pproaches 39608 M a r c h 2 0 0 7 N o t e N u m b e r 1 3 Output-based aid in health The Argentine Maternal-Child Health Insurance Program by Lars Johannes T o fight infant mortality in the poorest provinces of Ar- How the program works gentina, local authorities and the World Bank set up the To respond to the growing crisis in health, particu- Maternal-Child Health Insurance Program in 2004. larly in maternal and child health, the government of The program is administered by provincial governments, which Argentina and the World Bank in 2004 created the receive funding on the basis of the numbers of mothers and Maternal-Child Health Insurance Program, known to children enrolled and the performance on results-based "trac- the Argentine public as Plan Nacer. ers"--sets of indicators measuring service delivery and quality. The services are provided by existing health care facilities, which Targeting of subsidies receive a standard payment per patient and per service provided. Using an output-based aid approach, the program fo- The health care facilities compete on the basis of quality. cuses on providing basic health services to the poorest Even in the 1990s, when Argentina had strong groups in the poorest provinces of Argentina, located economic growth and ranked 17th in the world in per in the northern part of the country. The targeted prov- capita health spending, the country had difficulty in inces have an infant mortality rate significantly above providing access to health care for its poorer citizens. Its the national average. The goal is to reach 80 percent health system ranked 71st on a performance index cal- of the target population (600,000 beneficiaries) by the culated by the World Health Organization. The system end of the third year. was (and still is) fragmented and geared primarily to the All pregnant women and children up to age six who formally employed. Meanwhile, a large part of the poor live in a participating province and are not covered by population, working mostly in the informal sector, had the existing health insurance system are eligible to en- no health coverage. During Argentina's economic crisis roll free of charge. The package of services provided by in 1997­2002 the disparities grew as per capita health the program includes about 80 cost-effective interven- spending dropped from $687 in 1998 to $233 in 2002. tions targeting the main causes of infant and maternal Many more poor Argentines lost their health insur- mortality. Health services are contracted from indepen- ance during the crisis. Formal employment with health dent third-party providers (public and private), certified benefits was replaced by lower-paying jobs in the as able to provide the services included in the package. informal sector. The poverty rate rose by 20 percent- age points--reaching 50 percent of the population Implementation in 2004--and the number of extremely poor nearly The program is implemented in each participating doubled. The tough economic conditions made it diffi- province by a health service purchasing unit supervised cult for people to compensate for the loss of insurance by the provincial government, under an implementa- through out-of-pocket spending. At the same time tion agreement with the national government. The public health programs were unprepared for the big growth in demand for their services. All these constraints on health care showed up in Lars Johannes works for the Global Partnership on Output- health indicators. In 2004 infant mortality and the Based Aid. incidence of infectious diseases in mothers and chil- Special thanks for comments and contributions to Cristian dren began to rise for the first time in decades, with the Baeza, Task Team Leader on the Maternal-Child Health biggest increases occurring in the poorer provinces. Insurance Program. Supporting the delivery of basic services in developing countries pproaches health service purchasing units are responsible for en- Figure 1 Output-based program funding rolling participants in the program and for conducting awareness-raising activities for the target population. The program incorporates an output-based ap- Argentine World Bank government proach in two ways: 50% 50% · Funding is disbursed to the provincial governments Special account administered on the basis of two sets of results: the number of by government Disbursement patients enrolled and the performance on a set of triggered by output and outcome oriented tracers (health indi- Capitations paid verification of after verification of cators) in the target group. outputs outputs · As in a voucher scheme, independent service Auditor providers are reimbursed for eligible treatments provided to patients enrolled in the program. Since 60% on number 40% on achieved enrolled in target tracers (output) beneficiaries are free to choose their service provid- Provision of group (output) er, the providers compete for patients on the basis documentation on enrollment of quality. and service Program funding Provincial governments The program is funded jointly by the government of Health service Argentina and the World Bank. For the first three years purchasing units the government is to provide half the financing and the World Bank the other half. After three years, when the World Bank's financing begins to phase out, the provincial governments will gradually step in to finance its share. enrollment of patients, and the other 40 percent on the The program funds are held in a special account basis of performance on 10 output-oriented indica- from which disbursements are paid on the basis of tors called "tracers" (table 1). The tracers include both performance (figure 1). The health service purchasing targets on the output-level like numbers of vaccina- units receive 60 percent of payments for confirmed tions and on the outcome-level like newborns' average weight at birth. Enrollment and tracer targets are set in annual performance agreements between the national Table 1 Output-based tracers and provincial governments. Tracer Tying disbursement of a large share of the program funds to enrollment targets gives the health service 1. Timely inclusion of eligible pregnant women in prenatal purchasing units an incentive to achieve high enroll- care services ment numbers. Tying another substantial share to 2. Effectiveness of early neonatal and delivery care (Apgar health quality indicators gives them an incentive to Score) ensure high-quality service provision by contracted 3. Effectiveness of pre-natal care and prevention of pre- service providers and to encourage participants to use mature births (birth weight above 2.5 kilos) the services offered. Program funds are released only 4. Quality of pre-natal and delivery care (testing for STDs, after achievement of performance targets are verified vaccine program completed for pregnant women) by independent third-party auditors. 5. Medical Auditing of Maternal and Infant death Because the disbursements are capped by the 6. Immunization Coverage (measles vaccine) number of enrolled population, the provincial govern- 7. Sexual and Reproductive Health Care ments and their health service purchasing units bear 8. Well child care (1 year old or younger) the risk of budget overruns, giving them an incentive 9. Well child care (1 to 6 year old) to negotiate and manage their contracts with service 10. Inclusion Indigenous Populations providers efficiently. But because interventions are paid for on the basis of previously agreed fees, health service Supporting the delivery of basic services in developing countries pproaches purchasing units can pass on some of the operational services below the reimbursement price while still at- risks to the service providers. tracting patients will be able to sustain their participa- tion in the program. Pricing of services The program's market-based approach provides The per-capita payments for services (capitations), a better mechanism for allocating resources than the similar to insurance premiums, are based on an ac- historical approaches of financing inputs via line-item tuarial calculation much like those used by insurance budgets, more commonly used in public health care. companies and health maintenance organizations. For Service providers rely on first-hand information and each intervention the calculation includes a unit price consumer preferences in deciding how to provide and the risk of the target group requiring that interven- services. They are free to invest in more equipment or tion, based on historic reference data for a population better facilities, since only they bear the risk of that with similar characteristics. investment. Similarly, they are free to offer financial Unit prices for each intervention covered were incentives for good performance on top of salaries to determined through a costing exercise based on the attract qualified personnel. And because the scheme prices of inputs for the provider. To confirm that the gives participating providers an incentive to increase unit prices are reasonable, a benchmarking exercise productivity, it may help reduce absenteeism (an issue compared them with prices for comparable services by in Argentina's public health care system, according to existing providers in Argentina and with costs in com- Di Tella and Savedoff 2001). parable projects in the region. Results so far Service provision The results in the initial phase of the program have Service providers are reimbursed on the basis of the been promising. Between January 2005 when the pro- number of interventions performed. Beneficiaries' gram began to operate and March 2006 the number freedom to choose their service providers makes this reimbursement a demand-side subsidy, analogous to a voucher scheme for a pre-defined benefit package (for Figure 2 Output-based service provision more on demand-side subsidies in health, see World Bank 2005). The underlying principle of a demand-side subsidy is that "money follows the patient"--meaning that service providers compete for patients. Provincial governments Providers offering more attractive services and con- ducting better outreach to the target group therefore Certification stand to benefit more from the program than those Health service purchasing units offering services that the target group finds unsatisfac- tory. In addition, providers risk being excluded from the scheme if they neglect quality or are found to have defrauded the system. To be reimbursed, participating institutions must provide medical records for auditing, and these are verified by patient surveys (figure 2). The fixed fee-for-service payment means that ser- Fee for service Medical records Enrollment for verification in program vice providers are free to choose how to provide the service, as long as it meets the quality requirements of the program as set by the service purchasing depart- ments. That means dealing with a tradeoff between the Choice of provider quality and attractiveness of the service offered and the Health Target cost of providing that service. How providers do so is service group up to them. However the purchasing departments (in providers cooperation with auditors) audit key indicators on the Treatment (output) quality of service delivery included in agreed interven- tion protocols. Only providers efficient enough to offer Supporting the delivery of basic services in developing countries pproaches of people enrolled reached 369,559--some 45 percent The design has also allowed the government to of the eligible population and more than the target introduce significant structural changes to the health number. The positive experience resulted in an addi- system at provincial level such as linking financing to tional World Bank loan approved in November 06 to results (service delivery and outcomes), implement include additional 1.7m beneficiaries in 15 additional output and quality information systems with regular provinces. audits, improve costing and pricing capacity, and establish performance based incentives for health Conclusion personnel. Those are significant changes that contrib- The Argentine Maternal-Child Health Insurance Pro- ute to increased accountability, efficiency and choice gram uses an interesting approach of combining out- within the health sector. put-based contracting with an output-based funding The arm's-length relationships between partici- mechanism. Particularly innovative is the combination pants in the program help increase transparency and of enrollment numbers and performance indicators as accountability thus providing incentives for efficiency. a way to address the tradeoff between quantity and Together with reporting and documentation require- quality. ments, these arrangements also make it easier to iden- The program's transfer of performance risks to tify problems and adjust the system accordingly. health service purchasing units and service providers The achievements so far show the advantages of the helps increase accountability. Health service purchas- concept of combining output-based contracting with ing units bear risks related to factors mainly within output-based funding. Due to the program's continued their control, such as the marketing of the scheme to success, the World Bank and the Argentine government the target group and the design and enforcement of have decided to expand it to other provinces. contracts with service providers. Service providers bear operational risks more closely related to the provision References of services. Managing these risks effectively requires Argentina, Provincial Maternal-Child Health Investment gaining first-hand information from and experience Project (1st phase APL) P071025. World Bank. with the target group. And that helps in gearing ser- Di Tella, Rafael, and William D. Savedoff, eds. 2001. vices to the intended beneficiaries. Diagnosis Corruption: Fraud in Latin America's Public Designating a target group for enrollment helps Hospitals. Washington, D.C.: Inter-American Devel- improve the targeting of the subsidy. And decentralized opment Bank. management at the provincial level helps ensure that World Bank. 2005. A Guide to Cosmpetitive Vouchers in the subsidies reach only the intended beneficiaries. Health. Washington, D.C.: World Bank. About OBApproaches OBApproaches is a forum for discussing and dis- The case studies have been chosen and presented seminating recent experiences and innovations by the authors in agreement with the GPOBA for supporting the delivery of basic services to the management team, and are not to be attributed to poor. The series will focus on the provision of water, GPOBA's donors, the World Bank or any other af- energy, telecommunications, transport, health and filiated organizations. Nor do any of the conclusions education in developing countries, in particular represent official policy of the GPOBA, World Bank, through output, or performance,-based approaches. or the countries they represent. To find out more, visit www.gpoba.org The Global Partnership on Output-Based Aid Supporting the delivery of basic services in developing countries