Abstract 2317: Development and validation of a colorectal cancer prediction model: A nationwide cohort-based study

Abstract

Abstract Background: Colorectal cancer (CRC) ranks as the third most prevalent cancer globally and is a major cause of cancer-related mortality. Early diagnosis is critical to increase survival rates. While CRC screening has shown significant benefit, adherence remains low, and there is a need for better tools to identify high-risk patients. Risk prediction models were demonstrated to identify such patients. Aim: To establish an individualized risk prediction model for CRC diagnosis based on Electronic Health Records (EHR). Methods: This is a retrospective cohort study utilizing EHR data of Clalit Health Services (CHS) members aged 50-74 that were eligible for CRC screening, from January 2013 to January 2019. The model was trained to predict CRC diagnosis within two years using approximately 20,000 EHR features including socio-demographic information, laboratory results and medical history. Model performance, as a complementary screening method, was evaluated. Results: The study included 2679 subjects with CRC diagnosis and 1,133,713 subjects without CRC diagnosis. The model was trained on subjects from 2013-2017, and performance was validated on subjects from 2019 and a cohort of subjects that underwent fecal occult blood test (FOBT). Incidence values of CRC among subjects in the top 1% risk scores were higher than baseline (2.3% vs. 0.3%; lift 8.38; P-value <0.001). Characteristics of subjects by risk scores percentiles are presented in Table 1. Cumulative event probabilities increased with higher model scores, indicating a correlation between predicted and actual risk of CRC diagnosis. Model-based risk stratification among subjects with a positive FOBT, identified subjects with more than twice the risk for CRC compared to FOBT alone. Conclusions: We developed an individualized risk prediction model for CRC that can be utilized as a complementary decision support tool for healthcare providers to precisely identify patients at high risk for CRC and refer them to confirmatory testing. Citation Format: Samah Hayek. Development and validation of a colorectal cancer prediction model: A nationwide cohort-based study [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 2317.