Rifaximin for outcomes beyond hepatic encephalopathy in cirrhosis: An updated systematic review and meta-analysis

BACKGROUND Rifaximin is guideline-endorsed for hepatic encephalopathy (HE) recurrence prevention, but its effects on mortality and other non-HE complications remain uncertain across clinically heterogeneous trials.AIM To reassess non-HE outcomes of rifaximin-containing regimens in cirrhosis and examine sensitivity to comparators, masking, indication and inference.METHODS Four databases were searched through April 15, 2026; targeted supplementary verification was undertaken on September 8, 2026.Eligible randomized trials, including HE-directed studies, enrolled adults with cirrhosis and reported non-HE clinical or safety outcomes.Extractable events were pooled using DerSimonian-Laird risk ratios (RRs), with Hartung-Knapp-Sidik-Jonkman (HKSJ) sensitivity inference.Comparator-specific analyses were exploratory.Heterogeneous infection endpoints were reported separately.RESULTS Twenty-one trials were eligible; 15 contributed to pooled analyses.Twelve mortality comparisons (1979 participants) yielded RR 0.801 (95% confidence interval: 0.653-0.982;HKSJ: 0.641-1.001).Restricting to rifaximin as the sole randomized difference gave RR 0.808 (0.608-1.072).Masking and indication interactions were inconclusive.Spontaneous bacterial peritonitis-specific events yielded RR 0.518 (0.344-0.782; six trials); hepatorenal syndrome yielded RR 0.262 (0.090-0.767; two trials, 19 events), with an HKSJ interval of 0.028-2.497.One small trial assessed antimicrobial susceptibility, with substantial missing 4 / 55 microbiological data.Historical search limitations and departures from registered methods limit interpretation.CONCLUSION Mortality inference depends on analysis and comparator.Uncertain benefits for spontaneous bacterial peritonitis and hepatorenal syndrome do not justify routine expansion of rifaximin use.