Correlation between clinical and autopsy diagnoses at a general hospital

In the past years, the number of hospital clinical autopsies have significantly decreased, with the risk of loosing some of the functions of this procedure, among which the evaluation of the diagnostic accuracy, which has been considered as a health care quality index, stands out. In this paper, the consistency between clinical diagnosis and autopsies in a general hospital is studied, as well as its association to some potentially relevant factors. We review 91 autopsies performed in patients dying in the Service of Internal Medicine during a period of six years. Each case was classified though consensus among a clinician and a pathologist into: mayor clinicopathological discrepancy, minor discrepancy or complete consistency. The relationship between consistency and age, sex, average length of stay in the hospital and study period was analyzed. The 91 autopsies were performed on 64 men and 27 women. The mean age was 63 years and the average length of study in the hospital was 9.7 days. The global consistency was 71.4%. In 15 (16.5%) cases, a major discrepancy was observed and in 11 (12.1%), a minor discrepancy. No differences were observed in the degree of consistency with respect to age, sex, average length of stay and study period. Cardiovascular diseases showed the greatest proportion of major discrepancies at the expense of three cases of pulmonary thromboembolism. We conclude that the degree of concordance in this study is consistent with other studies and we discuss the relevance of clinical autopsies regarding several aspects of health care.