It was made an analysis of the medical certifications of death; it was analysed the fulfilling of the items of the certifications and the declaration of the underlying cause of death. A study had been made on a sample of the deaths occurred in hospitals in the period from March 1971 to February 1972. Additional information about each case was obtained through medical records; so, it was possible to know the real underlying cause of death and compare this one with that stated in the original certification. It was observed that, due to the lack of fulfilment of the corresponding items, only 4.0 per cent of the certifications brought together information about the time break between the onset of the disease and death and information about the laboratory data that confirmed the diagnosis. It was also observed that 65.5 per cent didn't show any information about the referred items. The most neglected item was that related to the time interval: 93.5 per cent of the certifications didn't carry it; 68.0 per cent didn't show informations about the laboratory data that confirmed the diagnosis; 31.4 per cent of the certifications didn't carry any statement about the real underlying cause of death; 6.5 per cent, though carrying it, were fulfilled in such an inadequate way that it became impossible to select the underlying cause of death for statistical purposes. It was concluded that the quality of the medical certifications of death of the city of S. Paulo is not satisfactory; comparing the results of this research with those of a similar one made in 1962/1963 it was verified that no improvement was observed in the last 10 years.
Death certificate; Vital statistics; Mortality
