The objective of this paper was to compare hospital admission associated factors in Brazil in 1998 and 2003. The data of the 2003 National Household Survey Health Supplement were analyzed using a logistic regression model with normalized weights, using Andersen's Behavioral Model, and the results were compared to those published in the PNAD 1998. There was a decrease in hospital admission coefficients for women in reproductive age and older adults, and an increase in the coefficient for children. For adults, there was an increase in inequality in hospital admissions according to the income. Among children, those with no income had more chances for hospital admission. Race and educational level had no influence in hospital admissions. Some general trends with respect to hospital admissions in Brazil could be verified in this study: the increase in admissions for surgical procedures, the decrease in admissions related to pregnancy, childbirth and puerperium, shorter in-patient care, the increase in admissions to public hospitals, the increase in admissions financed by the Brazilian Unified Health System (SUS) in comparison to admissions covered by private health insurances and out-of-pocket payment, the decrease in admissions financed by more than one system and the great proportion of users satisfied with the care they received.
Hospital services utilization; Social inequality; Income inequality
