Selection for necropsy: evaluation of 321 autopsied cases from 3041 consecutive deaths among residents of a defined geographical area

By undertaking to act as 'medico necroscopo' (the medical doctors in charge of authorizing every burial) in the Unità Socio-Sanitaria Locale (USSL) 70, a region containing 92,500 inhabitants, we were able to collect over a four-year period (1985-88) information on 3371 deaths among residents of the area. Of these, 68.1% died at home, 12.6% in hospital with no autopsy, 9.5% in hospital with an autopsy and 9.8% outside of the USSL region. For the analysis, age, sex and leading cause of death were considered as explanatory variables and place of death and proportion of autopsied cases as the response variables. The factors found to be associated with deaths occurring in hospital were non-neoplastic (particularly cardiovascular and digestive) pathological conditions in patients under the age of 60. People with malignant neoplasms usually died at home, especially when they had a gastric or breast cancer. Factors associated with autopsy were: death at a young age, short length of hospitalization and death from a cardiovascular disease. Sex did not appear to play a direct role in selection for necropsy. Our results indicate that, in our and similar regions, the cases selected for autopsy form a subset that is so specific that any extrapolation to the total deceased population is misleading.