Prognostic index in prediction of mortality from infective endocarditis

Abstract

Out of 176 patients with infective endocarditis complicating rheumatic (120) and congenital heart disease (38), mortality occurred in 35 patients (19.9%). Presence of leukocytosis, heart failure, major embolisation and isolation of certain organisms including Staphylococci, beta-haemolytic Streptococci, Pseudomonas and Klebsiella were ominous with higher mortality rates compared with those when they were absent (P less than 0.02). Stepwise logistic multiple regression was then applied and the four most important independent variables were identified. A prognostic index for the prediction of mortality for infective endocarditis was then constructed by the summation of the regression coefficients. By applying this index, patients with infective endocarditis could be divided into subgroups with increasing proportional mortality from 5.8 to 83.3%. It provides an objective assessment of the risk patients with infective endocarditis, and a more reliable evaluation of benefit of any new treatment regimen, including cardiac surgery, during the acute stage.