RESUMO
OBJECTIVE: To identify factors associated with short-term, intermediate and long-term outcome in patients with infective endocarditis (IE) and the need for treatment on intensive care unit (ICU). DESIGN AND SETTING: Retrospective analysis and long-term follow-up by questionnaire in the two medical ICUs of our university hospital. PATIENTS: We conducted a retrospective analysis of all consecutive patients with IE and need for ICU treatment in our department between 2002 and 2009. All patients fulfilled the modified Duke criteria for definite diagnosis of IE. MEASUREMENTS AND MAIN RESULTS: Data of 216 patients (aged 62 ± 14 years, 31 % female) were analyzed, 15.7 % of whom had prosthetic valve endocarditis. Infectious agent (IA) was identified in 74 % and surgery was performed in 57 %. 56 patients (24.9 %) died on ICU, 9 patients were sent to palliative care units and died several days later. During follow-up, another 44 patients died. Multivariate Cox-regression analysis identified the following independent risk factors: High initial SAPS II for 30d-, multiple organ failure and high maximum SAPS II for 100d- and high maximum leukocyte count for long-term mortality. Surgical intervention during ICU was an independent predictor of a better 30d outcome. CONCLUSIONS: In contrast to general IE populations, IA and the type of infected impaired valve are not main predictors of survival in critically ill IE-patients. Biomarker of acute infection and markers for severity of illness (scores and organ failure) are independent risk factors for mortality. The surgical clearance of infected valve, device or abscesses is an independent predictor of 30d outcome.