Predictors of guideline concordance for surveillance colonoscopy recommendations in patients at a safety-net health system

Purpose: Appropriate surveillance intervals for colorectal cancer (CRC) screening is one of the Centers for Medicare and Medicaid Services 2014 physician quality reporting system measures. Appropriateness of surveillance intervals will continue to be monitored closely, particularly as reimbursements become tied to quality measures. Aims: Quantify and identify predictors for guideline-concordant surveillance recommendations after adenoma polypectomy. Methods: We conducted a retrospective cohort study of patients who had colonoscopy with polypectomy at a safety-net health system between June 2011 and December 2013. Surveillance recommendations shorter and longer than guideline recommendations were defined as potential overuse and underuse. We used multivariate logistic regression to identify correlates of guideline-concordant surveillance recommendations, overuse, and underuse. Results: Among 1,822 patients with polypectomy, 1,329 had ≥1 adenoma. Surveillance interval recommendations were guideline-concordant in 1,410 (77.4 %) patients, potential overuse in 263 (14.4 %), potential underuse in 85 (4.7 %), and missing in 64 (3.5 %) patients. Predictors of guideline-concordant recommendations in multivariate analyses included age >65 years (OR 1.36, 95 % CI 1.02–1.80), incomplete resection (OR 3.58, 95 % CI 1.41–9.09), and good/excellent prep quality (OR 2.22, 95 % CI 1.72–2.86). Underuse recommendations were more likely in patients with ≥3 adenomas; overuse recommendations were more likely in patients with high-grade dysplasia or fair prep quality and less likely in those with piecemeal resection, ≥3 adenomas, age >65, or Hispanic ethnicity. Conclusions: Surveillance recommendations are not concordant with guidelines in one of four cases. Interventions to improve prep quality and guideline concordance of surveillance recommendations can improve cost-effectiveness of CRC screening.