Clinical diagnosis is not a source of bias in selection for necropsy

Recent reports have suggested a minimal difference in demographics between hospital deaths undergoing necropsy and those that do not. In order to assess whether the clinical diagnosis at death influenced the decision to request necropsy in deaths occurring in the hospital, a 10% sample of all adult necropsies at the University of Kansas Medical Center (Kansas City) from 1962 through 1981 was obtained. These patients (cases) were matched by age, race, sex, and date of death to hospital deaths not undergoing necropsy (controls). The charts were reviewed, and the primary diagnosis at death was ascertained and categorized into one of the 14 categories of the International Classification of Diseases. There were 412 case-control pairs, with 237 men and 175 women in each group. Most patients were classified under the headings Circulatory (142 cases vs 140 controls) and Neoplasm (143 cases vs 148 controls), followed by Respiratory (21 cases vs 33 controls) and Digestive (44 cases vs 28 controls). Overall, no statistically significant differences were observed between cases and controls. There were no gender-specific differences in all but two of the categories. In men, the controls had more patients classified under the heading injury (19 vs seven). In women, the control group had more patients classified under the heading Respiratory (14 vs four). Because of multiple comparisons, these findings could have occurred by chance. These data suggest that the primary diagnosis at death is not a source of bias in the selection for necropsy. Consequently, necropsy-selection bias, if it exists, may be due to other factors, such as the complexity of the patient's clinical course.