NAIAD

  • Research type

    Research Study

  • Full title

    A Nationwide study of Artificial Intelligence in Adenoma Detection during colonoscopy

  • IRAS ID

    292323

  • Contact name

    Bu'Hussain Hayee

  • Contact email

    b.hayee@nhs.net

  • Sponsor organisation

    King's College Hospital NHS Foundation Trust

  • Clinicaltrials.gov Identifier

    NCT05870332

  • Duration of Study in the UK

    2 years, 0 months, 0 days

  • Research summary

    The early detection (screening) of bowel cancer represents a significant health challenge. Colonoscopy, a flexible camera examination of the large bowel (the gold standard test), requires training to perform comfortably and to a high enough standard to detect adenomas (polyps; pre-cancerous growths). The demand for high quality colonoscopy is rising as the demand for bowel cancer screening rises. 

    Adenoma detection rate (ADR) is the % number of colonoscopies performed where at least one adenoma is found. It is a widely accepted quality measure units worldwide. We do not have a good estimate of the average ADR across the UK, but it is likely to be <20%. Expert colonoscopists in the national bowel cancer screening programme (<2% of the total workforce in the UK) typically have ADR >35%, but >10x more colonoscopies in 'at risk' patients are performed outside of this system, in the "two-week-wait" referral pathway. We urgently need to evaluate technologies that can produce rapid and significant improvements in performance, across as wide a range of colonoscopists as possible. 

    The most intriguing of these is artificial intelligence (or computer assisted detection: CADe). This is a computer system in a box that is plugged into existing colonoscopy equipment and highlights polyps on the video screen. The GI Genius (GIG) system from Medtronic is advantageous as it is compatible with any colonoscopy equipment and can be used straight away. 

    We know GIG improves ADR in 'expert' colonoscopists with ADRs that are already well above average, but we do not yet know whether this applies to colonoscopists with varying levels of performance. We will therefore study the effect of implementing GIG into 'real-world' practice. We will study performance before, during and after GIG to fully understand whether and how it works.

  • REC name

    London - Hampstead Research Ethics Committee

  • REC reference

    23/PR/0645

  • Date of REC Opinion

    27 Jun 2023

  • REC opinion

    Favourable Opinion