Evaluation of a risk index for advanced proximal neoplasia of the colon

Background A clinical risk index that uses distal colorectal findings at flexible sigmoidoscopy (FS) in conjunction with easily determined risk factors for advanced proximal neoplasia (APN) may be useful for tailoring or prioritizing screening with colonoscopy. Objective To conduct an external evaluation of a previously published risk index in a large, well-characterized cohort. Design Cross-sectional. Setting Teaching hospital and colorectal cancer screening center. Patients A total of 5139 asymptomatic persons aged 50 to 74 (54.9% women) with a mean age (± SD) of 58.3 (± 6.2) years. Interventions Between 2003 and 2011, all participants underwent a complete screening colonoscopy and removal of all polyps. Main Outcome Measurements Participants were classified as low, intermediate, or high risk for APN, based on their composite risk index scores. The concordance or c-statistic was used to measure discriminating ability of the risk index. Results A total of 167 persons (3.2%) had APN. The prevalence of those with APN among low-, intermediate-, and high-risk categories was 2.1%, 2.9%, and 6.5%, respectively. High-risk individuals were 3.2 times more likely to have APN compared with those in the low-risk category. The index did not discriminate well between those in the low- and intermediate-risk categories. The c-statistic for the overall index was 0.62 (95% confidence interval, 0.58-0.66). Limitations Distal colorectal findings were derived from colonoscopies and not FS itself. Conclusion The risk index discriminated between those at low risk and those at high risk, but it had limited ability to discriminate between low- and intermediate-risk categories for prevalent APN. Information on other risk factors may be needed to tailor, or prioritize, access to screening colonoscopy.