Predicting bowel prep quality
Research type
Research Study
Full title
Predicting Bowel Preparation Quality Before Colonoscopy: A Prospective Observational Study
IRAS ID
361358
Contact name
Laurence Lovat
Contact email
Sponsor organisation
University College London Hospital NHS Foundation Trust
Clinicaltrials.gov Identifier
Z6364106/2025/08/28, UCL Data Protection Office; 182923, EDGE
Duration of Study in the UK
1 years, 11 months, 30 days
Research summary
Colorectal cancer is a common and serious disease, but it can be prevented or detected early through colonoscopy, a procedure that examines the bowel. For a colonoscopy to work well, the bowel must be thoroughly cleaned beforehand. However, many people do not achieve good bowel preparation, which can reduce the accuracy of the test, cause delays, or lead to the procedure being cancelled or repeated. This can be distressing for patients and puts extra strain on NHS services.
This study aims to develop a way to predict how well a person’s bowel will be prepared for a colonoscopy, using modern data science tools such as machine learning. Participants will be initially recruited from University College London Hospital (UCLH) and other places through clinical pathways or online promotion. With their consent, we will collect information such as medical history, previous colonoscopy outcomes and stool images (taken by participants using a smartphone). This information will help us build and test computer models that can predict the likely quality of bowel preparation before the procedure takes place.
In the future, this could allow healthcare teams to intervene earlier if poor preparation is predicted, by changing the timing of the colonoscopy or adjusting the bowel preparation dosing. This may help avoid procedure cancellations, improve patient outcomes, and make better use of NHS resources.
REC name
London - Riverside Research Ethics Committee
REC reference
26/PR/0099
Date of REC Opinion
26 Feb 2026
REC opinion
Further Information Favourable Opinion