A predictive model based on quantitative fecal immunochemical test can stratify the risk of CRC in an organized screening program.

Abstract

Objective: A significant number of post-faecal immunochemical test (FIT) colonoscopies in European organized colorectal cancer (CRC) screening programs are performed beyond the recommended 31-day threshold due to overburdened colonoscopy services. We aimed to develop a simple predictive model to stratify CRC risk of FIT+ patients.

Design: In a cohort of screenees undergoing colonoscopy following a positive ( ≥ 20μg Hb/g. feces) OC-sensor FIT result between 2004 and 2019, we derived and validated logistic regression-based models including variables independently associated with CRC and advanced neoplasms. Odds ratios (ORs) and 95% confidence intervals (CIs) were reported.

Results: Overall, 40,276 patients (46% female, mean age 66+4 years) undergoing post-FIT colonoscopy were included. Variables independently associated with CRC were age ≥70 years (OR 1.20, 95%CI 1.03-1.40), male sex (OR 1.23, 95%CI 1.11-1.37), f-Hb level (50 to 199 μg/g.: OR 2.84, 95%CI 2.47-3.27; ≥200 μg/g: OR 6.91, 95%CI 5.99-7.98), and first round of FIT (OR 1.53, 95%CI 1.35-1.73). The discriminative ability of the model was good (AUROC 0.75, 95%CI 0.73-0.77) in the validation cohort. Applying the model would lead to over two thirds decrease in delayed CRC diagnoses, considering various scenarios of timely colonoscopy scheduling after FIT+.

Conclusions: We derived and validated a predictive model for risk stratification of patients with positive FIT in a large CRC screening cohort. Applying our model in screening practice would allow policy makers to effectively prioritize FIT+ individuals based on the risk of CRC, allowing to substantially reduce the rate of delayed CRC diagnosis.