Ketorolac Use and Anastomotic Leak in Elective Colorectal Surgery: A Detailed Analysis.

BACKGROUND: Recent population-level analyses have linked ketorolac use to adverse outcomes. However, its use is also associated with decreased opioids and faster return of bowel function. OBJECTIVE: This study aims to assess the association between ketorolac and anastomotic leak. We hypothesize that receiving at least 1 dose of ketorolac will not be associated with anastomotic leak in elective colorectal surgery. DESIGN: This is a retrospective, observational cohort study of a prospectively collected data base. Anastomotic leak rates and other patient outcomes were adjusted for patient-level factors and then compared via a multivariable logistic regression. A secondary analysis assessed a dose-response association with anastomotic leak. SETTING: This study was conducted at a tertiary care colorectal surgery service. PATIENTS: Consecutive patients undergoing elective colorectal surgery with a nondiverted anastomosis were identified from 2012 to 2016. INTERVENTION: Exposure was defined as any administration of ketorolac during the perioperative time period. MAIN OUTCOME MEASURES: The primary outcome measured was anastomotic leak. RESULTS: A total of 877 patients met inclusion criteria. Of these, 479 (54.6%) were women, and the median age was 55 years. Overall, 566 (64.5%) patients were exposed to ketorolac. In the cohort, 27 (3.1%) patients experienced an anastomotic leak. In an unadjusted analysis, there was no association between ketorolac exposure and anastomotic leak (ketorolac: 3.1% vs no ketorolac: 3.3%; p = 0.84). This persisted in a multivariable model (OR, 0.98; 95% CI, 0.38-2.57; p = 0.98). Neither AKI (OR, 3.24; 95% CI, 0.51-20.6; p = 0.21), return to the operating room (OR, 1.07; 95% CI, 0.40-2.85; p = 0.88), nor readmission (OR, 1.03; 95% CI, 0.59-1.80; p = 0.93) was associated with ketorolac use. In a secondary analysis of patients receiving ketorolac, there was no association between total ketorolac dosing and anastomotic leak (OR, 0.99; 95% CI, 0.99-1.00; p = 0.20). LIMITATIONS: This study was a retrospective review, and there was a low incidence of anastomotic leak. CONCLUSION: Ketorolac exposure was associated with neither anastomotic leak nor other important postoperative outcomes. See Video Abstract at http://links.lww.com/DCR/A784.