Tu1455 Comparison of Physician, Endoscopic and Patient Characteristics Between High and Low Polyp Detectors

Abstract

Prevention of colorectal cancer depends on the timely detection and removal of adenomatous colon polyps. While the adenoma detection rate (ADR) is a frequently used quality measure for colonoscopy, the overall polypectomy rate (PR) has recently been suggested as another quality indicator because it is easy to measure and has good correlation with the ADR. However, the PR varies widely among endoscopists, and the factors leading to this variation are not entirely known. To compare endoscopic, patient and physician characteristics between endoscopists with a high PR and a low PR. A retrospective chart review was completed for all patients who underwent colonoscopy between January and December 2009. Colonoscopy performed for anemia, bleeding and/or inflammatory bowel disease was excluded. The ADR was defined as the detection of at least one adenoma per colonoscopy, and the PR was defined as the proportion of procedures in which at least one polyp was removed. Based on recent studies, a 40% PR was determined to correspond to the benchmark ADR of 25%. ADR and PR for individual endoscopists were calculated, and endoscopists were categorized into two groups: a) Low PR ≤ 40% b) high PR ≥ 40%. Demographic, endoscopic and physician factors were compared between the two groups using chi-square, t-tests, and generalized linear models to control for clustered data. Data are presented as mean ± SD and proportions, in the format: (low PR vs high PR). Data were collected from 4574 colonoscopies performed by 19 endoscopists. Physician factors did not differ significantly between low PR and high PR groups: experience (17±10 vs 14±7 yrs, P=0.46), colonoscopy volume (211±107 vs 273±104, P=0.21) and endoscopists' gender (70% M vs 89% M, P=0.58). The patient sample did not differ in age (62 ± 12 vs 62 ± 12 yrs, P = 0.434) or BMI (29 ± 23 vs 28 ± 10, P=0.41), but included more males (48% vs 53%, P<0.001) in the high PR group. The proportion of patients with good tolerance (80% vs 83%, P<0.001) and good quality of bowel preparation (92% vs 88%, P <0.001) were significantly different between two groups, but there was no significant difference in incomplete colonoscopies (6% vs 7%, P =0.11) or the largest size of resected polyps (6±13 vs 6±7mms, P=0.53). The ADR (21% vs 30%, P<0.001) and withdrawal time (9 ± 6 vs 13 ± 8 mins, P<0.001) were significantly lower in the low PR group compared to high PR group. Data from this large patient sample showed that higher polypectomy rates among endoscopists were associated with better patient tolerance, male patient population and longer withdrawal times. Further prospective studies are needed to confirm our findings and derive the factors needed to improve adenoma and polyp detection.