastructure, 414­415 catastrophic impact of OOPs, 13­14, motivations for reform, 424 13f, 14t need for strategic planning, 426 copayments, 400 organization, 425 country summary, 50­53 reform evaluations, 425­426 coverage by MSP facilities, 407t reform timeline, 424­425 CSS insurance coverage rates, 396t regulation trends, 416 DALYs, 392t, 393, 393t workforce training and debt, 386t unemployment, 423 demographic context, 388­390, 391t, health care spending, 398t 392 and services and outcomes, 417 disease burden, 392t, 393, 393t indicators, 398, 399t economic and social approach, 432 public and private trends, 398 economy, 61, 385­386, 386t, 393 rising for households as share of total epidemiologic transition, 392­393 spending, 405­406 fertility, 391t, 392 sources, 398­401 financial contribution inequity trends by source, 400f comparison, 420t health care system, 51, 420, 422­424 General Union of Tunisian Laborers health inequalities comparison, 421t (UGTT), 423 health information system, 416 governance, 394 health insurance, 394­397 government-sponsored benefits, contribution rates, 431t 386­387, 387t hospital care fee, 431­432 health care access, 417­418, 420, 422 health insurance coverage linked to coverage type, 403 fragmentation, 403 health care benefits packages, 394, 397 health insurance reform, 423, 427 health care card, 406t, 435n implementation challenges, 431­432 health care costs rising, 400­401 lessons learned, 430­431 502 Good Practice in Health Financing remaining issues, 426­427, 430 nursing schools, admission to, 413 stakeholder analysis, 428t out-of-pocket spending (OOPS) subsystem priorities, 434­435 catastrophic impact in, 13­14, 13f, 14t health insurance system, 394, 432 parastatal health sector, 409 health outcomes, 50­51, 419t, 420 payment methods, 403, 404t, 405 health status, 385 pharmaceutical sector, 413, 415 hospital beds, 50, 410, 410t, 411t physicians, 51 hospitals, 409 population, 389, 389f, 391t household consumption by category, poverty database, 396 407t poverty gap, 386t household health spending, 408, 408f, poverty reduction, 387­388, 388t 408t, 418 primary health care coverage, 418t immunization, 417 private health care sector, 53, 410, 434 infant mortality, 41t7, 390, 391t and household spending, 408 life expectancy, 50, 390, 390f, 417t facilities, 411t managed care opposition, 424 private health spending, 399­400 mandatory health insurance (MHI), 52 public health spending, 399 medical assistance programs (AMG), 52, reforms, 51­53 396 regulatory and administrative capacity, medical education system, 413, 415 53 medical equipment, 414t, 415 reimbursement, 405, 406t medical technology, 413, 415, 423 RNS, 431t medical workforce, 410­411, 411t, 412t service fees, 405 military hospitals, 404t social development, 394 Ministère de la Santé Publique (MSP) social policy, 387 care levels, 409 Social Security funds, 397, 435n financing, 403f social security schemes, 396­397 Ministry of Social Affairs, Solidarity, and social security systems (CNSS and Tunisians Abroad (MASSTE), 396 CNRPS), 52 mortality, 391t, 392 socioeconomic transition, 388 mutual insurance companies and tax revenue, 386t, 387 employee associations, 397 trade agreements, 386, 393 National Health Insurance Fund Tunisia, National Pension and Social (CNAM), 422, 427, 432, 435n Protection Fund (CNRPS), 52, 405, National Program of Aid to Poor 418, 435n Families (PNAFN), 396 Tunisian Medical Store (PCT), 413 NFPSP, 402t contribution rates, 431t UCS. See Thailand, Universal Coverage mandatory and facultative schemes, Scheme 397 UF. See Chile, unidades de fomento NFSPS affiliates, 406t, 407t UGTT. See Tunisia, General Union of NFSS, 402t, 403, 404t Tunisian Laborers contribution rates, 431t universal health care coverage, 3, 222. See grants, 405 also country illness coverage, 397 UPH. See hospital production unit Index 503 VHCS. See Thailand, Voluntary Health monitoring and evaluation of policy, Card Scheme 469 VHW. See Vietnam, village health workers reform motivations, 439 Vietnam. See also country studies society resource mobilization, 449, 465 and health outcomes, 53 sources by type of service, 449t background, 439­440 trends, 447­449 benefits package uniformity, 473 Health Care Fund for the Poor (HCFP), budget allocation not aligned with local 448, 451, 468 needs, 447 health care quality, 439, 459 children, benefits for, 452, 453, 466, health care sector regulation, 440, 461 467­468 health care system, 457­459, 459­461 commune health stations (CHSs), 458, factors for success, 472 464 strengths and weaknesses, 469­470 contribution rates, 55, 452­453 health care, free for children under 72 country summary, 53­56 months, 452 Decision 139, 456­457 health exports, 462 demographics, 442 health information system, 471 disease burden, 445t health insurance, 454, 455, 465, 466 disease targeting, 444, 472 increasing use of health services, 466, District Health Centers (DHCs), 457 467 drug reimbursement, 454 voluntary scheme not up to economic growth, 61, 441, 470­471 expectations, 465 economic history and Doi Moi, 440 health insurance coverage, 456 economic indicators, 441t by type, 450t economic transition, 463 health insurance fund, 453, 454, 455 family planning, 443 health insurance programs, government government decentralization, 446­447 sponsored, 451­452 government revenue, 442 health insurance, compulsory and government subsidies, 454 voluntary, 450­451 group health insurance, 451 Health Sector Master Plan to the Year health as priority, 471, 473 2010 and Vision to Year 2020, 461 health card initiatives' impact, 467 health services use, 468 health care coverage, 54, 463 health services, inequitable use of, 456, and outcome indicators, 460t 469­470 reform health spending, 447­448 chronology, 463­466, 464t by type of service, 448, 449t enabling factors, 470­472 total per capita, 448t rationale for, 462­463 health workforce density, 462 replicability of, 472­474 hospital admissions and deaths, 446 health care delivery system, 461­462 hospital beds, 53, 458, 469 health care equity indicators for hospital care, reliance on, 469 outcomes and distributions, hospital service commercialization and 456­457 perverse incentives, 465 health care financing, 449­452, 468­469 hospitals, number of, 458 and payment, 452­455 household health care costs increase, 466 504 Good Practice in Health Financing human resources for health indicators, physicians, 53 459t political accountability, 440 immunization coverage, 460­461 political and governance environment, income poverty rates, 441 446­447 infant mortality, 443t political stability, 63 insurance coverage rates, 54­55 population, 442, 443t, 444, 444f irrational drug use and antibiotic preventive health programs, 449 resistance, 470 primary health care history, 458 Law on Protection and Care of Children, primary health care network, inherited, 452, 466 472 life expectancy, 53, 442 private health care sector, 54, 458­459, literacy, 53, 56, 442, 443t 463­464 management capacities, 472 regulation, 461 medical education, 461 provider reimbursement procedures, medical equipment upgrade, 465 453­454 Ministry of Health, 457, 461 provincial health bureaus, 457 morbidity patterns, 446 public debt, 442 mortality rates, 443t reforms, 54­56, 471 national health programs, 458 self-treatment, less reliance on, 467 official development assistance (ODA), service fees, 464 limited use of, 442 structure of society contributes to Ordinance on Private Practice of reform implementation, 471 Medicine and Pharmacy, 461 targeting the poor, 456, 474 out-of-pocket spending (OOPS), village health workers (VHWs), 458 reduction in, 467 workforce indicators, 459t payment mechanisms, 454 pharmaceutical sector, 462, 463­464 World Health Survey (2003), 12 F or humanitarian reasons and the concern for households' economic and health security, the health sector is at the center of global development policy. Developing countries and the international communityarescalinguphealthsystemstomeettheMillenniumDevelopment Goals (MDGs) and are improving nancial protection by securing long-term support for these gains. Yet money alone cannot buy health gains or prevent impoverishment due to catastrophic medical bills; well structured, results- based financing reforms are needed. Unfortunately, global evidence of "successful" health nancing policies that can guide the reform e ort is very limited and therefore the policy debate is o en driven by ideological, one- size- ts-all solutions. Good Practices in Health Financing: Lessons from Reforms in Low- and Middle- Income Countries attempts to begin to ll the void by systematically assessing health nancing reforms in nine low- and middle-income countries that have managed to expand their health nancing systems to both improve health status and protect against catastrophic medical expenses. e participating countries are: Chile, Colombia, Costa Rica, Estonia, the Kyrgyz Republic, Sri Lanka, ailand, Tunisia, and Vietnam. e study seeks to identify common enabling factors of their good performance. While the findings for each country are important, collectively they send a clear message to the global community that more attention is needed to de ne "good practice" and then to evaluate and disseminate the global evidence base. 978-0-8213-7511-2 SKU 17511