Objectives: To compare the causes of deaths recorded on death certificates with findings at autopsy and to identify other deficiencies in the completion of death certificates.
Design and setting: The causes of death (Part I and II) recorded on the death certificates of all patients undergoing autopsy at St Vincent's Hospital, Melbourne, in 1992 were compared with the autopsy findings. Additional clinical information about previous medical and surgical history and antemortem investigations was obtained from the clinical summary in the autopsy reports, and from hospital charts in some cases.
Patients: In 1992, 132 hospital autopsies were performed (autopsy rate of 24.2%). Of these patients, 68% were aged 65 years or over, and 30% were aged 75 years or over.
Results: Major discrepancies between the cause of death listed on the certificates and autopsy findings were found in 16 cases (12%). Other deficiencies of death certification included listing the mode of death (e.g., cardiac failure) without an underlying cause in 14 cases (11%); failure to cite recent major surgery in 17 of 20 cases (85%); failure to specify site or organism in 33 of 40 cases (82.5%) of infection or sepsis.
Conclusions: This study confirmed findings of previous studies with respect to missed major diagnoses, but identified other deficiencies in certification of causes of death, which could compromise accuracy of statistics obtained from death certificates. Mechanisms by which these deficiencies can be prevented or corrected are discussed.