SAMURAI-Fracture protocol: cluster-randomised cross-over trial of AI fracture detection in UK emergency departments and minor injury units

A multicentre cluster-randomised cross-over trial will evaluate whether an AI-assisted fracture detection tool (RBfracture) reduces inappropriate healthcare contacts after plain-film suspected-fracture imaging. Across 4 UK Trusts, alternating monthly 'AI on' vs 'AI off' periods…

• Trial design: prospective, cluster cross-over with sites randomised to start with 'AI on' or 'AI off' and then alternate monthly for a 6-month period across 4 UK EDs/MIUs.
• The composite primary endpoint is incidence of 'inappropriate healthcare contacts'—inappropriate referrals to fracture clinic, repeated hospital attendances, and subsequent follow-up communications regarding missed fractures.
• Secondary outcomes include patient-reported outcomes, clinician surveys, health economic evaluation (cost-effectiveness, budget impact), and diagnostic performance of the CE-approved RBfracture algorithm.

Automated summary

RadPigeon summaries are original and for information only. They are not clinical advice.