|| Checking for direct PDF access through Ovid
Tailoring the use of screening colonoscopy based on the risk of advanced proximal neoplasia (APN) has been advocated as a strategy for reducing demand and optimizing effectiveness. A 7-point index based on age, sex, and distal findings at sigmoidoscopy has been proposed that stratifies individuals into low, intermediate, and high-risk categories. The aim of this cross-sectional analysis was to determine the validity of this index, which was originally derived and validated among mostly whites, for black and Hispanic patients.Data, including age, sex, colonoscopic findings, and pathology, were collected retrospectively from 1,481 white, 1,329 black, and 689 Hispanic asymptomatic, average-risk patients undergoing screening colonoscopy between 2000 and 2005. Cumulative scores ranging from 0 to 7 were derived for each subject and categorized as low, intermediate, or high risk. Rates of APN were assessed for each risk category after stratification by race/ethnicity. Index performance was assessed using the C-statistic and compared across the three racial groups.Rates of APN among patients categorized as low, intermediate, or high risk increased from 1.0 to 2.8 to 3.7% for whites, 1.0 to 2.2 to 4.2% for blacks, and 0.6 to 1.9 to 3.7% for Hispanics. The index performed similarly for all three groups, but showed limited ability to discriminate low from intermediate-risk patients, with C-statistic values of 0.62 for whites, 0.63 for blacks, and 0.68 for Hispanics.A risk index based on age, sex, and distal endoscopic findings has limited ability to discriminate low from intermediate-risk white, black, and Hispanic patients for APN.