Development and validation of a gut motility based model for predicting bowel preparation quality

• Published in: Scientific reports Vol. 15; no. 1; pp. 28265 - 9
• Main Authors: Wu, Hongjiao, Wu, Rui, Zhang, Yunyun, Lu, Xinyi, Zhao, Wei, Xu, Biyun, Liu, Jun, Zhang, Nina
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 02.08.2025; Nature Publishing Group; Nature Portfolio
• Subjects: 692/4020/1394; 692/4020/1503; 692/499; 692/699/1503/2753/1400; Adult; Aged; Bowel preparation; Cathartics - administration & dosage; Colon; Colonoscopy; Colonoscopy - methods; Constipation; Diabetes; Female; Gastric motility; Gastrointestinal motility; Gastrointestinal Motility - physiology; Hospitals; Humanities and Social Sciences; Humans; Intestine; Male; Middle Aged; Missing data; Motility; multidisciplinary; Multivariate analysis; Nomograms; Parkinson's disease; Patients; Polyethylene glycol; Polyethylene Glycols; Prediction models; Predictive learning models; Regression analysis; Science; Science (multidisciplinary); Predictive learning models; Gastrointestinal motility; Bowel preparation; Colonoscopy
• ISSN: 2045-2322; 2045-2322
• DOI: 10.1038/s41598-025-13739-0

Inadequate bowel preparation negatively impacts the quality of colonoscopy, potentially resulting in missed lesions and the need for repeat procedures. Although gut motility plays a key role in bowel cleansing, it is often neglected in risk prediction. This study developed and validated a predictive model and simplified scoring system that integrates gut motility parameters to identify individuals at risk of inadequate bowel preparation. A total of 1,165 patients from two hospitals were enrolled, with 815 forming the training set and 350 forming the external validation cohort. The overall rate of inadequate bowel preparation was 9.8%. Multivariate analysis revealed that altered bowel movement frequency, stool consistency based on the Bristol Stool Scale, low polyethylene glycol (PEG) volume, and delayed last bowel movement were significant predictors of poor preparation. These variables were incorporated into a user-friendly scoring system that demonstrated good discriminative ability, with area under the curve values of 0.778 and 0.774 in the training and validation cohorts, respectively. A cutoff score of 3.0 yielded a sensitivity of 75.0%, specificity of 66.9%, and a negative predictive value of 96.3% in external validation. This model provides a practical, motility-informed approach for risk stratification and personalized preparation regimens, offering potential to enhance the effectiveness and efficiency of colonoscopy in diverse clinical settings.