Principal investigator: Viviane Pires Ribeiro
Article title: Evaluation of the impact of the Family Health Program on infant mortality in Brazil, 1990–2002
Article authors: James Macinko, Frederico C Guanais and Maria de Fátima Marinho de Souza
Location of the intervention: Brazil
Sample size: 27 Brazilian Federative Units
Main theme: Health Insurance
Type of Intervention: Effects of the Family Health Program on the infant mortality rate
Primary variable of interest: Infant mortality
Assessment method: Experimental Evaluation (RCT)
Evaluation Context
The Brazilian Unified Health System (SUS) was created and structured on the principle of universal health coverage as a right of all citizens, with an emphasis on decentralization, equity, comprehensiveness, hierarchy, community participation, and complementary participation of the private sector. Since 1990, Brazil has undergone considerable health reforms to implement this vision. The decentralization process, in particular, has advanced rapidly in the area of primary health care.
The Family Health Program (PSF) can be considered the most important structural change ever implemented in Brazilian public health. Since 2004, the program has covered approximately 66 million people, almost 40% of the entire population, providing a wide range of services offered by a team composed of a doctor, a nurse, a nursing assistant, and (usually) four or more community health workers. In some locations, the team also includes dental and social professionals. Each team is assigned to a geographic area and is responsible for recording and monitoring the health status of the population living in that area, providing primary care services and referring them, when necessary, to other levels of care. Each team is responsible for an average of 3450 people. Doctors and nurses generally provide services in health facilities located within the community, while community health workers provide health promotion services during home visits.
Intervention Details
Macinko et al. (2006) used publicly available secondary data to assess the impact of the Family Health Program on the infant mortality rate (IMR) over time. The study was conducted through a longitudinal analysis, using panel data from all 27 Brazilian Federative Units (comprising the 26 states and the federal district, Brasília). The period 1990–2002 was chosen by the authors because it includes three distinct periods: prior to the Family Health Program (1990–1994), during the development of the Family Health Program (1995–1998), and the final expansion of the Family Health Program (1999–2002).
Data on infant mortality, PSF coverage area (proportion of the state population served by the PSF program), and health resources were obtained from the Ministry of Health website. Data on other health determinants were based on annual population surveys conducted by the Brazilian Institute of Geography and Statistics (IBGE) and developed for state-level representativeness by the Institute for Applied Economic Research (IPEA).
Methodology Details
Macinko et al. (2006) used fixed-effects regression models for the 27 Federative Units. The infant mortality rate (expressed as the number of deaths of children under 1 year of age per 1.000 births in the same year) was used as the dependent variable. Infant mortality was used because improving child health is a priority of the program.
The independent variables included in the model due to their potential influence on the dependent variable were: socioeconomic indicators (proportion of the population with access to clean water supply and per capita income); indicators of women's development (female illiteracy and fertility rates); and health service indicators (doctors and nurses per 10.000 inhabitants and hospital beds per 1.000 inhabitants). Additional analyses were also conducted to examine the relationship between program coverage and infant mortality caused by diarrhea and acute respiratory infections.
Results
The results found by Macinko et al. (2006) indicate that during the period from 1990 to 2002, the Infant Mortality Rate declined from 49,7 to 28,9 per 1.000 births. During the same period, coverage of the Family Health Program increased from 0% to 36%. A 10% increase in program coverage was associated with a 4,5% reduction in the Infant Mortality Rate, controlling for all other health determinants. Access to clean water and hospital beds per 1.000 inhabitants were negatively associated with the Infant Mortality Rate, while female illiteracy, fertility rates, and average income were positively associated with the Infant Mortality Rate. Examining the relationship between program coverage and deaths from diarrhea, the results suggest that the program may reduce infant mortality. The impact of the program on deaths from acute respiratory infections was ambiguous.
Lessons in Public Policy
The Family Health Program is associated with a reduction in the infant mortality rate, suggesting that it is a significant, though not the only, contributor to decreasing infant mortality. That is, the main determinants of infant mortality in Brazil include: primary care and availability of hospital beds, clean water, income, female literacy, and fertility. Thus, existing secondary data provide an important tool for evaluating the effectiveness of health services.
The policy implication is that a comprehensive approach based on improving child health through primary health care can lead to considerable improvements in outcomes. To inform better policy, future studies could assess the cost-effectiveness of expanding the Family Health Program (FHP), its impact on health, and estimate the impacts at other levels of analysis (e.g., municipal and individual levels).
References
MACINKO, James; GUANAIS, Frederico C.; DE SOUZA, Maria de Fátima Marinho. Evaluation of the impact of the Family Health Program on infant mortality in Brazil, 1990–2002. Journal of Epidemiology & Community Health, v. 60, no. 1, p. 13-19, 2006.
