Dinâmica demográfica e internações hospitalares: uma visão prospectiva para o Sistema Único de Saúde (SUS) em Minas Gerais, 2007 a 2050

Abstract

The objective of the present study is to assess the effect of population aging alone and associated with changes in the admission rates and in the features of the health services on the volume of public hospital stays in the state of Minas Gerais, in 2020 and 2050. To project admission rates, three methodological approaches have been used: 1) the method of constant use rates, which tries to capture the isolated effect of changes in demographic characteristics size and age structure -; 2) time series method, which considers the trend observed in admission rates in the last decade. The trend in rates was approached in two ways, using the deterministic and the stochastic method. The deterministic method considers that the admission rates in 2020 will follow the pace observed by the variation in the annual average seen between 1996 and 2007. The stochastic method quantifies uncertainty associated with projection through the adjustment of Lee & Carters method (1992) to the admission rates projected; 3) method for panel data regression with trend term, which takes into account the heterogeneity in admission rates, in demographic and socioeconomic characteristics and the features of health services of the 75 health micro regions in Minas Gerais, between 1998 and 2007. Results show that the demographic effect alone has a considerable impact on the volume of admissions, but when the trend in rates is incorporated into the analysis the effect of reducing admission rates offsets the demographic effect. The analysis from the regression method showed that the variables with greater association with the admission rate correspond to the number of beds per capita and the GDP per capita. As for the need for hospital beds in the public health sector in Minas Gerais, the projection method using fixed rates presented a need for beds in 2020 50% higher than the number of beds observed in 2007. The remaining methods, which incorporated the historical behavior, presented a decrease in the need for beds, indicating that the amount of beds in 2007 can meet the demand for beds in 2020. The present study provides two major contributions. The first contribution is to present evidences of the effects of population aging in the Brazilian context, which lacks this type of study. The second contribution is methodological since different methods have been adopted to forecast admissions, with a discussion on the applicability of each method in different data contexts.