ECCO Topical Review on Predictive Models on Inflammatory Bowel Disease Disease Course and Treatment Response

Area of Science:

• Gastroenterology
• Clinical Epidemiology
• Predictive Modeling

Background:

• Inflammatory bowel disease (IBD) presents a significant clinical challenge due to unpredictable disease progression and varied treatment responses.
• Existing predictive models for IBD lack consistent validation and methodological rigor, limiting their clinical utility.
• This review addresses the current state of predictive models for IBD, evaluating their relevance and identifying implementation barriers.

Purpose of the Study:

• To critically review existing predictive models for inflammatory bowel disease (IBD) disease course and treatment response.
• To assess the relevance and limitations of current predictive models in clinical practice.
• To identify and discuss barriers hindering the clinical implementation of these predictive models.

Main Methods:

• An expert panel from the European Crohn's and Colitis Organisation, comprising gastroenterologists, surgeons, and clinical epidemiologists, conducted a review.
• Delphi methodology was employed to achieve consensus and develop practice position statements.
• Agreement was defined as at least 80% of participants concurring on a recommendation.

Main Results:

• Fourteen practice positions and two perspective points were established, detailing factors in predictive models and implementation barriers.
• Methodological standards for developing and validating predictive models were defined, alongside identified challenges.
• Perspectives on incorporating biomarkers and flexible study designs were outlined.

Conclusions:

• This review provides practice recommendations and strategic guidance for the development and clinical implementation of future predictive models for IBD.
• The findings aim to enhance the utility and reliability of predictive tools for managing IBD patients.
• Guidance is offered on overcoming methodological and practical hurdles to integrate predictive models into routine clinical care.