The influence of diverting loop ileostomy vs. colostomy on postoperative morbidity in restorative anterior resection for rectal cancer: a systematic review and meta-analysis.

Abstract

The aim of this systematic review and meta-analysis was to evaluate the morbidity of loop ileostomy (LI) and loop colostomy (LC) creation in restorative anterior resection for rectal cancer as well as the morbidity of their reversal. PubMed, EMBASE, MEDLINE via Ovid, and Cochrane Library were systematically searched for records published from 1980 to 2017 by three independent researchers. The primary endpoint was overall morbidity after stoma creation and reversal. Mantel-Haenszel odds ratio (OR) was used to compare categorical variables. Clinical significance was evaluated using numbers needed to treat (NNT). Six studies (two randomized controlled trials and four observational studies) totaling 1063 patients (666 LI and 397 LC) were included in the meta-analysis. Overall morbidity rate after both stoma creation and closure was 15.6% in LI vs. 20.4% in LC [OR(95%CI) = 0.67 (0.29, 1.58); p = 0.36] [NNT(95%CI) = 21 (> 10.4 to benefit, > 2430.2 to harm)]. Morbidity rate after stoma creation was both statistically and clinically significantly lower after LI [18.2% vs. 30.6%; OR(95%CI) = 0.42 (0.25, 0.70); p = 0.001; NNT(95%CI) = 9 (4.7, 29.3)]. Dehydration rate was 3.1% (8/259) in LI vs. 0% (0/168) in LC. The difference was not statistically or clinically significant [OR(95%CI) = 3.00 (0.74, 12.22); p = 0.13; NNT (95%CI) = 33 (19.2, 101.9)]. Ileus rates after stoma closure were significantly higher in LI as compared to LC [5.2% vs. 1.7%; OR(95%CI) = 2.65 (1.13, 6.18); p = 0.02]. This meta-analysis found no difference between LI and LC in overall morbidity after stoma creation and closure. Morbidity rates following the creation of LI were significantly decreased at the cost of a risk for dehydration.