Objective To assess the effectiveness of antibiotics in the treatment of hepatic encephalopathy (HE).

Data sources Cochrane Hepato-Biliary Group controlled trials register, Cochrane Controlled Trials Register, MEDLINE, LILACS, SCI Expanded, Embase and http://clinicaltrials.gov/ until January 2010; reference lists of relevant articles and authors were searched.

Review methods The methodological approach outlined in the Cochrane Handbook was followed. Randomised trials comparing antibiotics with placebo, no intervention, non-absorbable disaccharides, another antibiotic or any other active treatment for HE were included. The primary outcome measures were no improvement of HE and all-cause mortality.

Results Twenty-eight studies were included in this review; most had a small sample size and were methodologically of a low quality. Two studies (N=132) compared antibiotics with placebo or no intervention, reporting no differences in primary outcomes between groups. Twelve studies (N=708) compared antibiotics with non-absorbable disaccharides, reporting no difference in all-cause mortality (RR=1.11 (95% CI 0.60 to 2.07)) but a marginally significant reduction in the risk of “no improvement” (RR 0.83 (0.69 to 1.00)). In sensitivity analysis, this finding was significant in trials evaluating chronic encephalopathy (RR 0.50 (0.28 to 0.89)) and in trials with co-administration of cathartic agents (RR 0.82 (0.67 to 0.99)). Ten trials (N=336), compared different antibiotic regimens, most evaluating rifaximin and neomycin. Rifaximin reduced the risk of no improvement compared to neomycin, but not significantly so (RR 0.61 (0.34 to 1.09)).

Conclusions Insufficient evidence exists to support or refute the use of antibiotics for HE. Antibiotics were slightly superior to non-absorbable disaccharides at improving HE, with borderline significance and possible effect modification.