An evaluation of the Duke criteria in 25 pathologically confirmed cases of prosthetic valve endocarditis.

Six studies have compared the sensitivity of the Duke criteria with that of the Beth Israel criteria for the diagnosis of infective endocarditis without attempting to distinguish between native valve and prosthetic valve cases. After reviewing clinical data from 372 cases of suspected prosthetic valve endocarditis, we selected 25 cases in which pathological confirmation of prosthetic valve endocarditis was obtained. By using the Duke criteria and then the Beth Israel criteria, we reassigned a diagnostic classification to these 25 cases on the basis of clinical parameters alone. Using the Duke criteria, we correctly classified 19 cases (76%) as definite endocarditis and rejected none, while the Beth Israel criteria correctly classified six cases as probable endocarditis (24%) and rejected five cases (20%). Similar to previous investigators who examined the diagnostic sensitivity of the Duke criteria and Beth Israel criteria, we found that the Duke criteria appear more sensitive than the Beth Israel criteria in pathologically confirmed cases of prosthetic valve endocarditis.