Risks, time trends, and mortality of colonoscopy-induced perforation in hospitalized patients.

ABSTRACT

BACKGROUND AND AIM: Colonic perforation is a rare complication of colonoscopy and ranges from 0% to 1% in all patients undergoing colonoscopy. The aim of this study was to assess the time trends, risk factors, and mortality associated with colonoscopy-induced perforation (CIP) in hospitalized patients as the data are limited. METHODS: Data are obtained from the Nationwide Inpatient Sample database to identify hospitalized patients between 2005 and 2014 that had CIP. Various factors like age and gender were assessed for association with CIP, followed by univariate and multivariate regression analyses. RESULTS: A total of 2 651 109 patients underwent inpatient colonoscopy between 2005 and 2014, and 4567 (0.2%) of the patients had CIP. Overall, incidence of CIP has increased from 2005 to 2014 (0.1% to 0.3%) (P < 0.001). On multivariate analysis, the adjusted odds ratio (OR) for CIP was highest in Caucasian race (OR 1.49 [1.09, 2.06]), followed by after polypectomy, history of inflammatory bowel disease, end-stage renal disease, and age > 65 years (OR [95% CI] of 1.35 [1.23, 1.47], 1.34 [1.17, 1.53], 1.28 [1.02, 1.62], and 1.21 [1.11, 1.33], respectively) (all P < 0.05). CIP group had 33% less obesity (OR [95% CI] 0.77 [0.65-0.9], P = 0.002) and 13-fold higher mortality (0.5% vs 8.1%) (P < 0.001) as compared to patients without CIP. The CIP-associated mortality ranged from 2% to 8% and remained stable throughout the study period. CONCLUSIONS: Our study suggests that the risk of CIP was highest in elderly patients, Caucasians, those with inflammatory bowel disease, end-stage renal disease, and after polypectomy. Recognizing the factors associated with CIP may lead to informed discussion about risks and benefits of inpatient colonoscopy.