Oral Mechanical Bowel Preparation for Colorectal Surgery: Systematic Review and Meta-Analysis

BACKGROUND: Oral mechanical bowel preparation is often used before elective colorectal surgery to reduce postoperative complications. OBJECTIVE: The purpose of this study was to synthesize the evidence on the comparative effectiveness and safety of oral mechanical bowel preparation versus no preparation or enema. DATA SOURCES: We searched MEDLINE, the Cochrane Central Register of Controlled Trials, Embase, and CINAHL without any language restrictions (last search on September 6, 2013). We also searched the US Food and Drug Administration Web site and ClinicalTrials.gov and supplemented our searches by asking technical experts and perusing reference lists. STUDY SELECTION: We included English-language, full-text reports of randomized clinical trials and nonrandomized comparative studies of patients undergoing elective colon or rectal surgery. For adverse events we also included single-group cohort studies of at least 200 participants. INTERVENTIONS: Interventions included oral mechanical bowel preparation, oral mechanical bowel preparation plus enema, enema only, and no oral mechanical bowel preparation or enema. MAIN OUTCOME MEASURES: Anastomotic leakage, all-cause mortality, wound infection, peritonitis/intra-abdominal abscess, reoperation, surgical site infection, quality of life, length of stay, and adverse events were measured. We synthesized results across studies qualitatively and with Bayesian random-effects meta-analyses. RESULTS: A total of 18 randomized clinical trials, 7 nonrandomized comparative studies, and 6 single-group cohorts were included. In meta-analyses of randomized clinical trials, the credibility intervals of the summary OR included the null value of 1.0 for comparisons of oral mechanical bowel preparation and either no oral preparation or enema for overall mortality, anastomotic leakage, wound infection, peritonitis, surgical site infection, and reoperation. These results were robust to extensive sensitivity analyses. Evidence on adverse events was sparse. LIMITATIONS: The study was limited by weaknesses in the underlying evidence, such as incomplete reporting of relevant information, exclusion of non-English and relevant unpublished studies, and possible missed indexing of nonrandomized studies. CONCLUSIONS: Our results could not exclude modest beneficial or harmful effects of oral mechanical bowel preparation compared with no preparation or enema.