Marseille scoring system for empiric treatment of infective endocarditis

Abstract

Despite advances in medical, surgical, and critical care, infective endocarditis (IE) remains associated with considerable morbidity and mortality. We evaluated the performance of the Marseille score, including clinical data and biological tests obtained within 2 h, to identify patients at high risk of IE in order to initiate early antimicrobial treatment. This was secondarily confirmed using modified ESC criteria combined with molecular testing and ⁽¹⁸⁾fluorodeoxyglucose-positron emission tomography/computed tomography as diagnostic tools. In a prospective cohort study, we enrolled 484 patients with cardiovascular predisposition and clinical suspicion of IE from 2011 to 2013. The final diagnosis was definite IE in 123 patients and possible IE in 107. Marseille score was calculated adding one point for each present parameter (range 0-9). This score includes clinical, epidemiological (male, fever, splenomegaly, clubbing, vascular disease and stroke) and biological criteria (Leucocytes >10,000/mm3, sedimentation rate (SR) > 50/mm or C reactive protein >10 mg/L and hemoglobin <100 g/l). A score of 2 or more performed best in predicting IE in patients with predisposing heart lesions. Sensitivity was better on left-side heart lesions (94%) than on right-side heart lesions (85%) (p = 0.04) and better for valvulopathy (94%) than intra cardiac devices (84%) (p = 0.02). The predictive positive value of prosthetic valves was greater than that of native valves (p = 0.02). Using our simple Marseille score combined with our standardized diagnostic procedures would help improve IE management by focusing on early empiric treatment within 2 h of admission for patients with cardiac predisposition factors.