Occurrence of Delayed Non-GI Events Post-Colonoscopy and Patients With Identifiable Increased Risk

Abstract

Introduction: Track overall rates of delayed non-GI (cardiac, vascular, pulmonary) adverse events (AEs) 30-days post-colonoscopy in average risk and complex patients. Determine any increased risk (IR) of delayed non-GI for complex patients having colonoscopy versus those not having colonoscopy or other surgical procedures. Methods: An independent data analysis was done by IMS using the LifeLink health claims database, a fully adjudicated, longitudinal claims database representative of the national, commercially-insured population. Of the patients, 495,902 colonoscopy patients (Mar 2011-Feb 2012) were identified and year prior comorbid diagnosis codes identified complex patients. Post-colonoscopy AEs in emergency department or hospital settings were tracked for 30 days. A cohort of 44,640 complex patients without colonoscopy or other surgery matched 1:1 basis on age, gender, and pre-colonoscopy risk factors. Results: Events 30 days after colonoscopy occurred in 4.7% of average risk patients vs. 15.4% (p<0.0001) of antithrombotic (AT) patients and 6.7% (p<0.0001) of COPD/OSA patients. Bleeding was the largest portion of AEs for average risk patients, but delayed non-GI AEs were increased in AT patients (11.2%; p<0.0001) and COPD/OSA patients (2.2%; p<0.0001). Over 94% occurred <2 weeks of colonoscopy. AT patients had increased bleeding (9.3% vs. 4.0%), yet delayed non-GI AEs had IR (11.2% vs. 0.9%). As AT and COPD/OSA patients have IR of delayed non-GI AEs, a matched cohort analysis of having a colonoscopy vs. patients not undergoing another surgical procedure showed an IR odds ratio (OR) of delayed non-GI AEs in AT patients (3.3 [95% confidence interval (CI) 2.9-3.8]) and COPD/OSA (1.6 [95% CI 1.4-1.9]). Patients prescribed anticoagulants (AC)+AT (5.0 [3.9-6.3]), AC (2.6 [2.3-2.8]), COPD (2.4 [2.1-2.7]) had highest increased risk. Linear regression of 4 variables showed an R2=0.17 limited by infrequent non-GI AEs and pre-index conditions only occurring in 20% of the population. Non-GI ORs were increased in: males, increasing age, and number of preconditions, but smaller AEs for colonoscopy with intervention. The regression’s predictive value was based on age (44%) and precondition (47%). A second regression crossing age and precondition showed non-GI AE risk increases with age, yet each precondition had substantial risk for patients 50+ (Table 1).Table 1: Non-GI AE ORs for Age and PreconditionConclusion: Following colonoscopy, considerable 30-day risks of delayed non-GI CP or neurologic AEs exist in patients with identifiable comorbid conditions. Appropriate cognizance should be incorporated into the global risk benefit assessments and development of risk minimization strategies in patients considered for colonoscopy. Disclosure - All authors are consultants of Given Imaging (Covidien). Funds to acquire IMS data were provided by Given Imaging (Covidien). As part of study, IMS provided data analysis and statistical support.