Nationwide Study of Artificial Intelligence in Adenoma Detection for Colonoscopy

This is a case-control study comparing adenoma detection rate (ADR) in hospitals (and individual colonoscopists), before, during and after use with an artificial intelligence unit called GI Genius™ (GIG). GIG is a Computer-assisted detection (CADe) module that assists the human colonoscopist in real-time, by detecting and marking out polyps during colonoscopy. It has been shown to be effective in expert colonoscopists, but the effect in non-expert, general, colonoscopists is not known.

The investigator wish to deploy GIG into colonoscopy through the UK using a step-wedge design. Sites will be randomly allocated a start date for GIG deployment, collecting data for four months prior to this. In this way, all sites will have the active intervention and will provide their own case-control data. (4 months collection prior to activating GIG, 4 months with GIG, 4 months afterwards without GIG)

The study will concentrate on non-expert colonoscopists, to determine whether GIG can increase ADR. Patients will undergo the same colonoscopy that they would have had in any case, with no additional trial visits or interventions. There will be no alteration to the usual care pathway from the patient's perspective.

If the investigator can prove GIG increases ADR in this way, it will provide support to roll out this technology routinely to improve the quality of colonoscopy nationwide.