[Assessing the impact of health sector reform in Costa Rica through a quasi-experimental study].

OBJECTIVE: To assess the impact of health sector reform in Costa Rica on that country's child and adult mortality rates and on the people's access to primary health care.
METHODS: Health sector reform was initiated in Costa Rica in 1995 in some districts, but in others reforms were adopted later. This made it possible to perform a time series analysis, using a quasi-experimental study design, in which observations were made annually from 1985 through 2001 for each of the 420 districts that existed in Costa Rica in 1984. The time series were divided into three periods that allowed all districts to be grouped into three categories (pioneer, intermediate, and late) according to the year when they first implemented health sector reform: 1995-1996; 1997-2000; and 2001 or after, respectively. For each of these periods, mortality rates were broken down by cause (communicable, socially-determined, or chronic disease), sex, and age group. The status of the reform process in a particular district was described by two indicators: (1) the presence or absence of health sector reform during a given period and, wherever such reforms had been adopted, (2) the number of years that had transpired since their adoption. Eight variables were used to control for confounders. Vital statistics and demographic data were obtained from the National Institute for Statistics and Census' [Centro Nacional de Estadística y Censos] electronic database. Poisson multiple regression analysis with fixed effects was used to estimate the impact of reform on child and adult mortality from different causes. Assessment of the population's access to primary care before and after the reform was based on the percentage of people who lived within a 4 km radius of a health facility that offered patient visiting hours two or more days a week. This information came from a previous study that used census data from 2000 and geographic information systems to map health care facilities throughout the country.
RESULTS: Multiple regression showed that the reform was associated with an overall 8% reduction in deaths among children and with a 2% reduction in deaths among adults, both statistically significant. Also noted were a 14% reduction in deaths from communicable diseases or from conditions brought on by the presence of infectious processes, a 0% reduction in deaths from socially-determined causes, and a 2% reduction in deaths from chronic diseases. An estimated 120 child lives and 350 adult lives were saved by the reform in 2001 alone. Also, the percentage of people without equitable access to primary health services dropped by 15% between 1994 and 2000 in areas where health sector reform was implemented in 1995-1996, whereas areas that had not yet initiated health sector reform in 2000 experienced only a 3% reduction.
CONCLUSION: Health sector reform significantly reduced mortality in Costa Rica and put an end to a decade of stagnation, as shown by certain health indicators, such as life expectancy. Equity in access to primary care improved considerably, perhaps because the first reforms were implemented in less developed areas of the country.