LAUSR

OBJECTIVE The aim of this study was to compare ghost ileostomy (GI) and loop ileostomy (LI) in patients undergoing oncologic resection for rectal cancer in terms of postoperative morbidity. SUMMARY BACKGROUND DATA LIs are often fashioned to protect downstream anastomoses following oncologic resection for low rectal cancer at medium-to-high risk of anastomotic leak. More recently, GIs have been utilized in patients with low-to-medium risk anastomoses to reduce the rate of unnecessary stomas. METHODS Medline, Embase, and CENTRAL were systematically searched. Studies investigating the use of GI in patients undergoing oncologic resection for rectal cancer were included. The primary outcomes were anastomotic leak and postoperative morbidity. Secondary outcomes included stoma-related complications and length of stay (LOS). Pairwise meta-analyses were performed with inverse variance random effects. RESULTS From 242 citations, 14 studies with 946 patients were included. In comparative studies, 359 patients were undergoing GI and 266 patients were undergoing LI. Pairwise meta-analysis revealed no differences in the prevalence of anastomotic leak (OR 1.40, 95%CI .73-2.68, P = .31), morbidity (OR .76, 95%CI .44-1.30, P = .32), or LOS (SMD -.05, 95%CI -.33-.23, P = .72). International Study Group of Rectal Cancer anastomotic leak grades were as follows: Grade A (GI 0% vs LI 13.3%), Grade B (GI 80.9% vs LI 86.7%), Grade C (GI 19.1% vs LI 0%). CONCLUSIONS GI appears to be a safe alternative to LI following oncologic resection for rectal cancer. Larger, prospective comparative studies are warranted to evaluate the use of GI in patients deemed to be at low-to-medium risk of anastomotic leak.