Allison J Pang, Mohammed Alqahtani, Gabriela Ghitulescu, Julio Faria, Nancy Morin, Carol-Ann Vasilevsky, Marylise Boutros, MD. McGill University
INTRODUCTION: Temporary loop ileostomies are frequently performed during rectal cancer surgeries in order to protect the colorectal anastomosis. Although ileostomy reversal is a relatively simple operation, postoperative ileus (POI) still remains one of its most common complications, paralleling the incidence found after major colorectal resections. The literature suggests that this may be due to dysfunction of the de-functioned bowel downstream from an ileostomy due to the lack of stimulation. We hypothesize that longer time interval between ileostomy creation and closure will be associated with a greater incidence of POI.
METHODS: After IRB exemption, we retrospectively reviewed the National Readmissions Database (NRD) from 2010-2015. Adults who underwent (i) non-emergent proctectomies with a diverting loop ileostomy and (ii) ileostomy reversal within the same calendar year were identified using ICD codes. The cohort was truncated after the first 5 months of each year to allow for a 7-month follow-up period for each patient. The primary outcome was POI after ileostomy closure. Independent associations between covariates and POI were identified by multivariate logistic regression.
RESULTS: The study identified 17,549 patients who underwent proctectomy with diverting loop ileostomy. Of these, 2,596 underwent stoma closure within 7 months of their proctectomy, and 11.2% developed POI at closure. On univariate analysis, patients who experienced POI were more likely to be male (p=0.028), have greater DRG-severity of illness (p<0.001), suffer pre-operative weight loss (p=0.002), experience an anastomotic complication after proctectomy (p=0.001), have a longer length of stay during their index surgery, and have surgery at a non-teaching hospital (p=0.003). Time interval to closure was not significantly different (14.0 weeks (9.1-22.9) vs. 15.3 weeks (9.6-24.7)) with POI, (p=0.081). On multivariate analysis after accounting for confounders, time to closure (1.019 OR, 95%CI 1.002,1.037) was a significant predictor of POI. Other significant predictors were non-teaching status of the treating hospital and the presence of major DRG-severity of illness (1.4 OR, 95%CI 1.9-1.1 and OR, 95%CI 1.1-3.4, respectively).
CONCLUSION: In this national database, we observed that when the time interval to ileostomy closure lengthens, the incidence of POI increases. This could have important implications in the management of colorectal patients with regards to advocating for more prompt closure surgery and for the potential use of bowel stimulation prior to ileostomy closure.
View Poster
This abstract was accepted for Poster presentation at the 2020 SAGES Virtual Meeting in the Colorectal topic. Its program number was: P260 and its Abstract ID was: 102844