Infectious Diseases Detected at Autopsy
Previous studies have demonstrated significant discrepancy rates between clinical and autopsy diagnoses. However, infectious diseases have not received emphasis in these studies. We conducted a study to determine whether the clinical and autopsy diagnoses of infectious diseases are concordant or discrepant and to determine discrepancy rates.

Retrospective reviews of the records of 276 patients (adults, 182; fetuses and neonates, 94) who underwent autopsy during the years 1996 through 2001 were performed. Comparison of clinical and autopsy diagnoses was performed using the Goldman classification scheme.

Of 182 adult patients, 137 (75.3%) had an infectious disease at autopsy. In 59 (43.1%) of 137 patients, the infectious disease diagnoses were unknown clinically. Of 94 fetuses and neonates, 45 (48%) had an infectious disease at autopsy. In 26 (58%) of 45 patients, the infectious disease diagnoses were unknown before death.

There are substantial discrepancies between clinical and autopsy diagnoses of infectious diseases. In adults, acute bronchopneumonia is the infectious disease most often missed clinically; in fetuses and neonates, it is acute chorioamnionitis.

The autopsy remains a valuable method for confirming clinical diagnoses, resolving diagnostic questions, revealing unexpected findings, and determining the cause of death. It also has an important role in assuring the quality of care within health care organizations and for evaluating the safety and effectiveness of new drugs, medical devices, and diagnostic methods. Despite the benefits of the autopsy, autopsy rates have declined substantially since the 1950s, when up to 50% of patients dying in American hospitals underwent autopsy. Excluding autopsies performed for forensic purposes, the autopsy rate in American hospitals is the lowest ever, with rates as low as 5%.

A number of reasons for the decline in autopsy rates have been given, including physician fear of failure, concerns that autopsy results increase the likelihood of medical malpractice lawsuits, a perception that technologic advances in premortem diagnoses have decreased the number of diagnoses made only at autopsy, and lack of reimbursement. The perception that other diagnostic methods have made the autopsy obsolete is refuted by published data: it has been a persistent observation that autopsies continue to reveal unexpected findings despite advances in diagnostic technology and that a substantial number of persons have diseases that were not diagnosed before death. Published studies of the rate of discrepancy between clinical and autopsy findings, for both overall diagnostic accuracy and for specific diagnoses, have shown that a substantial number of diagnoses are made only at autopsy ( Table 1 ). A number of studies have shown that many patients have infectious diseases that were not diagnosed before death.

The purposes of this study were to determine the discrepancy rate for infectious diseases at an urban public hospital over a 6-year period and to categorize those findings in terms of their clinical importance using an established classification system.