Colonoscopy training with or without artificial intelligence for endoscopy residents

Colonoscopy Training With or Without Artificial Intelligence Among Endoscopy Residents: a Randomized Controlled Study

This study tests whether using an AI polyp-detection tool during colonoscopy training helps endoscopy trainees find more adenomas and miss fewer lesions.

Quick facts

• Phase: NA
• Study type: Interventional
• Enrollment: 10 (estimated)
• Sex: All
• Sponsor: University of Thessaly (other)
• Locations: 1 site (Larissa)
• Trial ID: NCT07420309 on ClinicalTrials.gov

What this trial studies

This is a randomized controlled, crossover study comparing conventional colonoscopy training to training assisted by AI polyp-detection software among endoscopy trainees who have at least one year of training remaining and no prior structured CADe exposure. Trainees are randomized to one approach, with primary outcome measured as adenoma miss rate (AMR) at the end of the first year and again after crossover. Secondary outcomes include polyp detection rate (PDR), adenoma detection rate (ADR), advanced ADR, polyp miss rate (PMR), withdrawal time, cecal intubation rate (CIR), and adverse events. Adult patients undergoing diagnostic, screening, or surveillance colonoscopy at the General University Hospital of Larissa are enrolled and, when possible, followed for interval cancer until their next colonoscopy or diagnosis.

Who should consider this trial

Good fit: Ideal candidates are endoscopy trainees with at least one year of training remaining and no prior structured exposure to CADe who perform adult diagnostic, screening, or surveillance colonoscopies and can give informed consent.

Not a fit: Trainees in their final year or those with prior systematic CADe exposure, and patients with contraindications to biopsy, pregnant or lactating women, known polyposis syndromes, or a history of inflammatory bowel disease are unlikely to benefit or are excluded.

Why it matters

Potential benefit: If successful, the approach could help trainees detect more adenomas and reduce missed lesions, potentially improving long-term colorectal cancer prevention.

How similar studies have performed: Several trials in practicing endoscopists have shown that CADe systems can increase ADR and polyp detection, but randomized crossover data specifically focused on trainee education is more limited.

Eligibility criteria

Inclusion Criteria:

* Trial participants: Endoscopy trainees at any time of their training, with at least one year of training ahead without previous structured exposure to CADe during their training.

Recruited patient population: Adult patients undergoing diagnostic, screening, or surveillance colonoscopy will be included.

1. Provision of signed and dated informed consent form understand the risks and benefits of the study
2. Aged ≥18 years old
3. Able to read and write in the provided language.
4. Participants capable to provide written informed consent and comply with the trial protocol.

Exclusion Criteria:

* Trial participants: Endoscopy trainees after their basic training, during the last year of training or with systematic exposure to CADe.

Recruited patient population:

1. Patients with a known contraindication for biopsy, bowel obstruction or perforation,
2. Pregnant or lactating females
3. Known polyposis syndromes, or who had a history of inflammatory bowel disease (IBD), CRC, or colorectal surgery.
4. Any patient whose cecum is not reached and who were highly suspected of having polyposis syndromes, IBD, intestinal tuberculosis, or CRC.
5. Incomplete bowel preparation (segmental Boston Bowel Preparation Score \<2, or total \<6)

Where this trial is running

Larissa

• General University Hospital of Larissa — Larissa, Greece (RECRUITING)

Study contacts

• Study coordinator: Apostolis Papaefthymiou

Conditions: Colonoscopy Training