Same evidence but different recommendations – Dilemma in computer-aided detection in colonoscopy

A recent Position Statement from the European Society of Gastrointestinal Endoscopy (ESGE) [1] highlights the paradoxes of modern guideline development. In March 2025, ESGE [1], the British Medical Journal (BMJ) [2], and the American Gastroenterological Association (AGA) [3] released guidelines on using computer-aided detection (CADe) during colonoscopy, all based on the same evidence from the EU-funded OperA project [4]. Despite this shared foundation, ESGE weakly recommended CADe, BMJ weakly recommended against it, and AGA made no recommendation.

This confusing discord was not due to differences in the evidence base, but to differing interpretations of it, driven by uncertainties of the data, different preferences and values, and selection of the panel members, all of which are considered core pillars when making trustworthy GRADE-based guidelines [5]. All three guidelines used a microsimulation model showing that CADe would prevent 111 colorectal cancers per 100 000 screened (0.1% absolute reduction), require 6366 additional colonoscopies (6.4% increase), and cause no change in harms [4]. Yet they reached contradictory conclusions.

Key uncertainties that shaped their judgments included limitations of microsimulation modelling, the unblinded nature of colonoscopy trials, variable CADe performance among endoscopists, opaque regulatory processes for CADe, and unclear patient values and preferences [2]. When benefits are marginal and evidence is imperfect, how a panel interprets these uncertainties plays a crucial role in the final recommendation. Panel composition also played a role: ESGE included only clinicians, BMJ included patient partners and methodologists, and AGA had clinicians and methodologists but not patients. These differences suggest that guideline outcomes are influenced not only by evidence but by who is interpreting it – and the values they bring to the table.

For clinicians, especially when recommendations are weak or conflicting, understanding this process is essential. In such cases, shared decision-making – openly discussing potential benefits, harms, and burdens with patients – is crucial to delivering patient-centered care.

Publikationsverlauf

Artikel online veröffentlicht:
24. September 2025

• References

• 1 Bretthauer M, Ahmed J, Antonelli G. et al. Use of computer-assisted detection (CADe) colonoscopy in colorectal cancer screening and surveillance: European Society of Gastrointestinal Endoscopy (ESGE) Position Statement. Endoscopy 2025; 57: 667-673
• 2 Foroutan F, Vandvik PO, Helsingen LM. et al. Computer aided detection and diagnosis of polyps in adult patients undergoing colonoscopy: a living clinical practice guideline. BMJ 2025; 388: e082656
• 3 Sultan S, Shung DL, Kolb JM. et al. AGA living clinical practice guideline on computer-aided detection-assisted colonoscopy. Gastroenterology 2025; 168: 691-700
• 4 Halvorsen N, Hassan C, Correale L. et al. Benefits, burden, and harms of computer aided polyp detection with artificial intelligence in colorectal cancer screening: microsimulation modelling study. BMJ Med 2025; 4: e001446
• 5 Guyatt GH, Oxman AD, Vist GE. et al. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ 2008; 336: 924-926