54104 RESULTS-BASED FINANCING FOR HEALTH Getting Health Results in Afghanistan BY BERYL LIEFF BENDERLY For many people who keep up with world events, the news from Afghanistan has long appeared to consist mainly of terror attacks, counter-insurgency campaigns and the opium trade. Behind the sensational report- ing, however, another major story is unfolding: how one of the world's poorest countries, after decades of bru- tal conflict that claimed more than a million lives and made millions more people into refugees, is working to build a modern health system. In a rugged land of mountain and desert that would present major obstacles even without the effects of war and widespread poverty, Afghanistan has made significant progress. Central to this achievement has been the collaboration of the Afghan government with international donors including the World Bank in fashioning a system suited to the nation's values, culture, priorities and geographic setting. Each of three major donors--the World Bank, the European Commission (EC), and the United States Agency for International Development (USAID)--has taken a different approach to funding and managing its contribution to the effort. The one chosen by the World Bank uses results-based financing (RBF), a system that provides rewards--in this case monetary ones--for improved performance. RBF has, some health experts believe, features well suited to the opportunities often present in countries with histories of conflict. Unlike more traditional management methods that concentrate on inputs, RBF focuses on results. Health care is tremendously labor-intensive, so health system results depend "crucially on the productivity and motivation of health workers," write Rena Eichler, Ruth Levine and colleagues.i In situations such as rural Afghanistan, where many workers carry out their duties in isolated locations with little day-to-day supervision, "performance incen- tives can motivate individual health workers to provide more services through increased effort with the same level of resources," they continue.ii In addition to connecting health organizations' level of income to their perform- ance, the Bank's approach in Afghanistan also gives them wide latitude for decision making by putting authority into the hands of the people who daily face the realities of getting care to residents of a poor, rugged, war-torn country and encourages them to figure out how they can more effectively meet their organization's goals. This has allowed innovative ideas to emerge to solve major challenges. Interim evaluations reveal that in only a few years, the health system has managed to reach many more citizens throughout the country and deliver care of a higher quality than before. How much of the credit for this progress belongs specifically to RBF as opposed to other factors is as yet uncertain, however. In addition, enormous chal- lenges remain before the nation's ambitious goals are met. RBF already appears, however, to be a productive basis for delivering health services in a country whose historical context, resources and geography would severely test any system's ability to provide a decent standard of care. 1 FEATURE A New Approach to a Troubled Past The fall of the Taliban regime in 2002 found much of Afghanistan's health system in ruins, many of its facilities and equipment destroyed and its trained personnel poorly situated to meet the people's needs. The disorganization and disarray that result from conflict, however, can make the building of institutions to support RBF in areas such as verification or community involvement less complicated, says Health Advisor Petra Vergeer. "In non-conflict situ- ations, systems are much more developed and fixed while in the recovery stage post-conflict there is an opportu- nity to innovate and try to develop new systems.... People are more open to change." A results-based system that does not supply overly detailed prescriptions for how to get things done can serve as a "catalyst" for new ideas appropriate to unprecedented conditions, she continues. "Health professionals are thinking about how are we going to get the results we want.... [They] are focused on results and what's the best way of achieving them." But spending substantial funds on new programs in a war-torn developing country also offers ample possibilities for corruption, favoritism and other abuses that waste resources and torpedo effectiveness. Avoiding such problems, as well perverse incentives and other unintended consequences, requires extreme care in designing program goals and measurements as well as rigorous, independent and transparent evaluation of results. Goals must be stated with clarity and must directly and clearly relate to the desired outcomes. Goals must also be realistically possible for workers to attain through their best efforts. Long before 1979 Soviet invasion and the armed conflict, civil disruption and oppressive Taliban rule that followed, however, Afghanistan's health performance already ranked near the bottom of world comparisons. Maternal and child health statistics trailed those of neighboring countries and rural Afghans generally went with- out modern health care.iii By 2002, a quarter of children were dying before their fifth birthdayiv and the maternal death rate, 1800 for every 100,000 live births, was the world's highest.v That rate was even worse in rural areas, reaching 6507 women dying for every 100,000 children born alive in the rugged mountains of Badakshan province in the nation's northeast.vi 2 FEATURE The state of health care delivery was equally grim. With 80% of the population in often-remote rural areas, health facilities clustered in cities and towns. What health services the rural people received came from non-governmental organizations (NGOs), many of them international groups working "cross-border" from neighboring countries. The nation's health care was a "Swiss cheese," according to Dr. Benjamin Loevinsohn of the World Bank, with areas in the "holes" receiving no services at all. Wardak province near Kabul had a primary care facility for every 9000 residents--with some facilities less than a kilometer apart--but only four facilities served the 320,000 residents spread across Samangan province's nearly 17,000 square kilometers.vii, viii In early 2002, a total of fewer than 500 facilities--or one for every 50,000 rural residents--served the country, according to World Health Organization data. A "fairly dire" misdistribution of health workers further complicated service delivery.ix The supply of trained per- sonnel suffered from what Loevinsohn terms the "3 wrongs--wrong gender, wrong skills and wrong location."x In a culture that frowns on women receiving health care--and, most decidedly, maternal and reproductive health care--from men, only one rural facility in four had even a single trained female provider on staff. The people desperately needed public health, reproductive and child health services, but health professionals generally specialized in other fields. The population overwhelmingly inhabited the countryside, but health workers over- whelmingly lived and worked in the cities or their outskirts.xi Relatively few people outside of cities or towns, therefore, received even rudimentary health services. A 2003 survey of more than 20,000 households in 6 cities and 32 of the nation's 36 provinces found only one child in five with the full course of diptheria-pertussis-tetanus vaccinations, only one child in four suffering symptoms of pneumonia receiving medical treatment, and fewer that one woman in ten reporting having a skilled attendant when she last gave birth.xii, xiii Building for the Future Improving this dismal situation was a top priority of the government that came to power in 2002. The Ministry of Health, known since 2004 as the Ministry of Public Health (MOPH), quickly undertook far-reaching reforms. To focus effort on the most pressing needs, it formulat- ed the Basic Package of Health Services, a list of 7 BOX 1 | THE BASIC PACKAGE OF HEALTH high-impact, high-priority primary care areas that a SERVICES xxi national expert consensus agreed were both basic Originally devised in 2003, the BPHS was revised in 2005. If con- to public health and feasible to deliver through a sists of seven elements. "Six are basic services and the seventh system built on primary care. The list emphasizes element is necessary for the six service elements to succeed," giving children a good start in life and helping their states the MOPH official description. The elements are: mothers maintain good health. (See Box 1) Maternal and newborn health "Everyone who needs care must receive care, Child health and immunization regardless of ability to pay. Quality of care must be Nutrition, including prevention, assessment the same for paying and nonpaying patients," says and treatment of malnutrition xiv the MOPH's official statement of BPHS. Control and treatment of communicable diseases, especially tuberculosis, malaria and HIV The document also laid out the structure of the Mental health services health care delivery system, specifying the types of Disability services health care facilities slated to bring the Basic Regular supply of essential drugs Package to communities nationwide. The plan 3 FEATURE specifies a ladder of facilities, each with its own defined competencies, responsibilities and staff BOX 2 | THE THE LADDER OF CARE xxii structure and the ability to refer patients needing Health posts, located in isolated rural communities and higher levels of care. Facilities delivering the BPHS often in private homes, are ideally each staffed by a male range from small community health posts in remote and a female community health worker (CHW) and serve villages to fully staffed and equipped regional hospi- 1000 to 1500 persons. tals. (See Box 2) Basic health centers, called "the foundation of the health system" by MOPH, serve 15,000 to 30,000 persons and The MOPH decided, rather than trying to provide generally have staffs including a physician, either male or female; a male nurse; a community health midwife; and services itself, to contract with NGOs, some of them two vaccinators. already working on the ground, to bring the Basic Comprehensive health centers, serving 30,000 to 60,000 Package to specific geographic areas. The Bank, people, are staffed by a male and a female physician, a the EC and USAID all backed this project, but each male and a female nurse, and 2 community midwives, adopted a management approach that differed from and have inpatient beds, a laboratory, a room for minor the others in a number of respects. (See Box 3). surgery, a delivery room, holding beds, a health education Only the Bank opted for performance-based mone- area, a wound dressing area and a pharmacy. tary incentives. MOPH recruited its own staff of District hospitals, serving between 100,000 and 300,000 skilled Afghans, who were paid market wages, to persons, include an operating theater, a recovery room, an manage the contracts. It also made large invest- emergency room, a nursery, inpatient beds, exams rooms, a pharmacy and a laboratory. Staff includes 2 female and 2 ments to pay for monitoring and evaluation by male general physicians, a surgeon, an anesthetist, a pedia- impartial third parties. trician, a dentist, a pharmacist, 5 male and 5 female nurses, 4 midwives, a community health supervisor, a pharmacist, 2 NGOs working under 42 separate MOPH contracts laboratory technicians, an x-ray technician, a pharmacist were given responsibility for health care in areas that and two vaccinators. are home to 90% of the country's rural population. Together these contracts provide primary care to about 18 million people.xv Just under half of the contractors are local Afghan NGOs, with the remainder split between international NGOs and consortia of organizations. Over time, however, Afghan NGOs have gained an increasing share of the contracts. Eight provinces are served by contracts supported by Bank funds between MOPH and NGOs, which include local Afghan organizations, inter- national organizations, and consortia of organizations. Three other provinces are served under contracts sup- porting by Bank funds in which elements of the Ministry of Public Health provide specific services. EC- and USAID-supported contracts cover 10 and 14 provinces, respectively. EC and USAID contracts reimburse for budgeted expenditures. The Bank-supported contracts, known as "per- formance-based partnership agreements" (PPAs), on the other hand, provide each NGO a lump sum to deliver the BPHS within the contracted area for a year. The NGO makes almost all spending decisions. The agreements, however, specify the population ratios and staffing of the various health facilities as well as salary standards for the various types of workers. The MOPH also determines the required standard of performance. Under the Bank-supported PPAs, NGOs earn bonuses of up to 10% a year for specific amounts of improvement in speci- fied performance indicators. Contracts going into effect in 2009 also link bonuses to the performance of specific facilities. Provincial health officials can also receive bonuses "to align incentives." xvi Inadequate per- formance can bring serious consequences, including termination, as has happened to one international NGO, and expiration without renewal, as happened to a local NGO. 4 FEATURE Surveys by an independent third party organization assess performance. Since 2004, Johns Hopkins University Bloomberg School of Public Health has evaluated over 600 facilities each year, scoring them zero to 100 on each of the 29 indicators in the "balanced scorecard," (BSC) a survey instrument developed with the participation of many stakeholders. Indicators include patient satisfaction; availability of staff, equipment and medications; health workers' knowledge; patient loads; quality of interactions between patients and providers, and much more.xvii Household surveys using other methodologies also evaluate citizens' health and health care experience. Healthy Progress "The results [of MOPH contracting] thus far have been encouraging," write Loevinsohn and Sayed.xviii "Through con- tracting...the MOPH could rapidly expand equitable and standardized health care services to nearly all Afghanistan," says Bashir Ahmad Hamid, M.D., Program Director for Care of Afghan Families, a local NGO. "Having clear, con- crete, measurable and achievable output indicators...had a great impact on the performance of the NGOs [in] achieving the output indicators and motivated the NGO staff to focus their efforts [on] achieving the targets." Between 2004 and 2007, quality of care, as measured by the BSC, improved 32%, even as the number of pri- mary care facilities more than doubled from 496 in 2002 to 1169 in 2007.xix, xx In 2002, only a quarter of these facilities had a female doctor, nurse or midwife, but over 80% did in 2007. Over that period the number of outpatient visits quadrupled, from one for every four residents each year, to almost one per resident. Household surveys also showed marked improvement, with the rate of skilled attendance at births more than dou- bling between 2003 and 2006, the prevalence of contraceptive use tripling, and the percentage of pregnant women receiving prenatal care more than quadrupling. The rate of immunization with DPT almost doubled and the under- five mortality rate fell from over 250 per 100,000 live births in 2002 to 191 in 2005. BOX 3 | APPROACHES TO FUNDING THE BPHS xxiii Professional evaluators are not the only ones sizing The European Commission awards competitive up the system and its design. The flexibility permitted contracts to NGOs and reimburses what they spend on by the PPAs has allowed NGO staff members to inputs. The system does not tie payments to performance. Each NGO determines its own performance indicators and suggest innovative solutions to challenges. Some submits an annual narrative report on progress to the EC argued, for example, that many rugged and sparsely and more frequent reports to the MOPH. This plan covers populated areas needed a new level of health facility, 10 provinces, with some contracts covering whole with capabilities greater than the village health post provinces and others covering clusters of communities. designed to serve 1000 persons but closer and easier USAID awards contracts covering 13 provinces through an to reach than the nearest basic center, which must be implementing agency that contracts with NGOs for health located in order to serve a population of 15,000 to services in specific districts and with the MOPH and other partners for technical assistance. The contracts provide 30,000 persons. Based on this suggestion, a pilot payment for budgeted line items, with only limited flexibility study is examining a new form of intermediate to make changes. Performance is judged according to a set sub-center designed to serve an intermediate-sized of indicators determined by the implementing agency. NGOs population of 3,000 to 7000. submit monthly reports. Not meeting standard can cause payment to be withheld, as can problems revealed by Finding qualified staff, and especially female staff, monitoring visits. Other monitoring mechanisms include for isolated rural facilities was another challenge that quarterly reports, household surveys, and semiannual and annual meetings. Failure to perform can result in cancella- inspired innovative solutions. One idea is recruiting tion of the contract. The system does not award bonuses. couples rather than individuals, and, if necessary, 5 FEATURE arranging suitable employment for the health worker's spouse. Another is recruiting female physicians in neigh- boring Tajikistan, where they are more plentiful and speak the same Dari language. And perhaps most significant of all, the shortage of qualified midwives for rural areas led to a program of recruiting and training local women to provide professional-level services. Unlike the untrained traditional birth attendants who serve at many rural deliveries, the new cadre of community midwives are rural women with at least 8th grade education who undergo an 18-month clinical training program run by NGOs under provincial supervision. They must pass an official provincial examination upon completing the program and must, before their training beings, provide both their own promise and their community's assurance that they will return home to work once they qualify. Despite the progress since 2002 on these and other fronts, however, the challenges still facing Afghanistan's health system remain as forbidding as the nation's rugged landscape. Nor can it be asserted that RBF is neces- sarily the best approach to attaining the goal of universal, equitable and effective health care for the country, because the projects funded by the other major donors and based on other management systems have also pro- duced significant improvements in services and care. As Afghanistan continues to build for its future, however, RBF appears to be a feasible and productive approach to motivating and empowering health workers to produce better results for their fellow citizens. 6 FEATURE Footnotes i Eichler, R., and Levine, R., "More Health for the Money," in Eichler, R., Levine, R., and the Performance-Based Incentives Working Group. Performance Incentives for Global: Potential and Pitfalls. Washington, DC: Center for Global Development, 2009. ii Eichler and Levine. 38 iii Sondorp, E., et al. "Afghanistan: Paying NGOs for Performance in Postconflict Setting," in Eichler, R., Levine, R., and the Performance-Based Incentives Working Group. Performance Incentives for Global: Potential and Pitfalls. Washington, DC: Center for Global Development, 2009. iv Sondrop et al. v Summary sheet for RBF Scheme vi Loevinsohn, B., and Sayed, G. "Lessons From the Health Sector in Afghanistan." JAMA. August 13, 2009. vii Loevinsohn, B., and Sayed, G. Development of the Health Sector in Afghanistan--Recent Progress and Lessons Learned. viii http://en.wikipedia.org/wiki/Samangan_Province ix Loevinson and Sayed, Development x Loevinsohn, B. Some Practical Lessons on Implementing RBF from Afghanistan, 2008. xi Loevinsohn and Sayed, Development xii Loevinsohn and Sayed, paper xiii Loevinsohn, PDF. xiv A Basic Package of Health Services for Afghanistan, Islamic Republic of Afghanistan, Ministry of Public Health, 2005/1384 (Muslim year) xv Loevinsohn and Sayed, Development xvi Loevinsohn, Practical Lessons xvii Loevinsohn, Practical Lessons xviii Loevinsohn and Sayed, JAMA. xix Loevinsohn, Practical Lessons xx Loevinsohn and Sayed, JAMA xxi All information from A Basic Package of Health Services for Afghanistan xxii All information from A Basic Package of Health Services for Afghanistan xxiii Sondorp et al. 7 FEATURE