RESUMO
INTRODUCTION: Autopsies are a fundamental moment of clinical audit that have been progressively decaying. MATERIALS AND METHODS: The autoptic rates of 6 hospitals of Piedmont Region (1639 autopsies, years 1995-97) were calculated. The Positive Predictive Value and Sensitivity (indicators of concordance between clinical and autoptic diagnosis) were calculated for each diagnosis; all discrepancies were classified as type I (adverse impact on patient's survival) and type II (uncertain impact) clinical errors. The diagnostic performance of medical, surgical and emergency staffs were compared with one another. RESULTS: Autopsies are rarely performed on the patients who died in hospital (7.13%). Independently from the diagnosis, the global concordance expresses a Positive Predictive Value of 37.22% and a Sensitivity of 47.71%. 401 type I errors and 307 type II errors were found. The best performance is reached by emergency staff. AIM: To evaluate the use and the meaning of autopsies in today's Italian healthcare organisations.