The objective of this study was to analyze admissions in the hospital health care system in Brazil in the year of 2006, starting from the admitted subjects. The initial data were the admissions SUS (SIH), non-SUS (CIH) in 2006 and 2007. The identification of the individual was by probabilistic method of associations of registrations (linkage) and, for the composition of the admission data, specific algorithms were applied to the data. A total of 12,391,990 admissions with at least one day in hospital in the year of 2006 were analyzed. The imprecise nature of the probabilistic method at a certain degree, the adoption of conservative parameters in order to avoid the inclusion of false positive, and the subnotification of CIH represent possible limitations of this study. The Hospital Services System in Brazil presented admission rate of 5.6% in total, 5.0% SUS and 1.6% non-SUS, with differences according to age and region. The non-SUS admissions data collected by CIH give important additional information to the analysis of the hospital utilization in the country. The SUS administrative data are valid for analyses of admissions and the algorithms for admission data composition enhance the analysis of the Hospital Services System in Brazil.
hospital information systems; hospitalization; unified health system; epidemiology
