Risk-Based Colonoscopy Prioritization Model for Advanced Colorectal Neoplasia: Development and Validation Study.

Area of Science:

• Gastroenterology and Oncology
• Epidemiology
• Health Services Research

Background:

• Colorectal cancer (CRC) incidence is rising in Asia, yet validated prediction models for advanced colorectal neoplasia (ACN) are lacking in Thailand.
• Limited colonoscopy capacity necessitates effective precolonoscopy risk stratification strategies.

Purpose of the Study:

• To develop and validate a Thai-specific prediction model for ACN in asymptomatic adults aged 40 years and older.
• To utilize routine clinical variables for risk stratification to optimize colonoscopy resource allocation.

Main Methods:

• A retrospective cross-sectional study involving 1374 participants aged ≥40 years undergoing colonoscopy at Pattani Hospital, Thailand.
• Development of a multivariable fractional-polynomial interaction (MFPI) logistic regression model using seven clinical predictors.
• Internal validation assessed using AUC, calibration, Brier score, and decision-curve analysis (DCA).

Main Results:

• The study identified ACN in 13.9% of participants.
• The Thai-specific MFPI model demonstrated good performance with AUCs of 0.76 (derivation) and 0.67 (validation).
• A 12% risk cutoff identified one ACN for every three colonoscopies, reducing colonoscopy volume by 48.4% and showing consistent net benefit via DCA.

Conclusions:

• The developed Thai-specific MFPI model is well-calibrated and validated, offering clinically significant utility.
• This model can optimize colonoscopy allocation and enhance precision precolonoscopy risk stratification, particularly in resource-limited healthcare settings.