Absence of demographic bias in selection for necropsy

The epidemiologic use of necropsy data has been inhibited by the belief that the information is demographically biased. This study was undertaken to investigate whether patients who have died and who have undergone a necropsy are representative of hospital deaths. In a review of all hospital and necropsied deaths from 1962 to 1981 in adult patients at the University of Kansas Medical Center, Kansas City, it was found that 7393 hospital patients died and 4122 underwent necropsy (56%), with similar proportions in both men and women. The mean age of those undergoing necropsy was slightly younger: 60 years vs 62 years in men and 57 years vs 59 years in women. The proportion of necropsies performed was relatively similar for whites (57%) and nonwhites (51%). Although the large sample size made all these differences "statistically significant," there were no clinically relevant differences with regard to gender, age, or race. These findings confirm a previous study at another institution and suggest that a demographic bias does not distort necropsy population data compared with data of the population of hospital deaths.