284 Factors That Underpin the Increase in Right Colon Adenoma Detection Rate: A Pooled Analysis of Three Reports Comparing Water Exchange, Water Immersion, and Air Insufflation Colonoscopy

Abstract

INTRODUCTION: Interval cancers due to missed adenomas tend to cluster in the right colon. Both water immersion (WI) and water exchange (WE) use water infusion to aid insertion. WI, which is easy to adopt at the discretion of the colonoscopist, removes water mainly during withdrawal. WE, which has evolved from WI, removes water mainly during insertion using new skills. WE, but not WI, provides insertion cleansing and improves bowel cleanliness. Reduced cleaning-related multi-tasking distractions from search for polyps, but not improved bowel cleanliness per se, is linked significantly to the improved adenoma detection rate (ADR) with WE (UEGJ Dec 5, 2018). In a pooled data analysis of three reports, WE significantly increased right colon adenoma detection rate (rADR) compared with WI and air insufflation (AI) (DDW 2019). However, the significant and independent factors that account for the increased rADR have not been described. We aim to assess the factors associated with an increase in rADR by multiple logistic regression analysis of the pooled individual patient data in the three aforementioned reports. METHODS: Three reports with data on rADR comparing WE with WI and AI were identified. The demographic and procedural data were extracted, pooled to optimize sample size and analyzed. Multiple logistic regression analysis was conducted using backward likelihood ratio selection procedure. RESULTS: Table 1 shows comparable pooled demographic data. Table 2 shows volume infused and suctioned confirm use of WE, WI and AI. WE has highest cleanliness scores, least use of adjunct maneuvers, and highest willingness to repeat. rADR of WE (22.4%) is significantly higher than those of WI (17.4%) (P = 0.009) and AI (15.7%) (P < 0.001); but WI and AI are similar (reported 2019 DDW). Table 3 shows multiple logistic regression analysis (unreported): older age, shorter insertion time, longer withdrawal time, and WE; but not WI (when compared with AI), are significant and independent factors associated with a higher rADR. CONCLUSION: WE, compared with AI, was significantly associated with a higher rADR, adjusting for other confounding factors. The improved quality adds value to WE colonoscopy, and justifies the call to incorporate WE into colorectal cancer prevention programs. Since WE = WI + suction of water during insertion, the higher rADR of WE provides incentive for transitioning from WI to WE. The impact of WE on right colon interval cancers warrants further assessment.