Trends and outcomes of infective endocarditis in cirrhosis: a propensity-matched national study

Background Cirrhosis is the most common cause of liver-related death and bacterial infection is a common comorbidity in cirrhosis. We aimed to study the trends and outcomes of infective endocarditis in cirrhosis. Methods A propensity-matched analysis of the National Inpatient Sample database was performed to assess outcomes of infective endocarditis in adult patients (>18 years) from 2004-2013 with and without cirrhosis. Various outcomes were assessed for outcomes of infective endocarditis in cirrhosis. Multivariate regression analysis was performed for predictors of mortality in infective endocarditis. Results There has been no significant change in incidence (3.3-3.6%, P = 0.27) and overall mortality (6.3-8.6%, P = 0.42) of infective endocarditis in cirrhosis. After propensity matching, patients with cirrhosis had significantly higher in-hospital mortality (15 vs. 10.6%, P < 0.001) and acute kidney injury (AKI) (31.8 vs. 28.5%, P < 0.001) as compared to no cirrhosis. Microbiological analysis revealed significantly higher rates of streptococci (35.3 vs. 31.9%, P < 0.001) and fungal infective endocarditis (0.03 vs. 0%, P < 0.001) and lower incidence of Gram-negative infective endocarditis (3.9 vs. 6.3%, P < 0.001) in cirrhosis. Cirrhosis patients had significantly less surgical intervention (10.2 vs. 30.3%, P < 0.001) along with overall total cost and length of stay as compared to no cirrhosis. On multivariate analysis, advanced age, AKI, shock and mechanical ventilation were positive predictors of mortality in infective endocarditis patients with cirrhosis patients. Conclusions Cirrhosis is an independent predictor of mortality in infective endocarditis with worse outcomes and less surgical intervention. Gram-negative infective endocarditis is lower in cirrhosis, whereas streptococci and fungal infective endocarditis are higher than noncirrhotic patients. Copyright (C) 2021 Wolters Kluwer Health, Inc. All rights reserved.