Abstract
BACKGROUND: Bacterial endocarditis in dialysis patients is associated with high mortality rates. The literature is limited on the long-term outcomes of valvular replacement surgery and the choice of prosthesis in dialysis patients with bacterial endocarditis. METHODS AND RESULTS: Dialysis patients hospitalized for bacterial endocarditis from 2004 to 2007 were studied retrospectively using data from the United States Renal Data System. Long-term survival of patients undergoing valve replacement surgery with tissue or nontissue valves was compared by use of the Kaplan-Meier method. A Cox proportional hazards model was used to identify independent predictors of mortality in patients undergoing valvular replacement surgery. During the study period, 11 156 dialysis patients were hospitalized for bacterial endocarditis and 1267 (11.4%) underwent valvular replacement surgery (tissue valve, 44.3%; nontissue valve, 55.7%). In the valve replacement cohort, 60% were men, 50% were white, 54% were 45 to 64 years of age, and 36% were diabetic. Estimated survival with tissue and nontissue valves at 0.5, 1, 2, and 3 years was 59% and 60%, 48% and 50%, 35% and 37%, and 25% and 30%, respectively (log-rank P=0.42). Staphylococcus was the predominant organism (66% of identified organisms). Independent predictors of mortality in patients undergoing valve replacement surgery included older age, diabetes mellitus as the cause of end-stage renal disease, surgery during index hospitalization, staphylococcus as the causative organism, and dysrhythmias as a comorbid condition. CONCLUSIONS: Valve replacement surgery is appropriate for well-selected dialysis patients with bacterial endocarditis but is associated with high mortality rates. Survival does not differ with tissue or nontissue prosthesis.