Adenoma Detection Rate (ADR) and Mean Adenoma Per Patient (MAP) With the Perfectly Clean Colon: Water Exchange (WE) vs Air Insufflation (AI) Colonoscopy

Introduction: Water exchange colonoscopy (WE) significantly reduces colonoscopy discomfort, amount of sedation medications, and has been reported to increase the adenoma detection rate (ADR) proximal to the splenic flexure when compared to gas (air or carbon dioxide) insufflation (AI). The explanation for the latter finding is a possible added cleaning effect by the infused water on the right colon, leading to uncovering of small adenomas. The question remains as to why WE is superior in ADR than AI. We sought to answer this question by analyzing ADR and Mean Adenoma per Patient (MAP) differences using WE vs AI when the colon is perfectly clean (Boston Bowel Preparation Score (BBPS) = 9). Methods: We conducted a retrospective data analysis of patients who underwent outpatient colonoscopies by one colonoscopist experienced in WE between November 2011 and December 2019 at an academic medical center. Patients were prospectively assigned to AI or WE using quasi-randomization. Those undergoing colonoscopy for screening or surveillance with a perfect BBPS score of 9 were included in this study. Pathology reports were reviewed and extracted. We calculated ADR and MAP and believed the latter to be a better marker for answering our question of interest. Statistical significance was calculated using Kruskal-Wallis rank sum test or Fisher’s Exact Test for Count Data. Results: As seen in Table 1, 659 of the 1,825 (36.1%) adults had a perfect bowel preparatiop (BBPS = 9). There were 352 (53.1%) patients assigned to the AI method and 311 (46.9%) to the WE method. The overall MAP was similar for both groups (0.75 and 0.74 for AI and WE, respectively, P = 0.49), as was the ADR. Analysis per anatomical location yielded similar results except for the rectosigmoid, where it was in favor of WE (0.26) vs AI (0.006, P = 0.034). There was no statistically significant difference in ADR in the proximal colon or the size of polyps found using the 2 methods, though there was a trend in favor of diminutive polyps found by WE (66.8%) vs AI (59%, P = 0.08). Conclusion: In this large cohort of patients, we were unable to find a significant difference in the ADR between the WE and AI methods when the colon was perfectly clean (BBPS = 9). Likewise, the MAP, a better reflection of the adenomas detected, was similar for both groups except for in the rectosigmoid colon. There was a trend towards finding more diminutive adenomas with the WE method. Table 1. - Demographic Information of Patients WITH a Boston Bowel Preparation Score of 9 Who Underwent Colonoscopies Using Air Insufflation (AI) and Water Exchange (WE) as Well as Mean Adenoma per Patient (MAP) by Anatomical Location and Total Number of Polyps by Size Found in the 2 Groups Total patients undergoing colonoscopy n = 1825 Total patients with BPPS of 9 (n = 659, 36.1%) AI (n = 351) WE (n = 308) P-value* Age, years, median (IQR) 60 (52, 65) 60 (53, 66) 0.3661 Sex, n (%) 0.0862 Female 194 (55.3%) 149 (48.4%) Male 157 (44.7%) 159 (51.6%) MAP** 0.749 0.736 0.4871 Anatomical Location MAP (AI) MAP (WE) P-value*1 Appendix 0.003 0 0.349 Ileocecal valve 0.003 0.006 0.488 Cecum 0.074 0.104 0.277 Ascending 0.208 0.179 0.587 Hepatic flexure 0.060 0.084 0.533 Transverse 0.174 0.146 0.760 Splenic flexure 0.014 0.016 0.835 Descending 0.066 0.065 0.790 Sigmoid 0.083 0.065 0.608 Rectosigmoid 0.006 0.26 0.034 Rectum 0.037 0.036 0.589 ADR in the proximal colon*** 0.353 0.347 0.935 Total number of polyps (%) AI (n = 256) WE (n = 226) P-value*2 Length Diminutive (≤5 mm) 151 (59.0%) 151 (66.8%) 0.089 Small (6-9 mm) 65 (25.4%) 48 (21.2%) 0.332 Advanced (≥10 mm) 40 (15.6%) 27 (11.9%) 0.291 *Statistically significant P <0.05 **mean number of adenomas per patient ***adenoma detection rate (ADR) in the proximal colon (appendix, ileocecal valve, cecum, ascending colon, hepatic flexure, and transverse colon) 1Kruskal-Wallis rank sum test 2Fisher’s Exact Test for Count Data.