P015 ILEOCOLIC RESECTION FOR CROHN’S DISEASE: CURRENT TRENDS FROM ANALYSIS OF 2 YEARS OF NSQIP IBD COLLABORATIVE DATA

Abstract Background Ileocolic resection (ICR) is performed for Crohn’s disease (CD) patients with terminal ileitis requiring surgery. Current National Surgical Quality Improvement Program (NSQIP) data is lacking specificity around IBD surgery, including stoma formation and biologic therapies. The NSQIP IBD Collaborative (NSQIP-IBD) is a multicenter working group formed to better collect and analyze perioperative data unique to IBD patients under the auspices of NSQIP. We present retrospective analysis of a multicenter cohort of ICR for CD to describe the current practice of ICR for CD across our collaborative and explore factors associated with rates of postoperative complications on behalf of NSQIP-IBD. Methods Review of NSQIP data from 10 participating sites was performed to select ICR cases for CD from March 2017 to March 2019. In addition to standard data from NSQIP, IBD-specific data regarding stoma formation, immunosuppressant use (biologics, steroids, and immune modulators), and dysplasia is included. Primary outcome was anastomotic leak measured in a 30-day postoperative window in undiverted patients. Secondary outcomes were total non-leak complications and total postoperative infections. Multivariable analysis was performed to adjust for confounding pre- and intraoperative confounders. Backward selection of covariates and factors was performed using a cutoff of p<0.2 for main effects. Results 506 ICR cases for CD were identified. 78 patients had stomas per our unique ileostomy NSQIP-IBD variable, compared to 38 found by querying generally available NSQIP data (48.7% of total stomas). ICR patients receiving stoma were more likely to have more severe ASA class and weight loss and had significantly lower albumin and hematocrit. Age, BMI, and sex were similar in either group. Intraoperatively, stoma patients were more likely have worse wound class, be emergent, and longer operative time. 421 cases had complete baseline and intraoperative data for multivariable analysis of leak rates. 422 were available for secondary outcomes analysis. Multivariable analysis of leak rates showed significant association with infections prior to operation (PATOS; OR=6.6, 95% CI 1.1–40, p=0.041). Significant predictors of total postoperative complication rate and infection rate are shown in Tables 1 and 2, respectively. Conclusions NSQIP-IBD data provides clearer, more detailed data than the NSQIP colectomy module alone in IBD patients. We show that CD ICR patients receiving stomas were more emergent, had intraoperative infection, or had more severe disease. This is consistent with most contemporary surgical practices. Total infections PATOS are associated with increased risk of postoperative anastomotic leak, non-leak complications, and infections. NSQIP-IBD data improves on existing NSQIP data to allow a more robust analysis of factors and outcomes unique to IBD cases. We anticipate with more time and greater numbers we will be able to obtain even more granular data.