[Autopsy and ill-defined cause of death in the state of São Paulo, Brazil].

OBJECTIVE

To investigate the proportion of deaths with an ill-defined cause in the Brazilian state of São Paulo between 1980 and 2002, taking into account the influence of autopsies on this proportion.

METHOD

Data on the number of deaths were obtained from the Brazilian Ministry of Health. The communities of the state of São Paulo were divided into three groups: (1) municipalities with a service to verify the cause of death (and that may conduct an autopsy), (2) municipalities without a cause of death verification service, and (3) the region of Baixada Santista, which had an extremely large increase in the proportion of deaths from ill-defined causes between 1980 and 1995. The impact of autopsies on the proportion of deaths with an ill-defined cause was defined based on the classification made by the first physician evaluating the cause of death, that is, the physician who referred the case to the verification service for autopsy or who completed the death certificate without referring the case to the verification service. Deaths from external causes were excluded, since autopsy is mandatory in these cases. The following were evaluated: (1) proportion of cases classified by the first evaluating physician as having an ill-defined cause, (2) proportion of autopsies in relation to the total number of deaths (except from external causes), and (3) proportion of deaths classified as ill-defined by the first evaluating physician but explained by the autopsy.

RESULTS

The proportion of deaths classified by the first evaluating physician as having an ill-defined cause increased over the 1980-2002 period in all three groups studied: the municipalities with a verification service, the municipalities without a verification service, and in the Baixada Santista region. For the state of São Paulo overall, the increase was almost 30% over that 1980-2002 period. For the 1998-2002 period, the average proportion of autopsies compared to the total number of deaths (except from external causes) was 21.2% in the municipalities with a verification service, 6.4% in the municipalities without a verification service, and 2.6% in Baixada Santista. The proportion of deaths in 1998-2002 initially classified as having an ill-defined cause but that was explained by autopsy was 92.9% in the municipalities with a verification service, 32.5% in the municipalities without a verification service, and 10.7% in Baixada Santista.

CONCLUSIONS

The performance of autopsies accounts for the difference in proportional mortality with an ill-defined cause in the three groups of communities studied. The increase in the number of deaths classified as ill-defined by the first evaluating physician occurring in the state of São Paulo over the 1980-2002 period suggests an important decline in the quality of death certificates completed at that first level. More research should be done on the reasons for that change.