Back to Annual Meeting Program

Predictors of successful outcomes following reoperative pelvic surgery are poorly defined. We aim to identify predictors of successful pelvic colorectal reoperations.

After IRB approval, we identified all patients from our prospective database who underwent re-operative pelvic surgery from 01/02-07/11. Patient demographics and clinical variables were confirmed by chart review. Primary outcome was the overall complication rate; including early (≤30 day) and late (>30 day) complications, and the secondary outcome was diverting stoma closure [SC]. Chi-square and student’s t test were performed. All variables with p<0.1 on univariate analysis were included in multivariate cox regression analysis.

254 pelvic reoperations (mean age 52 years, 47% male) were performed, including 104 with establishment of intestinal continuity. The most common diagnoses were: mucosal ulcerative colitis (35%), rectal cancer (24%), diverticulitis (18%), Crohn’s disease (6%) and familial adenomatous polyposis (6%). The most common initial operations performed were: total proctocolectomy with ileoanal pouch anastomosis [IPAA] (41%), resection with colorectal anastomosis (27%) and resection with coloanal anastomosis (17%). Indications for reoperation included: anastomotic leak (53%), fistula (14%), anastomotic stricture (10%), IPAA dysfunction (11%) and recurrent cancer (8%). The overall complication rate after reoperation was 20%; 8% early and 12% late. On multivariate regression, BMI≥25 kg/m2 (p<0.03) and anastomotic complications (leak, fistula or stricture) as the indication for reoperation (p<0.0001) were independent predictors of complications. Intraoperative complications during initial operation (p<0.002) and steroids at the time of reoperation (p<0.01) were independent predictors of early and late complications, respectively. 104 patients underwent reoperation with re-establishment of intestinal continuity with an overall complication rate of 19% (8% early and 11% late). 88% had an ileostomy at time of reoperation, of whom 71% underwent SC. Complications after reoperation delayed time to SC (380 vs. 196 days, p<0.05). On multivariate regression, IPAA (p<0.0001) and ASA class I (p<0.03) were independent predictors of SC; while rectal cancer (p<0.005) and diverticulitis (p<0.02) as the initial indication for operation, and coloanal anastomosis at initial operation (p<0.03) were independent negative predictors of SC.

Despite the complexity of pelvic colorectal reoperations, in experienced hands, the overall complication rate is low. Whenever possible, patients should be counseled to taper steroids and reach ideal BMI prior to undergoing pelvic reoperative surgery.