Clinical setting and extent of premortem evaluation do not predict autopsy discrepancy rates

Autopsy rates have been affected by a number of factors, including technological advances and clinician beliefs of the diminished value of the autopsy. Such factors have resulted in a cultural shift in medicine away from the autopsy. Despite this shift, a number of studies have shown significant differences between antemortem clinical diagnoses and postmortem findings. Surveys of clinician beliefs about the autopsy have pointed toward antemortem diagnostic advancements as an important factor in declining rates. No study to date has addressed the hypothesis that such perceptions in diagnostic certainty have been matched by an actual decay in the yield of valuable or new information obtained by the autopsy. To address this hypothesis, we retrospectively compared the class I and class II discrepancies identified in 284 patients who died in three clinical settings with differing intensities of antemortem diagnostic workup. Despite a significantly different intensity of antemortem workup for patients in each clinical setting, including patients on a medical intensive care unit, patients on a surgical service and patients in an affiliated nursing home, discrepancy rates were found to be similar. Overall discrepancy rates for the medical intensive care unit, surgery service and nursing home patients were 27.8, 32.7 and 31.3%, respectively (P=0.84). In addition, we found no statistical difference in the complexity of workup in discrepant and nondiscrepant cases in each clinical setting. Our study data refute the hypothesis that the intensity of antemortem diagnostic evaluation correlates with an actual decrease in the rate of major diagnostic discrepancies identified at autopsy.