AUTHOR=Koch Maurizio , Maraolo Alberto Enrico , Natoli Giuseppe , Corrao Salvatore 

TITLE=Preventing acute diverticulitis. any roles for non-absorbable antibiotics? in search of evidence: a systematic review, meta-analysis, and trial sequential analysis

JOURNAL=Frontiers in Gastroenterology

VOLUME=Volume 2 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/gastroenterology/articles/10.3389/fgstr.2023.1170271

DOI=10.3389/fgstr.2023.1170271

ISSN=2813-1169

ABSTRACT=SUMMARY 
Background
Hospital admissions for diverticulitis, a complication of diverticular disease, are very much on the increase. Prevention of diverticulitis could cut costs and save lives.
Aims
To identify whether the risk of the first episode of diverticulitis (primary prevention) or recurrence of diverticulitis (secondary prevention) can be reduced in patients with diverticular disease using non-absorbable antibiotics (mainly rifaximin).
Methods
The studies were identified by searching PubMed and CENTRAL from 1990 to May 2022. The methodological quality of each study was evaluated. The outcome of the meta-analysis was the occurrence of a first or subsequent episode of diverticulitis. In addition, a trial sequential analysis was performed to assess whether the results would be subject to type I or type II errors.
Results 
Primary prevention: the risk difference was statistically significant in favor of rifaximin (-0,019, or -1.9%, CI -0,6 to -3,3%). There was no evidence of heterogeneity (I2 0%). At one year, two years, and eight years of age, the NNT was 62, 52, and 42, respectively. The level of evidence had a moderate degree of certainty.
Secondary prevention: the risk difference was statistically significant in favor of rifaximin (- 0,24, or -24%, CI -47 to -2%). There was evidence of heterogeneity (I2 92%); NNT resulted in 5. The grade level was low.
Conclusions
Rifaximin can reduce the risk of a first episode of diverticulitis. However, the cost-benefit ratio currently appears too high. Rifaximin could also reduce the risk of a second episode, but the quality of the evidence is low.
PROSPERO database number [379258].