Cancer medicine
Core Self-Reported Variables for Next-Generation Colorectal Cancer Risk Prediction: An International Expert Consensus Study
By Dubois H, Torbjörnsson E, Norrbom J and 5 others.
Abstract
Background
Colorectal cancer remains a major global health burden, and although population-based screening is beneficial, current programs have notable limitations. Risk-stratified screening has been proposed, yet existing prediction models vary in performance and in the inclusion of predictors. Data-adaptive modeling (machine learning) may strengthen risk prediction but depends on high-quality input data. Expert consensus methods offer a structured way to identify clinically meaningful variables for future tools.
Objective
To establish expert consensus on the selection of relevant variables and refinement of self-reported questionnaire items for next-generation colorectal cancer risk-prediction tools.