The sensitivity and specificity of clinical diagnostics during five decades. Toward an understanding of necessary fallibility.

Abstract

Published studies encompassing more than 50,000 autopsies were assessed to determine the sensitivity and specificity of clinical diagnostics (the diagnostic process) in persons dying of 1 of 11 specific diseases during the period 1930 through 1977. The accuracy of clinical diagnostics, as reflected in these two determinations, appeared to improve over this period with respect to some of the diseases studied (rheumatic heart disease and leukemia), while for others it worsened (pulmonary tuberculosis, peritonitis, carcinoma of the lung, gastric carcinoma, and carcinoma of the liver and extrahepatic biliary tract) and for a significant number diagnostic accuracy seemed refractory to sustained change (pulmonary embolism, primary cirrhosis of the liver, gastric/peptic ulcer, and acute coronary thrombosis/myocardial infarction). The findings suggest a new way in which the autopsy can be used to monitor clinical diagnostics to identify possible sources of systematic weaknesses and provide data that can be used to approach the difficult subject of necessary fallibility.