Mechanical bowel preparation for elective colorectal surgery.

BACKGROUND: For more than a century the presence of bowel content during surgery has been related to anastomotic leakage. Mechanical bowel preparation has been considered an efficient agent against leakage and infections complications. This dogma is not based on solid evidence, but more on observational data and expert's opinions. OBJECTIVES: To determine the security and effectiveness of prophylactic mechanical bowel preparation for morbidity and mortality rates in colorectal surgery. The following hypothesis was tested: "The use of mechanical bowel preparation before elective colorectal surgery reduces the incidence of postoperative complications". SEARCH STRATEGY: All publications describing mechanical bowel preparation before elective colorectal surgery was sought through computerized searches of EMBASE, LILACS, MEDLINE, and Cochrane Library; by hand-searching in relevant medical journals, from major gastroenterological congresses, without limitation for date and language, using the search strategy described by the Colorectal Cancer Review Group. In addition, randomised clinical trials will be searched through personal communication with colleagues and from conference proceedings SELECTION CRITERIA: None STUDIES: All randomised, clinical trials, that were performed in order to answer the hypothesis. PARTICIPANTS: Patients submitted elective colorectal surgery. INTERVENTIONS: Any strategy in mechanical bowel preparation compared with no mechanical bowel preparation. PRIMARY OUTCOME MEASURES: 1. Anastomosis leakage- stratified for rectum and colon 2. Overall anastomotic leakage SECONDARY OUTCOME MEASURES: 3. Mortality 4. Peritonitis 5. Re operation 6. Wound Infection 7 Infectious extra-abdominal complication 8. Non-infection extra-abdominal 9. Overall surgical site infections DATA COLLECTION AND ANALYSIS: Data was independently extracted by two reviewers and cross-checked. The methodological quality of each trial was assessed by the same two reviewers. Details of the randomisation (generation and concealment), blinding, whether an intention-to-treat analysis was done, and the number of patients lost to follow-up was recorded. The results of each RCT was summarised in 2 x 2 tables for each outcome. For analysis the Peto-Odds ratio was used as defaults (no statistical heterogeneity was observed) MAIN RESULTS: Of the 1159 patients with anastomosis (6 RCTs), 576 were allocated for mechanical bowel preparation (groups 1) and 583 for no preparation (groups 2) before elective colorectal surgery. Of 1204 patients totally enrolled 595 were in groups 1 and 609 in groups 2. PRIMARY OUTCOMES: 1) Anastomotic leakage - stratified:A) Low anterior resection: 12.5% (6 of 48 patients in 1) compared with 12% (6 of 50 patients in 2); Peto OR 1.17, 95% CI: 0.35 - 3.96 (non-significant) B) Colonic surgery: 1.16% (2 of 172 patients in 1) compared with 0.6% (1 of 166 patients in 2); Peto OR 1.75, 95% CI: 0.18 - 17.02 2) Overall anastomotic leakage: 5.5% (32 of 576 patients in 1) compared with 2.9% (17 of 583 patients in 2); Peto OR 1.94, 95% CI: 1.09 - 3.43 (P=0.02) SECONDARY OUTCOMES: 3) Mortality: 0.6% (2 of 329 patients in 1) compared with 0% (0 of 326 patients in 2); Peto OR 7.95, 95% CI: 0.49 - 128.34 (non-significant) 4) Peritonitis: 5.1% ( 13 of 254 patients in 1) compared with 2.8% (7 of 252 patients in 2); Peto OR 1.90, 95% CI: 0.78 -4.64) (non significant) 5) Reoperation: 3.3% ( 11 of 329 patients) compared with 2.5% (8 of 326 patients); Peto OR 1.40, 95% CI: 0.56 - 3.49) (non-significant) 6) Wound infection: 7.4% (44 of 595 patients in 1) compared with 5.7% (35 of 609 patients in 2); Peto OR 1.34, 95% CI: 0.85 - 2.13 (non-significant) 7) Infectious extra-abdominal complication: 8.3% ( 14 of 168 patients in 1) compared with 9.4% (15 of 159 patients in 2); Peto OR, 95%: 0.87 (0.41 - 1.87) 8) Non-infection extra-abdominal complication: 8.0% ( 20 of 250 patients in 1) compared with 7.0% (17 of 246 patients in 2); Peto OR 1.19, 95% CI: 0.61 - 2.32 (non-significant) - 9) Surgical site infection: 9.8% (31 of 325 patients in 1) compared with 8.3% (27 of 322 patients in 2); Peto OR 1.20, 95% CI: 0.70 - 2.05 (non-significant) - REVIEWER'S CONCLUSIONS: The results failed to support the hypothesis that bowel preparation reduces anastomotic leak rates and other complications. There was no a priori hypothesis that bowel preparation may increase anastomotic leak rates, so this was not stated. Thus, the routine use of mechanical bowel preparation in patients undergoing elective colorectal surgery is questioned.