Risk prediction models for inadequate bowel preparation in colonoscopy: a systematic review and meta-analysis

Area of Science:

• Gastroenterology
• Clinical Epidemiology
• Health Informatics

Background:

• Numerous risk prediction models for inadequate bowel preparation (IBP) exist for colonoscopy.
• Concerns remain regarding the methodological quality and clinical applicability of these IBP models.

Purpose of the Study:

• To systematically review and meta-analyze the performance, methodological rigor, and applicability of current IBP prediction models.
• To assess the diagnostic accuracy and identify key predictors within existing IBP models.

Main Methods:

• Comprehensive database search (PubMed, Web of Science, Embase, Cochrane Library) up to December 2025.
• Data extraction using CHARMS checklist; risk of bias and applicability assessed with PROBAST.
• Random-effects meta-analysis of area under the curve (AUC) for validated models; publication bias evaluation.

Main Results:

• Nineteen IBP prediction models were included; most used logistic regression.
• Diabetes and constipation were common predictors; behavioral factors were underrepresented.
• Pooled AUC for 17 validated models was 0.73 (95% CI: 0.70-0.77), indicating acceptable discrimination, despite high risk of bias in all studies.

Conclusions:

• Existing IBP prediction models offer moderate discriminatory ability but are limited by significant methodological bias.
• Future IBP model development requires enhanced analytical rigor and inclusion of behavioral factors.
• Improved models can better identify high-risk patients, potentially enhancing colonoscopy success rates.