Use of risk scores surgical in patients with non-surgically treated infectious endocarditis at the university hospitalier centre Ibn Rochd of Casablanca, Morocco (about 192 cases)

Abstract

The accuracy of surgical scores in predicting hospital mortality in non-surgically treated patients with infectious endocarditis (IE) has not yet been explored. Patients with defined EI who did not undergo valve surgery were selected from the database of the cardiology department of the Ibn Rochd university hospitlier centre in Casablanca, Morocco. Patients were noted using (a) six systems specifically designed to predict mortality in hospital after surgery for IE, (b) three commonly used risk scores for cardiac surgery, and (c) a risk score to predict mortality at six months in IE after surgery or medical therapy. The calibration (Hosmer-Lemeshow test) and discriminating power (analysis of receptor operating characteristics [ROC]) were evaluated for each score. Areas below the ROC curves were compared one by one (Hanley-McNeil method). A total of 192 patients (average age, 65.2 ± 15.2 years) were considered for the analysis. There were 38 (19.8%) death in hospital. Age > 70 years (P–0.001), Staphylococcus aureus as a causal agent (P–0.05) and severe sepsis (P–0.027) were independent predictors of hospital mortality. Despite many differences in the number and type of variables, all but two of the scores studied showed good calibration (P > 0.66). However, the discriminating power was satisfactory (area below the ROC curve > 0.70) for only three of the EI-specific scores and two of the scores used to predict mortality after cardiac surgery. Of the 10 surgical scores evaluated in this study, five could be adopted to predict hospital mortality, even for IE patients receiving medical treatment only.