Abstract
BACKGROUND AND AIMS: Previous studies have not been able to determine whether non-selective beta-blockers (NSBB) reduce the risk of sepsis in cirrhosis. We aimed to examine this question with data from 1198 patients with cirrhosis and ascites included in clinical studies of satavaptan, a vasopressin receptor antagonist with no effect on infection risk. METHODS: Risk of sepsis was estimated for NSBB users vs nonusers. Patients were examined every four weeks, or in relation to hospitalization, for the one-year duration of the trials. We computed the cumulative risk of sepsis for patients who did vs did not use NSBB at baseline. We used Cox regression to compare hazard rates of sepsis between current users and nonusers, accounting for changes in NSBB use over time. We adjusted for patient sex and age, MELD-Na score, albumin, use of antibiotics, use of proton pump inhibitors, cirrhosis etiology, history of variceal bleeding or SBP, severity of ascites and HE, HCC, other cancers, and diabetes, while stratifying on geographical region. RESULTS: Of the 1198 patients, 54% used NSBB at some time. There were 56 sepsis episodes. The 1-year risk of sepsis was reduced to 5.7% (95% confidence interval [CI] 2.8-8.6) in baseline NSBB users vs 11.6% (95% CI 7.0-15.9) in baseline nonusers. The hazard ratio of sepsis for current NSBB users vs current nonusers was reduced to 0.5 (95% CI 0.3-0.8) and after adjustment to 0.7 (95% CI 0.4-1.3). CONCLUSION: NSBB use may reduce the risk of sepsis in patients with cirrhosis and ascites, but the precision of the estimate was limited by the number of episodes of sepsis.