Quality Indicators of Endoscopists for Both Index and Surveillance Colonoscopy are Associated With Risk of Metachronous Colorectal Neoplasia

INTRODUCTION: An association between higher adenoma detection rate (ADR) at index screening colonoscopy and lower risk of metachronous advanced neoplasia (AN, defined as colorectal cancer [CRC] or advanced adenoma [AA]) has been reported. However, the relationship between ADR at both index and surveillance colonoscopy and subsequent AN is unknown. We examined the association between ADR and withdrawal time (WT) at index and surveillance colonoscopy and risk of metachronous AN at surveillance colonoscopy. METHODS: We used GIQuIC, a repository of colonoscopies across the United States. Each patient has a unique ID at a participating site. Endoscopist national provider identifiers are associated with each examination. We included patients with 2 colonoscopies at least 3 years apart (index and surveillance) between 2011 and 2022 and calculated the ADR and average WT for the endoscopist performing the index and surveillance colonoscopies, respectively. We built a multivariable logistic regression model with metachronous AN as the outcome and ADR and WT as independent variables, controlling for patient age, sex, and race. RESULTS: We included 768,274 patients and 3,425 endoscopists. Mean patient age was 61 years and 48% were men; 66% were White, and 3% were Hispanic. Indication for index colonoscopy were screening (43.4%), surveillance (39.0%), and diagnostic (17.6%). ADR quartiles were ≤29.7%, >29.7%-37.2%, >37.2%-45.0%, and >45%. WT quartiles were ≤7.1 minutes, >7.1-8.2 minutes, >8.2-9.7 minutes, and >9.7 minutes. AN detection was lowest when low ADR endoscopists performed both index and surveillance examinations (5.4%, Table 1) and high ADR index examinations were followed by low ADR surveillance examinations (4.0%). Compared with low ADR endoscopists for both index and surveillance examinations, advanced neoplasia detection was significantly higher when both examinations performed by a high ADR endoscopist (AA 7.4%; OR [odds ratio] for AN 1.10 [1.05-1.16]) or low ADR index examinations were followed by high ADR surveillance examinations (AA 13.3%; OR for AN 1.448 [1.37-1.51]). Compared with short WT endoscopists for both examinations (AA 7.2%; CRC 0.3%), AN detection was higher when both examinations were performed by a long WT endoscopist or short WT index examinations were followed by long WT surveillance examinations (AA 7.0% P = 0.53 and 9.9%, P < 0.001) but similar CRC detection of 0.2% and 0.2% ( P 0.14). Other factors associated with finding of metachronous AN were older age (76 years and older vs 45-55 years OR 1.64; 95% CI 1.48, 1.82), male sex (Male vs female OR 1.15; 95% CI 1.10-1.19), White race compared with non-White (OR 1.10; 95% CI 1.06, 1.14), 7-10 years between examinations compared with 3-5 years between examinations (OR 1.24; 95% CI 1.11, 1.37), indication of surveillance vs screening for the index examination (OR 1.1.7; 95% CI 1.13, 1.22), AA or sessile serrated lesion finding on the index examination (OR 2.08; 95% CI 1.97, 21.9 and OR 1.23; 95% CI 1.16, 1.30 respectively). DISCUSSION: Our findings show endoscopist ADR and WT for both index and surveillance colonoscopy are associated with risk of metachronous neoplasia. Future studies on metachronous neoplasia should include both sets of quality indicators.