A decision support model for colorectal cancer screening.

Area of Science:

• Decision analysis
• Public health
• Medical informatics

Background:

• Current European colorectal cancer (CRC) screening programs predominantly use age-based, one-size-fits-all strategies.
• These programs exhibit minor variations across European nations.
• There is a need for more personalized approaches to CRC screening.

Purpose of the Study:

• To develop a decision analysis-based framework for personalized colorectal cancer (CRC) screening.
• To support decisions on screening methods and the necessity of colonoscopy for individuals.
• To evaluate and potentially redesign existing national CRC screening programs.

Main Methods:

• An influence diagram model was developed to characterize CRC risk based on various factors.
• Decision criteria included comfort, costs, complications, and information (assessed via information-theoretic measures).
• A multi-attribute utility model integrated these criteria, with optimal screening policies determined by maximizing expected utility.

Main Results:

• The model enables personalized CRC screening recommendations based on individual features beyond age.
• It provides a tool for assessing current age-based programs and designing new risk-based strategies.
• The framework suggests transitioning from age-based to personalized, risk-dependent screening strategies.

Conclusions:

• This research introduces a novel framework for personalized colorectal cancer (CRC) screening.
• The proposed approach offers improvements over existing age-based screening strategies.
• The study highlights opportunities for redesigning and optimizing CRC screening programs.