March 2008 · Number 123 A regular series of notes highlighting recent lessons emerging from the operational and analytical program of the World Bank`s Latin America and Caribbean Region. BRAZIL 44693 Innovative Approaches to Extending Family Health Services Gerard La Forgia The recently completed Family Health Extension Project1 While poor urban children had better vaccination rates (FHEP) has provided support to Brazil's Family Health and were more likely to be treated for diarrhea and acute Program (Programa de Saude da Familia, or PSF) and respiratory infections (ARI) than their rural counterparts, allowed dramatic improvements in the coverage of basic the prevalence of these diseases among the urban poor was health services. The project aimed to improve utilization higher. Although the urban poor had better access to health and quality of publicly-financed health services by (a) services during childhood than the rural poor, the higher expanding coverage of the PSF in about 100 large munici- urban infant mortality rates raised questions about the time- palities, establishing well-articulated referral and coun- liness, continuity and quality of care received and/or the se- ter-referral systems, and introducing performance-based verity of the underlying illness. In general terms, the poor financing and management arrangements; (b) Establishing had utilization rates 25 percent lower than the non-poor in family health as a core element of health professional both urban and rural areas. It was these inequities, that re- and para-professional training; and (c) Strengthening the flected higher transaction costs and behavioral barriers, that Ministry of Health's (MOH or Ministério da Saúde, MS) led the Government of Brazil to introduce proactive com- capacity to monitor and evaluate PSF health services, poli- munity and family health delivery systems such as the PSF, cies and training activities on a systematic basis. and to provide financial incentives for the country's 5,500 municipalities, which are the main players in the organiza- Inequities in Health Outcomes. tion and delivery of public health services, to adopt them. Estimates from 2000/2001 suggested that 45 percent of Brazil's population lived in poverty and 23 percent (35 The Family Health Program and its Challenges. million) lived in extreme poverty. Approximately half of The PSF was introduced in 1994, building on the strengths the poor lived in urban areas. Data from the 1996 Demo- of its predecessor the Programa de Agentes Comunitários graphic Survey showed that the health status of Brazil's (PACS), which had introduced the use of paid health poor was significantly worse than that of the non-poor. agents who reached out to the community in an organized Children of poor families had three times higher risk of fashion. PSF was seen at the center of the transformation dying before the age of five than children from the wealth- of the mainly curative and hospital-centered health care iest segments of the population. Fertility rates of poor system toward one that was oriented toward prevention, teenage girls were six times higher than those of better-off promotion and basic care. PSF organized community teenagers, and poor girls face much higher risks of mater- agents into teams, which typically included a general prac- nal morbidity and mortality. tice or family health physician, a nurse, one or two aux- iliary nurses and, often, dental health professional teams. The Demographic Health Survey data suggested that, at Building on the PAC, the PSF provided teams with equip- similar income levels, the urban poor had slightly worse ment (medical, transport, and basic information technol- health outcomes than the rural poor. Approximately 40 ogy) and a standardized system of work and reporting percent of Brazilians in the poorest wealth quintile and 75 requirements. Each team was assigned the responsibility percent in the second poorest quintile lived in urban areas. of providing care to a defined set of 600-1,000 families. 1 Extracted from the Implementation Completion and Results Report (ICR 693) of the Brazil: Family Health Extension Adaptable Lending Program (First Phase), October 2007 The aim of the FHEP was to expand coverage of the PSF regulations would qualify for higher federal transfers. from 25% of the population (38 million people in 3,124 · The regulations also strengthened management municipalities with 10,900 Family Health Teams) to 50% requirements for municipalities to qualify for "autono- coverage. At the time, most of PSF's expansion had mous management" status and created a stronger role taken place in rural and peri-urban areas, where munici- to the states in overseeing municipal resource man- pal primary providers were in short supply. The program agement. States and municipalities were required to faced a number of challenges: develop health plans and sign performance agreements · Implementation in urban areas was patchy, in part that specified performance targets. Municipalities that because the PSF faced the dual challenge of extending complied with the new regulations would be "certified" coverage and converting the traditional and often inef- by the MOH to receive increase per capita payments; fective municipal primary care delivery system to one · Drafted a human resource policy that supported the based on the family health strategy. development of personnel skilled in health services · The Program's expansion in large urban areas and management, family planning and family health strate- metropolitan regions had been slower than expected, gies and practices, while standardizing and regulating despite changes in financial incentives the contractual arrangements that affected PSF teams. · Municipal health authorities did not have clear criteria for prioritizing program interventions and geographi- Results cal areas, and instead applied a mix of risk criteria The first phase of the FHEP went well beyond what combined with community and political demands; was originally expected in terms of the Project's cover- · An inadequate referral system was in place because age, expanding the number of participating municipali- of the lack of a defined entry point for the population ties from 40 to 187 and reached on average 34.4 percent and the absence of a provider who assumed responsi- of population in those municipalities. bility for an individual's and family's total care that co-existed under the existing traditional PHC models; Achievements in the area of health service indicators, such · Services were of uneven quality due to shortages of as maternal and child health (prenatal coverage) and reduc- vaccines, medicines, laboratory services, equipment tion in hospital admissions for ARI and diarrhea were also and medical supplies and treatment protocols and analyzed based on the level of PSF coverage in participat- standards, quality control and training; ing municipalities. Municipalities with PSF coverage of · There were important human resource supply bottle- over 60 percent showed the highest increase in the num- necks in terms of quantity and quality, and varying skills ber of mothers with 7 or more prenatal visits--12 percent, among existing PSF teams. Contractual/employment or six times more than the average for all municipalities. arrangements relied heavily on temporary personnel. As The increase in the coverage of tetanus vaccination for a result, they were flexible, but unstable, and resulted pregnant women was up to three times higher in municipal- in legal and political challenges and the high rotation of ities with over 40 percent coverage of PSF than the average personnel which compromised program continuity and increase for all participating municipalities. As expected, the stability of the doctor/patient relationship, which was the average decline in hospital admissions for ARI and di- central to the PSF's family health model. arrhea was higher in municipalities with greater PSF cover- age. For example, municipalities with over 60 percent cov- To reach its goal of expanding PSF to provide coverage erage reported a decline of hospital admissions for ARI to half of Brazil's population by 2002, the Government: and diarrhea of 15 and 7 percent respectively between · Promoted primary health care interventions that were 2003 and 2006, compared 2 and 4 percent for municipali- most cost-effective and beneficial to the poor, and ties with less than 20 percent coverage. The weakest dif- prioritized actions that focused on communicable and ference among municipalities with different PSF coverage non-communicable illnesses that affected the poor; levels was evident in the indicator for children's vaccina- · Secured increased financing for the sector and for tion coverage, probably because of the already high base- the PSF (14 percent of the MOH's budget was assigned line for this indicator. In sum, increases in PSF coverage to PSF); approved new regulations that sought to correct are strongly associated with improvements in utilization observed deficiencies in the decentralization process while (e.g., pre-natal visits) and results (e.g., vaccination cov- strengthening institutional arrangements to facilitate PSF erage) as well as quality of spending (e.g., reduction in expansion. Municipalities and states that complied with the hospital admissions for conditions sensitive to primary care, such as ARI and diarrhea). 2 · March 2008 · Number 123 Most of the outcome and output indicators with respect governments' fiduciary capacity. These innovative features to improving the quality of family health service provi- facilitated the smooth flow of Bank financing to municipali- sion through developing and strengthening in-service ties, created a single oversight and reporting system rather and pre-service training of human resources in the PSF than parallel systems for Government and Bank reporting model were met or exceeded. While in 2003, less than purposes, reduced the size of the administrative person- half (48%) of AB/PSF staff had training in areas related nel needed to follow fiduciary requirements and reoriented to planning, management and health service organization Bank supervision to technical matters, including strengthen- and delivery, that proportion had increased to 86% only ing of the Government's fiduciary framework. two years after the start of project implementation. Performance-based Financing: The pooled funding ap- M&E Utilization and Sustainability. The Government's proach enabled the application of a performance-based vision of the health sector--one that has functional referral financing scheme to the mingled funds (government and and counter-referral networks and induces continuous qual- IBRD) directed to participating municipalities. The scheme ity improvement at all levels--requires excellent monitoring consisted of two parts. The first involved a bonus for mu- and evaluation (M&E) systems. When FHEP was prepared, nicipalities that complied with three performance criteria: the lack of an evaluation culture, as well as shortcomings in (i) execution of at least 90 percent of funds received from the methods and processes for monitoring and evaluation, the project and according to approved implementation were major constraints to the development of a quality- and plans, (ii) evidence of progress toward achievement of performance-oriented service delivery system. The Project performance indicators particularly in terms of PSF cover- supported the development of these M&E systems, and the age extension, and (iii) presentation of accounting ledgers establishment of systems for accreditation of family health detailing expenditures for at least 75 percent of funds re- units and training institutions. The PSF monitoring system ceived. Announced in mid-2005 and implemented in 2006, was developed, including the design of instruments and cor- the bonus was distributed as a lump-sum payment to thirty- responding indicators. A referral system for mother-child five (of 187) of participating municipalities. The bonus care was developed and implemented, and a performance- represented 50 percent of the value of each municipality's based management system in support of PSF was developed original grant. The second part consisted of a "performance and tested in a sample of participating municipalities. The prize" based on three criteria: (i) spending was aligned with next phase of FHEP, currently under preparation, would sup- approved implementation plans; (ii) municipality attained port further improvements in M&E at the municipal level, PSF population coverage of at least 70 percent; and (iii) the and strengthen further the state health secretariats' capacity to municipality complied with fiduciary benchmarks related to conduct supervision, monitoring and impact evaluation. procurement and financial management. Twelve participat- ing municipalities received the prize in 2006, sharing R$6 Adoption of Pooled Funding. Perhaps the most innovative million in additional transfers. The prize was awarded dur- aspects of the FHEP were its fiduciary arrangements. First, ing a formal ceremony attended by the federal health minis- the Project adopted a pooled funding approach in which loan ter and the mayors of the winning municipalities. funds were pooled with those of the Government to satisfy the Government's request that Bank financial participation Although modest in scope, this performance-based financing in the grant-based financing system through which the fed- was the first of its kind for the MoH which has included a eral government finances health services managed by states modified version of the scheme in a recently approved policy and municipalities (known as fundo-a-fundo). (2006), known as Health Convenants (Pactos de Saúde). The health convenants link future federal financing to states and The first phase FHEP also introduced innovative operational muncipalties with compliance with agreed benchmarks. arrangements under these pooled components, including: (a) financing a share of the government transfers for PSF Lessons Learned instead of a share of each transaction (as is normal practice); The pooled funding approach offered several advan- (b) disbursements against Statements of Transfers issued by tages that make it worth replicating, given similar con- the MOH; (c) annual certification of expenditures by states ditions. It, (i) allows for a single procurement, financial and municipalities through internal and independent audits; management and disbursement system, which reduces (d) the application of national rules and systems for procure- transaction costs of Bank financing and ensures that sup- ment and financial management; (e) the carrying out of fidu- port under a project for improving Government's systems ciary risk assessments; and (f) strengthening of sub-national March 2008 · Number 123 · 3 brings results; (ii) directs Government and Bank efforts instrument; (ii) integrate the family health program into to technical and fiduciary oversight, monitoring and municipal policy making and planning; (iii) implement net- strengthening of the larger Government program rather work arrangements that link family health teams to diagnos- than ring-fencing a much smaller Bank-financed project; tic and specialized care centers as well as logistic and trans- (iii) allows the Bank's technical specialist(s) to focus port systems; (iv) promote healthy behaviors and integrated advice during supervision on technical aspects, while management of chronic diseases; and (v) improve the avail- financial management and procurement specialist address ability of essential pharmaceuticals for poor populations. compliance with the Bank's fiduciary requirements; (iv) inserts the Bank-financed project into a high-priority A final lesson is that it would be useful for large programs programmatic and policy framework; and (v) creates the that are implemented among three levels of government, vehicle to link both Government and Bank financing to such as the PSF, to establish a mechanism to facilitate the results. In effect, the pooling approach under the FHEP stability of personnel at the federal level that are hired benefited both the Government and the Bank by redefin- to provide technical assistance to states and municipali- ing a partnership to focus on results. ties with the appropriate accountabilities for productivity and quality. This may be difficult to ensure given staffing The inclusion of a very large number of municipalities changes resulting from changes in administrations and with varying institutional and technical capacities offered other reasons. Still, it might be possible to issue a con- many lessons. First and foremost, it is necessary to have tract, for the duration of the program, with a private firm in place both monitoring systems with a defined number or academic institution to recruit, hire and pay the person- of easily verifiable indicators to monitor and assess per- nel that will be providing technical assistance. formance and technical assistance arrangements in the form of a permanent cadre of central- and state-based Conclusions technical and fiduciary personnel to provide support to The health status of Brazil's population has improved the municipalities, especially those with lower capacity. significantly over the last decade: infant mortality has decreased by 47 percent in 14 years, mortality rates from To determine municipalities' capacities, it is necessary vaccine-preventable diseases in children are negligible to conduct in-depth technical and fiduciary risk assess- and diarrhea diseases are the cause of less than 7 percent ments and categorize municipalities according to their of deaths among children under five years of age. Still, technical and institutional capacity to plan, supervise and significant health system issues remain. Substantial dis- monitor family health care. Based on these assessments, parities still exist in health status, health financing and it is important to craft customized plans and interven- service utilization among regions, states and municipali- tions to: (i) extend services; (ii) strengthen planning, ties, income groups and between urban and rural areas. management and supervisory capacities for primary care; Also, Brazil's health care system, where the Government (iii) strengthen fiduciary capacity; and (iv) support mu- is an important actor, remains hospital- and specialty- nicipalities with change management. Also, eligibility centered, compromising the affordability and sustain- criteria should be established for continued participation ability of the delivery systems. Only 45 percent of Bra- in the program, and special support mechanisms should zilians have regular access to more affordable primary be introduced for municipalities with low capacity. health care, and coverage continues to be lower in large cities. Non-communicable diseases (NCDs) are now the The first phase of FHEP also provided important lessons main cause of death and disease in Brazil, placing a sig- on technical design issues for this particular model of fam- nificant and growing impact on the economy and requir- ily health program, especially in a decentralized environ- ing greater interaction between the different levels and ment. For the program to be successful, it is important to components of the health system, and a well functioning strengthen the capacity of family health teams to respond primary health care facility as the central point from to patient demand and to foster the stability and sustain- where other health facilities are referred. ability of family health physicians. Also, the program should strengthen interventions to; (i) expand quality certification systems applied to municipalities and fam- About the Author ily health teams, based on results of a quality assessment Gerard La Forgia is a Lead Health Specialist (LCSHH) in the Latin America and the Caribbean of the World Bank "en breve" is produced by the Knowledge and Learning Team of the Operations Services Department of the Latin America and the Caribbean Region of the World Bank - http://www.worldbank.org/lac · March 2008 · Number 123