Latin America & Caribbean Region 76894 Environment & Water Resources Occasional Paper Series Environmental Health in Nicaragua Addressing Key Environmental Challenges © 2013 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. The Environment and Water Resources Occasional Paper Series was developed under the direction of Karin Kemper, Sector Manager for Environment and Water Resources in the Latin America and Caribbean Region (LCSEN) of the World Bank. The publications in this Series were designed and produced by GRC Direct under the supervision of Emilia Battaglini and Rachel Pasternack (LCSEN). A list of the most recent papers is on the back cover of this publication. For electronic copies of all our LAC Environment & Water Resources Occasional Papers please visit our website: www.worldbank.org/lac Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any 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. All images courtesy of Thinkstock/Getty Images and The World Bank Environment and Water Resources LCSEN Occasional Paper Series Foreword The Latin America and Caribbean (LAC) region has water resources management and the pursuit of a unique mix of qualities and challenges when greener and more inclusive growth. The papers it comes to the environment. It is exceptionally seek to bring to a broader public – decision makers, endowed with natural assets, with globally development practitioners, academics and other significant biodiversity and valuable crops, and partners - lessons learned from World Bank- also harbors the world’s greatest carbon sink in financed projects, technical assistance and other the Amazon. At the same time, however, the region knowledge activities jointly undertaken with our registers the highest rates of urbanization in the partners. The series addresses issues relevant to the developing world with pollution, overuse of its water region’s environmental sustainability agenda from and natural resources and detrimental impacts on water resources management to environmental the health of people, especially the poor, and the health, natural resource management, biodiversity environment. conservation, environmental policy, pollution management, environmental institutions and Over the past twenty years, the LAC region has governance, ecosystem services, environmental made impressive gains in tackling these issues. financing, irrigation and climate change and their It leads the developing world in biodiversity linkages to development and growth. conservation and natural resource management and is at the forefront in reducing urban pollution. In this particular paper, we present you the findings of The World Bank has often been the partner of a study looking at three fundamental environmental choice for those countries in the region that have health risks in Nicaragua, notably inadequate water had the initiative to pioneer innovative policies for and sanitation, indoor and outdoor air pollution. environmental protection and natural resource The results are striking in that these three risks management, strengthen institutions responsible alone amount to an estimated 2.4 percent of for environmental management, enhance the country’s GDP, affecting primarily the poorer environmental sustainability, and introduce new segments of the population. The study proceeds to approaches to water resources management. Such look at priority investments and solutions, including initiatives include fuel and air quality standards by ranking potential interventions in terms of their in Peru, carbon emission reduction in Mexico, costs and the expected benefits they will generate. payment for ecosystem services in Costa Rica, It is clear that solutions exist and that the returns participatory and integrated water resources to environmental improvements can be very high in management in Brazil, and new approaches to terms of health improvements and better quality of irrigation management in Mexico. life, particularly for children. In this context, it is our pleasure to introduce the We hope that this paper, just as the entire series, Environment & Water Resources Occasional Paper will make a contribution to knowledge sharing Series, a publication of the Environment and within the LAC Region and globally. Water Resources Unit (LCSEN) of the Sustainable Development Department in the World Bank’s Latin Karin Kemper America and the Caribbean Region. The purpose of Sector Manager, Environment & Water the series is to contribute to the global knowledge Resources exchange on innovation in environmental and Sustainable Development Department Latin America and the Caribbean Region Table of Contents Acronyms and Abbreviations..................................................................................................................... v Acknowledgements...................................................................................................................................vii Abstract........................................................................................................................................................ix Executive Summary....................................................................................................................................1 I. Overview..................................................................................................................................................9 II. Water, Sanitation, and Hygiene..........................................................................................................13 Water, Sanitation and Hygiene Interventions..................................................................................17 Summary Assessment of Interventions Relating to Water, Sanitation and Hygiene.....................20 Recommendations for Actions on Water, Sanitation, and Hygiene................................................21 III. Indoor Air Pollution.............................................................................................................................23 Indoor Air Pollution Interventions.....................................................................................................24 Summary Assessment of Interventions to Address Indoor Air Pollution .......................................25 Recommendations for Actions on Indoor Air Pollution ...................................................................26 IV. Urban Air Pollution..............................................................................................................................29 Summary Assessment of Actions on Urban Air Pollution................................................................31 Recommendations for Actions on Urban Air Pollution....................................................................32 V. Conclusions..........................................................................................................................................33 Annex A. Methodology for Health Cost Estimation...........................................................................35 Annex B. Methodology for Cost-Benefit Analysis...............................................................................51 References.................................................................................................................................................85 Tables Table ES 1: Aggregate Environmental Health Costs for Nicaragua........................................................3 Table I.1: Aggregate Environmental Health Costs for Nicaragua......................................................10 Table I.2: Annual Cost of Environmental Damage – Low and High Estimates................................10 Table II.1: Estimated Annual Health Effects from Water, Sanitation, Hygiene..................................14 Table II.2: Estimated Annual Cost of Diarrheal Illness.......................................................................15 Table II.3: Baseline Data for Estimation of Costs Due to Water Boiling............................................15 Table II.4: Estimated Annual Household Avertive Expenditures........................................................16 Table II.5: Costs and Benefits of Reductions in Diarrheal Morbidity and Mortality..........................18 Table II.6: Benefits and Costs of a Hand-Washing Program (Rural and Urban)................................19 Table II.7: Benefits and Costs of a Rural Drinking Water Boiling Program........................................20 Table III.1: Estimated Annual Health Effects of Indoor Air Pollution..................................................23 Table III.2: Estimated Annual Costs (million NIO) of Indoor Air Pollution in Nicaragua.....................24 Table III.3: Interventions........................................................................................................................25 Table III.4: Benefits and Costs of Indoor Air Pollution Control in Nicaragua......................................26 i Table IV.1: Measured Average Annual Concentration of Pollutants in Managua .............................29 Table IV.2: Estimated Health Impact of Urban Air Pollution................................................................30 Table IV.3: Estimated Annual Cost of Health Impacts (Million NIO)...................................................31 Table A.1: Urban Air Pollution Dose-Repsonse Coefficients...............................................................35 Table A.2: Mortality Risk Associated with a 10 μg/m3 Change in PM2.5 ..........................................36 Table A.3: DALYs for Health Effects......................................................................................................37 Table A.4: Calculation of DALYs Per Case of Health Effects...............................................................37 Table A.5: Estimated Unit Cost by Health End-Point...........................................................................38 Table A.6: Baseline Data for Estimation of Cost of Health Impact Due to Air Polution....................39 Table A.7: Baseline Data on Health.....................................................................................................41 Table A.8: Calculation of DALYs Per Case of Health Effects...............................................................42 Table A.9: Baseline data for estimating health Effects Due to Lack of Improved Water Sanitation and Hygiene............................................................................................42 Table A.11: Health Risks of Indoor Air Pollution....................................................................................44 Table A.12: Baseline Data for Estimating Health Impacts of Indoor Air Pollution..............................44 Table A.13: Calculation of DALYs per Case of Health Effects From Indoor Air Pollution....................45 Table A.14: Baseline Data for Estimation of Costs of Health Impacts from Indoor Air Pollution.......46 Table A.15: Cost of Mortality (Per Death) Using HCA............................................................................47 Table A.16: Estimated Value of Statistical Life in Nicaragua...............................................................49 Table A.17: A Comparison of HCA and VSL Estimates Applied to Nicaragua......................................49 Table B.1: Summary of Meta-Analysis by Fewtrell and Colford (2004).............................................52 Table B.2: Selected Exposure Scenarios.............................................................................................53 Table B.3: Reductions in Diarrheal Illness Used by Prüss et al.........................................................54 Table B.4: Exposure Scenario Application to Nicaragua.....................................................................55 Table B.5: Water Supply and Sanitation (% of households): Different Sources................................56 Table B.6: Water Supply and Sanitation in Nicaragua (% of households) as in JMP 2006.............56 Table B.7: Scenario Allocation Rules...................................................................................................57 Table B.8: Rural Water Supply and Sanitation in Nicaragua (% of Rural Population).......................57 Table B.9: Urban Water Supply and Sanitation in Nicaragua (% of Urban Population)....................57 Table B.10: Water Treatment at Point of Use in Nicaragua (% of Households)...................................58 Table B.11: Estimated Annual Cases of Diarrheal Illness in Nicaragua in 2005................................58 Table B.12: Estimated Annual Cases of Diarrheal Illness Per Person in Urban Nicaragua................59 Table B.13: Estimated Annual Cases of Diarrheal Illness Per Person in Rural Nicaragua.................59 Table B.14: Benefits of Reductions in Diarrheal Morbidity and Mortality in Rural Nicaragua...........60 Table B.15: Benefits of Reductions in Diarrheal Morbidity and Mortality in Urban Nicaragua..........61 Table B.16: Key Parameters and Outcomes in a Benefit-cost Analysis of hand washing..................62 Table B.17: A Review of costs and effectiveness of hand-washing programs....................................63 Table B.18: Benefits and Costs of a Rural Hand-Washing Program.....................................................64 Table B.19: Benefits and Costs of a Urban Hand-Washing Program...................................................65 ii Table B.20: Benefits and Costs of a Rural Drinking Water Boiling Program........................................67 Table B.21: Benefits and Costs of a Urban Drinking Water Boiling Program......................................68 Table B.22: Benefits and Costs of a Rural Drinking Water Chlorination Program at the Point of Use.....69 Table B.23: Benefits and Costs of an Urban Drinking Water Chlorination Program at the Point of Use..70 Table B.24: Relative Risks for Strong and Moderate Health Outcomes..............................................71 Table B.25: 24-hour Exposure Concentrations for Cooks and Noncooks Among Solid Fuel Users Across Kitchn Configurations..................................................................72 Table B.26: Fuels, Stove Technology and Pollution Scenarios.............................................................73 Table B.27: Base Case Estimation of Scenario-Specific Relative Risks in Nicaragua........................ 74 Table B.28: Particulate (PM) Concentrations From Cooking Stoves....................................................75 Table B.29: Odds Ratios for ARI..............................................................................................................75 Table B.30: Estimated Annual Household Consumption of Fuelwood.................................................78 Table B.31: Annual Recurrent Cost of Complete Fuel Substitution.....................................................79 Table B.32: Baseline Parameters for Estimating the Cost of Fel Substitution....................................79 Table B.33: Estimates of Improved Wood Stoves and LPG Stoves......................................................80 Table B.34: Interventions........................................................................................................................80 Table B.35: Benefits and Costs of Indoor Air Pollution Control in Rural Nicaragua............................81 Table B.36: Benefits and Costs of Indoor Air Pollution Control in Urban Nicaragua...........................82 Figures Figure ES 1: Costs of Environmental Degradation...................................................................................2 Figure ES 2: Ranking of Interventions to Reduce WSSH Cost in Nicaragua...........................................6 Figure ES 3: Ranking of Interventions for IAP Cost Reduction in Nicaragua.........................................7 Figure I.1: The Burden of Mortality Related to Environmental Causes..............................................11 Figure I.2: Health Cost of Environmental Damage in Selected Central American Countries...........11 Figure II.1: Annual Costs of Inadequate Water, Sanitation and Hygiene (Million NIO)......................16 Figure II.2: Ranking of Interventions to Reduce WSSH Costs in Nicaragua.......................................21 Figure III.1: Ranking of Interventions for IAP Cost Reduction in Nicaragua........................................27 Figure B.1: Estimated Benefit-Cost Ratios (20% Program Effectiveness Target)...............................65 Figure B.2: Ranking of Interventions to Reduce WSSH Cost in Nicaragua.........................................70 Figure B.3: Stove Efficiency and Capitol Costs.....................................................................................77 Figure B.4: Stove Efficiency From a Study in Colombia........................................................................78 Figure B.5: Ranking of Interventions for IAP Cost Reduction in Nicaragua........................................83 BOXES Box ES 1: Estimation of Environmental Health Costs for Nicaragua..................................................3 iii Acronyms and Abbreviations ARI: Acute respiratory infections CEA: Country Environmental Analysis COI: Cost of Illness COPD: Chronic obstructive pulmonary disease DALYs: Disability adjusted life years EHS: Environmental Health Safeguards ENACAL: Empresa Nicaragüense de Acueductos y Alcantarillados Sanitarios (Nicaraguan National Water Utility Company) ENDESA: Encuesta Nacional de Demografía y Salud (National Survey on Demographics and Health) GDP: Gross domestic product GoN: Government of Nicaragua HCA: Human Capital Approach IAP: Indoor air pollution LPG: Liquefied petroleum gas MARENA: Ministerio de Ambiente y Recursos Naturales (Ministry of Environment and Natural Resources) MINSA: Ministerio de Salud (Ministry of Health) MDGs: Millennium Development Goals NGO: Non-governmental organization NIO: Nicaraguan Córdoba NO2: Nitrogen dioxide PM10: Particulate matter less than 10 microns in size PM2.5: Particulate matter less than 2.5 microns in size TSP: Total suspended particles UNDP: United Nations Development Programme VSL: Value of Statistical Life WHO: World Health Organization WSH: Water Sanitation and Hygiene WSSH: Water Supply Sanitation and Hygiene WSP: Water Sanitation v Environmental Health in Nicaragua Addressing Key Environmental Challenges Irina Klytchnikova, Senior Economist, World Bank Juan Carlos Belausteguigoitia, Lead Environmental Economist, World Bank Elena Strukova, Consultant, World Bank Anjali Acharya, Senior Environmental Specialist, World Bank Francisco Carranza, Water Sanitation Specialist, World Bank Nelson Antonio Medina Rocha, Coordinator for Nicaragua, Water Sanitation Specialist, World Bank Gerardo Sánchez, Environmental Specialist, World Bank Acknowledgements The authors would like to thank Yewande Awe (Sr. addition, workshops and other meetings were held Environmental Engineer, ENV) and Ernesto Sán- with key officials from MARENA and MINSA, as well chez-Triana (Lead Environmental Specialist, SASDI) as other key stakeholders; their assistance with for their comments on this study. Assistance by data and information has been critical for this re- Santiago Sandoval, Linda Castillo and editing by port. The authors would like to thank MARENA Vice Janice Molina are gratefully acknowledged. The Minister Roberto Araquistain, MINSA Vice Minister study was carried out under the overall guidance Nora Orozco, Messrs. Jesús Marín Ruíz and Boan- of Laura Frigenti (Country Director), Karin Kemper erges Castro (MINSA), members of the Steering (Sector Manager since December 2009), Laura Committee for the Study, composed of high-level Tlaiye (Sector Manager until November 2009), decision makers, and the inter-institutional Tech- Gregor Wolf (Sector Leader), Joseph Owen (Country nical Working Group, composed of technical ex- Manager), and the guidance of the Steering Com- perts, for their overall support for, and guidance of, mittee for the study, composed of high-level deci- this analysis. sion makers in Nicaragua. Financial support by the The study was completed and disseminated in Ni- Government of the Netherlands through the Bank caragua in June 2010, and published in electronic Netherlands Partnership Program (BNPP) is grate- form. The report was welcomed by MARENA, which fully acknowledged. also provided additional comments that future In order to provide country-specific estimates, studies would need to address. This is the paper workshops were held with the Technical Working version of the electronic publication dated June Group which was set up to provide inputs to, and 2010. comment on, this environmental health analysis. In vii Abstract In Nicaragua, poor access to water and sanitation, tion coverage, and proper waste disposal. Indoor high use of fuelwood for cooking in rural areas, air quality, especially in rural areas where biomass and growing vehicular use in urban areas are re- is used for cooking, and outdoor air quality, espe- sulting in environmental health risks, especially in cially in urban areas such as Managua, are grow- children under five years of age. The Government ing and important concerns. These environmental of Nicaragua requested the World Bank to under- health risks are especially important for the most take this study to answer two fundamental sets of vulnerable subgroups of Nicaragua’s population, questions on environment-related health problems including children under age five, women, the el- in Nicaragua. The first set helps to demonstrate the derly and the poor. significance of environment-related health prob- lems: (a) How much disease and how many deaths Several key messages have emerged from the are attributable to environmental risks? (b) Which process of putting together this study: (i) environ- subgroups bear the highest burden of disease? (c) mental health risks impose a significant burden on What are the economic costs of the health impact Nicaragua’s economy, amounting to 2.6 billion NIO of these environmental risks? The second set re- or 2.4 percent of the country’s GDP, and result in fers to the appropriate choice of interventions that premature deaths and infections, especially in chil- the GoN can undertake to reduce these environ- dren under five; (ii) cost-effective interventions to mental risks to human health: (a) Which interven- address these environmental health risks exist and tions are the most cost-effective? should be prioritized in Nicaragua; (iii) country-spe- cific health and environmental data are somewhat This study estimates that environmental health limited, especially in the case of air quality, and costs represent a significant burden on Nicara- data collection and monitoring need to be further gua’s economy. While progress is being made strengthened; and (iii) the capacity of MARENA on interventions to address environmental risks, and MINSA staff to conduct environmental health much more attention and resources need to be in- costing analysis needs to be strengthened through vested in water supply coverage, improved sanita- training. ix Executive Summary Globally, an estimated 24 percent of the disease unimproved latrines). In urban areas, the targets burden (healthy life years lost) and an estimated are to increase access to public water supply from 23 percent of all deaths (premature mortality) are 72 to 86 percent, and to sanitation from 36 to 48 attributable to environmental risks (WHO 2006). percent over the same period. In the area of air The burden of disease is unequally shared, with the quality management, the GoN is preparing guide- children and the poor being particularly affected. lines for an air quality policy, to be adopted in the Among children between the ages 0 and 14, the future. The findings and recommendations of this proportion of deaths attributable to environmental study are directly relevant to the GoN’s efforts to risks—such as poor water and sanitation, indoor air address the environmental causes of diseases. pollution and vector-borne diseases—is estimated In Nicaragua, poor access to water and sanitation, to be as high as 36 percent (WHO 2006). high use of fuelwood for cooking in rural areas, and In its 2009–2011 Updated National Human De- growing vehicular use in urban areas are resulting velopment Plan (NHDP), the Government of Nica- in environmental health risks, especially in chil- ragua (GoN) has placed strong emphasis on pre- dren under five years of age. The Government of ventive health measures and on addressing the Nicaragua requested the World Bank to undertake root causes of ill health. The Plan has identified a a study to answer two fundamental sets of ques- series of measures to help Nicaragua achieve the tions on environment-related health problems in MDG targets on child mortality and the environ- Nicaragua. The first set helps to demonstrate the ment. Recognizing the crucial role of improving the significance of environment-related health prob- urban and rural populations’ access to improved lems: (a) How much disease and how many deaths water and sanitation services, the GoN has also set are attributable to environmental risks? (b) Which ambitious targets to expand the provision of those subgroups bear the highest burden of disease? (c) services. This strategy has been reflected in the What are the economic costs of the health impact investment plans for the water supply and sanita- of these environmental risks? The second set re- tion sector. The NHDP states the targets to expand fers to the appropriate choice of interventions that access to public water supply in rural areas from the GoN can undertake to reduce these environ- 56 percent in 2007 to 64 percent by 2011, and to mental risks to human health: (a) Which interven- expand rural sanitation from 73 percent to 81 per- tions are the most cost-effective? cent (including all types of sanitation and including 1 Environmental Health in Nicaragua Using established methodology recommended what limited, especially in the case of air quality, by the World Health Organization and others, this and data collection and monitoring need to be fur- study attempts to provide information and raise ther strengthened; and (iii) the capacity of MARENA awareness about the importance of environmental and MINSA staff to conduct environmental health health interventions in addressing health, specifi- costing analysis needs to be strengthened through cally child health, issues in developing countries proper training. such as Nicaragua. This study concentrates on three main environmental health risk factors: inad- equate water supply and sanitation, indoor air pollu- Assessing the Economic Burden of tion and urban air pollution. Recognizing the impor- Environmental Health Risks tance of using country-specific data, this analysis Environmental health costs represent a significant has explicitly included data and figures provided by burden on Nicaragua’s economy. While progress is the Ministry of Environment and Natural Resources being made on interventions to address environ- (MARENA) and by the Ministry of Health (MINSA). mental risks, much more attention and resources The preparation of this analysis was launched by a need to be invested in water supply coverage, im- consultation workshop, which was attended by rep- proved sanitation coverage, and proper waste dis- resentatives of MARENA, MINSA, academia, non- posal. Indoor air quality, especially in rural areas governmental organizations and donor agencies. where biomass is used for cooking, and outdoor Drafts of the analysis were shared with the Tech- air quality, especially in urban areas such as Ma- nical Working Group that was created to review nagua, are growing and important concerns. These the findings, and comments were elicited. In June environmental health risks are especially impor- 2009, the Bank team organized a targeted one-day tant for the most vulnerable subgroups of Nicara- training course for selected staff in the Ministry of gua’s population, including children under age five, Health, aimed at building local capacity to carry women, the elderly and the poor. This study follows out this analysis. The draft report with results of a methodology validated around the world for con- the study were shared with MINSA, MARENA and ducting analyses of the health burden of environ- high-level decision makers from other agencies mental degradation (Box ES 1). in March 2010, and the feedback received during these consultations has been incorporated in this Figure ES 1: Costs of Environmental Degradation version. 4% 2500 Several key messages have emerged from the process of putting together this study: (i) environ- 2000 3% mental health risks impose a significant burden GDP per capita (USD) CED as % of GDP on Nicaragua’s economy, amounting to 2.6 billion 1500 2% NIO or 2.4 percent of the country’s GDP, and result 1000 in premature deaths and infections, especially in 1% children under five; (ii) cost-effective interventions 500 to address these environmental health risks exist 0% 0 and should be prioritized in Nicaragua; (iii) country- Guatemala El Salvador Nicaragua Honduras specific health and environmental data are some- GDP per capita Cost of Environmental Damage (CED) Source: CEA studies, World Bank 2005–2009. 2 Box ES 1. Estimation of Environmental Health Costs The analyses to assess the health impacts and subsequent costs attributed to risks from urban air pollution, poor water and sanitation, and indoor air pollution are based on World Health Organization (WHO) standard methodologies and on international research. Water, sanitation and hygiene: Data on Nicaragua’s population and on diarrheal mortality and prevalence in children under five are used to estimate the mortality and illness attributed to poor water and sanitation (WHO estimates that about 90 percent of diarrheal illness is attributable to inadequate water sanitation and hygiene). Estimates are made also for adult morbidity. Costs associated with this risk factor include expenses to treat diarrheal illness, such as doctor fees, laboratory tests, drugs and bed charges when hospitalization is needed. Other costs include lost productivity when adults fall sick and stay home from work, or when primary caregivers have to take care of sick children (and potentially lose wages). Indoor air pollution: Data on population using fuelwood for cooking in Nicaragua and on child mortality and prevalence of acute respiratory illnesses (ARIs) is used with odds ratios from international research evidence to estimate the premature deaths and cases of illness from ARIs in children and adults and chronic obstruc- tive pulmonary disease (COPD) in adult women. For COPD mortality and morbidity incidence, WHO regional estimates are used in the absence of Nicaraguan data. Urban air pollution: With Nicaraguan data on particulate matter and urban population figures, this analysis uses dose-response coefficients from international research evidence on long-term air pollution impacts on health to estimate the number of premature deaths and sickness attributed to urban air pollution. Using WHO methodologies, these health outcomes are then translated into disability adjusted life years (DALYs), and are also valued to arrive at cost estimates. The cost of mortality for adults is based on the value of statistical life (VSL) as a high bound and the Human Capital Approach (HCA) as a low bound, while that for children is based on the HCA. In addition, treatment cost represents private sector (unsubsidized) health care services, while the value of time for adults (for lost work days due to sickness or caregiving) in Nicaragua is estimated at 75 percent of average hourly wages. Source: Background report for this study (Strukova 2009). Similar to other countries in Central America, the hygiene amounted to 0.92 billion NIO, or 0.9 per- environmental health implications of poor water cent of the country’s GDP; (ii) indoor air pollution supply and sanitation and of indoor and urban air amounted to 870 million NIO, or 0.85 percent of pollution dramatically impact Nicaragua’s ability GDP; and (iii) urban air pollution amounted to 780 to achieve targets for reducing child mortality, million NIO, or 0.75 percent of the country’s GDP. improving maternal health, combating diseases, and other Millennium Development Goals (MDGs). Table ES 1: Aggregate Environmental Health Costs Overall, the analysis carried out in this study shows for Nicaragua that Nicaragua’s economic costs associated with NIO lack of water and sanitation and with indoor and (billions) Percent of GDP urban air pollution amount to nearly 2.6 billion Lack of water and NIO, or 2.4 percent of the country’s gross domestic 0.8–1.0 0.8–0.9 sanitation product (GDP) (see Table ES 1). Among the three Indoor air pollution 0.8–0.9 0.7–0.9 categories of environmental health risks, damages Urban air pollution 0.3–1.2 0.3–1.2 due to (i) inadequate water supply, sanitation and Total 1.9–3.1 1.8–3.0 3 Environmental Health in Nicaragua The costs of these environmental problems have Applying an established methodology used by WHO been estimated in this report to help policy makers and other international agencies and also used in in Nicaragua appreciate the magnitude of these this study, it is estimated that 240 children under issues and better integrate environmental health age five in urban areas and 320 in rural areas die considerations into economic development de- from diarrheal diseases attributed to poor water, cision making. These costs include not only the sanitation and hygiene. Furthermore, among chil- medical costs of treatment and lost productivity dren under age five, more than a million annual due to sickness and caregiving, but also provide cases of diarrhea in rural and urban areas, respec- an estimate of the value of pain and suffering from tively, result from poor water, sanitation and hy- premature death and disease. This analysis covers giene. The estimated number of cases of diarrheal only a limited number of diseases attributed to the illness in children under age five is about 1.3 times three environmental risk factors in question—inad- higher in rural areas.2 equate water and sanitation, indoor air pollution, The annual cost of diarrheal mortality and mor- and urban air pollution—and therefore underesti- bidity from inadequate water, sanitation and hy- mates the economic burden these environmental giene is estimated at 345 million NIO in urban risk factors place on Nicaragua. Furthermore, the areas and 455 million NIO in rural areas. The cost study does not include the indirect impacts of poor of morbidity includes the cost of illness (medical water, sanitation and hygiene which are mediated treatment, medicines and value of lost time). In through malnutrition. addition, the cost of averting expenditures in Ni- Poor water quality, sanitation and hygiene: Ac- caragua—associated with the purchase of bottled cording to the latest WHO/UNICEF Joint Monitoring water and disinfecting water through boiling or Program (JMP/2006), 90 percent of the urban chlorination to avoid health risks—is estimated to population and 56 percent of the rural population range from 80 to 150 million NIO per year. The total in Nicaragua have access to improved sources of estimated cost associated with inadequate water drinking water (piped water, public tap, borehole/ supply, sanitation and hygiene ranges from 830 tubewell, protected well, protected spring or rain- million to 1 billion NIO per year, with a mean of 915 water). However, even in connected households, million NIO, amounting to 0.9 percent of the coun- the quality of service is low (WSP 2008). In terms of try’s GDP. sanitation, 56 percent of the urban population and Indoor air pollution: Acute respiratory infections 34 percent of the rural population have access to (ARI) mainly affect women and children in rural improved sanitation (are connected to public ser- areas in Nicaragua, where nine out of ten house- vice or have a septic tank).1 About 20 percent of holds (92 percent) burn fuelwood in inefficient urban and 30 percent of rural households disinfect stoves in poorly ventilated areas.3 The vulnerable drinking water (WSP 2008) in Nicaragua. subgroups in the population include children under 1 These figures are comparable to the estimated shares of population in rural and urban areas with access to piped water and sanitation in all cases except for sanitation in rural areas (data from official Nicaraguan sources [ENACAL, FISE] and the 2009–2011 Updated National Human Development Plan). The figures for sanitation in rural areas are higher in the official statistics because all types of sanitation are included in these estimates, while this study excludes unimproved latrines. 2 These estimates are made by assuming the under-age-five child mortality rate of 31 and 40 per 1,000 live births in urban and rural areas, respectively, based on 2006 data provided by the 2006–2007 ENDESA survey, which inquired about deaths of children under age five during the five-year period prior to the survey. The same source reports that for the ten-year period prior to the survey, the mortality rate for children under age five 35 and 47 per 1,000 live births in urban and rural areas. 3 According to the 2006–2007 ENDESA survey, about 38 percent of households in urban and 92 percent in rural areas used fuelwood for cooking. 4 age five (ARI mortality and morbidity among chil- concentration for Managua was estimated at 67 dren) and women over age 30 (COPD mortality µg/m3, and in other Nicaraguan cities with popu- and morbidity; ARI morbidity), because these sub- lations over 100,000 at 43 µg/m3. Urban air par- groups are exposed to smoky kitchens. ticulate pollution is estimated to cause around 420 Each year, an estimated 140 to 200 children under premature deaths annually, although there is sig- age five die from ARIs in rural areas, and an addi- nificant uncertainty about this estimate because of tional 40 to 70 children die in urban areas in Nica- the limited data on emission levels and additional ragua. Among children under age five, more than uncertainty in the estimation of mortality rates half a million annual cases of acute respiratory from illnesses related to urban air pollution in Nica- infections in rural areas, and more than 200,000 ragua.4 The number of new cases of chronic bron- cases in urban areas, can be linked to indoor air chitis is estimated at about 520 per year. Annual pollution. Among females over age 30, each year hospitalizations due to pollution are estimated at there are nearly 285,000 cases of indoor-air-pollu- close to 630, and emergency room visits/outpa- tion-related ARI morbidity in rural areas and nearly tient hospitalizations at 26,500 per year. 162,000 cases in urban areas. Indoor air pollution In estimating the costs of health impacts from also causes chronic obstructive pulmonary disease particulate matter in Nicaragua, health conditions (COPD) in females over age 30: nearly 120 women such as premature mortality, hospital admissions, die annually from COPD in urban and rural areas, restricted activity days, and emergency visits have and about 3,000 new cases of COPD can be attrib- been considered. In the absence of proper data on uted to indoor air pollution each year. treatment costs, informed estimates have been The total estimated annual cost of indoor air pol- provided by medical experts in Managua. The lution ranges from 540 million 1.2 billion NIO with mean estimated annual cost of urban air pollution a mean cost of 870 million NIO. These costs ac- due to PM ranges from 305 million to 1.25 billion counted for about 0.8 percent of GDP in 2007. NIO, with a mean of about 780 million NIO. The low ARI in children represents the largest share (40 and high estimates are obtained by applying the percent) of the costs. The rural poor are especially HCA and the VSL approaches to the valuation of vulnerable and account for 56 percent of the total the costs of mortality, respectively. This represents costs of indoor air pollution. about 0.7 percent of the country’s GDP. Around 70 percent of the costs are due to mortality, and Urban air pollution: The total urban population ex- the remaining 30 percent are associated with mor- posed to air pollution was estimated to be about bidity. 1.95 million, or 62 percent of the total Nicaraguan population in 2007. Data relating to urban air quality concentrations are extremely limited and Prioritizing Interventions Through very dated. Furthermore, there is no emissions in- ventory, and very little city-specific data exist. Using Cost-Benefit Analyses the latest available data for 2001, adjusted using Recognizing the importance of setting priorities, World Bank estimates, the annual average PM10 subsequent follow-up work included cost-benefit 4 An estimate of the annual incidence of chronic bronchitis (CB) is required in order to apply the dose response coefficients, which translate air pollution measures into the incidence of illness. In the absence of data on CB incidence for Nicaragua, this study has used the CB rate from WHO (2001) and Shibuya et al. (2001) for the AMRO D region of WHO, which includes Nicaragua. For details, see Annex A. 5 Environmental Health in Nicaragua analyses of various environmental health interven- the cumulative cost curve equals the unit cost of tions to address inadequate water sanitation and an intervention per one percentage point reduction hygiene, and indoor air pollution. These analyses of health damages; the total area below each seg- would help the government determine the most ment is equal to the total cost of this intervention cost-effective interventions to implement in terms (see Figure ES 2). Although several of these inter- of addressing key environmental health risks in Ni- ventions have significant time-savings benefits, caragua. only health benefits were considered in this study. Water, sanitation, hygiene: Interventions related to Several assumptions were also made relating to improving the quality and quantity of water supply, the sustainability of these programs from a behav- improving access to improved sanitation, and pro- ioral perspective. grams encouraging better hygiene practices (in- Findings from this analysis revealed that in terms cluding hand washing) have potential health ben- of water, sanitation and hygiene interventions, efits. For Nicaragua, this study estimated the cost- highest priority should be given to drinking water benefit ratios for improvements in water supply disinfection and hand-washing programs (in rural and sanitation, hand-washing programs, and and urban areas) because they had the lowest drinking water disinfection programs (both boiling cost-benefit ratios. and chlorination). The height of each segment of Figure ES 2: Ranking of Interventions to Reduce Water Supply, Sanitation and Hygiene Cost in Nicaragua 80 Water supply (urban, rural) Marginal costs and marginal benefits (Million NIO) 70 Improved Cumulative 60 sanitation cost Disinfection – water boiling (urban) 50 Household Improved 40 chlorination sanitation (urban) (rural ) 30 Disinfection – water boiling (rural) 20 Hygiene Household chlorination 10 (rural) 0 0 5 10 15 20 25 30 35 40 45 50 Reduction of health costs (% of total health costs in the sector) 6 Indoor air pollution: Interventions are available to to LPG) is much more uncertain and depends sub- reduce indoor air pollution and associated health stantively on the costs and benefits used. effects. These interventions typically include ad- Urban air pollution: For urban air pollution, a similar dressing (a) source of pollution, including moving cost-benefit analysis was not carried out because from traditional stoves to improved stoves, and the paucity of air quality data is a significant con- switching to cleaner fuels such as LPG; (b) the straint to prioritizing interventions to reduce urban living environment, such as chimneys and smoke air pollution in Nicaragua. The study recommends hoods (with flues); and (c) user behavior, such as the development of an emissions inventory to iden- keeping young children away from smoke. For Nica- tify the main sources of air pollution in Managua ragua, cost-benefit ratios were only estimated for and León. Furthermore, the study recommends interventions related to moving from unimproved to as a priority the development of an air quality improved stoves, better ventilation, and switching monitoring system for Managua and León so as to to cleaner fuels (see Figure ES 3). monitor emissions in the main urban areas. Work The analysis for Nicaragua recommends shifting recently initiated by the World Bank to develop an households with unimproved stoves to improved emissions inventory for Managua and León will ones, and improving ventilation. The relative at- contribute to the identification of relevant interven- tractiveness of other interventions (such as from tions to reduce urban air pollution in Nicaragua. unimproved stoves to LPG or from improved stoves Figure ES 3: Ranking of Interventions for IAP Cost Reduction in Nicaragua Unimproved wood 35 stoves to gas stoves Marginal costs and marginal benefits (Million NIO) (rural) 30 Unimproved Cumulative 25 wood stoves to cost gas stoves (urban) 20 15 Unimproved Unimproved to improved Improved to improved wood stoves Improved 10 ventilation wood stoves (rural) ventilation (rural) (urban) (urban) 5 0 0 5 10 15 20 25 30 35 40 Reduction of health costs (% of total health costs in the sector) 7 I. Overview Environmental pollution and inadequate environ- water and sanitation services, the GoN has also mental services, such as improved water supply and set ambitious targets to expand the provision of sanitation, impose significant costs to Nicaragua’s these services. This strategy has been reflected economy in the form of ill health, lost income, and in the investment plans for the water supply and increased poverty and vulnerability. While progress sanitation sector. The NHDP states the targets to is being made on interventions to address environ- expand access to public water supply in rural areas mental risks, much more attention and resources from 56 percent in 2007 to 64 percent by 2011, need to be invested in water supply coverage and and to expand rural sanitation from 73 to 81 per- in improved sanitation coverage. Indoor air quality, cent (including all types of sanitation and unim- especially in rural areas where biomass is used proved latrines). In urban areas, the targets are to for cooking, and outdoor air quality, especially in increase access to public water supply from 72 to urban areas such as Managua, are growing con- 86 percent, and to sanitation from 36 to 48 per- cerns. These environmental health risks are espe- cent over the same period. In the area of air quality cially important for the most vulnerable subgroups management, the GoN is preparing guidelines for of Nicaraguan population, including children under an air quality policy, to be adopted in the future. age five, women, the elderly and the poor. This sec- The findings and recommendations of this study tion provides overall estimates of the health costs are directly relevant to the GoN’s efforts to address of environmental pollution in Nicaragua. the environmental causes of diseases. In its 2009–2011 Updated National Human De- As the evidence shows, the environmental health velopment Plan (NHDP), the Government of Nica- implications of poor water supply and sanitation ragua (GoN) has placed strong emphasis on pre- and of indoor and urban air pollution dramatically ventive health measures and on addressing the impact Nicaragua’s ability to achieve targets for re- root causes of ill health. The Plan has identified a ducing child mortality, improving maternal health, series of measures to help Nicaragua achieve the combating diseases, and other Millennium Devel- MDG targets on child mortality and the environ- opment Goals (MDGs). Overall, the analysis shows ment. Recognizing the crucial role of improving the that Nicaragua’s economic costs associated with urban and rural populations’ access to improved lack of water and sanitation and with indoor and 9 Environmental Health in Nicaragua urban air pollution amount to nearly 2.6 billion giene is in large part associated with uncertainties NIO, or 2.4 percent of the country’s gross domestic regarding estimates of diarrheal child mortality and product (GDP) (see Table I.1). Among the three cat- morbidity. In the case of urban air pollution, two egories of environmental health risks, (i) damages different valuation techniques for estimating the due to inadequate water supply, sanitation and social cost of mortality have been applied; these hygiene amounted to 0.92 billion NIO, or 0.9 per- yielded the “low� and “high� estimates. The range cent of the country’s GDP; (ii) indoor air pollution for indoor air pollution is mainly from the uncer- amounted to 870 million NIO, or 0.85 percent of tainty about the level of exposure to indoor smoke GDP; and (iii) urban outdoor air pollution amounted from the use of fuelwood, and thus a range has to 780 million NIO, or 0.75 percent of the country’s been applied for the level of health risk. GDP. Distributional impacts: The costs of damages as- sociated with environmental health are distributed Table I.1: Aggregate Environmental Health Costs unevenly across the population. Losses due to out- for Nicaragua door air pollution were estimated for the inhabit- NIO Percent of ants of the cities with populations over 100,000; (billion) GDP costs of inadequate water supply, sanitation and Lack of water and hygiene were estimated for the entire population 0.8–1.0 0.8–0.9 sanitation of Nicaragua; indoor air pollution costs were esti- Indoor air pollution 0.8–0.9 0.7–0.9 mated for households that use solid fuel for cooking Urban air pollution 0.3–1.2 0.3–1.2 (about 60 percent of households). Outdoor urban Total 1.9–3.1 1.8–3.1 air pollution has the highest cost on a per-person basis, followed by damages due to indoor air pollu- In addition to the mean estimates, “low� and “high� tion. Among the estimated cost items, inadequate estimates of annual costs are presented in Table water supply and sanitation has the lowest cost per I.2. The range for water supply, sanitation and hy- capita. Vulnerable subgroups: About 2 percent of adult Table I.2: Annual Cost of Environmental Damage – mortality is attributed to outdoor and indoor air pol- Low and High Estimates (million NIO per year) lution, and about 15 percent of under-age-five child mortality is attributed to inadequate water supply, Environmental Mean “Low� “High� sanitation and hygiene and to indoor air pollution. Categories Estimate Figure 1.1 presents the burden of mortality related Water supply, to environmental causes as a percent of total mor- sanitation and 832 917 1,001 hygiene tality among adults and children under age five in Nicaragua. Indoor air 772 872 972 pollution Central American Context: The cost of environ- Outdoor air mental degradation in Nicaragua is comparable 309 779 1,249 pollution with other countries at similar income levels. In all Total costs 1,913 2,568 3,222 the selected countries in Central America, these Costs as percent costs range from two to three percent of the re- 1.8% 2.4% 3.1% of GDP in 2007 10 Figure I.1: The Burden of Mortality Related to a environmental risks to health, no negative impacts Environmental Causes on productive assets from environmental degrada- tion, and no damages to health, infrastructure and 12% housing from natural disasters. The estimates of damage costs presented above can therefore be 8% used as a starting point to evaluate the benefits of interventions to improve environmental quality and 4% reduce environmental damage. From a socioeconomic point of view, the well-being of society will improve if interventions that provide 0% Outdoor air pollution Indoor air pollution Water supply, greater benefits than costs are implemented. A sanitation and hygiene % Adult mortality % Under 5 child mortality cost-benefit analysis would provide a sense of pri- oritization among various interventions. In terms of spective country’s GDP. In Nicaragua, the health water supply, sanitation and hygiene, a cost-benefit cost of environmental damage—including costs analysis would involve looking at various interven- of inadequate water, sanitation and hygiene, and tions such as hand washing, drinking water dis- infection, improved sanitation and water supply Figure I.2: Health Cost of Environmental Damage options, etc. Similarly, for indoor air pollution a in Selected Central American Countries cost-benefit analysis would consider alternative scenarios of indoor cooking that include the use 4% 2500 of improved or unimproved stoves, and the use of solid or clean fuel (LPG). For outdoor air pollution, 3% 2000 this analysis would assess the costs and benefits GDP per capita (USD) CED as % of GDP 1500 of various particulate matter (PM) emissions con- 2% trol options in urban areas. However, the identifi- 1000 cation of appropriate interventions is constrained 1% by the lack of air quality monitoring data and the 500 absence of an emissions inventory in Nicaragua. 0% 0 In the following sections, this study provides esti- Guatemala El Salvador Nicaragua Honduras GDP per capita Cost of Environmental Damage (CED) mates for the costs associated with environmental Source: CEA studies, World Bank 2005–2009. health risks from inadequate water, sanitation and hygiene, and from indoor and urban air pollution. costs of indoor and outdoor air pollution—amounts Therefore, in order to help the GoN prioritize among to 1.8–3.1 percent of the country’s GDP. available interventions, this study also attempts to The cost of environmental damage is equivalent undertake cost-benefit analyses for each of these to the benefits that society would enjoy if environ- environmental health risks in the context of Nica- mental quality were improved to a condition with no ragua. 11 II. Water, Sanitation, and Hygiene According to the latest WHO/UNICEF Joint Moni- Health impacts: The estimation for diarrheal mor- toring Program (JMP/2006), 90 percent of the tality in Nicaragua is based on total child mortality. urban population and 56 percent of the rural Nicaragua has relatively high child mortality in gen- population in Nicaragua have access to improved eral, although the estimates are quite uncertain. sources of drinking water (piped water, public tap, According to the latest demographic and health borehole/tubewell, protected well, protected spring survey (Encuesta Nicaragüense de Demografía y or rainwater). However, even in connected house- Salud, ENDESA), mortality among children under holds, the quality of service is low (WSP 2008). In age five is 35 per 1,000 live births, using the 2006 terms of sanitation, 56 percent of the urban popu- child mortality rate as the starting point in the es- lation and 34 percent of the rural population have timation (ENDESA 2006–2007). A similar estimate access to improved sanitation (are connected to of child mortality in 1998 was 51 deaths for 1,000 public service or have a septic tank). live births (ENDESA 1998). The main reason for Inadequate quantity and quality of potable water the uncertainty and year-to-year variation in the es- supply, sanitation facilities and practices, and hy- timates is that not all births are registered (Alarcón, giene conditions are associated with various ill- Robles 2007). nesses both in adults and children. The major The estimated health effects from inadequate health effects of inadequate water quality and water, sanitation and hygiene are based on back- quantity, sanitation facilities and practices, and hy- ground health data, taking into account the WHO giene are diarrheal morbidity and mortality.5 WHO estimate that 88 percent of diarrheal illness in estimates that about 90 percent of diarrheal illness Nicaragua is attributable to water, sanitation and is attributable to these factors (WHO 2002). While hygiene.6 The percent of diarrheal cases among diarrheal illness is generally not as serious as some children under age five is estimated from ENDESA other waterborne illnesses, it is more common and 2006–2007, and in the age group older than age affects a larger number of people. five it is estimated from the Encuesta de Hogares sobre Medición del Nivel de Vida, EMNV (2005).7 5 Hygiene refers to personal hygiene (such as hand washing), domestic hygiene and food hygiene. 6 Estimated applying methodology presented in Fewtrell et al., 2007. 7 EMNV (2005) reports significantly lower diarrheal prevalence among people over 5 in comparison with other AMRO D countries. This lower diarrheal prevalence among people over 5 explains relatively low estimates of WSSH losses on Nicaragua. WHO GBD (2004) predicts higher rates of diarrheal illness in AMRO D than those reported in EMNV (2005). 13 Environmental Health in Nicaragua The same method was applied to estimate treated health benefits of drinking boiled water, and poorer cases because only the percentage of treated access to energy are also determinants of higher cases among children under age five was available diarrheal illness in rural children. from ENDESA 2006–2007. Economic costs: Inadequate access to water and Each year, an estimated 240 children under age sanitation leads to premature deaths and dis- five in urban areas and 320 in rural areas die from ease, which in turn impose costs on Nicaragua’s diarrheal diseases attributed to poor water, sani- economy. These costs include the expenses in- tation and hygiene. Furthermore, among children curred to treat illness from diseases attributed to under age five, more than a million annual cases poor water and sanitation, doctor fees, laboratory of diarrhea in rural and urban areas, respectively, tests, drugs, and bed charges when hospitaliza- result from poor water, sanitation and hygiene. The tion is needed. Other costs include lost productivity estimated number of cases of diarrheal illness in when adults fall sick and stay home from work, or children under age five is about 1.3 times higher in when primary caregivers have to take care of sick rural areas. children (and potentially lose wages). Furthermore, Although the rural population in Nicaragua com- the pain and suffering from premature death and prises about 44 percent of the total population, the illness can also be valued by calculating the burden estimated number of cases of diarrheal child mor- of disease in disability adjusted life years (DALYs)8 tality is about 40 percent higher there because the and by costing these DALYs at GDP per capita. share of children in the population as well as the The annual cost of diarrheal mortality and morbidity diarrheal prevalence are also substantially higher from inadequate water, sanitation and hygiene is in rural areas. The larger share of children in the estimated at 345 million NIO in urban areas and rural population, lower public awareness of the 455 million NIO in rural areas (Table II.2). The cost of diarrheal child mortality is based on the Human Table II.1: Estimated Annual Health Effects from Capital Approach (HCA) discussed in Annex A. The Water, Sanitation, Hygiene cost of morbidity includes the cost of illness (med- ical treatment, medicines, and value of lost time). Urban cases Rural cases About 50 percent of these costs are associated Children with the value of time lost to illness (including care (under age five 240 320 giving), and 50 percent are from cost of treatment – increased mortality and medicines. Children (under Avertive Expenditures: In the presence of perceived age five) – health risks, people often take averting measures 1,000,000 1,200,000 increased to avoid these risks. If people perceive there is a morbidity risk of illness from the municipal water supply or Population from other sources of water supply that they use, over age five 700,000 650,000 some of them are likely to purchase bottled water – increased morbidity for drinking purposes, boil their water, filter or chlo- 8 The Disability Adjusted Life Year (DALY) is a health gap measure that extends the concept of years of life lost due to premature death to include equivalent years of healthy life lost due to poor health or disability. DALYs for a disease or health condition are calculated as the sum of the years of life lost due to premature mortality (YLL) in the population and the years lost due to disability (YLD) for incident cases of the health condition. 14 Table II.2: Estimated Annual Cost of Diarrheal Ill- drinking water per person is 0.5 to 1.0 liters among ness (million NIO) households that use boiled water. The residential cost of energy is estimated based on data from ex- Urban Rural perts, using the average stove efficiency for natural Mortality: Children 135 185 gas and wood fuel (see Annex A for details). Base- under age 5 line data are presented in Table II.3. The annual Morbidity: Children cost of boiling water for these households is esti- under age 5, 210 270 mated at 20 million NIO per year. population over age 5 About 20 percent of urban and 30 percent of rural Total Annual Costs 345 455 households disinfect drinking water (WSP 2008) in Nicaragua. Chlorination is predominantly used rinate it. These averting expenditures associated to disinfect drinking water.9 Braghetta (2006) cites with the purchase of bottled water and boiling the the cost of chlorinating drinking water at the point water account for the costs of health risks. of use in Nicaragua at US$10 per household per year. It is estimated that rural households spend 24 No reports were identified on the use of bottled million NIO and urban households spend 15 million water in Nicaragua. Bottled water use in Nicaragua NIO annually for water chlorination. Data on filter was estimated at 23,000 m3 (23 million liters) use are sketchy. Although MINSA has a program per year in 2004, with information on pricing and for household filter distribution and about 80,000 costs of production derived from Gómez 2008. Ac- filters are reported to have been distributed, there cording to surveys, nearly five to seven percent of is no information on actual use. WSP (2008) esti- urban households and one to three percent of rural mates that about 5 percent of urban households households in Nicaragua boil their drinking water, use filters; this would translate into annual costs of either all the time or sometimes (WSP 2008). It is about 7 million NIO.10 assumed that the average daily consumption of Table II.3: Baseline Data for Estimation of Costs Due to Water Boiling Indicator Data Notes % of households that boil their drinking water 1–7% WSP 2008 Liters per person per Average daily consumption of drinking water 0.5–1.0 day % of households using electricity (urban-rural) 0.9–0.3% % of households using LPG (urban-rural) 65.6–7.3% ENDESA 2006–2007 % of households using fuel wood (urban-rural) 30.9–91.4% % of households using coal (urban-rural) 0.9–0.1% Energy requirement for heating water (100% efficiency) 4,200 Joules/ltr/1 degree C Average stove efficiency for heating water 25–50% Varies by type of stove Average time of boiling water (after reaching boiling point) 10 Minutes 9 WSP (2008) reports that 90 percent in rural areas and 55 to 60 percent in urban areas of those who disinfect choose to chlorinate water. However, other sources, such as Ramírez (2008), indicate that water filtration is a significant source of disinfected water in Nicaragua. 15 Environmental Health in Nicaragua Table II.4: Estimated Annual Household Avertive impacts represents an estimated 87 percent, and Expenditures avertive expenditures represent about 13 percent of total costs. While rural areas account for 52 per- Total Annual Cost (million NIO) cent of total cost, urban areas account for the re- maining 48 percent. “Low� “High� Bottled water Inadequate water and sanitation as well as poor hy- 25 75 consumption giene contribute directly to health risks (diarrheal Boiling of drinking water 10 30 diseases), especially in children under age five. Chlorination 38 38 However, given the linkages among environmental health, malnutrition and disease, there are addi- Filtration 7 7 tional malnutrition-mediated (indirect) health im- Total annual cost 80 150 pacts associated with inadequate water and sani- tation provisions and with improper hygienic prac- Total avertive expenditures for Nicaragua are esti- tices (Fewtrell, Prüss-Üstün et al. 2007). Further- mated to range from 80 to 150 million NIO (Table more, the consequences of malnutrition in children II.4) per year. This represents about 10 to 17 per- under age five, in terms of longer term cognition cent of the total estimated annual cost associated and learning impacts, can also be partly attributed with inadequate water supply, sanitation and hy- to inadequate water, sanitation and hygiene. Re- giene. cent analyses in Ghana and Pakistan have shown The total estimated cost associated with inade- that by adding these indirect impacts (both in the quate water supply, sanitation and hygiene ranges short term and long term), the health costs attrib- from 830 million to 1 billion NIO per year, with a uted to poor water, sanitation and hygiene nearly mean of 915 million NIO (Figure II.1). The damages double (World Bank 2008). Estimating these addi- include health impacts, mortality and morbidity and tional impacts requires much more detailed anal- averting expenditures which mostly include house- yses; it is recommended as follow-up work but has hold boiling of drinking water. The cost of health not been included within the scope of this study. Figure II.1: Annual Costs of Inadequate Water, Sanitation and Hygiene (million NIO) 500 Urban cost Rural cost 400 300 Million NIO 200 100 0 Diarrheal Diarrheal Bottled water Boiling of Chlorination Filtration Morbidity Mortality consumption drinking water 10 Filter distribution and pricing data are from communications with B. Castro. MINSA, March 2009. 16 using information about incremental service in- Water, Sanitation, and Hygiene Interventions crease and corresponding investment needs from Interventions related to improving the quality and the Estrategia Sectorial de Agua Potable y Sa- quantity of water supply, improving access to im- neamiento 2005. This is based on a ten percent proved sanitation, and programs encouraging annual discount rate, an annual five percent for op- better hygiene practices (including hand washing) eration and maintenance (O&M) and five percent have potential health benefits. Based on studies for promotion/water source protection costs, and from several countries around the world, estimates US$6 in annual sewage costs.11 of the benefits associated with different interven- Benefits associated with improvements in water tions to reduce damage arising from unsafe water supply and sanitation include reductions in mor- and sanitation have been developed (see Annex A tality and morbidity from diarrheal diseases (see for details). These international studies estimate Table II.5). Improved water supply is estimated the percentage reductions in the incidence of di- to avert 200,000 diarrheal cases (and 40 child arrheal morbidity and mortality following interven- deaths) per year in rural areas, and 70,000 diar- tions to improve the infrastructure that provides rheal cases (and 12 child deaths) per year in urban drinking water or basic sanitation, or interventions areas. Similarly, improvements in sanitation are es- that improve personal hygiene. The results are ap- timated to avert 300,000 diarrheal cases (and 60 plied to available Nicaraguan health data and to child deaths) per year in rural areas, and 100,000 estimates of the costs of providing such improve- diarrheal cases (and 17 child deaths) per year in ments in rural and urban areas of Nicaragua. urban areas. Details on the assumptions and cal- A. Cost-Benefit Analysis of Water and Sanitation culations can be found in Annex B. The morbidity Improvements costs, based on the costs of treatment and value Providing piped water supply to all rural house- of lost time, are US$8–14 per case of diarrhea. The holds is likely to be very expensive. A realistic ob- mortality costs are calculated based on the Human jective might be to at least provide improved water Capital Approach (HCA). supply (protected wells or boreholes) and sanita- The analysis estimates that interventions (pro- tion facilities (improved pit latrines or pour-flush la- grams) that improve sanitation in Nicaragua will trines). The aim of the infrastructure interventions have a benefit-to-cost ratio greater than 1.0 when is to improve water supply and sanitation, largely in the time savings of improved water are included. rural areas. Two programs are analyzed: one that For water supply programs, the benefit cost ratio provides 1.5 million people with improved sanita- is less than 1. For water supply/sanitation pro- tion, and one that provides 1.2 million people with grams in urban areas, the costs are higher and the an improved water supply both in rural and urban benefits are lower due to lower diarrheal mortality areas. among children under age five and lower diarrheal The annualized investment costs per capita are prevalence. Marginal cost12 would be higher than estimated at about US$12–19 for improved sani- marginal benefits13 (health damage reduction) for tation and US$30–40 for improved water supply, all four investment programs considered. 11 Per capita investment costs represent average costs in South America (WHO/UNICEF 2000). 12 Approximated by average cost per one percent of WSSH cost reduction. 13 Value of one percent of WSSH cost reduction, estimated at about 9 million NIO. 17 Environmental Health in Nicaragua Table II.5: Costs and Benefits of Reductions in Diarrheal Morbidity and Mortality Rural Areas Urban Areas Improved Improved Improved Improved Sanitation Water Sanitation Water Facilities Supply Facilities Supply Population (million) receiving improved sanitation 1.0 0.5 Population (million) receiving improved water supply 0.9 0.2 % reduction in diarrheal illness /person (Fewtrell and Colford 32% 25% 32% 25% 2004) Diarrheal cases (million) averted per year 0.3 0.2 0.1 0.07 Deaths in children averted per year 60 40 17 12 Annual Health Benefits of Improved Services (US$ million) 4 3 2 1 Annualized Costs of Service Provision (US$ million) 12 24 9 10 Benefit-Cost Ratio (health benefits only) 0.35 0.1 0.2 0.1 Marginal cost (million NIO per 1% reduction) in WSSH health cost 26 69 45 70 B. Benefit-Cost Analysis of Hygiene Improve- 0.6 in the “high� scenario. These ratios are higher ments for the urban population since medical treatment In many studies, the single most effective hygiene costs and annual wages are higher in urban areas. intervention is found to be hand washing after def- Apart from children under age five, the general ecation, before preparing meals, and before eating. population over age five can also benefit from a For Nicaragua, this analysis estimates the benefits hand-washing program. However, the benefit-cost and costs of hand-washing programs in both rural ratios are found to be below one, primarily due to and urban areas. The costs of improved hand- the substantially lower diarrheal incidence rates in washing practices include (i) the costs of program the sub-population above age five. preparation and implementation, and (ii) the pri- C. Benefit-Cost Analysis of Drinking Water Disin- vate (household) costs associated with increased fection water and soap consumption. Benefits from a hand- The benefit-cost analysis for drinking water disin- washing program include the percentage reduction fection interventions (boiling water, which is more in diarrheal illness per child. These benefits are de- common in Nicaragua, and chlorination programs) pendent on the effectiveness and sustainability of has also been estimated. There are no estimates of the hand-washing program in terms of behavioral program costs to promote drinking water disinfec- change at the household and individual levels. As- tion at point-of-use; therefore, the same costs as sumptions made, and details of methodology, can for hand-washing programs (and three scenarios of be found in Annex B. effectiveness ranging from 10 to 20 percent) have In terms of the sustainability of the hand-washing been applied. The program costs are estimated programs, benefit-cost ratios are estimated for per household, while the private cost of boiling three scenarios of behavioral change (low, me- drinking water is estimated at US$5 per year for dium, high). The analysis estimates that the ben- urban households using commercial fuels and efit-cost ratios for hand-washing programs in Nica- about US$3 for rural households using fuelwood ragua range from 4.5 in the “low� scenario to about collected by household members. The cost of water 18 Table II.6: Benefits and Costs of a Hand-Washing Program (Rural and Urban) Rural Households Urban Households with Children with Children Under Age 5 Under Age 5   Low Med High Low Med High Program target (million households)* 0.3 0.3 0.3 0.3 0.3 0.3 Program response (% of households with behavioral change) 10% 15% 20% 10% 15% 20% Percent reduction in diarrheal illness per child 45% 45% 45% 45% 45% 45% (Fewtrell and Colford 2004) Total program cost (US$ million) 0.1 0.4 1.7 0.1 0.4 1.6 Private costs of water and hygiene products per year (US$ 0.1 0.2 0.2 0.1 0.2 0.2 million) Total costs from hand-washing program 0.2 0.6 1.9 0.2 0.6 1.8 Cases of diarrheal illness averted per year (thousands) 30 45 60 30 50 60 Deaths in children averted per year 9 13 18 9 14 19 Total benefits from hand-washing program 0.5 0.8 1.0 0.7 1.0 1.4 Benefit-Cost Ratios             LOW: If behavioral change lasts 1 year 2.2 1.4 0.6 2.9 1.9 0.8 MEDIUM: If behavioral change lasts 2 years** 3.0 2.1 1.0 3.9 2.8 1.3 HIGH: If behavioral change lasts 3 years** 3.4 2.6 1.3 4.5 3.4 1.7 Marginal cost for medium scenario (million NIO per 1% of WSSH 10 7 health cost reduction) * There are about 0.6 million rural children under age five in Nicaragua. It is assumed there is one child under age five in each household (thus the program target is 0.6 million households). However, the estimated benefit-cost ratio is higher for households with more than one child under age five. There are about 0.4 million urban children under age five in Nicaragua. It is assumed there is one child under age five in each household (thus the program target is 0.4 million households). However, the estimated benefit-cost ratio is higher for households with more than one child under age five.** Benefits and costs in the second and third years are discounted at an annual rate of 10 percent. chlorination is estimated at US$3 per household program cost, and with improved hand washing based on an average drinking water consumption only sustained for one year, the benefit-cost ratio is of 0.75 liters per person per day (Lantagne et al. 3.2. Benefit cost ratios are lower, but well above 1 2005). in urban areas. In urban areas, the disinfection pro- In rural areas of Nicaragua, disinfection programs grams are estimated to avert 70,000 to 130,000 are estimated to avert 110,000 to 220,000 cases cases of diarrhea and 10 to 25 deaths in children of diarrhea and 25 to 50 deaths in children per per year. Drinking water chlorination programs in year. The benefit-cost ratio for the central estimate both rural and urban areas have high benefit-cost in water boiling programs is 1.9 and for water chlo- ratios, because associated private costs are low. rination 4.8, corresponding to a 15 percent pro- Detailed tables for disinfection in urban areas, as gram response rate with drinking water disinfec- well as for chlorination programs in both rural and tion sustained for two years. Even for the “high� urban areas, are in Annex B. 19 Environmental Health in Nicaragua Table II.7: Benefits and Costs of a Rural Drinking Water Boiling Program   “Low� “Medium� “High� Target population – rural population not practicing disinfection (millions) 1.5 1.5 1.5 Target households (millions) 0.3 0.3 0.3 Program response (% of households with behavioral change) 10% 15% 20% Percent reduction in diarrheal illness per person (Fewtrell and Colford 2004) 47% 47% 47% Program costs (US$ million) 0.1 0.3 1.4 Private costs of boiling drinking water per year (US$ million)* 0.3 0.4 0.5 Total costs of boiling drinking water 0.4 0.7 1.9 Cases of diarrheal illness averted per year (thousands) 45 65 90 Deaths in children averted per year <15 <20 25 Total benefits from boiling drinking water (US$ million) 0.7 1.1 1.5 Benefit-Cost Ratios       LOW: If behavioral change lasts 1 year 1.9 1.5 0.7 MEDIUM: If behavioral change lasts 2 years** 2.2 1.9 1.1 HIGH: If behavioral change lasts 3 years** 2.3 2.1 1.4 Marginal cost for medium scenario (million NIO per 1% of WSSH health cost reduction) 5 * Estimated based on efficiency of LPG and wood stoves, cost of LPG, fuelwood collection time of 30 minutes per day and 10 percent of fuelwood used for water boiling, and per-person water consumption of 0.75 liter per day. ** Benefits and costs in the second and third years are discounted at an annual rate of 10 percent. percentage point reduction of health damages. Summary Assessment of Interventions The total area below each segment is equal to the Relating to Water, Sanitation and Hygiene total cost of this intervention. For example, the first As part of this analysis, interventions relating to intervention to reduce water pollution is chlorina- sanitation and water supply, various forms of dis- tion of water at the point of use in rural areas. It infection (boiling water and chlorination, as well is depicted by the first segment of Figure II.2. This as hand-washing programs) were examined. The intervention reduces five percent of total health results for these interventions can be summarized damage associated with inadequate water supply, in terms of their contribution to reduced environ- sanitation and hygiene at the cost of 2 million NIO mental damages and the costs per one percent of per one percentage point of this damage reduction. health damage reduction (marginal cost). The total cost of this intervention is 10 million NIO. The segments are ranked in the order of increasing The length of each segment depicted in Figure II.2 unit costs. represents percentage point reduction of health damage (morbidity, mortality) from the implemen- The disinfection and hygiene programs are esti- tation of a corresponding intervention. The height mated to have the greatest potential health bene- of each segment of the cumulative cost curve fits, but only if at least 20 percent of the population equals the unit cost of an intervention per one responds favorably to the program and improves hand-washing practices. Hygiene improvement and 20 disinfection of drinking water at point of use have change being sustained for two years. The ratios a substantial potential to reduce diarrheal illness would be higher (lower) if, as a result of promotion and mortality. However, the challenge is to develop programs, households sustain improved behavior and deliver programs that induce sustained behav- for longer (less) than two years. This figure does not ioral response at a large scale while containing pro- consider the possible interaction effects between gram costs at an affordable level. different interventions (i.e., how the impacts of a The benefit-cost ratios for hand washing and first intervention affect those of a second interven- drinking water disinfection are based on behavioral tion), because data constraints preclude a sound analysis of such effects. Figure II.2: Ranking of Interventions to Reduce Water, Sanitation and Hygiene Costs in Nicaragua 80 Water supply (urban, rural) Marginal costs and marginal benefits (Million NIO) 70 Improved Cumulative 60 sanitation cost Disinfection – (urban) water boiling 50 Household Improved 40 chlorination sanitation (urban) (rural ) 30 Disinfection – water boiling (rural) 20 Hygiene Household chlorination 10 (rural) 0 0 5 10 15 20 25 30 35 40 45 50 Reduction of health costs (% of total health costs in the sector) benefits in excess of costs under most assump- Recommendations for Actions on Water, tions. The programs are also justified because the Sanitation, and Hygiene benefits are concentrated primarily among the From the analysis presented here, it is clear that poor. The highest priority should be given to the in Nicaragua, most measures to improve the water drinking water disinfection and hand-washing pro- supply and sanitation facilities in rural areas yield grams in both rural and urban areas. 21 III. Indoor Air Pollution About 2.4 billion people worldwide burn biomass tion and 32 percent of the urban population use (wood, crop residues, charcoal and dung) for solid fuel for cooking in Nicaragua (ENDESA 2006– cooking and heating. The smoke created from 2007). WHO (Desai et al. 2004) suggests using a burning these fuels turns the kitchens of the world’s ventilation coefficient of 0.25 for households that poorest countries into death traps. Indoor air pollu- use improved stoves or have outside kitchens. tion from the burning of solid fuels kills over 1.6 The national survey of solid fuel use in Nicaragua million people, predominantly women and children, (ENDL 2006–2007) estimates that kitchens are lo- each year (World Bank 2002b). Smoke in the home cated outside the house in 50 percent of rural and is one of the world’s leading child killers, claiming 75 percent of urban households. the lives of nearly one million children each year Health effects: Acute respiratory infections (ARIs) (ITDG 2004). The strongest links between indoor and COPD are the most common diseases asso- smoke and health are for lower respiratory infec- ciated with indoor air pollution in Nicaragua. The tions, chronic obstructive pulmonary disease vulnerable subgroups in the population include (COPD), and cancer of the respiratory system. Of children under age five (ARI mortality and morbidity all the respiratory diseases associated with indoor among children) and women over age 30 (COPD smoke, lower respiratory infections account for mortality and morbidity; ARI morbidity), because about 37.5 percent, COPD accounts for 22 COPD, these subgroups are exposed to smoky kitchens. and cancer of the respiratory system accounts for The health estimates attributed to indoor air pol- about 1.5 percent globally (WHO 2002b). lution are based on background health data on Rural energy use: Indoor air pollution is associated ARI and COPD prevalence, available from ENDESA with use of traditional fuels (mainly cotton stalks in 2006–2007 and international sources, taking into Nicaragua). About 92 percent of the rural popula- account odds ratios (see Annex A for details). Table III.1: Estimated Annual Health Effects of Indoor Air Pollution Rural areas Urban Areas “Low� “High� “Low� “High� Mortality: Acute respiratory infections in children under age 5 140 200 40 70 Chronic obstructive pulmonary disease in women over age 30 45 70 20 50 Morbidity: Acute respiratory infections in children under age 5 586,900 1,049,800 208,000 627,200 Acute respiratory infections in women over age 30 284,800 412,700 161,800 273,900 Chronic obstructive pulmonary disease 1,280 2,060 530 1,160 23 Environmental Health in Nicaragua Each year, an estimated 140 to 200 children under can be attributed to indoor air pollution each year age five die from ARIs in rural areas, and an addi- (Table III.1). tional 40 to 70 children die in urban areas in Nica- 66. Economic costs: The total estimated annual ragua. Among children under age five, more than cost of indoor air pollution ranges from 540 mil- half a million annual cases of acute respiratory lion to 1.2 billion NIO with a mean cost of 870 mil- infections in rural areas, and more than 200,000 lion NIO (Table III.2). These costs account for about cases in urban areas, can be linked to indoor air 0.83 percent of GDP in 2007. ARI in children repre- pollution. Among females over age 30, each year sents 40 percent of cost; respiratory child mortality there are nearly 285,000 cases of indoor-air-pollu- represents 12 percent; COPD mortality in adult fe- tion-related ARI morbidity in rural areas and nearly males and ARI morbidity in adult females represent 162,000 cases in urban areas. Indoor air pollution 11 and 12 percent of the total costs, respectively. also causes COPD in females over age 30: nearly The rural poor are especially vulnerable and ac- 120 women die annually from COPD in urban and count for 56 percent of the total costs of indoor air rural areas, and about 3,000 new cases of COPD pollution. Table III.2: Estimated Annual Costs (million NIO) of Indoor Air Pollution in Nicaragua   “Low� “High� Acute Respiratory Illness (ARI): Children (under age 5) – increased mortality 104 157 Children (under age 5) – increased morbidity 249 558 Adult females – increased morbidity 158 248 Chronic obstructive pulmonary disease (COPD): Adult females – increased mortality 12 212 Adult females – increased morbidity 17 30 Total 539 1205 In order to estimate the cost of COPD mortality for adults, the Value of Statistical Life (VSL) is used as Indoor Air Pollution Interventions the upper bound and HCA as the lower bound (see A wide range of interventions are available to re- Annex A for details on this methodology). In the duce indoor air pollution and associated health case of children, HCA is used to estimate the cost effects. These interventions can be classified ac- of mortality. The cost of morbidity includes the cost cording to the level at which they are effective: (a) of illness (medical treatment and value of time lost interventions on the source of pollution, including due to illness). moving from traditional stoves to improved stoves, 24 and switching to cleaner fuels such as LPG; (b) Five intervention scenarios considered for Nica- interventions on the living environment, such as ragua represent stylized situations commonly chimneys and smoke hoods (with flues); and (c) found in most developing countries representing interventions on user behavior, such as keeping reduction of pollution loads from solid fuel use. young children away from smoke. For purposes However, actual pollution exposure can vary sub- of this economic analysis for indoor air quality in stantially in each scenario, and depends on addi- Nicaragua, only those interventions relating to the tional factors such as household ventilation as well source of pollution have been considered, that is, as other characteristics and household behavior moving from unimproved to improved stoves, better (Table III.3). ventilation, and switching to cleaner fuels. Table III.3: Interventions Scenarios: Description: From I (50% HH) to II From unimproved stove inside to better ventilation From I (50% HH) to III From unimproved stove inside to improved stoves From III to V From improved stove to LPG From IV to VI From unimproved stove and LPG mix to LPG only From V to VI From improved stove and LPG mix to LPG only In the benefit-cost analyses of selected interven- tions, improved ventilation and replacement of Summary Assessment of Interventions to unimproved stoves with and improved ones in indi- Address Indoor Air Pollution vidual households are found to have substantially The range of interventions is summarized in terms higher benefits than costs. The benefit-cost ratio is of their contribution to reduced environmental estimated at above 3 for all four of these interven- damages and costs per one percent of health tions. The benefit-cost ratio is lower than 1 if there damage reduction (marginal cost)14 (Figure III.1). is switching to LPG due to high LPG prices. For Similar to the case of water pollution, the length households with improved stoves, the health ben- of each segment depicted represents a percentage efits alone are not large enough to outweigh the point reduction of health damage (morbidity, mor- cost of switching to LPG. Although promotion of im- tality) from the implementation of a corresponding proved stoves is a very attractive intervention, the intervention. The unit costs reflect only program merits of promoting LPG in individual rural house- costs and private household costs, without taking holds are uncertain. LPG prices would have to be into account time and fuelwood savings. reduced dramatically for the estimated benefits to Improved ventilation and household replacement exceed costs. Therefore, it appears that LPG will of unimproved stoves with improved stoves re- have a chance of success only in better-off house- sults in the largest reduction in damages in rural holds. areas. This is followed by households switching 14 Approximated by the average cost to implement intervention per one per cent of IAP cost reduction. 25 Environmental Health in Nicaragua Table III.4: Benefits and Costs of Indoor Air Pollution Control in Nicaragua Rural Areas Urban Areas Better Improved Better Improved ventilation stoves from LPG from ventilation LPG from stoves from from unimproved improved from improved unimproved unimproved stoves stove unimproved stove stoves inside stove inside inside stove inside Population receiving intervention (million) 0.5 0.5 0.07 0.1 0.1 0.03 ARI cases averted/year (thousand) 400 400 50 80 80 20 ARI deaths in children averted/year 36 36 5 10 10 <5 COPD cases averted per year 110 110 20 50 50 15 COPD deaths averted per year 10 10 <5 <10 <10 <5 Annual health benefits (million NIO) 96 96 14 50 50 10 Program cost (million NIO) 13 13 2 4 4 1 Annualized stove cost (million NIO) 3 22 4 1 6 2 Annual cost of LPG (million NIO) 0 0 40 0 0 20 Total annual costs (million NIO) 18 37 47 5 10 23 Benefit-cost ratio (health benefits only)* 6.0 2.7 0.3 10.6 4.9 0.5 Marginal cost (million NIO per 1% of IAP health cost 1.5 3.2 30 0.8 1.8 17 reduction) Note: *rounded off to LPG alone from improved stoves in urban and rural areas. Each of these measures contributes Recommendations for Actions a smaller amount of reduction in environmental on Indoor Air Pollution damages; marginal costs for these two interven- The analysis presented here recommends shifting tions slightly exceed marginal benefits. In total, the households who have unimproved stoves to im- first four interventions reduce the cost of health proved ones, and to improve ventilation. The rela- effects by about 37 percent per year. This reflects tive attractiveness of other interventions (such as better ventilation and switching to improved stoves switching from unimproved stoves to LPG or from in 43 percent of rural households and in 8 percent improved stoves to LPG) is much more uncertain of urban households, and switching to LPG from and depends substantively on the costs and ben- improved stoves in 3 percent of rural and 1 percent efits used. of urban households. 26 Figure III.1: Ranking of Interventions for Indoor Air Pollution Cost Reduction in Nicaragua Unimproved wood 35 stoves to gas stoves (rural) Marginal costs and marginal benefits (Million NIO) 30 Unimproved Cumulative wood stoves to cost 25 gas stoves (urban) 20 15 Unimproved Unimproved to improved Improved to improved wood stoves 10 Improved ventilation wood stoves (rural) ventilation (rural) (urban) (urban) 5 0 0 5 10 15 20 25 30 35 40 Reduction of health costs (% of total health costs in the sector) 27 iv. Urban Air Pollution Air quality: Managua, the capital of Nicaragua, has city-specific data are available. Furthermore, the a population of about 1.2 million. It is located in estimates relating to health effects from urban air the western part of Nicaragua on the southwestern pollution relate to only Managua and other cities shore of Lake Managua. Urban (outdoor) air pol- with populations over 100,000. Although the urban lution is emerging as a major problem in Nicara- population in other smaller cities in Nicaragua is gua’s urban centers, particularly in Managua. Total also exposed to ambient air pollution, these cities urban population exposed to air pollution is about were not included in the analysis due to the lack 1.95 million, or 62 percent of the total Nicaraguan of monitoring data and estimated concentrations. population15 in 2007. Data relating to urban air MARENA provided the monitored data for annual quality concentrations are extremely limited and average concentrations of major pollutants in Ma- very dated (last available in 2001, see Table IV.1). nagua in 1995–2001. No monitoring of PM10 and There is no emissions inventory, and very little TSP was conducted after 2001. Table IV.1: Measured Average Annual Concentration of Pollutants in Managua (1996–2001) Units 1996 1997 1998 1999 2000 2001 Average Standard NO2 µg/m3 40 34 31 34 22 23 31 40 Ozone µg/m 3 88 71 75 119 42 23 70 60 PM10 µg/m 3 66 67 62 52 65 70 64 50 TSP µg/m 3  n/a 288 341 241 300 253 285 75 Source: MARENA 2009 15 Estimated from http://www.mongabay.com/igapo/2005_world_city_populations/Nicaragua.html, adjusted to 2007 by applying the urban population growth rate in Nicaragua from http://www.makingcitieswork.org/files/pdf/latin-am-carribean/Nicaragua.pdf for all cities with populations over 100,000. 29 Environmental Health in Nicaragua Table IV.2: Estimated Health Impact of Urban Air The annual health effects of ambient particulate Pollution air pollution in Nicaragua are calculated using the Health end-points Total Cases information on applicable concentration-response Premature mortality 429 coefficients that link annual average PM pollution with additional cardiopulmonary mortality and Chronic bronchitis 514 various morbidity end-points (for details of the Hospital admissions 1,352 methodology, see Annex A). Urban particulate air Emergency room visits/ pollution is estimated to cause around 420 prema- 26,518 Outpatient hospital visits ture deaths annually. The number of new cases of Restricted activity days 4,478,531 chronic bronchitis is estimated at about 520 per Lower respiratory illness in year. Annual hospitalizations due to pollution are 58,749 children estimated at close to 630, and annual emergency Respiratory symptoms 14,253,413 room visits/outpatient hospitalizations at 26,500. Cases of less severe health impacts are also pre- Vehicle growth: However, since the number of ve- sented in Table IV.2. In terms of annual DALYs lost, hicles nearly doubled in Nicaragua in 2000–2007 mortality accounts for an estimated 45 percent, (Figueroa de la Vega 2009) and the transportation chronic bronchitis around 16 percent of the total, fleet is a major polluting source in the urban areas, Restricted Activity Days (RADs) 20 percent, and re- annual average concentrations of PM10 have not spiratory symptoms 15 percent. decreased since 2001. Using the latest available monitoring data, and then adjusting them in rela- Economic costs: Increasing air pollution in Nica- tion to World Bank estimates, the annual average ragua’s urban areas, especially in Managua, is PM10 concentration for Managua was estimated at imposing a negative economic impact from pre- 67 µg/m3; in other Nicaraguan cities with popula- mature deaths, illness, medical costs and lost pro- tions over 100,000 it was estimated at 43 µg/m3.16 ductivity. Costs of health impacts from particulate matter have been assessed, and health conditions Health impacts: The most significant health im- such as premature mortality, hospital admissions, pact of outdoor air pollution has been associated restricted activity days, and emergency visits have with particulate matter and, to a lesser extent, with been considered. In the absence of proper data on ground-level ozone. Particles smaller than 10 mi- treatment costs, informed estimates have been crons in size (PM10), and especially those smaller provided by medical experts in Managua. than 2.5 microns (PM2.5), penetrate deep into human lungs and cause health impacts such as The mean estimated annual cost of urban air pol- acute respiratory infection (both upper and lower lution due to PM ranges from 305 million to 1.25 respiratory tract infections), COPD (especially bron- billion NIO, with a mean of is about 780 million NIO. chitis), asthma attacks, cardiovascular disease and This represents about 0.7 percent of the country’s lung cancer. Certain population subgroups, such as GDP. Around 70 percent of the cost is due to mor- the elderly, children, and individuals with existing tality, and the remaining 30 percent is associated respiratory or cardiovascular diseases, are at in- with morbidity (Table IV.3). The cost of mortality, creased risk from exposure to particulate matter. based on the Human Capital Approach (HCA) and the Value of Statistical Life (VSL), ranges from 1 16 www.worldbank.org/nipr/Atrium/mapping.html.url. 30 Table IV.3: Estimated Annual Cost of Health Impacts (Million NIO) % of Total Cost Health Categories Total Annual Cost (Mean) Mortality 55–1,000 68% Morbidity: Chronic bronchitis 7 1% Hospital admissions 6 1% Emergency room visits/Outpatient hospital visits 15 2% Restricted activity days (adults) 202 26% Lower respiratory illness in children 20 3% Respiratory symptoms (adults) 0 0% Total costs of morbidity 250 32% Total costs (mortality and morbidity) 305–1,250 100% billion to 55 million NIO. The measure of the wel- is complex, and estimates of the benefits and costs fare cost of morbidity is often based on the will- of control options therefore need to be accompa- ingness-to-pay (WTP) to avoid or reduce the risk of nied by a careful sensitivity analysis of key parame- illness. It is often found to be several times higher ters of the estimates. Because Nicaragua lacks an than the cost of medical treatment and the value emissions inventory, it is difficult to conduct these of time losses (Cropper and Oates 1992), and re- related economic analyses. flects the value that individuals place on avoiding At a general level, interventions to control primary pain and discomfort. However, there is an insuffi- and secondary particulate emissions from mobile cient number of WTP studies from Central America. sources can be broadly classified into: (a) market- Thus, the Cost-of-Illness (COI) approach (mainly based instruments, such as fuel pricing and taxa- medical costs and value of time losses) has been tion, vehicle taxation, and emission taxes; (b) applied to estimate morbidity (see Annex A). vehicle technology standards and regulations, in- cluding in-fleet technology retrofitting, inspection and maintenance programs, as well as auto import Summary Assessment of Actions regulations; (c) fuel quality improvements and fuel on Urban Air Pollution use regulations, such as low-sulfur diesel and con- Health effects of air pollution are a function of am- version to compressed natural gas; and (d) traffic bient air quality. Estimating the benefits of urban management and urban planning, including public air pollution control therefore requires: (a) an emis- transportation policies (Larsen 2005). Several of sions inventory, (b) a relationship between the in- these measures may be potential options for the ventory and ambient air quality, and (c) an estimate Government of Nicaragua to consider in light of of emissions reductions from individual pollution worsening air quality in its cities. control measures. Each of these three dimensions 31 Environmental Health in Nicaragua An air quality indicator system should be estab- Recommendations for Actions lished to monitor emissions in the main urban on Urban Air Pollution areas of Managua and León. Moreover, as part of Rapidly increasing urban populations and accom- improving the Nicaraguan air quality management panying vehicle growth, coupled with inadequate system, a detailed and integrated emissions inven- regulations and monitoring, will impact the air tory for Managua and León should be developed. A quality in cities such as Managua unless appro- World Bank analytical study ( funded by TFESSD) is priate actions are taken. The lack of adequate under way to carry out an emissions inventory for data on air quality in Nicaragua cities, and of sub- Managua and León. Results from this work will be sequent analysis about the sources of urban pol- very useful inputs to determine the specific kinds of lution, constrain the analysis of potential interven- interventions that can address urban air pollution. tions to address this issue. 32 v. Conclusions Several key messages have emerged from the pro- it is recommended that the government invest in cess of putting together this study and its results: such programs, in both in rural and urban areas. (i) environmental health risks impose a significant However, infrastructure investments to expand ac- burden on Nicaragua’s economy, amounting to 2.6 cess to water and sanitation are also extremely billion NIO, or 2.4 percent of the country’s GDP, and important, especially in rural Nicaragua, because resulting in premature deaths and infections, espe- they also provide significant non-health benefits cially in children under age five; (ii) cost-effective in- (convenience, time-savings, etc.). terventions to address these environmental health Interventions that aim to improve air quality risks exist and should be prioritized in Nicaragua; through improved cookstoves, switching to cleaner (iii) country-specific health and environmental fuels, and better ventilation can help reduce the in- data are somewhat limited, especially in the case cidence of respiratory diseases in Nicaragua, espe- of air quality, and data collection and monitoring cially among women and young children. This anal- need to be further strengthened; and (iii) the ca- ysis recommends that the GoN invest in improved pacity of MARENA and MINSA staff to carry out en- cookstove programs, especially for the numerous vironmental health costing analyses needs to be rural households where unimproved stoves are cur- strengthened through proper training. rently being used. Switching to cleaner fuels (such Diarrheal and other waterborne diseases result as LPG) is also recommended; however, the GoN from inadequate water, sanitation and hygiene in needs to review the relative costs of these fuels to Nicaragua. The significant economic burden at- ensure affordability. tributed to health costs from these diseases can Rapidly increasing urban populations and accom- be alleviated through various water, sanitation panying vehicle growth, coupled with inadequate and hygiene interventions. For Nicaragua, water regulations and monitoring, are continuing to im- disinfection and hygiene programs are estimated pact the air quality in cities such as Managua. How- to have the greatest potential health benefits, and ever, the lack of adequate data and the absence of 33 Environmental Health in Nicaragua an emissions inventory make it difficult to identify health risks and contribute to a healthier Nicaragua. the appropriate interventions for improving urban As part of this study, the Bank team helped to build air quality in Nicaragua. The analysis endorses the capacity within MINSA and MARENA to carry out ongoing work to carry out an emissions inventory such environmental health costing work in future for Managua and León, and recommends that the exercises. Apart from the ongoing analytical work GoN take a more integrated approach to air quality on developing an emissions inventory, the GoN management. has also expressed interest in a study that would Finally, discussions with the GoN and other stake- estimate health costs associated with dengue and holders have provided feedback about the useful- identify appropriate environmental management ness of such analyses in helping to identify the ap- solutions for Nicaragua. propriate interventions to address environmental 34 Annex A. Methodology for Health Cost Estimation This annex provides technical details on the meth- consequences because they can react with other odology used to estimate health impacts and substances in the atmosphere to form particulates. subsequent costs associated with environmental Dose Response Coefficients health risks (urban air pollution, inadequate water In order to estimate health impacts, we have used supply and sanitation, and indoor air pollution). work by Ostro (1994), Abbey et al. (1995) and Pope et al. (2002). Ostro and Abbey et al. used dose re- sponse coefficients to analyze morbidity effects Methodology: Urban Air Pollution and provide estimates of dose response of PM Particulate Matter Pollution (PM10) to chronic bronchitis. A survey of the current Research in the United States in the 1990s and status of worldwide research shows that the risk most recently by Pope et al. (2002) provides strong ratios or dose-response coefficients by Pope et al. evidence that it is even smaller particulates (PM2.5) (2002) are likely to be the best available evidence that have the greatest health effects. Gaseous pol- of the mortality effects of ambient particulate pol- lutants (SO2, NOx, CO and ozone) are generally not lution (PM2.5). These coefficients were applied thought to be as damaging as fine particulates. by WHO in the World Health Report 2002, which However, SO2 and NOx may have important health provided a global estimate of the health effects of Table A.1: Urban Air Pollution Dose-Response Coefficients Per 1 µg/m3 annual Dose-response Annual Health Effect average ambient Coefficient concentration of: Mortality (% change in cardiopulmonary and lung cancer 0.8% PM2.5 mortality) Chronic bronchitis (% change in annual incidence) 0.9% PM10 Respiratory hospital admissions (per 100,000 population) 1.2 PM10 Emergency room visits (per 100,000 population) 24 PM10 Restricted activity days (per 100,000 adults) 5,750 PM10 Lower respiratory illness in children (per 100,000 children) 169 PM10 Respiratory symptoms (per 100,000 adults) 18,300 PM10 Source: Pope et al. (2002) for the mortality coefficient. Ostro (1994) and Abbey et al. (1995) for the morbidity coefficients. environmental risk factors. The mortality and mor- relationship between air pollution and mortality. bidity coefficients from the above studies are pre- The study confirms and strengthens the evidence sented in Table A.1. of the long-term mortality effects of particulate Pope et al. (2002) provide the most comprehen- pollution found by Pope et al. (1995) and Dockery sive and detailed research study to date on the et al. (1993). Pope et al. (2002) used ambient air 35 Environmental Health in Nicaragua quality data from metropolitan areas across the levels of PM2.5 and all other causes of mortality (see United States for 1979–1983 and 1999–2000, Table A.2). and information on certified causes of mortality The share of cardiopulmonary and lung cancer of adults from the American Cancer Society (ACS) deaths in total mortality sometimes varies substan- database over a period of 16 years. The ACS data- tially across countries. Therefore, it may reasonably base has individual specific information for more be expected that the risk ratios for cardiopulmo- than one million adults that was obtained through nary and lung cancer mortality provide more reli- questionnaires. The study could therefore control able estimates of mortality from PM2.5 than the risk for a large set of factors that may also affect varia- ratio for all-cause mortality when the risk ratios are tions in mortality rates across metropolitan areas, applied to countries other than the United States. such as age, smoking behavior, education, marital The former two risk ratios are therefore used in status, body weight, occupational risk factors and this report. The mortality coefficient in Table A.1 dietary indexes. is a combination of the cardiopulmonary and lung The study found a statistically significant relation- cancer mortality risk ratios in Table A.2. ship between levels of PM2.5 and mortality rates, In order to use the mortality coefficients in Table A.1 controlling for all the factors discussed above. All- to estimate mortality from urban air pollution in Ni- cause mortality was found to increase by 4–6 per- caraguan cities, baseline data on total annual car- cent for every 10 µg/m3 increase in PM2.5. The in- diopulmonary and lung cancer deaths are required. crease in cardiopulmonary mortality was 6–9 per- Data from the Nicaraguan Ministry of Health, com- cent, and 8–14 percent in lung cancer. No statis- bined with WHO data by cause of deaths, were also tically significant relationship was found between used. Data from the above two sources showed an Table A.2: Mortality Risk Associated with a 10 µg/m3 Change in PM2.5 Adjusted Relative Risk Ratios (RR) Cause of Mortality 1979–1983 1999–2000 Average All-cause 1.04 1.06 1.06 Cardiopulmonary 1.06 1.08 1.09 Lung cancer 1.08 1.13 1.14 All other cause 1.01 1.01 1.01 Reproduced from Pope et al. (2002). average rate of 30 percent of total deaths due to An estimate of annual incidence of chronic bron- cardiopulmonary and lung cancer. To estimate the chitis (CB) is required in order to apply the CB coef- mortality effects, following the procedure used by ficient in Table A.1. In the absence of data on CB the 202 World Health Report (WHO), a threshold incidence for Nicaragua, we have used the rate level of 7.5 µg/m3 of PM2.5 has been applied, below from WHO (2001) and Shibuya et al. (2001) for which it is assumed there are no mortality effects. the AMRO D region of WHO, of which Nicaragua is No threshold level has been applied for morbidity. a part. Therefore, it has not been possible to use city-specific CB incidence rates. Other morbidity 36 health end-points considered are hospital admis- tion subgroup that is asthmatic and the frequency sions of patients with respiratory problems, emer- of asthma attacks is not readily available for Nica- gency room visits (or hospital outpatient visits), ragua. restricted activity days, lower respiratory infections The health effects of air pollution are then con- in children, and respiratory symptoms. The coeffi- verted to disability adjusted life years (DALYs) to cients are expressed as cases per 100,000 in the facilitate a comparison to health effects from other absence of incidence data for Nicaragua, though environmental risk factors. DALYs per 10,000 it would be preferable to have incidence data and cases of various health end-points are presented use coefficients that reflect the percentage change in Table A.3. in incidence. Increases in asthma attacks among Table A.4 presents the disability weights and av- asthmatics have also been related to air pollution erage duration of illness used in this report to calcu- in many studies. But these data are on the popula- late DALYs, as presented in Table A.5. The weights Table A.3: DALYs for Health Effects for lower respiratory illness (LRI) and chronic bron- DALYs chitis (CB) are average disability weights.17 Dis- lost per ability weights for the other morbidity end-points Health Effect 10,000 are not readily available, and are estimated by cases Larsen (2004a) based on weights for other com- Mortality 75,000 parable illnesses.18 Average duration of CB is esti- Chronic bronchitis (adults) 22,000 mated based on age distribution in Nicaragua and Respiratory hospital admissions 160 age-specific CB incidence in Shibuya et al. (2001). Emergency room visits 45 Years lost to premature mortality from air pollu- Restricted activity days (adults) 3 tion is estimated from age-specific mortality data Lower respiratory illness in children 65 for cardiopulmonary and lung cancer deaths, and have been discounted at three percent per year. Av- Respiratory symptoms (adults) 0.75 Note: DALYs are calculated using a discount rate of 3 percent and full erage duration of illness for the other health end- age weighting based on WHO tables. Estimates of DALYs for the mor- points is from Larsen (2004a). bidity end-points are from Larsen (2004a, b). Table A.4: Calculation of DALYs Per Case of Health Effects Disability Weight Average Duration of Illness Mortality 1.0 (7.5 years lost) Lower respiratory Illness – children 0.28 10 days Respiratory symptoms – adults 0.05 0.5 days Restricted activity days – adults 0.1 1 day Emergency room visits 0.30 5 days Hospital admissions 0.40 14 days* Chronic bronchitis 0.2 20 years * Includes days of hospitalization and recovery period after hospitalization. 17 See: http://www.dcp2.org/pubs/GBD 18 The disability weight for mortality is 1.0. 37 Environmental Health in Nicaragua Estimated Health Impacts opportunity cost of time lost due to illness. In the The total annual costs associated with the health case of non-income-earning individuals, the above effects described above were computed from the rate is justified because most of these individuals total cost per case (see Table A.5) and the number provide a household function that has a value as- of estimated cases. For morbidity, the cost of ill- sociated with it; equally, they could choose to join ness alone has been estimated. In the case of re- the paid labor force. spiratory symptoms, the estimated cost per occur- Overall, there is very little information about the rence is zero and hence the total annual estimated frequency of doctor visits, emergency visits and cost due to respiratory symptoms in Table A.5 is hospitalization for CB patients in developing coun- zero. tries. However, Schulman et al. (2001) and Nie- Table A.5: Estimated Unit Cost by Health End-point derman et al. (1999) provide some information on frequency of doctor visits, emergency visits and Cost-of-Illness Health categories hospitalization for CB patients in the United States Per Case (NIO) and Europe.20 In the absence of country-specific Chronic bronchitis 13,060 data, results from the above studies have been ap- Hospital admissions 4,800 plied to Nicaragua. Lost workdays per year are es- Emergency room visits/ timated based on frequency of estimated medical 560 Outpatient hospital visits treatment plus an additional seven days for each Restricted activity days hospitalization and one extra day for each doctor 330 (adults) visit and emergency visit. These days are added to- Lower respiratory illness in 45 gether to reflect the total time needed for recovery children from illness. Respiratory symptoms 0 To estimate the cost of a new case of CB, the (adults) medical cost and value of time losses have been discounted over a period of 20 years of illness. An The baseline data used to estimate the cost per annual real increase of two percent in medical cost case of illness is presented in Table A.6. The oppor- and value of time lost has been applied to reflect tunity cost of time lost because of illness for adults an average expected increase in annual labor pro- is estimated based on urban wages. The average ductivity and real wages. The costs are discounted urban wage in Nicaragua is about 5,000 NIO per at three percent per year, a rate commonly applied month.19 In the case of both income-earning and by WHO for health effects. non-income-earning individuals, 75 percent of the average urban wage rate has been imputed as the 19 http://www.bcn.gob.ni/estadisticas/indicadores/3-4.htm 20 CB is a major component of COPD, which is the focus of the referenced studies. 38 Table A.6: Baseline Data for Estimation of Cost of Health Impact Due to Air Pollution Baseline Source: Cost Data for All Health End-Points: Cost of hospitalization (NIO per hospitalization) 500 Per consultations with Cost of emergency visit (NIO) – urban 200 medical service providers Cost of doctor visit (NIO) (mainly private doctors) – urban 150 and health authorities 75% of urban wages in Value of time lost to illness (NIO per day) 180 Nicaragua Chronic Bronchitis (CB): Based on Shibuya et al. Average duration of Illness (years) 20 (2001) % of CB patients being hospitalized per year 1.5% Average length of hospitalization (days) 10 From Schulman et al. (2001) Average number of doctor visits per CB patient per year 1 and Niederman et al. (1999) % of CB patients with an emergency doctor/hospital 15% outpatient visit per year Estimated based on Estimated lost work days (including household work days) per frequency of doctor visits, 2.6 year per CB patient emergency visits, and hospitalization Annual real increases in economic cost of health services 2% Estimate and value of time (real wages) Applied by WHO for health Annual discount rate 3% effects Hospital Admissions: Average length of hospitalization (days) 6 Average number of days lost to illness (after hospitalization) 4 Emergency Room Visits: Average number of days lost to illness 2 Estimates Restricted Activity Days: Average number of days of illness (per 10 cases) 2.5 Lower Respiratory Illness in Children: Number of doctor visits 1 Estimated at 1–2 hours per Total time of caregiving by adult (days) 1 day 39 Environmental Health in Nicaragua among infant and child mortality rates and piped Methodology: Inadequate Water Supply, water supply, flush toilet, maternal education, ac- Sanitation and Hygiene cess to electricity, medical services, oral rehydra- Inadequate water supply, sanitation and hygiene tion therapy (ORT), vaccination, dirt floor in house- result in a variety of diseases, such as diarrhea, hold dwelling, fertility rates and malnutrition. Simi- that result in deaths (predominantly in children larly, Larsen (2003) provides a regression analysis under age five) as well as illness. Esrey et al. of child mortality, using national data for year 2000 (1991) provide a comprehensive review of studies from 84 developing countries representing 95 per- documenting this relationship for diseases such as cent of the total population in the developing world. schistosomiasis (bilharzia), intestinal worms, diar- A statistically significant relationship between child rhea, etc. Fewtrell and Colford (2004) provide a mortality and access to improved water supply, meta-analysis of studies of water supply, sanitation safe sanitation, and female literacy is confirmed. and hygiene that updates the findings on diarrheal Baseline health data for estimating the health im- illness by Esrey et al. (1991). Water, sanitation and pacts of inadequate water supply, sanitation and hygiene factors also influence child mortality. Esrey hygiene are presented in Table A.7. WHO estimates et al. (1991) find in their review of studies that the that 12 percent of child mortality was due to diar- median reduction in child mortality from improved rhea in 2006. This is used for diarrheal mortality water and sanitation is 55 percent. Shi (1999) pro- estimation. In the case of diarrheal morbidity, it is vides econometric estimates of the impact of po- very difficult or practically impossible to identify all table water and sewerage connection on child mor- cases of diarrhea because a substantial share of tality, using a data set for about 90 cities around cases are not treated or do not require treatment the world. Literacy and education level are also at health facilities, and therefore are never re- found to be important determinants of parental corded. Furthermore, those cases that are treated protection of child health against environmental by private doctors or clinics are often not reported risk factors. Esrey and Habicht (1988) report from to public health authorities. Therefore, household a study in Malaysia that maternal literacy reduces surveys provide the most reliable indicator of total child mortality by about 50 percent in the absence cases of diarrheal illness. At the same time, the of adequate sanitation, but only by 5 percent in the surveys only reflect diarrheal prevalence at the presence of good sanitation facilities. time of the survey. Because there is often high vari- Literacy is also found to reduce child mortality by ation in diarrheal prevalence across seasons of the 40 percent if piped water is present, suggesting year, extrapolation to an annual average will result that literate mothers are better at taking advantage in either an over- or underestimate of total annual of water availability for hygiene purposes to protect cases; correcting this bias is often difficult without child health. Findings from the Demographic and knowledge of seasonal variations. Health Surveys from various developing countries ENDESA 2006–2007 provides data on diarrheal further confirm the role of literacy in child mortality prevalence (preceding two weeks) in children under reduction. Rutstein (2000) provides a multivariate age five, at rates of 13.2 percent in urban areas regression analysis of infant and child mortality and 17.6 percent in rural areas. This rate is used in developing countries, using Demographic and to estimate the number of annual cases per child Health Survey data from 56 countries from 1986 under age five, and then the total annual cases in to 1998. The study finds a significant relationship all children under five. The procedure applied is to 40 multiply the two-week prevalence rate by 52/2.5 average duration of diarrheal illness is assumed to arrive at an approximation of the number of an- to be three to four days. This implies that the two- nual cases per child. The prevalence rate is not week prevalence captures a quarter of the diar- multiplied by 26 two-week periods (i.e., 52/2), but rheal prevalence in the week prior to and a quarter multiplied by 52/2.5 for the following reason: the in the week after the two-week prevalence period. Table A.7: Baseline Data on Health Urban Rural Source: Under-age-5 child mortality rate in 2006 31 40 ENDESA 2006–2007 (per 1000 live births) Diarrheal 2-week prevalence in children under age 5 13.2% 17.6% ENDESA 2006–2007 Estimated annual diarrheal cases per child under Estimated from ENDESA 2.8 3.7 age 5 2006–2007 Estimated annual diarrheal cases per person Estimated from EMNV 2005 0.3 0.3 (> age 5) and Fewtrell et al. 2007 % of diarrheal cases attributable to inadequate water Estimated from Fewtrell et al. 87% 88% supply, sanitation and hygiene 2007 Since ENDESA does not provide information on The burden of disease methodology is used to es- diarrheal illness in the population above age five, timate the number of DALYs lost because of inad- estimations from another survey (EMNV 2005)21 equate water supply, sanitation and hygiene (see and approximations from Fewtrell et al. 2007 were Table A.8). The disability weight for diarrheal mor- used. The survey provides data on the prevalence bidity is 0.119 for children under age five and 0.086 of diarrhea by age during last 30 days. for the rest of the population, and the duration of Diarrheal illness sometimes requires hospitaliza- illness is assumed to be the same (i.e., three to four tion. However, there are no available centralized days). However, the DALYs per 100,000 cases of di- records in Nicaragua that provide data on the an- arrheal illness are much higher for the population nual number of diarrheal hospitalizations. Thus, over age five. This is because DALY calculations in- the number of hospital admissions due to diarrhea volve age weighting that attaches a low weight to was estimated using information on hospitalization young children and a higher weight to adults, cor- rates provided in Baschieri and Falkingham (2007). responding to physical and mental development A hospitalization rate of 0.5 percent was applied stages.22 For diarrheal child mortality, the number to all cases of diarrhea estimated from ENDESA of DALYs lost is 34. This reflects an annual discount 2006–2007 to equal the figure on total number of rate of three percent of life years lost. hospitalizations due to diarrheal diseases. 21 EMNV 2005 estimates are somewhat higher than those in other LAC countries for populations over age five. We applied these estimates in this report. This explains relatively lower health losses in Nicaragua associated with inadequate WSSH than in other LAC countries. 22 It should be noted that some researchers elect not to use age weighting, or report DALYs with and without age weighting. 41 Environmental Health in Nicaragua Table A.8: Calculation of DALYs Per Case of Health Effects Urban Rural Source: DALYs per 100,000 cases of diarrhea in 40 40 Estimated from WHO tables using children under age 5 age weighting and an average DALYs per 100,000 cases of diarrhea in 130 130 duration of illness of 4 days, and persons > age 5 age weighting and 3% discount rate DALYs per case of diarrheal mortality in for mortality 34 34 children under age 5 Baseline data used for the estimation of the cost (EMNV 2005). The cost of medical services reflects of morbidity are presented in Table A.9. The per- the cost of estimated private health care (WHO- centage of diarrheal cases in the population older CHOICE) and is a better indication of the economic than five years treated at medical facilities is es- cost of health services than public services, which timated from the percentage of treated cases are subsidized in Nicaragua. among children (ENDESA 2006–2007) and adults Table A.9: Baseline Data for Estimating Health Effects Due to Lack of Improved Water Sanitation and Hygiene Urban Rural Source: % of diarrheal cases treated at medical facilities 46% 41% ENDESA 2006–2007 (children < 5 years) % of diarrheal cases treated with ORS 97% 95% ENDESA 2006–2007 (children < 5 years) % of diarrheal cases (children < 5 years) treated with 46% 41% ENDESA 2006–2007 use of pharmacy % of diarrheal cases treated at medical facilities Estimated from EMNV 45% 40% (population > 5 years) and with medicines 2005 Average cost of health services (NIO per visit) 150 130 Per consultations with Average cost of medicines for treatment of diarrhea 70 70 pharmacies, medical (NIO) service providers and Average cost of ORS per diarrheal case in children health authorities 10 10 (NIO) Average duration of diarrheal illness in days 4 4 Assumption (children and adults) Hours per day of caregiving per case of diarrhea in 2 2 Assumption children Hours per day lost to illness per case of diarrhea in 2 2 Assumption adults Value of time for adults (caregiving and ill adults) 75% of urban and rural 20 10 NIO/hour wages in Nicaragua % of diarrheal cases attributable to inadequate Estimated from Fewtrell et 87% 88% water, sanitation and hygiene al. 2007 42 The value of time lost for adults is imputed based lution. A range of “low� to “high� ratios is presented on urban and rural wages. The analysis used 75 in Table A.11, which reflects the review by Desai et percent of urban and rural wages in Nicaragua as al. (2004). the imputed value for both income-earning and Studies around the world have also found link- non-income-earning adults. It may be noted here ages between indoor air pollution from traditional that non-income-earning adults are engaged in a fuels and increased prevalence of tuberculosis and household function that has value and they also asthma. It is also likely that indoor air pollution have an opportunity to join the paid labor force.23 from such fuels can cause an increase in ischemic heart disease and other cardiopulmonary disor- ders. As discussed in the methodology for urban air Methodology: Indoor Air Pollution pollution, Pope et al. (2002) and others have found There are two main steps in quantifying the health that the largest effect of urban fine particulate pol- effects of indoor air pollution. The first step is to lution on mortality is for the cardiopulmonary dis- estimate the number of people or households ex- ease group. Because indoor smoke from traditional posed to pollution from solid fuels, and to mea- fuels is high in fine particulates, the effect on these sure the extent of pollution, or concentration. The diseases might be substantial. However, more re- second step is to quantify the health impacts from search is required in order to draw a definite con- the exposure based on epidemiological assess- clusion about the linkage and magnitude of effect. ments. Once the health impacts are quantified, the The odds ratios in Table A.11 have been applied to value of this damage can be estimated. young children under the age of five (for ARI) and Data from ENDESA 2006–2007 on household use adult females (for ARI and COPD) to estimate the of fuels for cooking is used for this estimation. Ac- increase in mortality and morbidity associated with cording to this survey, about 38 percent of urban indoor air pollution.24 It is these population groups and 92 percent of rural households used fuelwood who suffer the most from indoor air pollution. This for cooking in 2006–2007 in Nicaragua. is because they spend much more of their time at Desai et al. (2004) provide a review of research home, and/or more time cooking than older chil- studies around the world that have assessed the dren and adult males. magnitude of health effects of indoor air pollution To estimate the health effects of indoor air pollu- from solid fuels. The odds ratios for ARI and COPD tion from the odds ratios in Table A.11, baseline are presented in Table A.11. The odds ratios repre- data for ARI and COPD need to be established (see sent the risk of illness for those who are exposed Table A.12). Data on COPD mortality and especially to indoor air pollution, compared to the risk for on the incidence of morbidity, according to interna- those who are not exposed. The exact odds ratio tional disease classifications, are not readily avail- depends on several factors such as concentration able for Nicaragua. Regional estimates from WHO level of pollution in the indoor environment and the (2001) and Shibuya et al. (2001) for the AMRO D amount of time individuals are exposed to the pol- region are therefore used.25 23 Some may argue that the value of time based on wage rates should be adjusted by the unemployment rate to reflect the probability of obtaining paid work. 24 Although Desai et al. (2004) present odds ratios for lung cancer. This effect of pollution is not estimated in this report because the incidence of lung cancer among rural women is generally very low. The number of cases in rural Nicaragua associated with indoor air pollution is therefore likely to be minimal. 25 Nicaragua belongs to the AMRO-D region of WHO. 43 Environmental Health in Nicaragua Table A.11: Health Risks of Indoor Air Pollution Odds Ratios (OR) “Low� “High� Acute respiratory illness (ARI)a 1.9 2.7 Chronic obstructive pulmonary disease (COPD) 2.3 4.8 Source: Desai et al. (2004). The national average of the two-week prevalence Estimated Health Impacts rate of ARI in children under age five reported in Annual new cases of ARI and COPD morbidity and ENDESA 2006–2007 was used to estimate total mortality (Di) from fuelwood smoke were estimated annual cases of ARI in children under age five. The from the following equation: procedure applied is to multiply the two-week prev- Di = PAR *DiB ; (A.1) alence rate by 52/3 to arrive at an approximation where DiB is baseline cases of illness or mortality, of the annual cases of ARI per child.26 The EMNV i (estimated from the baseline data in Table A.12), 2005 provides information on ARI prevalence in and PAR is given by: the last month by age in adults. ARI mortality in children under age five is 14 percent of total esti- PAR = PP*(OR-1)/(PP*(OR-1)+1) (A.2) mated child mortality (WHO 2006). where PP is the percentage of population exposed Table A.12: Baseline Data for Estimating Health Impacts of Indoor Air Pollution Urban Rural Source 1.4– Female COPD mortality rate (% of total female deaths) 1.4–2.1% WHO GBD (2004) and 2.1% Shibuya et al. (2001) Female COPD incidence rate (per 100,000) 94 94 ARI 2-week prevalence in children 28% 30% ENDESA 2006–2007 4.9 Estimated from ENDESA Estimated annual cases of ARI per child under age 5 5.2 2006–2007 1.7 Estimated from EMNV Estimated annual cases of ARI per adult female (> 30 years) 2.6 2005 ARI mortality in children under age five (% of child mortality) 14% 14% WHO 2006 26 A factor of 52/3 is applied for the following reason: the average duration of ARI is assumed to be about seven days. This implies that the two- week prevalence captures half of the ARI prevalence in the week prior to and the week after the two-week prevalence period. 44 Table A.13: Calculation of DALYs per Case of Health Effects From Indoor Air Pollution Baseline Source Urban Rural DALYs per 100,000 cases of ARI in children under 5 165 165 DALYs per 100,000 cases of ARI in female adults (>30) 700 700 Estimated from DALYs per case of ARI mortality in children under 5 34 34 WHO tables DALYs per case of COPD morbidity in adult females 2.25 2.25 DALYs per case of COPD mortality in adult females 6.8 6.8 to fuelwood smoke (38 percent of urban and 92 Estimating Costs percent of rural households according to ENDESA Treatment costs are estimated based on costs of 2006–2007), and OR is the odds ratios (or relative health care services offered by the private sector, risk ratios) presented in Table A.11. because these are likely to better reflect the true The methodology described above is used to es- economic costs. The percentage of ARI cases in the timate the number of DALYs lost associated with age group older than five years treated at medical cases of ARI and COPD attributed to indoor air pol- facilities is estimated from EMNV (2005). lution. The disability weight for ARI morbidity is the The time lost in the case of adults is valued at 75 same for children and adults (0.28), and the du- percent of average hourly wages. There is very ration of illness is also assumed to be the same little information about the frequency of doctor (seven days). However, the DALYs per 100,000 visits, emergency visits and hospitalization for cases of ARI are much higher for adults. This is COPD patients in developing countries. Since such because DALY calculations involve age weighting data are not available for Nicaragua, data derived that attaches a low weight to young children, and from Schulman et al. (2001) and Niederman et al. a higher weight to adults that corresponds to phys- (1999) on the United States and Europe are used ical and mental development stages.27 For ARI in this analysis. The lost workdays per year are es- child mortality, the number of DALYs lost is 34. This timated based on the frequency of estimated med- reflects an annual discount rate of three percent of ical treatment, an additional seven days for each life years lost. hospitalization and one extra day for each doctor DALYs lost per case of COPD morbidity and mor- visit and emergency visit. These days are added to tality are based on life tables and age-specific inci- reflect time needed for recovery from illness. dence of the onset of COPD reported by Shibuya et To estimate the cost of a new case of COPD, the al. (2001) for the AMRO-D region. A disability weight medical cost and value of time losses are dis- of 0.2 has been applied to COPD morbidity.28 A dis- counted over a 20-year duration of illness. An an- count rate of three percent is applied to both COPD nual real increase of two percent in medical costs morbidity and mortality. and value of time has been applied to reflect an average expected increase in annual labor produc- 27 It should be noted that some researchers elect not to use age weighting, or report DALYs with and without age weighting. 28 See: http://www.dcp2.org/pubs/GBD 45 Environmental Health in Nicaragua Table A.14: Baseline Data for Estimation of Costs of Health Impacts From Indoor Air Pollution Rural Urban Source % of ARI cases treated at medical facilities (children < 5 60% 68% ENDESA 2006–2007 years) %age of cases with use of pharmacy (children < 5 years) 60% 68% ENDESA 2006–2007 Cost of medicines for treatment of acute respiratory Per consultations with 70 70 illness (population < 5 years) (NIO) pharmacies % of ARI cases treated at medical facilities (females > 30 30% 30% EMNV (2005) years) % of COPD patients being hospitalized per year 1.5 1.5 Assumption based on % of COPD patients with an emergency doctor/hospital 15 15 Schulman et al. (2001) and outpatient visits per year Niederman et al. (1999) Avg. number of doctor visits per COPD patient per year 1 1 Estimated based on freq. Estimated lost workdays (including household work days) 2.6 2.6 of doctor visits, emergency per year per COPD patient visits and hospitalization Cost of doctor visit (NIO per visit) 150 130 Per consultations with Cost of hospitalization (NIO per day) 500 500 pharmacies, medical service providers and Cost of emergency visit (NIO per visit) 200 20029 health authorities Average duration of ARI in days (children and adults) 7 7 Assumption Hours per day of caregiving per case of ARI in children 2 2 Assumption Hours per day lost to illness per case of ARI in adults 3 3 Assumption Value of time for adults (caregiving and ill adults) – NIO/ 10 20 75% of wages in Nicaragua hour Average length of hospitalization for COPD (days) 10 10 Larsen (2004b) tivity and real wages. The costs are then discounted lost due to illness in the case of adults is valued at at three percent per year, a rate commonly used by 75 percent of the average hourly wage for urban WHO for health effects. and rural workers. The rationale for this valuation Baseline data used to estimate the cost of mor- of time is discussed in the previous sections. bidity are presented in Table A.14. The treatment Methodology: Valuation of Mortality costs are based on costs of private sector health Two distinct methods of valuation of mortality are care services, because these are likely to better re- commonly used to estimate the social cost of pre- flect economic cost. Cost of mortality is discussed mature death: the Human Capital Approach (HCA) below. The percentage of ARI cases in the age and the Value of Statistical Life (VSL). Although both group older than five years, treated at medical fa- approaches are being used, the VSL approach has cilities, is estimated using EMNV (2005). The time 29 These costs include charges for a bed and costs of medicines including what the government incurs for these services 46 been more commonly used in the last few decades. n. In the case of children, we may have i {20, 65}, In this report, the HCA has been applied as a lower assuming the lifetime income on average starts at bound and the VSL approach as the higher bound age 20 and ends at retirement at age 65. An an- in estimating the cost of adult mortality. However, nual growth of real income and the discount rate the HCA has been used in the case of child mor- used in this study are two percent and three per- tality. cent, respectively. Human Capital Approach Two important issues are often raised with respect The HCA is based on the economic contribution of to the application of the HCA. The first issue relates an individual to society over the lifetime of the indi- to the application of this valuation approach to in- vidual, and death results in an economic loss that dividuals who do not participate in economic activi- is approximated by the loss of all future income of ties, i.e., to individuals such as the elderly, family the individual. Future income is discounted to re- members taking care of the home, and children flect its value at the time of death. The discount who do not earn an income. One may think of an rate commonly applied is the rate of time prefer- extension of the HCA that recognizes the value of ence. Thus the social cost of mortality, according non-paid household work at the same rate as the to the HCA, is the discounted future income of an average income earner, or at a rate equal to the individual at the time of death. If the risk of death, cost of hiring a household helper. In this case, the or mortality risk, is evenly distributed across in- HCA can be applied to the death of non-income come groups, average expected future income is earners and children (whether or not children will applied to calculate the social cost of death. Math- become income earners or take care of the home ematically, the present value of future income is during their adult life). In the case of the elderly, the expressed as follows: HCA would not assign an economic value to older i n individuals who have either retired from the work- PV 0 ( I )   I 0 (1  g ) i /(1  r ) i (A.3) force or do not make significant contributions to i k household work. This obviously is a serious short- where PV0 (I) is present value of future income (I) coming of the HCA approach. in year 0 (year of death), g is annual growth in real income, and r is the discount rate (rate of time pref- The second issue regarding the HCA is that the so- erence). As can be seen from (1), the equation al- cial cost of mortality is limited to the economic con- lows for income to start from year k, ending in year tribution of an individual or the value of household work if the individual takes care of the home. Alter- Table A.15: Cost of Mortality (per death) Using HCA Average Number Thousand of Years Lost NIO Adults: Mortality from urban air pollution 7 132 Mortality from indoor air pollution 6.8 128 Children: Mortality from indoor air pollution 65 580 Mortality from diarrheal illness 65 580 47 Environmental Health in Nicaragua native approaches (such as the value of statistical mortality risk. Mathematically it can be expressed life) to the valuation of mortality, or social cost of as: mortality, have therefore been developed and have VSL = WTPAve * 1/ R (A.4) increasingly been applied in the past few decades. where WTPAve is the average willingness to pay The estimated cost of mortality in Nicaragua based (NIO per year) per individual for reducing the mor- on HCA, using Eq. A.3, is presented in Table A.15. tality risk of magnitude R. In the example above, Average annual income is approximated by GDP R=1/10.000 (or R=0.0001) and WTPAve= 10 NIO. per capita, at about 18,800 NIO per year. Thus, if 10 individuals die each year from the health Value of Statistical Life (VSL) risk illustrated above, the cost to society is 10* VSL While the HCA involves valuation of the death of = 10*100,000 NIO = 1 million NIO. an individual, the VSL is based on valuation of Estimating VSL mortality risk. Everyone in society is constantly The main two approaches used to estimate the VSL facing a certain risk of dying. Examples of such are the revealed preference and stated preference risks are occupational fatality risk, risk of traffic ac- analyses. Most of the studies that use revealed cident fatality, and environmental mortality risks. preferences are hedonic wage studies, which es- It has been observed that individuals adjust their timate labor market wage differentials associated behavior and decisions in response to such risks. with differences in occupational mortality risk. The For instance, individuals demand a higher wage stated preference studies use contingent valuation (a wage premium) for a job that involves a higher- methods (CVM), which estimate individuals’ willing- than-average occupational risk of fatal accident, ness to pay for mortality risk reduction. individuals may purchase safety equipment to re- Mrozek and Taylor (2002) provide a meta-analysis duce the risk of death, and/or individuals and fami- of VSL estimates from labor market studies from lies may be willing to pay a premium or higher rent around the world. They identify a “best-practice� for properties (land and buildings) in a cleaner and sample and control for industry characteristics less polluted neighborhood or city. other than occupational mortality risk that also af- Therefore, by observing individuals’ choices and fects inter-industry wage differentials. The study willingness to pay for reducing mortality risk (or min- concludes that a lower estimate of US$2 million imum amounts that individuals require to accept a for VSL can be reasonably inferred from labor higher mortality risk), it is possible to measure or market studies when “best practice� assumptions estimate the value to society of reducing mortality are used. However, it should be noted that the risk, or, equivalently, to measure the social cost of VSL range inferred by Mrozek and Taylor (2002) is a particular mortality risk. For example, if a certain substantially lower than average VSL estimated in health hazard has a mortality risk of 1/10,000, i.e., some other meta-analyses where the mean VSL is one individual (on average) for every 10,000 indi- as high as US$6 million. As a higher bound for VSL, viduals dies every year from that particular health a mean estimate of VSL meta- analysis of US$5.4 hazard. If each individual on average is willing to million was applied. The latest meta-analysis of pay 10 NIO per year for eliminating this mortality VSL estimates is by Kochi et al. (2006) and uses risk, then every 10,000 individuals are collectively the empirical Bayes pooling method to combine willing to pay 100,000 NIO per year for eliminating and compare estimates of VSL data from 40 se- the mortality risk. This is the VSL for eliminating the 48 lected studies published between 1974 and 2002, However, the most appropriate income elasticity containing 197 VSL estimates. to use for low-income countries such as Nicaragua Benefit Transfer remains uncertain because the income level in Ni- No studies on VSL have been undertaken in Nica- caragua falls far outside the range of income in the ragua; the overwhelming majority of VSL studies sample of countries from which the income elastici- have been conducted in countries with substantially ties of VSL are estimated in the empirical literature. higher income levels. Hence, the VSL estimates A prudent approach might be to apply an elasticity from these countries must be adjusted through a of 1.0 in order to reduce the risk of overstating the benefit transfer to suit the conditions in Nicaragua. cost of mortality in Nicaragua. One commonly used approach in benefit transfer is Table A.16 presents the VSL for Nicaragua from to use income elasticities.30 Viscusi and Aldi (2002) benefit transfer based on the range of VSL re- estimate an income elasticity of VSL in the range of ported by Mrozek and Taylor (2002) as the lower 0.5 to 0.6 from a large sample of VSL studies. How- bound, and that reported by Kochi et al. (2006) as ever, the range in income elasticity is influenced by the upper bound, and an income elasticity of 1.0. three unusually high estimates of VSL from labor These figures are substantially higher than the market data in one state in India. Leaving out ones from the HCA, especially for adult mortality these three studies provides an income elasticity due to urban air pollution and indoor air pollution. of about 0.80. A comparison is presented in Table A.17. Table A.16: Estimated Value of Statistical Life in Nicaragua “High� “Low� Source: Kochi et al. (2006), Average VSL in high-income countries (US$ million) 5.4 2 Mrozek and Taylor (2002) Average GDP/capita in high-income countries (US$) 30 000 30 000 World Bank* GDP per capita in Nicaragua (US$ in 2007) 1020 1020 WDI 2009 Income elasticity 1.0 1.0 Estimated VSL in Nicaragua (million NIO) 3.39 1.26 Benefit transfer * Weighted average GDP per capita, based on the sample in Mrozek and Taylor (2002). Table A.17: A Comparison of HCA and VSL Estimates Applied to Nicaragua Ratio of VSL/HCA Adult mortality 18 Children mortality 4 30 The income elasticity is the percentage change in VSL per percentage change in income. 49 Annex B. Methodology for Cost-Benefit Analysis the estimated benefits and costs represent orders I. Introduction of magnitude and gross averages at the national The objective of this report is to provide estimates level. They should therefore be subject to sensitivity of benefits and costs of water-sanitation-hygiene analysis and can undoubtedly be improved by more improvements and control of indoor air pollution extensive data collection and in-depth analysis. from solid fuels. A full range of potential interventions could concep- tually be evaluated in terms of their benefits and II. Water, Sanitation and Hygiene costs to society. However, data and resource limita- This section presents the damage cost of inad- tions constrain the level of detail and the number equate water supply, sanitation and hygiene and of interventions that can be evaluated practically provides estimates of the benefits and costs of in- in a relatively short period of time. Interventions terventions to reduce this damage cost. evaluated in this report are therefore confined to a relatively small number of interventions, and are aggregated to a level that reflects the limited data, Effectiveness of Improved Water, Sanitation and time and resource availability for the study. Nev- Hygiene ertheless, it is hoped that the interventions evalu- In a seminal review of international studies, Esrey ated in this report by and large are consistent with et al. (1991) reported mean estimates of reduc- interventions that are generally believed to be the tions in diarrheal illness around the world from most effective in improving environmental condi- improvements in water supply, sanitation and tions. The report can also be utilized for further as- household hygiene practices. Since then, Curtis sessment of the effectiveness, benefits and costs and Cairncross (2003) presented results from a of interventions at local levels in the areas of inad- meta-analysis of studies that have investigated re- equate water supply, sanitation and hygiene, and ductions in diarrheal illness from hand washing. In indoor air pollution that produce the major relative 2004, Fewtrell and Colford provided a systematic health damage associated with environmental fac- review and meta-analysis of water, sanitation and tors in Nicaragua. hygiene interventions on diarrheal illness. Environmental cost-benefit analysis (CBA), the The main results of Fewtrell and Colford (2004) are primary focus of this report, is extensively used presented in Table B.1 for developing countries. in environmental policy, for example in the United The relative risk ratio (RR) is in relation to a non- States. It has also been undertaken in developing intervention situation. The percentage reduction countries, but most often for a particular environ- in diarrheal illness from intervention is therefore mental issue. In contrast, this report provides an 1.0-RR. The relative risk ratio for hand-washing in- opportunity to compare benefits and costs of inter- tervention in Table B.1 is very similar to the results ventions that span several environmental issues or from the meta-analysis by Curtis and Cairncross categories. However, it should be recognized that (2003). 51 Environmental Health in Nicaragua The single most effective hygiene intervention is and rivers, and water provided by vendor or tanker hand washing after defecation, before preparing trucks. In terms of water quality improvement, the meals, and before eating. This intervention is studies reviewed by Fewtrell and Colford that have therefore reported in Table B.1. Improved sanita- assessed the reduction in diarrheal illness from tion refers to facilities for safe and hygienic removal source water treatment are not very conclusive. of excreta, such as flush toilets, pour-flush latrines, The pooled study results suggest a mean reduc- ventilated improved pit latrines (VIP) and simple pit tion in diarrheal incidence of 10 percent, but with latrines. Unimproved sanitation includes open pit no statistical significance. In contrast, point-of-use latrines, public latrines, service or bucket latrines, drinking water treatment (i.e., household drinking and the absence of any facilities. Improved water water treatment) seems very effective in reducing supply refers to house connections, standpipes, diarrheal illness. Point-of-use treatment refers to boreholes, protected wells or springs, and collected non-chemical (e.g., boiling of water) and chemical rain water. Unimproved water supply includes un- treatment (e.g., chlorination) and seems to be most protected wells or springs, open surface water effective in rural areas. Table B.1: Summary of Meta-Analysis by Fewtrell and Colford (2004) Relative Risk Confidence Intervention (RR) Interval (95%) Improved hygiene (hand washing) 0.556 0.334–0.925 Improved sanitation 0.678 0.529–0.868 Improved water supply 0.749 0.618–0.907 Water quality improvement (source treatment) 0.891 0.418–1.899 Water quality improvement (point-of-use treatment; rural) 0.534 0.392–0.727 Water quality improvement (point-of-use treatment; urban) 0.771 0.725–0.819 Water quality improvement (point-of-use chemical treatment) 0.605 0.443–0.828 Water quality improvement (point-of-use non-chemical treatment) 0.534 0.379–0.752 Note: Summarized from Table 22 in Fewtrell and Colford (2004). To evaluate the benefits and costs of the interven- efit-cost analysis. With regard to infrastructure or tions in Table B.1, it is important to distinguish be- hardware (water supply and sanitation facilities), tween interventions that involve changes in house- improvements are predominantly functions of pro- hold behavior and interventions that involve infra- vision and are likely to be utilized by households structure or hardware improvements. Interventions if design and service delivery reflect demand and that involve changes in household behavior are provide convenience. Uncertainties regarding be- improved hygiene and water treatment at point-of- havioral change in relation to water supply and use. While public authorities can promote these sanitation therefore tend to be less important in a behaviors, actual changes in behavior is beyond benefit-cost analysis than for hygiene improvement their control. It is therefore important to explicitly and point-of-use treatment of drinking water. account for this behavioral component in a ben- 52 A Benefit-Cost Analysis Framework three to five times higher than in North America Prüss et al. (2002) provided a framework for esti- and high-income countries in Europe, and as much mating the burden of disease from water, sanita- as six times higher in Sub-Saharan Africa. These tion and hygiene. This is presented in Table B.2. figures are relatively consistent with Table B.2 and Prüss et al. applied this framework to estimate the suggest that most developing countries are some- global burden of diarrheal disease, but it can also where in the range of Scenario 4 to Scenario 6. be applied conveniently to estimate the benefits These figures represent averages, and it should be and costs of improved water supply and sanitation. made clear that there are larger variations within each developing country, with some parts of the According to the Global Burden of Disease 2002 population being closer to Scenario II. (WHO), diarrheal incidence (cases per person per year) in most developing regions of the world is Prüss et al. derived the relative risks of diarrheal illness from international literature (Table B.3). Table B.2: Selected Exposure Scenarios Scenario/ Pathogen Relative Risk Description Situation Load (RR) NO IMPROVED WATER SUPPLY AND NO BASIC SANITATION in a VI country that is not extensively covered by those services, and Very High 11.0 where water supply is not routinely controlled IMPROVED WATER SUPPLY and no basic sanitation in a Vb country that is not extensively covered by those services, and Very High 8.7 where water supply is not routinely controlled BASIC SANITATION but no improved water supply in a country Va that is not extensively covered by those services, and where High 6.9 water supply is not routinely controlled IMPROVED WATER SUPPLY AND BASIC SANITATION in a IV country that is not extensively covered by those services, and High 6.9 where water supply is not routinely controlled IV and improved access to drinking water (generally piped to IIIc High - household) IIIb IV and improved personal hygiene High 4.5 IIIa IV and drinking water disinfected at point of use High 3.8 Medium Regulated water supply and full sanitation coverage, with II partial treatment for sewage, corresponding to a situation to 2.5 typically found in developed countries Low Ideal situation, corresponding to the absence of transmission I Low 1.0 of diarrheal disease through water, sanitation, and hygiene Based on Prüss et al. 2002. 53 Environmental Health in Nicaragua Prüss et al. derived the relative risks of diarrheal illness from international literature (Table B.3). Table B.3: Reductions in Diarrheal Illness Used by Prüss et al. Scenario Reduction Source Progression in Diarrheal Illness From VI to Vb Providing improved water supply 20.8% Esrey (1996) From VI to Va Providing basic sanitation facilities 37.5% Esrey (1996) Providing improved water supply and basic From VI to IV 37.5% Esrey (1996) sanitation facilities From IV to IIIb Improved personal hygiene 35% Huttly et al. (1997) Disinfection of drinking water at point of From IV to IIIa 45% Quick et al. (1999) use Combined results Regulated water supply and full sanitation from Huttly et al. From IV to II 65% coverage, with partial treatment of sewage (1997) and Quick et al. (1999) Absence of transmission of diarrheal Using results from From II to I disease through water, sanitation and 60% Mead et al. (1999) hygiene Benefit-Cost Analysis of Water and Sanitation piped water supply that is treated at source (water Improvements treatment plant) and those with piped water that The framework by Prüss et al. (2002) is applied in is not treated at source. Scenario IIId is therefore this report to estimate the benefits and costs of added in Table B.4. interventions to reduce diarrheal illness and diar- To allow for a comparison to Table B.2, the relative rheal mortality in Nicaragua. The relative risks in risk for Scenario VI without point-of-use disinfec- Table B.2, or reductions in diarrheal illness in Table tion is also 11.0 in Table B.4. The relative risks in B.3, are modified to reflect the more recent find- Vb and Va is derived by multiplying the relative risk ings of relative risks in the meta-analysis study by in VI by the relevant relative risk ratios in Table B.1. Fewtrell and Colford (2004), and the meta-analysis As in Prüss et al., there is no difference between Va study of hand washing by Curtis and Cairncross and IV. The difference between IV and IIId is a rela- (2003). tive risk ratio of 0.9, reported in Fewtrell and Col- The first modification to the framework presented ford (2004) for household water supply connection. by Prüss et al. is to distinguish between households The difference between IIId and IIIc is the relative that disinfect their drinking water at point of use risk ratio of 0.891 presented in Table B.1. and those households that do not disinfect their The difference between with and without point-of- drinking water. This distinction is made for each use disinfection for Scenarios VI to IIId is a relative of the scenarios from III to VI. The second modifi- risk ratio of 0.53. This corresponds to the relative cation is to distinguish between households with risk ratio for rural water supply in Table B.1. This 54 ratio may be considered more appropriate to apply 2006–2007 household survey with data on the than the urban relative risk ratio because Scenarios water and sanitation situation in urban and rural IV–VI are typically found in rural areas, and water areas, Censos Nacionales (National Censuses) quality in Scenarios IIId to VI is on average likely 2005, and WHO/UNICEF Joint Monitoring Program to involve higher disease risk than urban treated (JMP 2008) as well as household drinking water piped water supply. The relative risk for Scenarios disinfection data, and data from Nicaraguan water IIIc and IIId is assumed to be the same if drinking supply authorities on the piped water network and water is disinfected at point of use. water treatment plants (Estrategia Sectorial de Scenario II is not included in Table B.4. Scenario II is Agua Potable y Saneamiento 2005). the situation typically found in developed countries The water supply and sanitation situation is pre- (see Table B.2). The provision of this level of service sented in Tables B.5 and B.6. As discussed in rela- (including partial sewage treatment) to the entire tion to Tables B.2 to B.4, unimproved or no basic urban and rural population in developing countries sanitation mainly refers to households with no san- is likely to be very costly. The benefit-cost analysis itation facilities or with open pit latrines. Similarly, in this report therefore focuses on improved water no improved water supply refers mainly to surface supply and basic sanitation for those segments of water, tanker trucks and unprotected well or spring the population without these services. water. To estimate the health benefits of water and sanita- The latter was selected as the basis for the anal- tion interventions, it is necessary to provide an esti- ysis. In order to use the data in Table B.6 to es- mate of the Nicaraguan population shares in each timate the population shares in each of the sce- of the scenarios. Three sources of data from Nica- narios in Table B.4, a set of allocation “rules� were ragua are used for this purpose, i.e., the ENDESA applied. These “rules� are presented in Table B.7. Table B.4: Exposure Scenario Application to Nicaragua WITHOUT WITH Scenario/ Description Point-of Use Point-of-Use Situation Disinfection Disinfection RR RR VI No improved water supply and no basic sanitation 11.0 5.8 Vb Improved water supply and no basic sanitation 8.2 4.4 Va Basic sanitation but no improved water supply 7.5 4.0 IV Improved water supply and basic sanitation 7.5 4.0 IV and water supply piped to household (no source IIId 6.7 3.6 treatment) IIIc IV and water supply piped to household (source treatment) 5.9 3.6 Modified from Prüss et al. 2002. Note: RR is relative risk of diarrheal disease. 55 Environmental Health in Nicaragua Table B.5: Water Supply and Sanitation (% of Households): Different Sources ENDESA 2006–2007 Censos Nacionales 2005 Water Service by: Total Urban Rural Total Urban Rural Pipe inside HH 30.7 49.5 5.1 39.7 62.7 10.4 Pipe outside HH in lot 32.0 41.4 19.0 20.6 24.1 16.0 Public fountain 3.7 1.0 7.5 3.2 1.5 5.4 Public/private well 14.9 3.3 30.9 16.7 5.6 30.8 Purchase/tanker 0.8 1.0 0.8 River, watershed 13.1 0.6 30.2 13.2 0.5 29.3 Other 4.8 3.2 6.5 6.7 5.6 8.1 Sanitation Service by: Toilet w/sewer connection 56.7 44.9 72.9 60.0 53.0 68.0 Sewage connection in HH 23.2 39.5 0.9 17.5 33.0 0.0 Toilet w/septic tank 7.7 12.1 1.7 6.0 11.0 2.0 River 0.2 0.3 0.1 0.5 0.3 0.1 No sanitation 12.2 3.2 24.4 16.0 3.0 30.0 Source: ENDESA 2006–2007; Resumen Censal (Census Summary) 2005 Table B.6: Water Supply and Sanitation in Nicaragua (% of Households) as in JMP 2006 Urban (58% of the Rural (42% of Total population) the population) Broad definition 79 90 63 Water House connections 60 84 27 Broad definition 47 56 34 Sanitation Sewerage 13 22 0 Source: WHO/UNICEF Joint Monitoring Program (JMP/2006). Data from Nicaragua’s 2005 Estrategia Sectorial with piped water supply that is treated. According de Agua Potable y Saneamiento (Sectoral Strategy to these data, 100 percent of drinking water was for Drinking Water and Sanitation) were used to disinfected at the source in urban and rural areas provide an estimate of the urban population share in 2005. 56 Table B.7: Scenario Allocation Rules Scenario/ Allocation rule Situation VI Lesser population share without improved water supply and without basic sanitation. Vb Diff. between population share w/o basic sanitation and w/o improved water, if difference is > 0. Diff. between population share w/o improved water supply and w/o basic sanitation, if Va difference is > 0 Lesser population share with improved water supply and basic sanitation minus population IV share with piped water. IIId Population share with piped water supply. IIIc Table B.8: Rural Water Supply and Sanitation in Nicaragua (% of Rural Population) Share Scenario/ Description of rural Situation population VI No improved water supply and no basic sanitation 37% Vb Improved water supply and no basic sanitation 2% Va Basic sanitation but no improved water supply 0% IV Improved water supply and basic sanitation 34% IIId IV and water supply piped to household (no source treatment) 27% IIIc IV and water supply piped to household (source treatment) Table B.9: Urban Water Supply and Sanitation in Nicaragua (% of Urban Population) Share Scenario/ Description of urban Situation population VI No improved water supply and no basic sanitation 10% Vb Improved water supply and no basic sanitation 6% Va Basic sanitation but no improved water supply 0% IV Improved water supply and basic sanitation 28% IIId IV and water supply piped to household (no source treatment) 56% IIIc IV and water supply piped to household (source treatment) 57 Environmental Health in Nicaragua Table B.10: Water Treatment at Point of Use in Nicaragua (% of Households) Type of water treatment Urban31 Rural Boiling 5 1.5 Chlorination 10 40 Filtering 5 1.5 Source: Estimated based on World Bank 2008 The source does not provide the share of house- Strukova (2009) presents the estimated number holds using disinfection in relation to their type of of diarrheal illness cases in Nicaragua based on water supply and sanitation. The data on disinfec- the ENDESA 2006–2007 (Table B.11). The total tion are therefore applied uniformly to each of the number of annual cases was estimated at close to scenarios. seven million, or 1.2 cases per person. Table B.11: Estimated Annual Cases of Diarrheal Illness in Nicaragua in 2005 National Urban Rural 2-week diarrheal prevalence (children < 5 years) 21.7% 13.2% 17.6% 2-week diarrheal prevalence (population > 5 years) 3.7% 1.2% 1.5% Annual diarrheal cases in children under 5 (000) 2970 898 1212 Annual diarrheal cases in population > 5 (000) 3778 697 664 Total annual diarrheal cases 6748 1595 1876 Diarrheal cases per person (all population) 1.21 0.51 0.76 Source: Strukova 2009 Estimated cases of diarrheal illness per person per person average 0.40 in households with piped per year in Nicaragua are estimated for Scenarios water supply and basic sanitation that practice dis- IIIc to VI from the relative risks in Table B.4, the infection of drinking water, and 1.2 to 1.3 per year scenario population distribution in Tables B.8–B.9, in households that do not have improved water and the average diarrheal cases per person in supply, have no basic sanitation and do not prac- Table B.11. Tables B.12–B.13 indicate that cases tice drinking water disinfection. 31 For estimates of periurban settlements. 58 Table B.12: Estimated Annual Cases of Diarrheal Illness per Person in Urban Nicaragua WITHOUT WITH Scenario/ Description Point-of Use Point-of-Use Situation Disinfection Disinfection VI No improved water supply and no basic sanitation 1.31 0.69 Vb Improved water supply and no basic sanitation 0.98 0.52 Va Basic sanitation but no improved water supply 0.89 0.47 IV Improved water supply and basic sanitation 0.89 0.47 IV and water supply piped to household (no source IIId 0.8 0.42 treatment) IV and water supply piped to household (source IIIc 0.71 0.42 treatment) Table B.13: Estimated Annual Cases of Diarrheal Illness per Person in Rural Nicaragua WITHOUT WITH Scenario/ Description Point-of Use Point-of-Use Situation Disinfection Disinfection VI No improved water supply and no basic sanitation 1.22 0.65 Vb Improved water supply and no basic sanitation 0.92 0.49 Va Basic sanitation but no improved water supply 0.83 0.44 IV Improved water supply and basic sanitation 0.83 0.44 IV and water supply piped to household (no source IIId 0.75 0.40 treatment) IV and water supply piped to household (source IIIc 0.67 0.40 treatment) Providing piped water supply to all rural households Using information about incremental service in- is likely to be very expensive. A realistic objective crease and corresponding investment needs from might be to at least provide improved water supply the 2005 Estrategia Sectorial de Agua Potable y (protected wells or boreholes) and sanitation fa- Saneamiento (Sectoral Strategy for Drinking Water cilities (improved pit latrines or pour-flush latrines). and Sanitation), annualized per capita investment The aim of the infrastructure interventions is to im- costs are estimated at about US$12–19 for im- prove water supply and sanitation, largely in rural proved sanitation and US$30–40 for improved areas. Two programs are investigated: one that water supply. This is based on a ten percent annual provides 1.5 million people with improved sanita- discount rate, annual five percent O&M and five tion, and one that provides 1.2 million people with percent promotion/water source protection cost, an improved water supply both in rural and urban and US$6 in annual sewage cost.32 The population areas. receiving improved water supply and sanitation is 32 Per capita investment costs represent average costs in South America (WHO/UNICEF 2000). O&M refers to operations and maintenance. 59 Environmental Health in Nicaragua calculated from Tables B8 and B.9. Diarrheal cases Approach (HCA) as presented in Strukova 2009. averted are calculated from Table B.13. Deaths However, there are strong reasons to believe that averted are calculated based on an estimated the HCA approach underestimates the value of a case fatality rate of 0.3 per 1,000 cases in children lost life; hence, the figures reported here should under age five, using data presented in Strukova be taken as lower bounds. Finally, the programs 2009. The key assumptions in deriving the benefits generate savings in time, which is an important relate to the costs of morbidity and mortality and to ingredient in the calculations. It is based on data the value of time saved. The morbidity costs, based for households more than a 15-minute walk from a on the costs of treatment and value of lost time, water source, and time saved is valued at 75 per- are US$8–14 per case of diarrhea. The mortality cent of the average rural wage (Strukova 2009). costs are calculated based on the Human Capital Table B.14: Benefits of Reductions in Diarrheal Morbidity and Mortality in Rural Nicaragua Improved Improved Sanitation Water Facilities Supply Population (million) receiving improved sanitation* 0.96 Population (million) receiving improved water supply** 0.9 Percent reduction in diarrheal illness per person (from Fewtrell and Colford 32% 25% 2004) Diarrheal cases (million) averted per year 0.3 0.2 Deaths in children averted per year 60 40 Annual health benefits of improved services (US$ million) 4 3 Annual value of time savings from improved services (US$ million) 26 5 Annualized cost of service provision (US$ million) 12 24 Benefit-cost ratio (health benefits only) 0.35 0.1 Marginal cost (million NIO per 1% of WSSH health cost reduction) 26 69 Benefit-cost ratio (health benefits and time savings) 2.6 0.32 The data reveal that programs to improve sanita- the benefits lower due to lower diarrheal mortality tion have a benefit-to-cost ratio greater than 1.0 among children under age five and to lower diar- when the time savings of improved water are in- rheal prevalence. cluded. For water supply program benefit cost ra- Marginal cost33 would be higher than marginal ben- tion is less than 1. For water supply/sanitation efits34 (health damage reduction) for all four invest- programs in urban areas the costs are higher and ment programs considered. 33 Approximated by average cost per one percent of WSSH cost reduction. 34 Value of one percent of WSSH cost reduction, estimated at about nine million NIO. 60 Table B.15: Benefits of Reductions in Diarrheal Morbidity and Mortality in Urban Nicaragua Improved Improved Sanitation Water Facilities Supply Population (million) receiving improved sanitation* 0.5 Population (million) receiving improved water supply** 0.2 Percent reduction in diarrheal illness per person (from Fewtrell and Colford 32% 25% 2004) Diarrheal cases (million) averted per year 0.1 0.07 Deaths in children averted per year 17 12 Annual health benefits of improved services (US$ million) 2 1 Annual value of time savings from improved services (US$ million) 16 2 Annualized cost of service provision (US$ million) 9 10 Benefit-cost ratio (health benefits only) 0.2 0.1 Marginal cost (million NIO per 1% of WSSH health cost reduction) 45 70 Benefit-cost ratio (health benefits and time savings) 1.9 0.3 Benefit-Cost Analysis of Hygiene Improvements of 45 percent is therefore applied in the benefit- In many studies, the single most effective hygiene cost analysis in this report for all age groups. intervention is found to be hand washing after def- A benefit-cost analysis of hygiene improvement ecation, before preparing meals, and before eating. (hand-washing programs) involves an assessment Curtis and Cairncross (2003) provide a meta-anal- of several key parameters and outcomes. These ysis of close to 20 hand-washing studies and report are listed in Table B.16. The costs of improved hand a mean reduction in diarrheal illness of about 47 washing practices are twofold. First, a program percent. In their meta-analysis, Fewtrell and Colford to encourage behavioral change (improved hand- (2004) report a mean reduction in diarrheal illness washing) has a cost that should be fully captured. of about 45 percent from hand-washing interven- This includes the cost of program preparation and tions (Table B.1). About two-thirds of the studies re- implementation. Second, improved hand-washing viewed in the two meta-analysis studies assessed practices have a private cost that includes the cost the effect of hand washing on diarrheal illness in of increased water and soap consumption. The children under age five. The meta-analyses do not most uncertain and critical parameter is the effec- report the effect of hand washing on diarrheal ill- tiveness of the hand-washing program in terms of ness in children under age five versus older chil- changing household and individual behavior, and dren and adults. A pooled analysis of the studies the lasting effect of changed behavior (sustain- reviewed in the two meta-analyses was therefore ability). This is likely dependent on several dimen- undertaken in this report, but no statistically signifi- sions and will vary from country to country. It will cant difference in diarrheal reduction was found in also depend on the design, duration and overall children and adults. A reduction in diarrheal illness magnitude of the hand-washing program. The ex- 61 Environmental Health in Nicaragua pected benefit of the program can be estimated Colford (2004), and the monetized benefits (or from the diarrheal illness risk reductions reported costs avoided per case of diarrheal illness reduc- in Curtis and Cairncross (2003) and Fewtrell and tion) presented in Strukova 2009 for Nicaragua. Table B.16: Key Parameters and Outcomes in a Benefit-Cost Analysis of Hand Washing Key Parameters: Outcomes: Program cost Overall cost of hand-washing program Behavioral change in target population (% of population that improves or starts Program effectiveness regular hand-washing) Program sustainability Lasting effect of the program Costs of hand washing in the group with behavioral change (increased water Private cost and soap expenditures) Percent reduction in diarrheal illness from hand-washing in group with Program benefits behavioral change Monetized benefits of reduced diarrheal illness A review of three hand-washing programs that pro- therefore have a major impact on the overall cost of vide program costs and behavioral change is pre- hygiene programs that aim to achieve substantial sented in Table B.17. The program in Guatemala reductions in the overall number of cases of diar- was national in scope and targeted households rheal illness in a country. with children under age five (Saade et al. 2001). The program cost per target household with behav- The program in Thailand focused on all households ioral change is the most relevant unit cost in Table in a set of rural villages and involved a different B.17 for a benefit-cost analysis. This cost can then level of program intervention in two subsets of the be compared to the reduction in (and thus the ben- villages (Pinford and Horan 1996). The program efits of) diarrheal illness in the target population in Burkina Faso involved one city and targeted with behavioral change. households with children under age three (Borghi A benefit and cost analysis of rural and urban et al. 2002). As seen in the table, the percentage hand-washing programs is provided for diarrheal of the target population that changed behavior reduction in children under age five. The “low� to (i.e., started regular hand washing or improved “high� scenarios for children represent: (a) a pro- hand-washing practices) range from 10 to 18 per- gram effectiveness of 10–20 percent in terms of cent, and costs per target household range from the percent of households (or primary caretakers US$0.36 to US$5.03. While the studies are too of children) that start regular hand washing or im- few to draw a definite conclusion, the results do prove hand-washing practices for the protection of suggest that program costs per unit of behavioral child health; and (b) a program cost ranging from change (per percentage point increase in popula- US$0.4 to US$5.0 per targeted household or pri- tion with behavioral change) may increase sub- mary caretaker (US$4.0–US$25.0 per house- stantially if the objective is behavioral change in a hold or primary caretaker with behavior change). large share of the target population. This issue may This range of program effectiveness and costs is 62 Table B.17: A Review of Costs and Effectiveness of Hand-Washing Programs Guatemala Thailand Burkina Faso Target area National 25 rural villages One City Intervention level “Low� “High� HH w/children HH w/children Target population All HHs All HHs under age 5 under age3 Number of target households 1,570,000 10,000 6,550 38,600 3–4 3–4 Duration of program implementation 1 year 3 years months months Behavioral change (% of target population) 10% 11% 16% 18% Program cost (US$) 561,400 5,960 7,715 194,000 Program cost per target household (US$) 0.36 0.60 1.18 5.03 Program cost per target household with 3.58 5.42 7.36 27.92 behavioral change (US$) Source: Derived from Saade et al. (2001), Pinfold and Horan (1996), and Borghi et al. (2002). based on the figures in Table B.17. The “high’ sce- experts in Nicaragua. The rural and urban price of nario corresponds to the experience in one city in water is communicated by experts from Nicaragua, Burkina Faso. However, it is possible that a national and the cost of soap is based on a spot survey of program will benefit from economies of scale and retail prices for soap in Nicaragua. therefore achieve a 20 percent effectiveness at a A benefit and cost analysis of rural and urban hand- lower unit cost than US$5.0 per primary caretaker washing programs is presented in Tables B.18– of children under age five. It is therefore possible B.19. Three scenarios are provided for diarrheal that the “high� scenario represents a higher bound reduction in children under age five. of program cost. The diarrheal illness baseline data The estimated reduction in annual cases of diar- and cost of illness per case of diarrhea and diar- rheal illness is about 0.1 million in children under rheal mortality are from Strukova (2009). Avoided age five. With morbidity and mortality costs esti- cost of illness is the program benefit per case of re- mated in Strukova 2009, the total program benefit duced or averted diarrheal morbidity and mortality. is about US$2 million. The total first year program Mortality is valued by the human capital approach, cost is about US$1 million for the “medium� sce- as in the section on water and sanitation improve- nario. ment in this report. The percent reduction in di- arrheal illness (45 percent) in children with care- The total estimated benefits and costs result in a takers who start regular hand washing or improve benefit-cost ratio that ranges from 4.5 in the “low� hand-washing practices is from Curtis and Cairn- scenario to about 0.6 in the “high� scenario. Ben- cross (2003) and Fewtrell and Colford (2004), as efit-cost ratios are higher for the urban population presented in Table B.1. Regarding the private cost since medical treatment cost and annual wages of hand washing, the quantity of increased water are higher in urban areas. and soap consumption reflects the assessment of 63 Environmental Health in Nicaragua Table B.18: Benefits and Costs of a Rural Hand-Washing Program Rural Households with Children Under Age 5   “Low� “Medium� “High� Program Effectiveness       Program target (million households)* 0.3 0.3 0.3 Program response (% of households with behavioral change) 10% 15% 20% Percent reduction in diarrheal illness per child (Fewtrell and Colford 45% 45% 45% 2004) Program Cost       Total program cost (US$ million) 0.1 0.4 1.7 Private Costs       Cost of water and hygiene products per year (US$ million) 0.1 0.2 0.2 Program Benefits       Cases of diarrheal illness averted per year (thousands) 30 45 60 Deaths in children averted per year 9 13 18 Total annual health benefits (US$ million) 0.5 0.8 1.0 Benefit-Cost Ratios       LOW: If behavioral change lasts 1 year 2.2 1.4 0.6 MEDIUM: If behavioral change lasts 2 years** 3.0 2.1 1.0 HIGH: If behavioral change lasts 3 years** 3.4 2.6 1.3 Marginal cost for medium scenario (million NIO per 1% of WSSH 10 health cost reduction) * There are about 0.3 million rural children under the age of five in Nicaragua. It is assumed there is one child under five in each household (thus the program target is 0.3 million households). However, the estimated benefit-cost ratio is higher for households with more than one child under five. ** Benefits and costs in the second and third years are discounted at an annual rate of 10 percent. Not only can children under age five benefit from assess the sustainability of behavioral change, ben- a hand-washing program, but the population over efit for only one year is clearly a very conservative age five can benefit as well. However, these pro- assumption. If benefits are sustained for two years, grams were found to have lower than one benefit the estimated benefit-cost ratios would increase. cost ratio. This low ratio, even at zero incremental Note that the benefit-cost ratio does not change program cost, is due to the fact that diarrheal in- for the population over five years of age. This is be- cidence is on average substantially lower in this cause of the assumption that behavioral change population group than in children under age five. takes place at no incremental program cost. Figure One very important aspect of the benefit-cost anal- B.1 presents benefit-cost ratios for children for this ysis presented above should be noted. It is implicitly target in a program with sustainability of behavioral assumed that the benefit of the program is only re- change lasting from one to three years. alized for one year, or that behavioral change (hand washing) only lasts one year. While it is difficult to 64 Table B.19: Benefits and Costs of an Urban Hand-Washing Program Urban Households with Children Under Age Five   “Low� “Medium� “High� Program Effectiveness       Program target (million households)* 0.3 0.3 0.3 Program response (% of households with behavioral change) 10% 15% 20% Percent reduction in diarrheal illness per child (Fewtrell and Colford 45% 45% 45% 2004) Program Cost       Total program cost (US$ million) 0.1 0.4 1.6 Private Costs       Cost of water and hygiene products per year (US$ million) 0.1 0.2 0.2 Program Benefits       Cases of diarrheal illness averted per year (thousands) 30 50 60 Deaths in children averted per year 9 14 19 Total annual health benefits (US$ million) 0.7 1.0 1.4 Benefit-Cost Ratios       LOW: If behavioral change lasts 1 year 2.9 1.9 0.8 MEDIUM: If behavioral change lasts 2 years** 3.9 2.8 1.3 HIGH: If behavioral change lasts 3 years** 4.5 3.4 1.7 Marginal cost for medium scenario (million NIO per 1% of WSSH 7 health cost reduction) * There are about 0.4 million urban children under the age of five in Nicaragua. It is assumed there is one child under five in each household (thus the program target is 0.4 million households). However, the estimated benefit-cost ratio is higher for households with more than one child under five. ** Benefits and costs in the second and third years are discounted at an annual rate of 10 percent. Figure B.1: Estimated Benefit-cost Ratios (20% Program Effectiveness Target) 65 Environmental Health in Nicaragua It is important to note that some authors link hand This equation provides an estimated diarrheal washing and ARI prevention (Cairncross 2003). Al- incidence of 0.9 in the rural population not prac- though there is a lack of meta studies in this area, ticing disinfection, compared to a rural average of some evidence is already collected. Some suggest 0.8 from Strukova 2009, in urban areas an esti- that about 20 percent of ARI could be reduced if mated diarrheal incidence is 0.54 for populations a hand-washing program is implemented. If one not practicing disinfection, compared to the urban takes this into account, then benefit from hygiene average of 0.51 from Strukova 2009. improvement may be increased at least 30 percent. There are no estimates of program costs to pro- It would make it by far the most efficient interven- mote drinking water disinfection at point of use. tion. To take a conservative approach, we did not The same costs as for hand-washing programs include this additional benefit while awaiting addi- (and for the same three scenarios of effectiveness tional scientific evidence. ranging from 10 to 20 percent) have therefore been Benefit-Cost Analysis of Drinking Water Disin- applied in the cost-benefit analysis. The program fection cost, instead of per primary caretaker of children, Table B.10 presented data on the share of house- is now per household, with the assumption that holds in Nicaragua that disinfect drinking water. one person in the household is primarily respon- The source does not provide the share of house- sible for the boiling of drinking water. The private holds using disinfection in relation to their type of cost of boiling drinking water is estimated at US$5 water supply and sanitation. The data on disinfec- per year for urban households using commercial tion is therefore applied uniformly to each of the fuels and about US$3 for rural households using scenarios. fuelwood that is collected by the household mem- bers. Collection time is estimated at 75 percent of According to the survey, the most common method the average wage. The cost of water chlorination of disinfection is boiling water and is therefore the is estimated at about US$3 per household based method considered in the benefit-cost analysis. on an average drinking water consumption of 0.75 Fewtrell and Colford (2004) report from their meta- liters per person per day (Lantagne et al. 2005). analysis that disinfection of drinking water at point The disinfection program benefits are estimated of use reduces diarrheal illness by an average of the same way as for hand-washing programs. 47 percent in rural areas and 23 percent in urban areas. In rural areas the disinfection programs are esti- mated to avert 110,000 to 220,000 cases of diar- In order to estimate the reduction in the number rhea and 25 to 50 deaths in children per year (Ta- of cases of diarrheal illness, it is necessary to es- bles B.20 and B.22). The benefit-cost ratio for the timate the diarrheal incidence in the population central estimate in water boiling programs is 1.9 share that does not practice point-of-use disinfec- and for water chlorination is 4.8, corresponding to tion of drinking water. This is given by the following a 15 percent program response rate with drinking equation: water disinfection sustained for two years. Benefit- Psd + (1-Ps)d(1-r) = dA (B.1) cost ratios are lower, but well above one in urban where Ps is the population share not practicing areas. In urban areas the disinfection programs disinfection; d is diarrheal incidence in Ps; r is re- are estimated to avert 70,000 to 130,000 cases duction in diarrheal incidence from disinfection; of diarrhea and 10 to 25 deaths in children per and dA is the national average diarrheal incidence. year. Drinking water chlorination programs both in 66 Table B.20: Benefits and Costs of a Rural Drinking Water Boiling Program “Low� “Medium� “High� Program Effectiveness       Target population – rural population not practicing disinfection (millions) 1.5 1.5 1.5 Target households (millions) 0.3 0.3 0.3 Program response (% of households with behavioral change) 10% 15% 20% Percent reduction in diarrheal illness per person (Fewtrell and Colford 47% 47% 47% 2004) Program Cost       Total program cost (US$ million) 0.1 0.3 1.4 Private Costs       Cost of boiling drinking water per year (US$ million)* 0.3 0.4 0.5 Program Benefits       Cases of diarrheal illness averted per year (thousands) 45 65 90 Deaths in children averted per year <15 <20 25 Total health benefits (US$ million) 0.7 1.1 1.5 Benefit-Cost Ratios       LOW: If behavioral change lasts 1 year 1.9 1.5 0.7 MEDIUM: If behavioral change lasts 2 years** 2.2 1.9 1.1 HIGH: If behavioral change lasts 3 years** 2.3 2.1 1.4 Marginal cost for medium scenario (million NIO per 1% of WSSH health 5 cost reduction) * Estimated based on efficiency of LPG and wood stoves, cost of LPG, fuelwood collection time of 30 minutes per day and 10% of fuelwood used for water boiling, and per person water consumption of 0.75 liter per day. ** Benefits and costs in the second and third years are discounted at an annual rate of 10 percent. rural and urban areas have high benefit cost ratios rination (for background data, see Ramírez 2008). since private program costs are low. Preliminary es- As in the case of the hand-washing program, one timates for the filtration program suggest that the very important aspect of the analysis should be benefits of this program are at the same level as noted. It is implicitly assumed that the benefit of for chlorination, since annual maintenance costs the disinfection program is only realized for one are negligible and the diarrheal reduction rate is at year, or that behavioral change (boiling of drinking the same level as for chlorination (for background water) only lasts one year. While it is difficult to as- data, see Ramírez 2008). sess the sustainability of behavioral change, ben- Preliminary estimates for the point-of-use filtration efits for only one year are clearly a very conserva- program suggest that the benefits of this program tive assumption. If benefits are sustained for three are at the same level as for chlorination, since an- years, the estimated benefit-cost ratios would in- nual maintenance costs are negligible and the diar- crease in the same way as for hygiene programs. rheal reduction rate is at the same level as for chlo- 67 Environmental Health in Nicaragua Table B.21: Benefits and Costs of an Urban Drinking Water Boiling Program “Low� “Medium� “High� Target households – urban population not practicing disinfection 0.5 0.5 0.5 (millions) Percent reduction in diarrheal illness per person (Fewtrell and Colford 23% 23% 23% 2004) Program Cost       Total program cost (US$ million) 0.2 0.2 2.7 Private Costs       Cost of boiling drinking water per year (US$ million)* 0.2 0.4 0.5 Program Benefits       Cases of diarrheal illness averted per year (thousands) 30 50 65 Deaths in children averted per year <10 <10 <15 Total health benefits (US$ million) 0.6 0.9 1.2 Benefit-Cost Ratios LOW: If behavioral change lasts 1 year 1.2 0.9 0.4 MEDIUM: If behavioral change lasts 2 years* 1.6 1.2 0.6 HIGH: If behavioral change lasts 3 years* 1.8 1.4 0.8 Marginal cost for medium scenario (million NIO per 1% of WSSH health 8 cost reduction) * Benefits and costs in the second and third years are discounted at an annual rate of 10 percent. The different interventions discussed above can as indicated in Table B.3. However, the challenge is be summarized in terms of their contribution to re- to develop and deliver programs that induce sus- duced environmental damages and the costs per tained behavioral response at a large scale, while one percent of health damage reduction (marginal containing program costs at an affordable level. cost). This is done in Figure B.2. It shows the per- For sanitation, benefits exceed costs with time- cent of reduction of environmental damage on the saving benefits taken into account. Finally, the horizontal axis and the marginal cost as explained benefits of hand washing among adults are not above on the vertical axis. The graph then plots presented at the graph above. These interventions the relative values of these two pieces of informa- were estimated to have significantly higher costs tion for a number of interventions. The disinfection than benefits and corresponding benefit-cost ratios and hygiene programs are estimated to have the lower than one. All interventions that have marginal greatest potential health benefits, but only if at costs lower than marginal benefits could reduce least 20 percent of the population responds favor- total WSSH health costs by about 50 percent. The ably to the program and improves hand-washing benefit-cost ratios for hand washing and drinking practices. Hygiene improvement and disinfection water disinfection are based on behavioral change of drinking water at point of use have substantial being sustained for two years. The ratios would be potential to reduce diarrheal illness and mortality, higher (lower) if, as a result of promotion programs, 68 Table B.22: Benefits and Costs of a Rural Drinking Water Chlorination Program at the Point of Use “Low� “Medium� “High� Program Effectiveness       Target households (millions) 0.3 0.3 0.3 Percent reduction in diarrheal illness per person (Fewtrell and Colford 2004) 47% 47% 47% Program Cost Total program cost (US$ million) 2.1 5.3 26.5 Private Costs Cost of chlorinating drinking water per year (US$ million)* 0.1 0.1 0.2 Program Benefits Cases of diarrheal illness averted per year (thousands) 70 100 130 Deaths in children averted per year 10 20 25 Total health benefits (US$ million) 0.9 1.4 1.9 Benefit-Cost Ratios       LOW: If behavioral change lasts 1 year 4.5 3.3 1.2 MEDIUM: If behavioral change lasts 2 years** 6.1 4.8 2.0 HIGH: If behavioral change lasts 3 years** 6.9 5.7 2.6 Marginal cost for medium scenario (million NIO per 1% of WSSH health cost 2 reduction) ** Benefits and costs in the second and third years are discounted at an annual rate of 10 percent. households sustain improved behavior for longer tion facilities in rural areas yield benefits in excess (less than) than two years. This figure does not of costs under most assumptions. The programs consider the possible interaction effects between are also justified because the benefits are concen- different interventions (i.e., how the impacts of a trated primarily among the poor. These measures first intervention affect those of a second interven- include drinking water disinfection, hand washing, tion), because data constraints preclude a sound improved rural water supply and safe rural sani- analysis of such effects. tation. The highest priority should be given to the From the analysis presented here, it is clear that drinking water disinfection and hand-washing pro- most measures to improve water supply and sanita- grams. 69 Environmental Health in Nicaragua Table B.23: Benefits and Costs of an Urban Drinking Water Chlorination Program at the Point of Use “Low� “Medium� “High� Target households – urban population not practicing disinfection (millions) 0.5 0.5 0.5 Percent reduction in diarrheal illness per person (Fewtrell and Colford 2004) 23% 23% 23% Program Cost Total program cost (US$ million) 0.2 0.5 2.7 Private Costs Cost of chlorinating drinking water per year (US$ million)* 0.15 0.2 0.3 Program Benefits Cases of diarrheal illness averted per year (thousands) 30 50 65 Deaths in children averted per year <10 <10 <15 Total health benefits (US$ million) 0.6 0.9 1.2 Benefit-Cost Ratios LOW: If behavioral change lasts 1 year 1.6 1.2 0.4 MEDIUM: If behavioral change lasts 2 years* 2.3 1.8 0.7 HIGH: If behavioral change lasts 3 years* 2.6 2.1 0.9 Marginal cost for medium scenario (million NIO per 1% of WSSH health cost 5 reduction) *Benefits and costs in the second and third years are discounted at an annual rate of 10 percent. Figure B.2: Ranking of Interventions to Reduce WSSH Cost in Nicaragua 80 Water supply urban, rural 70 Marginal costs and marginal benefits (Million NIO) Marginal Cost 60 50 40 Improved Improved sanitation sanitation rural urban 30 Disinfection boiling urban HH clorination urban 20 Disinfection Marginal boiling rural Benefit Hygiene 10 HH clorination rural 0 0 5 10 15 20 25 30 35 40 45 50 Per cent CED WSSH reduction 70 Smith et al. (2004) and Desai et al. (2004) report III. Indoor Air Pollution results of health effects from biomass smoke (fuel- This section presents the damage cost of indoor wood, etc.) and coal smoke based on a meta-anal- air pollution and provides estimates of the benefits ysis of available studies. The results are presented and costs of interventions to reduce this damage in Table B24. The relative risks (RRs) represent the cost. The damage cost is from the health effects of risk of health effect or illness relative to the use smoke from solid fuels used in the household envi- of clean fuels such as LPG. The RR for households ronment, as presented by Strukova (2009). using LPG is therefore 1.0. Health Effects of Indoor Air Pollution The strongest evidence of health effects is for It is well documented from studies around the acute lower respiratory illness (ALRI) in children world that indoor air pollution from solid fuels used under age five, COPD in adult females, and lung for cooking and heating in the indoor environment cancer in adult females from coal smoke. Smith et has substantial respiratory health effects. Women al. and Desai et al. do not report a relative risk ratio and young children appear to be most affected be- (RR) for acute respiratory illness in groups above cause they tend to spend more time indoors and/ age five. This is because most of the studies have or closer to the cooking areas. However, studies concentrated on children under age five. However, have also found health effects in men. Ezzati and Kammen (2002) from a study in Kenya Table B.24: Relative Risks for Strong and Moderate Health Outcomes Evidence Health Outcome Group RR CI ALRI Children <5 years 2.3 1.9–2.7 COPD Strong Women >30 years 3.2 2.3–4.8 Lung cancer Women > 30 years 1.9 1.1–3.5 (from coal smoke) COPD Men > 30 years 1.8 1.0–3.2 Moderate-I Lung cancer Men > 30 years 1.5 1.0–2.5 (from coal smoke) Lung cancer (from biomass Women > 30 years 1.5 1.0–2.1 smoke) Children 5–14 years 1.6 1.0–2.5 Asthma Moderate-II All > 15 years 1.2 1.0–1.5 Asthma All > 15 years 1.3 1.0–1.7 Cataracts All > 15 years 1.5 1.0–2.4 Tuberculosis Source: Desai et al. (2004). Notations: RR= relative risk. CI= confidence interval. ALRI=acute lower respiratory infection. COPD=chronic obstructive pulmonary disease. 71 Environmental Health in Nicaragua find that the RR for the 5–49 age group is quite that even with outdoor cooking, the exposure level similar to the RR for children under age five at var- is substantial. Even when cooking is done outside ious levels of pollution levels. the house, the resulting indoor levels of PM and ex- WHO recommends that ventilation level be taken posure of all family members greatly exceed health into account while assessing exposure level in the guidelines for ambient air. However, resulting PM household. Balakrishnan (2004) demonstrated concentrations are lower than with indoor cooking (Table B.25). Table B.25: 24-Hour Exposure Concentrations for Cooks and Noncooks Among Solid Fuel Users Across Kitchen Configurations Type of kitchen N Mean Std. error of mean Cooks Enclosed kitchen Enclosed indoor kitchen with partitions 76 520 56 Enclosed indoor kitchen without partitions 73 540 50 Separate enclosed kitchen outside the house 83 439 52 Outdoor kitchens Outdoor cooking 70 259* 23 Noncooks Enclosed kitchen Enclosed indoor kitchen with partitions 232 264 17 Enclosed indoor kitchen without partitions 155 280* 17 Separate enclosed kitchen outside the house 264 178 11 Outdoor kitchens Outdoor cooking 220 175* 10 * F-statistic significant at P<0.05 as compared to other kitchens. * Significantly different as compared to noncooks in other types of enclosed kitchens Source: Balakrishnan (2004) Table B.25 shows that exposure concentrations A Benefit-Cost Analysis Framework are about 50 percent lower for cooks and 30 per- A benefit-cost analysis of interventions to reduce cent lower for non-cooks. Due to the non-linear indoor air pollution from solid fuels represents a character of the concentration-response function, challenge for many reasons. The relative risks re- WHO suggests the use of a 0.25 coefficient for the ported in Table B.24 represent averages from many estimation of exposure to account for ventilation studies, and do not necessarily reflect the pollution for the households with outside kitchens (Desai exposure situation in households using solid fuels 2004). in Nicaragua. Moreover, the pollution load from 72 solid fuels is not uniform across households. Some However, actual pollution exposure can vary sub- households use unimproved stoves or open fires stantially in each scenario and can depend on while others use improved stoves with chimneys; additional factors such as household ventilation some households use a combination of solid fuels practices, housing characteristics, and household and clean fuels such as LPG. Substantial numbers behavior.Because data on these factors are not of households practice cooking outside the house. readily available at national level, a sensitivity anal- A benefit-cost analysis framework should therefore ysis of relative risk will need to be undertaken in be flexible enough to accommodate these differ- order to assess the likely influence of these factors ences and allow for a sensitivity analysis of param- on the benefit-cost ratios of interventions. eters that will influence the benefits and costs of The next step is to assign population shares and interventions to reduce pollution loads and/or ex- relative risks to each of the stylized situations in posure. Table B.26. According to ENDESA 2006–2007,35 For purposes of this report, six scenarios were se- about 92 percent of the rural population in Nica- lected that represent six stylized situations com- ragua relies on fuelwood and charcoal as their monly found in most developing countries (Table main fuel for cooking. In urban areas, about 32 to B.26). These stylized situations represent the pol- 38 percent of the population relies on these fuels. lution loads from solid fuel use reasonably well. There is a total annual fuelwood consumption of Table B.26: Fuels, Stove Technology and Pollution Scenarios Stylized Situation Stylized Description Relative Risk I. Unimproved wood stoves or open Low energy efficiency. No chimney or ventilation fire with kitchen located inside the Very High device. Very high indoor pollution load. house II. Unimproved wood stoves or open Low energy efficiency. Natural ventilation. fire with kitchen located outside the High Relatively high pollution load. house Relatively low energy efficiency. Chimney (or other ventilation device) taking much of the III. Improved wood stoves smoke outdoors. Still relatively high indoor High pollution load if stove/chimney is not well maintained. IV. Unimproved wood stoves and LPG Pollution load reduced in proportion to the use of Medium (or other clean fuel) LPG (relative to situation I.). V. Improved wood stoves and LPG (or Pollution load reduced in proportion to the use of Medium to Low other clean fuel) LPG (relative to situation II.). VI. LPG or other clean fuel Absence of smoke from solid fuels. Low Note: The framework in this table is very similar to the exposure scenario framework presented in Prüss et al. (2002), which was applied in the water-sanitation-hygiene section of this report. 35 Estimates are slightly different in Censo 2005 and Prolenia 2005. 73 Environmental Health in Nicaragua 2.6 million tons (Prolenia 2005) A total of nearly proved stoves. While there is limited guidance from 0.6 million households use fuelwood and charcoal the international literature, several studies from as their primary energy for cooking. around the world present measurements of par- Very little information is available in Nicaragua to ticulate exposure from solid fuel use, particularly assign relative risks to each of the stylized situ- from cooking. Several results from Latin America ations in Table B.26. While some information is are presented in Table B.28. The 24-hour average available from Prolenia 2005) and international lit- concentration levels from open fire or traditional erature, the situation-specific evidence is less firm stoves are many times higher than urban air quality than for water and sanitation. It is therefore neces- standards in most countries. Improved stoves, such sary to apply a base case of relative risks to each as the plancha, produce PM2.5 or PM3.5 levels that of the stylized situations and perform a sensitivity are often only 20 percent of concentration levels analysis. from an open fire, and are even found to be less than 10 percent of that of an open fire in a study The base case for population shares and rela- in Guatemala by McCracken and Smith (1998). Ac- tive risks is presented in Table B.27. It is almost cording to the figures in Table B.28, the reduction in exclusively the rural population in Nicaragua that PM2.5 seems to even larger than reductions in PM10. uses solid fuels as its main cooking fuel. Rural and While the indoor concentration levels of PM can be urban population shares are applied in Table B.27. substantially reduced by using improved stoves, Shares of improved stoves and outdoor kitchens it is not clear that the health benefits are propor- were estimated using Prolenia 2005. tional. The concentration levels of PM, even with an An important factor is the reduction in “excess� improved stove, are still substantially higher than risk from using improved stoves instead of unim- those found in most outdoor urban environments. Table B.27: Base Case Estimation of Scenario-Specific Relative Risks in Nicaragua Rural Urban Stylized Situations RR RR Pop. RR Pop. RR for COPD for COPD Share for ARI Share for ARI in Women in Women I. Unimproved wood stoves or open fire 43% 2.8 4.1 8% 2.8 4.1 inside II. Unimproved wood stoves or open fire 44% 1.9 2.6 24% 1.9 2.6 outside III. Improved wood stoves 3% 1.9 2.6 7% 1.9 2.6 IV. Unimproved stoves and LPG (or other 0% 1.5 2.5 5% 1.5 2.5 clean fuel) V. Improved stoves and LPG (or other 2% 1.0 1.8 5% 1.0 1.8 clean fuel) VI. LPG or other clean fuel 8% 1.0 1.0 76% 1.0 1.0 Weighed average risk of I to VI 2.2 2.0 1.3 1.0 74 Table B.28: Particulate (PM) Concentrations from Cooking Stoves Open fire/ Improved LPG Traditional Stove Stove 24-hour PM3.5 1,930 330 Guatemala. Albalak et al. (2001). 24-hour PM10 1,210 520 140 Referenced in Albalak et al. (2001), 24 hour PM2.5 520 88 45 adapted from Naeher et al. (2000). 24-hour PM2.5 868 152 PM10 600–1,000 300 50 Mexico. Saatkamp et al. (2000) Table B.29: Odds Ratios for ARI PM10 (µg/m3) Children under age five Age group 5–49 <200 1.0 1.0 200–500 2.42 3.01 500–1,000 2.15 2.77 1,000–2,000 4.30 3.79 2,000–3,500 4.72 2,000–4,000 4.49 >3,500 6.73 4,000–7,000 5.40 >7,000 7.93 Source: Ezzati and Kammen (2002). Table B.29 provides the results of a study by Ezzati With regard to chronic respiratory illness, Albalak et and Kammen (2002) that presents odds ratios for al. (1999) find a 60 percent reduced risk of chronic ARI in relation to PM10 indoor concentrations levels. bronchitis from outdoor cooking with solid fuels A halving of PM10 levels from 1,000–2,000 to 500– compared to indoor cooking with the same fuels 1,000 µg/m3 can correspond to the concentration in a study from Bolivia. However, it should be noted levels associated with the use of an unimproved that outdoor cooking is not free from fine particu- stove versus an improved stove. This reduction in late exposure. Studies have found that those en- concentration levels is associated with a reduction gaged in outdoor cooking activities are exposed to in the odds ratio of about 50 percent for children elevated levels of pollution. Dennis et al. (1996) in a under age five and 25 percent for the 5–49 age study from Bogotá find that individuals who did not groups. The 50 percent risk reduction has there- use solid fuels in the household had an almost 75 fore been applied in the base case in Table B.27 for percent lower prevalence rate of (chronic) obstruc- ARI to establish the relative order of magnitude of tive airways disease than those who had lived in the RR in situation I and III. households with solid fuel use. Similar results were 75 Environmental Health in Nicaragua found in a study from Mexico for chronic bronchitis, Stove Efficiency with an even greater difference in prevalence in in- Figure B.3 presents an energy efficiency ladder for dividuals with longer life exposure to solid fuel pol- stoves and their typical costs; this ladder is often lution (Pérez-Padilla et al. 1996). In a cost-benefit cited in the research literature on fuel use and in- analysis, a 50 percent “excess� risk reduction in door air pollution (e.g., Baranzini and Goldemberg COPD is applied for switching from open fire/tradi- 1996; Luo and Hulscher 1999; and Saatkamp et tional stove to an improved stove with chimney or to al. 2000). The stove efficiency ladder provides a the transfer of a kitchen outdoors. generic perspective on potential energy savings Using the parameter values discussed above, the from improved wood and charcoal stoves and kero- estimated RRs are presented in Table B.27. Only sene, LPG, and electric stoves in comparison to tra- COPD in women and ARI are included. While there ditional stoves. According to Figure B.3, improved is moderate evidence for an increased risk of COPD wood and charcoal stoves are about 50 percent in men, this is not included in this report. For ARI, more efficient than traditional stoves, and LPG and however, children under age five and women over electric stoves are twice as efficient than improved age 15 are included, although ARI in adult women is wood and charcoal stoves. not included in the meta-analysis results presented A study of stove efficiency in Colombia was con- in Smith et al. (2004) and Desai et al. (2004). How- ducted by the National University of Colombia’s De- ever, the study by Ezzati and Kammen (2002) pro- partment of Physics. The results of the study were vides some evidence of similar ARI risk ratios for provided by UPME for the purposes of this report, children under age five and the 5–49 age group. and are presented in Figure B.4. The study esti- The population shares and the relative risk ratios mated the stove efficiency for three different sizes in Table B.27 allow a base case estimation of the of pots/kettles. The efficiency for electric plates health benefits of interventions such as improved was around 80 percent for all three pot sizes. LPG ventilation, improved stoves and switching to clean and natural gas stoves had an efficiency of 45 to fuels (Table B.26). However, in order to estimate 55 percent for the two smallest pot sizes. The ef- benefit-cost ratios, the costs of interventions need ficiency for the largest pot size was more than 75 to be estimated. In addition to costs of improved percent. For a “wood firebox,� an improved wood stoves and LPG stoves (or stoves for other clean stove, the efficiency was less than 10 percent for fuels), this involves an assessment of household the two smallest pots, but about 30 percent for the energy consumption for cooking in order to esti- largest pot. The efficiency of “fuelwood,� or a tradi- mate the recurrent cost of LPG requirements or tional open stove, was less than 5 percent for the other clean fuels. Moreover, for many households smaller pots, and 20 percent of the largest pot. the time savings from less fuelwood collection may For the most part, the results from this Colombian be an important benefit of switching to improved study are consistent with the generic efficiency stoves or clean fuels. Fuelwood consumption must ladder in Figure B.4. A notable difference is the therefore be estimated for the various stylized situ- wood stove efficiencies for the two smallest pot ations or scenarios in Table B.27. A first step in es- sizes. However, a study of the plancha woodstove timating the costs of interventions is therefore to in Guatemala found an efficiency of around 10 per- consider stove efficiencies. cent, very similar to the efficiency of using open fire. 76 Figure B.3: Stove Efficiency and Capital Costs 80 70 60 and Stove Cost (US$) Stove Efficiency (%) 50 40 Efficiency 30 Stove Cost 20 10 0 Traditional wood stove Improved wood stove Karosene wick stove Karosene presure stove LPG stove Electric hot plate Traditional charcoal stove Imporoved charcoal stove A modified version of the plancha had an efficiency also removes about 90 percent of toxic smoke of around 12 percent (Boy et al. 2000). McCracken from the kitchen (Conway et al. 2006). In 2004, and Smith (1998) found an efficiency of 13 to 14 with a grant from the United States Environmental percent for the plancha and open fire. Saatkamp et Protection Agency (US EPA), and with technical as- al. (2000) presented estimates of stove efficiencies sistance from Aprovecho, TWP and AHDESA began in Jaracuaro, Mexico, and found that the efficiency developing commercial models based on the ex- of the lorena woodstove is not much different than isting Rocket and EcoStove designs. Efficiency of that of traditional open stoves. the Justa stoves introduced through this project in Trees, Water, & People (TWP), the Honduran As- Nicaragua was not estimated, but for the purpose sociation for Development (AHDESA), and the of the study was considered in the same range as Aprovecho Research Center are undertaking the in Colombia. “Micro-Enterprise Stove Project� to introduce safer The Justa woodstoves have been adopted by many and more environmentally friendly cook stoves communities in Latin America. While these stoves for the people of Honduras. The team came to- are often not very efficient, they are equipped with gether after Hurricane Mitch devastated Honduras a chimney that removes the smoke from the indoor in 1998 to adapt the Rocket stove technology to environment and therefore have important health cooking conditions there. Working along with local and quality-of-life benefits. Several versions and Honduran women, Aprovecho stove technicians offsprings of these stoves have also been devel- came up with the Justa stove design in 1999. Justa oped in order to improve stove efficiency (Conway stoves save an average of 50 to 70 percent in fire- et al. 2006). wood compared to traditional stoves. The Justa 77 Environmental Health in Nicaragua Figure B.4: Stove Efficiency From a Study in Colombia 90 80 70 60 Stove Efficiency (%) 50 2.9 ltr pot 5.5 ltr pot 40 7.2 ltr pot 30 20 10 0 Electric LPG Natural gas Wood Firebox Fuel Wood Source: Study conducted by the National University of Colombia, Department of Physics. The results of the study were provided by UPME. Fuelwood Consumption Table B.30: Estimated Annual Household Among households that use fuelwood, there are Consumption of Fuelwood typically three types of situations. Some house- Fuelwood Technology: holds use open fire/traditional woodstoves. Others consumption (tons) use improved woodstoves. Finally, some house- Unimproved stoves 3.8 holds use fuelwood in combination with LPG or Improved stoves 2.6 another type of cleaner fuel. Based on data from Unimproved stoves and 1.9 UPME, it is estimated that the average household LPG stoves that uses fuelwood has a consumption of 3.01 tons Improved stoves and LPG 1.3 stoves per year. However, household consumption, for any given household size, will tend to vary depending tion that the groups of households have on average on the efficiency of the woodstove and whether or the same demand for cooking energy. not the household also uses other types of fuels for cooking. It is important to assess the realism of fuelwood consumption in these four cases. Saatkamp et al. Fuelwood consumption was estimated for different (2000) estimates fuelwood consumption in Jarac- scenarios using the corresponding share of popula- uaro, Mexico, in the range of 0.5–1.0 kg per meal tion with each stove type and stove efficiency from per person for the lorena stove and open fire. For Figure B.4. The estimates are presented in Table two meals a day and an average household size of B.30. These estimates are based on the assump- five, this implies an annual fuelwood consumption 78 Table B.31: Annual Recurrent Cost of Complete Cost of Interventions Fuel Substitution Estimated annual recurrent costs of switching from Annual Cost Per fuelwood to commercial energy, or from a combi- Household (000 NIO) nation of fuelwood and LPG to LPG only, are pre- LPG Electricity sented in Table B.31. At current prices (in 2008) (25 lb bottles) (urban only) of energies in Nicaragua, LPG and electricity are Unimproved stoves 2.8 2.5 found to be very close price options. Using elec- Improved stoves 2.8 2.5 tricity to replace fuelwood is found to be slightly Unimproved stoves and less expensive than LPG, but it is difficult to ensure 1.4 - LPG stoves adequate access to electricity in rural areas. The Improved stoves and LPG stoves 1.4 - cost is the same for households using unimproved and improved woodstoves (with 10 and 15 percent of 1.8–3.6 tons per year. For three meals a day, the efficiency). This is because of the assumption that annual consumption would be 2.7–5.4 tons per they have the same effective energy demand. LPG year. This is in the range of the estimated fuelwood therefore seems to be the most viable option. consumption in Table B.30. Table B.32: Baseline Parameters for Estimating the Cost of Fuel Substitution Fuelwood Average gross energy content of fuelwood (MJ per kg) 19 Average net energy content of fuelwood (MJ per kg) 11 Propane (LPG) Average stove efficiency (propane) 55% Average energy content of propane (Btu/gallon) 92,000 Cost per bottle (25 lb bottles) - NIO 175* Electricity Average stove efficiency 80% Cost of electricity (NIO./kWh) 2** Sources: Energy contents of LPG and natural gas are based on data from UPME. Gross and net energy content of fuelwood (gross and net heating content) and stove efficiencies are from a study by the National University of Colombia and provided by UPME; * Instituto Nicaragüense de Energía (INE) 2008; **IPC-BCN, Canasta Básica 2008. 79 Environmental Health in Nicaragua It should be noted that the estimated costs in Benefit-Cost Analysis Table B.31 are likely to represent underestimates Five intervention scenarios considered for Nica- of the costs of providing commercial energy to dis- ragua represent four stylized situations commonly tant rural areas. Currently less than 1 percent of found in most developing countries. These stylized rural households in Nicaragua are already using situations represent reduction of pollution loads LPG and other commercial energies, according to from solid fuel use reasonably well. However, actual ENDESA 2006–2007. Thus, supplying these fuels pollution exposure can vary substantially in each to other rural households may not be substantially scenario, and can depend on additional factors more expensive than to those households already such as household ventilation practices, housing using commercial fuels. characteristics and household behavior. Table B.32 presents the baseline parameters that Table B.34: Interventions were used to estimate the cost of switching to Scenarios: Description: commercial energy. These parameters are based on data from UPME and results of a study by the From unimproved stove From I (50% HH) to II inside to better ventilation National University of Colombia, as well as Nicara- From unimproved stove guan national statistics. From I (50% HH) to III inside to improved stoves Capital costs and annualized costs of improved From III to V From improved stove to LPG stoves and LPG stoves are presented in Table B.33. From unimproved stove and From IV to VI Capital costs are adapted from Prolenia (2005).36 LPG mix to LPG only While simple improved stoves used in many devel- From V to VI From improved stove and oping countries are found to be quite inexpensive, LPG mix to LPG only as presented in Figure B.2, the type of improved stoves commonly used in Latin America is far more A benefit-cost analysis is undertaken for the three expensive, as indicated in Table B.33. Annualized of five interventions presented in Table B.34. They costs are calculated by using a useful life of 10 are based on the stylized situations in Table B.26. years and an annual discount rate of 10 percent. The estimated benefits and costs of these inter- ventions are presented in Table B.35 and B.36 for Table B.33: Cost Estimates of Improved Wood rural and urban areas. Avoided cases of ARI and Stoves and LPG Stoves COPD are estimated from the relative risk ratios in Table B.24 and baseline estimates of annual cases NIO of ARI and COPD. Baseline cases are presented in Capital Cost: Low High Strukova (2009). The monetary benefits of avoided Traditional stove 170 cases are calculated from the estimated unit costs Improved stove 165 220 of ARI and COPD morbidity and mortality, also pre- LPG stove 315 315 sented in Strukova (2009). Unit costs of morbidity include medical treatment cost, and value of time Annualized Cost: losses (at 75 percent of wages). Child mortality is Traditional stove 26 valued by using the human capital approach (HCA) Improved stove 40 240 of discounted life earnings losses. Adult mortality is 36 Some traditional stove prices are presented in Estufas Tradicionales (1995) http://www.gtz.de/de/dokumente/es-estufas-en-imag2-1995.pdf. 80 valued by the HCA as a lower bound and the value Improved ventilation and switching from unim- of statistical life (VSL) as a higher bound. VSL is de- proved and improved stoves in individual house- rived from benefit transfer from the United States holds is found to have substantially higher benefits and other high-income countries using an income than costs (Tables B.35 and B.36). The benefit-cost elasticity of 1.0, based on the mid-point value from ratio is estimated at above 3 for all four of these the range of US$2 to US$5.4 million, as in Strukova interventions. The benefit-cost ratio is lower than 1 (2009). The total health benefits of the interven- if there is switching to LPG due to high LPG prices. tions are greatly influenced by the choice of valua- For households with improved stoves, the health tion techniques for adult mortality. benefits alone are not large enough to outweigh Annual costs of interventions include program cost, the cost of switching to LPG. Although promotion stove cost and LPG cost. LPG represents 80 to 90 of improved stoves is a very attractive interven- percent of total cost in interventions (3). Interven- tion, the merits of promoting LPG in individual tions (4) and (5) from Table B.34 are not assessed rural households are uncertain. LPG prices would since LPG costs greatly outweigh all benefits from have to be reduced dramatically for the estimated morbidity and mortality reduction. A tentative esti- benefits to exceed costs. Therefore, it seems that mate of the program cost of promoting and imple- LPG will have a chance of success only in better-off menting improved stoves and LPG fuel switching, households. and sustaining a stove inspection and maintenance The various interventions are summarized in terms program, is also included in Tables B.35 and B.36. of their contribution to reduced environmental Only health benefits are included for the benefit- damages and costs per one percent of health dam- cost ratio estimation. ages reduction (marginal cost),37 as was done for Table B.35: Benefits and Costs of Indoor Air Pollution Control in Rural Nicaragua Better ventilation Improved stoves LPG from from unimproved from unimproved improved stove inside stoves inside stoves Population receiving intervention (million) 0.5 0.5 0.07 ARI cases averted per year (thousand) 400 400 50 ARI deaths in children averted per year 36 36 5 COPD cases averted per year 110 110 20 COPD deaths averted per year 10 10 <5 Annual health benefits (million NIO) 96 96 14 Program cost (million NIO) 13 13 2 Annualized stove cost, (million NIO) 3 22 4 Annual cost of LPG (million NIO) 0 0 40 Benefit-cost ratio (health benefits only) 6.0 2.7 0.3 Marginal cost (million NIO per 1% of IAP 1.5 3.2 30 health cost reduction) 37 Approximated by the average cost to implement intervention per one percent of IAP cost reduction. 81 Environmental Health in Nicaragua Table B.36: Benefits and Costs of Indoor Air Pollution Control in Urban Nicaragua Better ventilation Improved stoves LPG from from unimproved from unimproved improved stove inside stoves inside stove Population receiving intervention (million) 0.1 0.1 0.03 ARI cases averted per year (thousand) 80 80 20 ARI deaths in children averted per year 10 10 <5 COPD cases averted per year 50 50 15 COPD deaths averted per year <10 <10 <5 Annual health benefits (million NIO) 50 50 10 Program cost (million NIO) 4 4 1 Annualized stove cost, (million NIO) 1 6 2 Annual cost of LPG (million NIO) 0 0 20 Benefit-cost ratio (health benefits only) 10.6 4.9 0.5 Marginal cost (million NIO per 1% of IAP 0.8 1.8 17 health cost reduction) the water and sanitation programs (Figure B.5). by about 37 percent per year. This reflects better Marginal cost reflects only program cost and pri- ventilation and switching to improved stoves in 43 vate household cost, without taking into account percent of rural households, and eight percent of time and fuelwood savings. Marginal benefit is ap- urban households; switching to LPG from improved proximated by the value of one percent of indoor stoves in three percent of rural and one percent of air pollution cost reduction, estimated at 9 million urban households. NIO. Improved ventilation and household switching The analysis presented here supports the unquali- from unimproved to improved stoves have the fied recommendation to shift households that largest reduction in damages in rural areas. This have unimproved stoves to improved ones and to is followed by households switching to LPG alone improve ventilation. The results of other interven- from improved stoves in urban and rural areas. tions, such as from unimproved stoves to LPG or Each of these measures contributes a smaller from improved stoves to LPG, depend on the costs amount of reduction in environmental damages; and benefits used. Therefore, a more detailed marginal costs for these two interventions slightly analysis needs to be carried out for such changes, exceed marginal benefits. In total, the former four looking at specific cases and taking into account interventions reduce the cost of health effects other benefits. 82 Figure B.5: Ranking of Interventions for IAP Cost Reduction in Nicaragua 35 U->I: From unimproved to improved stoves I->C: From improved stoves to LPG I->C rural Marginal costs and marginal benefits (Million NIO) 30 Marginal Costs 25 20 I->C 15 10 Marginal Health Benefits Improved ventilation U->I rural 5 U->I Improved ventilation 0 0 5 10 15 20 25 30 35 40 CED IAP reduction (%) 83 References Abbey, D. et al. (1995). Long-Term Ambient Concentra- Boy, E., Bruce, N., Smith, K., and Hernández, R. (2000). tions of Particulates and Oxidants and Development of Fuel efficiency of an improved wood-burning stove in Chronic Disease in a Cohort of Nonsmoking California rural Guatemala: Implications for health, environment Residents. Inhalation Toxicology, Vol 7: 19–34. and development. Energy for Sustainable Development. Vol. IV, No. 2. August 2000. Ahmed, K., Awe, Y., Barnes, D., Cropper, M., and Kojima, M. (2005). Environmental Health and Traditional Fuels Braghetta, A. (2006). Drawing the Connection between Use in Guatemala. Directions in Development. ESMAP. Malnutrition and Lack of Safe Drinking Water in Guate- World Bank. mala. www.livingwatersfortheworld.org/docs/drawing- the-connection-en.doc Alarcón, D., and Robles, M. (2007). The Challenges of Measuring Child Mortality when Birth Registration is Brugman, A. (2004). Diseño de un Programa de Instru- Incomplete. Global Forum on Gender Statistics. Rome, mentos Económicos para el Manejo y Control de la Con- Italy. http://unstats.un.org/unsd/Demographic/meet- taminación Atmosférica Urbana en Colombia. Prepared ings/wshops/Gender_Statistics_10Dec07_Rome/ for MAVDT, Colombia. December 2004. docs/8.1_Alarcón.pdf Conway, S., Africano, S., Hao, J. (2006). Nicaragua Micro- Albalak, R., Bruce, N., McCracken, J., Smith, K., and De Enterprise Stove Project: A Market-Based Pilot Project Gallard, T. (2001). Indoor Respirable Particulate Matter for Increasing the Use of Safer Cook Stoves. Partnership Concentrations from an Open Fire, Improved Cook Stove, for Clean Indoor Air Bulletin, Oct. 2006, No. 9. and LPG/Open Fire Combination in a Rural Guatemalan Community. Environmental Sciences and Technology. Countness, R. (2003). Reconciling Fugitive Dust Emis- Vol. 35, No. 13. sion Inventories with Ambient Measurements. Pre- sented at the 12th International Emission Inventory Con- Albalak, R., Frisancho, A.R., and Keeler, G.J. (1999). Do- ference. San Diego, CA, April 29–May 1, 2003. mestic biomass fuel combustion and chronic bronchitis in two rural Bolivian villages. Thorax. Vol. 54. Cropper, M. and Oates, W. (1992). Environmental Eco- nomics: A Survey. Journal of Economic Literature. Vol. Balakrishnan, B. et al. (2004). Exposure assessment for XXX, pp. 675–740. respirable particulates associated with household fuel use in rural districts of Andhra Pradesh, India. Journal of Curtis, V. and Cairncross, S. (2003). Effect of Washing Exposure Analysis and Environmental Epidemiology 14, Hands with Soap on Diarrhoea Risk in the Community: S14–S25. A Systematic Review. Lancet Infectious Diseases. Vol 3. Baranzini, A. and Goldemberg, J. (1996). Desertifica- Dennis, R., Maldonado, D., Norman, S., Baena, E., and tion, energy consumption, and liquefied petroleum gas Martínez, G. (1996). Woodsmoke Woodsmoke exposure use, with an emphasis on Africa. Energy for Sustainable and risk for obstructive airways disease in women. Chest Development. Vol. II, No. 5. January 1996. 109, pp. 115-9. Blössner, M., de Onis, M.(2005). Malnutrition: quanti- Desai, M.A, Mehta, S., and Smith, K. (2004). Indoor fying the health impact at national and local levels. WHO Smoke from Solid Fuels: Assessing the Environmental Environmental Burden of Disease Series, No. 12. Ge- Burden of Disease at National and Local Levels. En- neva, World Health Organization, 2005. vironmental Burden of Disease Series, No. 4. World Health Organization. Borghi, J., Guinness, L., Ouedraogo, J., and Curtis, V. (2002). Is Hygiene Promotion Cost-Effective? A Case Dockery D.W., Pope C.A. III, Xu X. et al. (1993). An associ- Study in Burkina Faso. Tropical Medicine and Interna- ation between air pollution and mortality in six US cities. tional Health, Vol. 7 No. 11. November 2002. New England Journal of Medicine, 329: 1753–1759. Encuesta de Hogares sobre Medición del Nivel de Vida, EMNV (2005). INIDE. Nicaragua 85 Environmental Health in Nicaragua Encuesta Nacional de Leña, ENDL, (2006–2007). Multi- Fishman S., Caulfield L.E., de Onis M., Blossner M., Mul- consult & Cia. Ltda. lany L., Black R.E. (2004). Childhood and maternal un- derweight. In: Ezzati M., López A.D., Rodgers A., Murray Encuesta Nicaragüense de Demografía y Salud, ENDESA C.J.L., eds. Comparative quantification of health risks: 2006–2007, Nicaragua global and regional burden of disease attributable to selected major risk factors. Geneva, World Health Orga- Esrey, S.A., Potash, J.B., Roberts, L., and Shiff, C. (1991). nization. Effects of Improved Water Supply and Sanitation on As- cariasis, Diarrhoea, Dracunculiasis, Hookworm infec- Gómez, O., (2008). Producción y distribución son dos tion, Schistosomiasis, and Trachoma. WHO Bulletin, Vol. gotas de agua envasada. http://www.elnuevodiario. 69, No. 5: 609 – 621. com.ni/especiales/20749 Esrey, S.A., and Habicht, J.P. (1988). Maternal literacy Huttly, S., Morris, S. and Pisani, V. (1997). Prevention modifies the effect of toilets and piped water on infant of Diarrhoea in Young Children in Developing Countries. survival in Malaysia. American Journal of Epidemiology, WHO Bulletin. Vol. 75, No. 2. 127(5): 1079–1087. Kochi, I., Hubbell, B., and Kramer, R. (2006). An Em- Esrey, S.A. (1996). Water, Waste and Well-being: A Mul- pirical Bayes Approach to Combining and Comparing ticountry Study. American Journal of Epidemiology, Vol. Estimates of the Value of a Statistical Life for Environ- 143, No. 6. mental Policy Analysis. Environmental and Resource Economics. 34 (3): 385–406 Estrategia Sectorial de Agua Potable y Saneamiento, 2005–2015 (2005). República de Nicaragua. Lantagne, D., Quick, R. and Mintz E. (2005). Household Water Treatment and Safe Storage Options in Devel- Estufas Tradicionales (1995) http://www.gtz.de/de/do- oping Countries: A Review of Current Implementation kumente/es-estufas-en-imag2-1995.pdf Practices (CDC). http://s189535770.onlinehome.us/ Ezzati, M. and Kammen, D. (2002). The Health Impacts pottersforpeace/wp-content/uploads/household_ of Exposure to Indoor Air Pollution from Solid Fuels in water_treatment.pdf Developing Countries: Knowledge, Gaps, and Data Larsen, B. (2003). Hygiene and Health in Developing Needs. Discussion Paper No. 02–24. Resources for the Countries: Defining Priorities through Cost-Benefit As- Future. Washington DC. sessments. International Journal of Environmental Exposure and Risk for Obstructive Airway Disease among Health Research. Vol 13: S37–S46. Women. Chest 109: 115–119. 1996. Larsen, B. (2004a). Cost of Environmental Degradation: Fewtrell, L. and J. Colford Jr. (2004). Water, Sanitation A Socio-Economic and Environmental Health Assess- and Hygiene: Interventions and Diarroea – A systematic ment in Damietta, Egypt. Prepared for SEAM II, Cairo, review and meta-analysis. HNP Discussion Paper. World Egypt. Bank. Larsen, B. (2004b). Colombia. Cost of Environmental Fewtrell, L., Prüss-Üstün, A., Bos, R., Gore, F., Bartram, Damage: A Socio-Economic and Environmental Health J. (2007). Quantifying the health impact at national and Risk Assessment. Final Report. Prepared for the Ministry local levels in countries with incomplete water supply of Environment, Housing and Land Development of the and sanitation coverage. Environmental Burden of Dis- Republic of Colombia. ease Series, No. 15. WHO. Geneva. Larsen, B. (2005). Cost of Environmental Damage: Anal- Figueroa de la Vega, F. (2009) Diagnóstico Energético ysis by Department. Prepared for MAVDT, Colombia. del Sector Transporte y Proyección del Consumo de Larsen, B., Strukova, E. (2005). Peru. Cost of Environ- Combustibles y Biocombustibles y Emisiones GEI en mental Damage: A Socio-Economic and Environmental Centroamérica al año 2020. Proyecto CEPAL/República Health Risk Assessment. Prepared for the World Bank. de Italia. Luo, Z. and Hulscher, W. (1999). Woodfuel Emissions. RWEDP, Bangkok. May 1999. 86 McCracken, J. and Smith, K. (1998). Emissions and Ef- Prüss, A., Kay, D., Fewtrell, L. and Bartram, J. (2002). Es- ficiency of Improved Woodburning Cookstoves in High- timating the Burden of Disease from Water, Sanitation land Guatemala. Environment International. Vol. 24, No. and Hygiene at the Global Level. Environmental Health 7. Perspectives. Vol. 110, No. 5. Mead, P. et al. (1999). Food-Related Illness and Death in Quick, R.E. et al. (1999). Diarrhoea Prevention in Bo- the United States. Emerg Infect Dis. Vol. 5, No. 5. livia through Point-of-use Treatment and Safe Storage: A Promising New Strategy. Epidemiol Infect. Vol. 122. Mrozek, J. and Taylor, L. (2002). What Determines the Value of Life? A Meta Analysis. Ramírez, C. (2008). Informe Estudio de Mercado Filtros Caseros Sostenibles. World Bank. WSP. Managua, Nica- Journal of Policy Analysis and Management. Vol 21 (2): ragua. 253–270. Rockhill B., Newman B., Weinberg C. (1998). Use and Naeher LP, Leaderer BP, Smith KR. (2000). Particulate Misuse Of Population Attributable Fractions. Am J Public matter and carbon monoxide in highland Guatemala: in- Health 88: 15–19. door and outdoor levels from traditional and improved wood stoves and gas stoves. Indoor Air. Vol. 10, No. 3. Rutstein, S.O. (2000). Factors Associated with Trends in Infant and Child Mortality in Developing Countries Ostro, B. (1994). Estimating the Health Effects of Air Pol- during the 1990s. Bulletin of the World Health Organiza- lution: A Method with an Application to Jakarta. Policy tion. Vol. 78(10): 1256–1270. Research Working Paper, World Bank. Saade, C., Bateman, M. and Bendahmane, D. (2001). Ostro B. (2004). Outdoor air pollution: Assessing the The Story of a Successful Public-Private Partnership in environmental burden of disease at national and local Central America: Handwashing for Diarrheal Disease levels. Geneva, World Health organization. WHO Environ- Prevention. Published by BASICS II, EHP, Unicef, USAID, mental Burden of Disease Series, No 5. and World Bank. Pérez-Padilla, R. et al. (1996). Exposure to biomass Saatkamp, B., Masera, O. and Kammen, D. (2000). En- smoke and chronic airway disease in Mexican women: ergy and health transitions in development: Fuel use, A case-control study. Am. J. Respir. Crit. Care Med. Vol. stove technology, and morbidity in Jaracuaro, Mexico. 154, No. 3. Energy for Sustainable Development. Vol. IV, No 2. Au- gust 2000. Pinfold, J. and Horan, N. (1996). Measuring the Effect of a Hygiene Behaviour Intervention by Indicators of Scarborough J, Clinton N, Gong P (2002). Creating a Behaviour and Diarrhoeal Disease. Transactions of the Statewide Spatially and Temporally Allocated Agricul- Royal Society of Tropical Medicine and Hygiene, Vol 90. tural Burning Emissions Inventory Using Consistent Emission Factors. Prepared for the Air Resources Board, Pope C. A. III, Thun M.J,. Nambudiri M.M. et al. (1995). California EPA. Particulate air pollution as a predictor of mortality in a prospective study of US adults. American Journal of Re- Schulman, Ronca and Bucuvalas, Inc. (2001). Con- spiratory and Critical Care Medicine, 151: 669–674. fronting COPD in North America and Europe: A Survey of Patients and Doctors in Eight Countries. Pope C.A. III, Burnett R.T., Thun M.J. et al. (2002). Lung cancer, cardiopulmonary mortality, and long-term expo- Shi, A. (1999). The Impact of Access to Urban Potable sure to fine particulate air pollution. Journal of the Amer- Water and Sewerage Connection on Child Mortality: City- ican Medical Association, 287: 1132–1141. Level Evidence, 1993. Economists’ Forum, Vol. 1, World Bank 1999. Prüss-Üstün A., Mathers C., Corvalán, Woodward A. (2003). Introduction and methods: assessing the envi- Shibuya, K., Mathers, C., and López, A. (2001). Chronic ronmental burden of disease at national and local levels. Obstructive Pulmonary Disease (COPD): Consistent Geneva, World Health Organization (WHO Environmental Estimates of Incidence, Prevalence, and Mortality by Burden of Disease Series, No. 1). WHO Region. Global Programme on Evidence for Health Policy. World Health Organization. November 2001. Prolenia, Informe Final Tarea A, 2005 87 Environmental Health in Nicaragua Smith, K., Mehta, S. and Feuz. M. (2004). Indoor Air WHO (2002b). The World Health Report 2002. The Pollution from Household Use of Solid Fuels. In: Ezzati World Health Organization. M., Rodgers A.D., López A.D., and Murray C.J.L. (eds): Comparative Quantification of Health Risks: Global and WHO/UNICEF. (2008) Joint Monitoring Programme Regional Burden of Disease due to Selected Major Risk for Water Supply and Sanitation (JMP). Progress on Factors. World Health Organization, Vol. 2. Drinking Water and Sanitation: Special Focus on Sanita- tion. UNICEF, New York and WHO, Geneva. Strukova, E. (2009). Nicaragua. Health Costs of Environ- mental Damage. Working Paper. World Bank. World Bank (2002a). Egypt Cost Assessment of Envi- ronmental Degradation. Report No. 25175-EGT. Wash- University of the Andes (2004). Diseño e Implemen- ington DC. tación de un Modelo de Calidad de Aire para Bogotá. Informe Semestral 6. Civil Engineering and Environment World Bank (2008) El estado del saneamiento en Nica- Department. CIIA. ragua. Resultados de una evaluación en comunidades rurales, pequeñas localidades y zonas periurbanas. Varley, Tarvid and Chao (1998). A reassessment of the WSP-LAC. cost-effectiveness of water and sanitation interventions in programmes for controlling childhood diarrhoea. Bul- World Bank (2009). World Development Indicators letin of the World Health Organization, 76(6):617–631. 2009. Washington DC. Viscusi, W.K. and Aldi, J.E. (2002). The Value of a WSP. (2008). El estado del saneamiento en Nicaragua. Statistical Life: A critical review of market estimates Resultados de una evaluación en comunidades rurales, throughout the world. Discussion Paper No. 392. Har- pequeñas localidades y zonas periurbanas. World Bank. vard Law School. Cambridge, MA. United States. http://www.who.int/evidence/nbd WHO (2001). Global Burden of Disease 2001. The World http://www.worldbank.org/nipr/Atrium/mapping.html. Health Organization. url WHO (2002a). Global Burden of Disease 2002. The World Health Organization. 88 Publications from the LCSEN Occasional Paper Series Environment & Water Resources n Empowering Women in Irrigation Management: The Sierra in Peru (2013) n Environmental Health in Nicaragua: Addressing Key Environmental Challenges (Originally Published in 2010, Republished in 2013) (Available in Spanish and English) n Expanding Financing for Biodiversity Conservation: Experiences from Latin America and the Caribbean (Available in English (2012) and Spanish (2013)) n Overcoming Institutional and Governance Challenges in Environmental Management. Case Studies from Latin America and the Caribbean Region (2013) n Policy and Investment Priorities to Reduce Environmental Degradation of the Lake Nicaragua Watershed (Cocibolca) (Originally Published in 2010, Republished in 2013) (Available in Spanish and English) n Uncertain Future, Robust Decisions; The Case of Climate Change Adaptation in Campeche, Mexico (2013) To find copies of these publications, please visit our website: www.worldbank.org/lac 89 Latin America & Caribbean Region Environment & Water Resources Occasional Paper Series