Mortality statistics play a key role in the estimation of health indexes. In Brazil, the source of these data is the System on Mortality Information of the Ministry of Health (SIM/MS), created in 1975/1976, which covers almost 85% of the deaths. However, there are some important gaps in our knowledge on the distribution of deaths according to their underlying causes. Some examples of these events are deaths classified as the consequence of ill defined conditions, incomplete diagnoses, maternal causes and Aids. The objective of this study was to evaluate the quality of the information of the SIM/MS and the gain that could be attained by using a very simple method of information recovery. The studied population was comprised of 618 deaths classified in any of the four referred situations, and that had occurred in the last three months of 2000, in 15 Brazilian cities with heterogeneous demographic, social and economic characteristics, in the states of Sergipe, São Paulo, Mato Grosso. Interviews and medical appointments were conducted in houses, hospitals, and forensic institutes using specific forms. Of 294 deaths with an ill defined condition, excluding 38 losses, there was the possibility of having a well defined underlying cause in 71% ; from the 172 deaths with an incomplete diagnosis, 51.2% had a well defined condition. It was possible to aggregate seven new maternal deaths to the four already known (permitting to estimate an adjustment factor equal to 2.75 for maternal deaths), and three new cases of deaths due to Aids were informed. The results provide evidence that the quality of the SIM/MS data can be improved, by adopting a very simple, accessible and executable methodology.
Quality of information; Mortality; Ill-defined underlying cause of death; Incomplete diagnosis; Maternal death; Aids