Value of autopsy in Internal Medicine: discrepancies between clinical and postmortem diagnosis

Introduction
The number of autoptic controls is progressively dropping worldwide during the last decades. Several social, cultural and normative causes may underlie this phenomenon. However, a hidden cause may be represented by the scepticism of physicians about the gain in information given by this practice. There is in fact a general thinking that available diagnostic techniques may be adequate for a correct diagnosis, even in the case of a mortal outcome. Nevertheless, several studies have shown quite relevant discrepancies between premortem and postmortem diagnosis. In this study, the authors evaluate the accuracy of clinical diagnosis in diseases leading the patient to death.

Materials and methods
Retrospective analysis including all patients died in the Department of Internal Medicine of a General Hospital in Rome during a three year time period (1st January 2007-31 December 2009). Age, sex, period of hospitalization, clinical diagnosis, autoptic diagnosis, cause of death, and level of discrepancy between clinical diagnosis and autoptic control have been collected. Main diagnoses have been classified as follows: (1) discrepancy with a possible influence on survival; (2) discrepancy with no or questionable influence on survival.

Results
Ninety-two cases (42 males, 50 females; mean age 79.3 years) have been included. Thirty-four main diagnoses (36.9%) have been classified as discrepant (15.2% classified as type 1 discrepancy and 21.7 as type 2).

Discussion
A substantial discrepancy between clinical diagnosis and autoptic control is confirmed by the current study, which supports the role of autopsy as a tool for the improvement of medical practice.