Impact of Community Referral on Colonoscopy Quality Metrics in a Veterans Affairs Medical Center.

INTRODUCTION: The Veterans Access, Choice, and Accountability Act of 2014 expands the number of options veterans have to ensure timely access to high-quality care. There are minimal data currently available analyzing the impact and quality of colonoscopy metrics in veterans receiving procedures within the Department of Veterans' Affairs (VA) vs community settings. METHODS: All patients at our academic VA medical center who were referred to a community care colonoscopy (CCC) for positive fecal immunochemical testing, colorectal cancer screening, and adenoma surveillance from 2015 to 2018 were identified and matched for sex, age, and year of procedure to patients referred for a VA-based colonoscopy (VAC). Metrics measured included time to procedure measured in days, adenoma detection rate (ADR), advanced ADR (AADR), adenomas per colonoscopy, sessile serrated polyp detection rate, cecal intubation rate, bowel preparation quality, and compliance with guideline recommendations for surveillance. Patient comorbidities were also recorded. Variable associations with adenoma detection and compliance with surveillance guidelines were analyzed with univariate and multivariate logistic regression. RESULTS: In total, 235 veterans (mean age, 64.6 years, and 95.7% male) underwent a CCC and had an appropriately matched VAC. ADR in the community was 36.9% compared with 62.6% for the VAC group (P < 0.0001). The mean number of adenomas per procedure in the community was 0.77 compared with 1.83 per VAC (P < 0.0001). CCC AADR was 8.9% compared with 18.3% for VAC (P = 0.003). The cecal intubation rate for community colonoscopies was 90.6% compared with 95.3% for VA colonoscopies (P = 0.047). Community care compliance with surveillance guidelines was 74.9% compared with 93.3% for VA (P < 0.0001). This nonconformity was primarily due to recommending a shorter interval follow-up in the CCC group (15.3%) compared with the VAC group (5.5%) (P = 0.0012). The mean time to procedure was 58.4 days (±33.7) for CCC compared with 83.8 days (±38.6) for VAC (P < 0.0001). In multivariate regression, CCC was associated with lower ADR (odds ratio 0.39; 95% confidence interval, 0.20-0.63) and lower compliance with surveillance guidelines (odds ratio 0.21; 95% confidence interval, 0.09-0.45) (P < 0.0001 for both). DISCUSSION: Time to colonoscopy was significantly shorter for CCC compared with VAC. However, compared with VA colonoscopies, there was significantly lower ADR, AADR, and surveillance guideline compliance for services rendered by community providers. This impact on quality of care should be further studied to ensure that colonoscopy quality standards for veterans are not compromised by the process of care and site of care.