Mo1674 DIFFERENCES IN DETECTION OF POLYP PATHOLOGY BASED ON ENDOSCOPIST CHARACTERISTICS AND EXAMINATION OF THE ROLE OF WITHDRAWAL TIME IN EXPLAINING THESE DIFFERENCES

Abstract

Previous studies have demonstrated adenoma detection (AD) to be operator dependent. In this study we examined endoscopist characteristics not only as they relate to AD, but also to sessile serrated adenoma detection (SSPD), and advanced adenoma detection (AAD--defined as ≥1cm diam., having villous component or high-grade dysplasia). Further, we examined the role of withdrawal time (amount of time spent viewing as colonoscope is withdrawn) as a possible explanation of these differences. Outpatient screening colonoscopies performed on average risk patients by providers who have performed ≥100 colonoscopies in the Cleveland Clinic between January 2015-August 2017 were reviewed. Patient and colonoscopic characteristics along with various endoscopist characteristics were obtained. Descriptive statistics were performed to compute differences in SSPD, AD, AAD, and withdrawal times (WT) based on endoscopist characteristics. This was followed by multivariate (MV) analysis adjusting for various endoscopist characteristics, as well as various patient and colonoscopy based characteristics: age, gender, ethnicity, insurance, BMI, comorbidities (diabetes, cirrhosis, dementia, stroke, constipation, CAD, CHF), aspirin use, opiate use, quality of bowel preparation, location/timing (AM v. PM) of colonoscopy, and number of polyps seen. MV analysis was repeated with inclusion of WT as a variable. N=18,501 colonoscopies were performed by 122 providers. Baseline characteristics and differences in SSPD, AD, AAD, and WT based on various endoscopist characteristics, were obtained (Table 1). MV analysis excluding the effect of WT but adjusting to afore mentioned variables (Table 2a) and including the effect of WT (Table 2b) is presented. Most notable of these associations include: After controlling for WT, odds of SSPD compared to gastroenterologists (GI) has increased from 56% to 67% in general surgeons(GS) and from 53% to 61% in colorectal surgeons (CS). Whereas GS had 17% decreased chance of AD compared to GI, this difference has become insignificant after adjusting to WT. Odds of AD for CS improved by 14% after adjusting for WT. Female providers noted to have higher WT (table-1) have seen 5% decreased chance of SSP after adjusting to WT. Compared to academic physicians, the odds of SSPD among private physicians increased 4% after adjusting to WT. After adjusting for multiple known risk factors, our study revealed significant differences in AD, SSPD, and AAD based on endoscopist characteristics. Comparison of these characteristics before and after controlling for WT demonstrated that many of the differences polyp detection may be partially or fully related to WT. This study emphasizes the need to educate endoscopists about appropriate WT during colonoscopy—especially in providers with characteristics associated lower polyp detection.Table 2- Adjusted analysis of factors associated with sessile serrated polyp detection rate (SSP), adenoma detection rate (AD), advanced adenoma detection rate (AAD)View Large Image Figure ViewerDownload Hi-res image Download (PPT)