Comparison of the von Reyn and Duke criteria for the diagnosis of infective endocarditis: a critical analysis of 161 episodes.

The newly proposed diagnostic criteria for infective endocarditis (Duke criteria) were evaluated in 161 consecutive episodes treated for suspected infective endocarditis (IE) at one institution over a 5-year period. A significantly higher proportion of episodes were diagnosed as definite endocarditis by the new Duke criteria compared with a diagnosis as definite or probable endocarditis by the older von Reyn criteria (68% vs 56%; p < 0.05). If all 161 episodes were to be reclassified, excluding pathological data, which are seldom available at the start of treatment, the Duke criteria classified significantly more episodes as 'definite' compared with the analogous category 'probable' endocarditis by the von Reyn criteria (60% vs 44%; p < 0.01). Forty-four pathologically proven episodes were reclassified in the same way, and 73% of these episodes were classified as 'definite' IE by the Duke criteria compared to 55% classified as 'probable' IE using the von Reyn criteria In 33 (20%) episodes no heart murmur could be detected on admission and the Duke criteria provided an initial diagnosis of 'definite' IE in 58% of these episodes compared with only 6% classified as 'definite' or 'probable' IE by the von Reyn criteria (p < 0.0001). The newly proposed Duke criteria are an improvement on the older von Reyn criteria in the clinical diagnosis of IE, especially in initial phase of treatment. However, the sensitivity when establishing a correct clinical diagnosis of 'definite' IE for the pathologically proven cases was only 73%. The category of 'possible' IE by the Duke criteria is confusing, since it does not say anything of the likelihood on an actual IE; the only objective fact is that no alternative diagnosis has been proven.