Four RCTs were included in the review (n=1,660 patients, 970 in the rifaximin group and 690 in the control group). One trial was blinded. Three trials had adequate sequence generation. All trials had adequate allocation concealment. Incomplete outcome data was addressed and all included trials were considered to be free of selective reporting and other bias. The Jadad scores ranged from 2 to 4 points.
A significant increase in symptom relief was found in favour of rifaximin plus fibre supplementation compared with control groups (RD 29.0%, 95% CI 24.5 to 33.6; four RCTs; NNT=3).
A statistically, but not clinically, significant reduction in complication rate was found in the rifaximin group compared with the control group (RD -1.7%, 95% CI -3.2 to -1.5; four RCTs; NNT=59). No evidence of statistical heterogeneity was found (I2=0%). When only acute diverticulitis was considered, the pooled risk difference in the control group for complication rate was -1.9% (95% CI -3.4 to -0.57; four RCTs; NNT=50).
Three out of four trials reported side effect data; the authors reported that no significant differences were found between treatment and control groups.
No evidence of publication bias was found.