New diagnostic criteria for infective endocarditis: A study of sensitivity and specificity

Abstract

The purpose of this study was to determine the sensitivity and specificity of new criteria proposed by Duke University for case definition of infective endocarditis as compared to the previously accepted Von Reyn criteria. A total of 143 consecutive suspected cases of infective endocarditis in 137 febrile patients were included. Of these, 69 had infective endocarditis, pathologically proven in 28, but with only a clinical diagnosis in 41. In the remaining 74 cases, the diagnosis of infective endocarditis was rejected after a follow-up of at least 3 months. The sensitivity of Duke's criteria was significantly higher, both when patients with possible infective endocarditis were considered as true-positive (definition 1; 100% vs 69%, P < 0.001) and when possible cases were considered as rejected (definition 2; 76% vs 51%, P < 0.01). Specificity was very high with both criteria: 92% Von Reyn vs 88% Duke (ns) with definition 1 and 99% Von Reyn vs 97% Duke (ns) with definition 2. The overall accuracy of the Duke criteria in the entire population was significantly higher with both definitions (0.94 vs 0.81 definition 1, P < 0.001; 0.87 vs 0.75, P = 0.015 definition 2). Duke's criteria for defining infective endocarditis has been shown to be more sensitive than previously adopted criteria, while maintaining a high degree of specificity. Therefore, they must be accepted as a substitute for previous criteria.