go z O_,u1tput=Based flea¶h Care Rena Eichler, Paul Paying for Performance in Haiti Auxila, andJohni Pollock In 1999 the U.S. Agency for International Development introduced Renia Eichler performance-based contracting in an effort to improve the (reichler@msh.org) is a health economist at effectiveness of some of Haiti's nongovernmental organizations in Manaemen Scincesfor providing basic health services, such as immunization and prenatal Health (MSH), a U.S.- based nongovertnmental and maternal care. These providers had been operating under a organization workiig in payment system that reimbursed their expenses up to a ceiling. The developing cou ntries. She is based in MSH's new system set performance targets and withheld a portion of their Washington, D.C., office historical budget, allowing them to earn back the withheld amount and has worked on z performance-based plus a bonus if they met the targets. A one-year pilot involving three paymnent schemes in Haiti providers showed some marked improvements in performance. .,e and Kenya. Paul Auxila, o also at MSH, leads the Those paying for health care services in develop- ing the capacity of NGOs to deliver primary Haiti project. John ing countries typically have not required the health care services. A key part of this effort was Pollock provides U.S.- providers to guarantee their performance. Public the introduction of a performance-based pay- __ based support to the payers tend to fund public institutions to main- ment system. The challenge was to develop a sys- : Haiti project from MSH's tain capacity (paying salaries and recurrent costs) tem based on attainment of goals without central office itn Boston, rather than to ensure that consumers receive imposing an excessive burden of monitoring Massachutsetts. high-quality services. Any contracts with private and reporting requirements. providers generally have not held them account- Following competitive tenders, USAID t: able for performance. Donors have tended to awarded funding for the two-phase, US$92 mil- adopt similar practices, providing lump sum lion project to Management Sciences for Health z grants or reimbursing public providers and non- (MSH), a U.S.-based NGO operating in devel- governmental organizations (NGOs) for docu- oping countries. MSH manages and disburses mented expenditures. As a result, providers tend the funds. During the first five-year phase, -J to focus on securing funds rather than improving beginning in 1995, the project provided fund- O efficiency or the quality of care. ing to 23 NGOs, an established group that had In this context, in 1995 the U.S. Agency for received USAID support in the past. For the sec- International Development (USAID) launched ond five-year phase, beginning in 2000, the a 10-year project in Haiti aimed at strengthen- number of NGOs increased to 33. OUTPUT-BASED HEALTH CARE PAYING FOR PERFORMANCE IN HAITI When the project began, the immediate need performance-based system NGOs receive an up- was to develop rapid mechanisms for funding front payment and then a quarterly sum rather NGOs so that they could provide critical basic than submitting their expenditures every month. health services, including maternal and child At the end of a defined period-one year in this health and family planning services. Initially, and case-performance is measured and the size of in line with general practice, NGOs were re- the bonus determined. imbursed for expenses up to a ceiling that was To ensure that the NGOs viewed the change essentially a negotiated budget. Under this as advantageous, MSH used a collaborative - expenditure-based financing NGOs submit a approach in designing the new system. NGOs proposed annual budget and a plan showing how demonstrating the leadership and institutional they intend to ensure the delivery of a basic pack- capacity to respond to the system were invited age of services. Then each month they submit to meetings to express their views about the cost reports with detailed documentation of their pilot. Because these meetings occurred after expenditures for reimbursement. NGOs are free NGOs had signed contracts for fiscal 1999 to set their own fees for services. Most charge (October 1998-September 1999), they were patients for drugs and some for consultations. willing to renegotiate only if the new contract could make them better off. Switching to performance-based contracts The meetings led to agreement on a new con- A 1997 population-based survey to review the tract that would pay 95 percent of the budget existing system found that NGO performance under the expenditure-based contract-but was extremely uneven. In vaccinations a good would also pay a bonus of as much as 10 percent performer reached 70 percent of the target pop- of that budget. The NGOs thus assumed a finan- ulation, while the worst performer reached only cial risk: if they failed to attain performance tar- 7 percent. One NGO made sure that 80 percent gets, they would lose 5 percent of the budget of women knew how to prepare an oral rehy- under the original contract. But they were will- dration solution; another educated only 44 per- ing to do so because they also had the possibility cent. Some NGOs provided the minimum two of earning 5 percent more than the budget. prenatalvisitsto43 percentof pregnantwomen; Seven performance indicators were chosen, others reached only 21 percent. These wide- and a target was negotiated for each indicator ranging results were not correlated with costs and linked to a share of the bonus (table 1). (average costs per patient visit ranged from (Negotiating with MSH, each NGO then trans- US$1.35 to US$51.93). lated the general targets into specific targets.) So in 1999 MSH decided to test a new Five indicators related to improving health approach-performance-based payment. The impact, one to increasing consumer satisfaction new payment system was expected to lead to effi- by reducing waiting time, and one to improving cient delivery of high-quality services in several community participation and coordination with ways: the Ministry of Health. * Because institutions receive a bonus if they Another goal of the project was to improve achieve performance targets, they feel strong institutional sustainability. To facilitate learning incentives to attain those targets. and sharing, the project helped create a net- * Because institutions assume financial risk for work of local NGOs. Regular meetings encour- improving performance, they feel strong aged NGOs to share strategies that have incentives to use resources efficiently and succeeded or failed in the challenging Haitian effectively. environment. The project also provided techni- * Because institutions are paid on the basis of cal assistance, to help NGOs review their pricing results, they face strong incentives to policies and develop a plan to generate revenue improve management, motivate staff, and through sources unrelated to health services. innovate. CORE, a cost and revenue analysis tool, was Three NGOs, serving about 534,000 people, used to help NGOs identify unit costs, revenues, participated in a one-year pilot study. Under the and staff utilization (MSH 1998). The goal was to promote a culture of information-based deci- US$40,000, less than US$1 per person benefit- sionmaking to improve efficiency. ing from the project. Measuring performance The results one year later Since payment is tied to performance, the The most striking result was the increase in NGOs agreed that reporting on their own per- immunization coverage in all three NGO ser- formance would create incentive problems. vice areas (table 2). In two of the three areas the MSH contracted a neutral third party- share of mothers who reported using oral rehy- l'Institut Haitien de l'Enfance (IHE), a local dration therapy increased-and so did the survey research firm-to measure baseline and share who reported using it correctly. end-of-pilot performance. Performance was relatively weak in meeting pre- Using the standard cluster sampling method- natal care and contraception targets, probably ology recommended by the World Health because of the need for ongoing counseling and Organization (WHO 1991), IHE sampled behavioral change. The availability of modern households in each NGO's service area to meas- contraceptive methods increased substantially. ure immunization coverage, based on both Waiting time wasjudged to be an invalid indi- immunization cards and reports from caretak- cator of quality because people who have to ers. IHE determined the percentage of women travel long distances to obtain lab tests might using oral rehydration solution to treat diarrhea wait an entire day for results rather than come through exit interviews at clinics with women back. A new indicator of client satisfaction is who brought children in for other reasons. It being developed for the next phase. And reviewed a sample of medical records to find out because no easily measurable and verifiable what share of pregnant women had had three indicator could be devised for community par- or more prenatal visits. Discontinuation rates ticipation and collaboration with the Ministry of for oral and injectable contraceptives were Health, the bonus linked to this performance determined by reviewing family planning regis- indicator was given to all three NGOs. ters to identify women who had discontinued All the NGOs in the pilot received more rev- use, had not chosen another method, and had enue than they would have under the not expressed a desire to have a child. And aver- expenditure-based scheme-and all supported age waiting time was determined through meas- continuing performance-based payment. The ures in a sample of institutions at different shift fromjustifying expenditures to focusing on intervals. results inspired them to question their model of This survey needs to be done annually, first service delivery and experiment with changes. to set up a baseline and then to check perform- The NGOs' possibility of earning bonuses ance against this baseline. The annual cost is sharpened staffs focus on achieving goals and 1 Indicator Target Share of bonus Women using oral rehydration therapy to treat diarrhea in children 15% increase 10% Children ages 12-23 months receiving full vaccination coverage 10% increase 20% Pregnant women receiving at least 3 prenatal visits 20% increase 10% Discontinuation rate for oral and injectable contraceptives 25% reduction 20% Clinics with at least 4 modern methods of family planning; 100% of clinics; outreach points with at least 3 50% of outreach points 20% Average waiting time for attention to a child 50% reduction 10% Participation in local health organizing committee (UCS) and coordination with the Ministry of Health UCS defined 10% Source: Authors' compilation. OUTPUT-BASED H EALTH CARE PAYING FOR PERFORMANCE IN HAITI - ~ ~ ~ ~ ~ 4:-dV1A rS WA II 1BliliD C) NGO I NGO 2 NGO 3 2 Indicator Base Target Results Base Target Results Base Target Results Percentage of women using oral rehydration therapy 43 50 47 56 64 50 56 64 86 ViewpOint Percentage of women using oral rehydration therapy correctly 71 80 81 53 59 26 61 67 74 Immunization coverage is an open forum to (percentage of children encourage dissemination of ages 12-23 months) 40 44 79 49 54 69 35 38 73 public policy innovations for Prenatal visits (percentage of private sector-led and pregnant women with at market-based solutions for least 3) 32 38 36 49 59 44 18 21 16 development. The views Contraceptive discontinuation published are those of the rate (percent) 32 24 43 43 32 30 26 20 12 pubhed are those othe authors and should not be Clinics with 4+ modern family planning methods 6 9 9 2 5 5 0 5 5 Bank or any other affiliated organizations. Nor do any of Note: Base data refer to September 1999, results to April 2000. the conclusions represent Source: Authors' compilation. official policy of the World Bank or of its Executive led to innovation, including greater efforts to MSH staff and the NGOs in the pilot will Directors or the countries involve the community. work together to develop new indicators and they represent. To motivate staff to focus on results, two of improve processes for measuring and validating the three NGOs introduced bonus schemes for performance. More NGOs will join the To order additional copies staff. And one introduced a scheme for com- performance-based payment system each year. contact Suzanne Smith, munity health agents, halving their salary and And to increase performance incentives, a managing editor, reserving the rest for bonuses. But when it model being considered for fiscal 2002 would Room 19-017, found that transferring this much risk to rela- reduce the share of payment based on historical The World Bank, tively low-paid staff lessened their motivation, it budgets and phase in a capitation (a fixed pay- 1818 H Street, NW, increased the fixed share of payment. Another ment for providing defined services to an Washington, DC 20433. NGO set up a bonus scheme for local organiza- enrolled patient) combined with rewards for tions with which it works, and all considered results. Telephone: 00 1 202 458 728 1 allocating a share of any bonus they earn to clin- Fax: ics in their network on the basis of performance. 00 1 202 522 3181 The performance-based payment also moti- This Note is based on a longer paper by the sane Email: vated NGOs to request assistance in strength- authors, "Performance-BasedPayment to Improve thelInpact of ssmith7@worldbank.org ening their strategic planning, strategic pricing, Health Services: Evidence from Haiti," availaible at cost and revenue analysis, human resource , ,,.',!.,,1....e,.., ' . '.. l t, ..... ,! ,, d management, and measurement of client per- Printed on recycled paper ceptions of quality. References MSH (Management Sciences for Health). 1998. The future CORE: A 7oolfor Cost and Revenue Analysis. Boston. The results of the pilot suggest that performance- WHO (World Health Organization). 1991. 7he EPI based payment is a powerful way to hold NGOs Coverage Sumrey. WHO/EPI/MLM/91.10. Geneva. accountable for results. The challenge is to define indicators that relate directly to health impact, client satisfaction, and institutional sustainability and to measure and monitor performance in a way that is not prohibitively expensive. This Note is available online: www.worldbank.org/html/fpd/notes/