Abstract PR001: The use of risk prediction for patients with abnormal fecal testing (FIT) to prioritize patient navigation

Abstract

Abstract Context: “Precision delivery involves using an individual’s electronic health data to predict risk and personalize care to substantially improve value” (Parikh). Rates of completion of the follow-up colonoscopy following an abnormal FIT are low in community clinics. Tools like predictive analytics can be used for population management and more precise delivery of care to those who need it the most. Objective: We describe 3 studies use of predictive analytics after abnormal FIT testing to prioritize patient navigation to complete subsequent colonoscopies. Study design and analysis: The Strategies and Opportunities to Prevent Colorectal Cancer (STOP CRC) trial developed an initial risk prediction model identifying patients unlikely to complete a follow-up colonoscopy on their own. Kaiser Permanente Northwest (KPNW) then redeveloped the model for use with Navigation. The Predicting and Addressing Colonoscopy Non-adherence in Community Settings (PRECISE) study then redeveloped the model for use with Navigation in a large Federally Qualified Health Center (FQHC). Results: The STOP CRC model included 8 variables, 1596 patients had an abnormal FIT result, 34.8% had a recorded colonoscopy within 6 months. The model shows an adequate separation of patients across risk levels for non-adherence to follow-up colonoscopy (bootstrap-corrected C-statistic > 0.63). The KPNW model included 14 variables, and the performance of the model was adequate with a bootstrap corrected C-statistic of 0.65. The PRECISE model included 12 variables and a bootstrap corrected C-statistic of 0.58. Here we will describe each of the models and the differences in the application of the models. Conclusions: Risk prediction can be used to identify patients unlikely to complete colonoscopies, and as a population management tool to tailor interventions to specific populations. Tailoring interventions can save money, and increase patient satisfaction. Citation Format: Amanda F. Petrik. The use of risk prediction for patients with abnormal fecal testing (FIT) to prioritize patient navigation [abstract]. In: Proceedings of the AACR Special Conference on Colorectal Cancer; 2022 Oct 1-4; Portland, OR. Philadelphia (PA): AACR; Cancer Res 2022;82(23 Suppl_1):Abstract nr PR001.