AI skin-lesion assessment in Bradford: early reports on capacity and biopsy use

The reported service change

According to BBC News, St Luke’s Hospital in Bradford introduced the DERM AI system in April to assess suspicious skin lesions. BBC News reports that every image is also checked by a clinician. The reported arrangement therefore involves both AI assessment and human review, rather than an AI-only pathway.

For surgeons interested in technology and innovation, the account raises questions about how an AI-supported assessment process may affect biopsy workload, specialist referrals and service capacity. It also provides a starting point for discussing clinical oversight, while leaving important details about the service’s operation and evaluation unclear.

Capacity per session

BBC News reports that the service was seeing 32 patients per session following introduction of the system, compared with 24 previously. These are session-level figures: they describe reported activity, but do not by themselves establish why that activity changed.

It is not yet clear from the reported information whether session length, staffing, patient mix or other aspects of the pathway were comparable across the two periods. Without that context, the figures cannot establish how much of the reported increase was attributable to AI, rather than to other changes in the service.

Biopsy use

According to the trust, as reported by BBC News, the number of unnecessary biopsies has fallen by about 10%. This is the trust’s reported service finding, rather than evidence that AI alone produced the reduction.

The meaning of that figure requires care. It is not yet clear from the reported information how “unnecessary” was defined, what comparison period was used, or which patients and procedures were included. The available account also leaves unclear the underlying biopsy counts and whether the percentage describes a change in the number of procedures or a change relative to patient activity.

For surgical audiences, the distinction between biopsy numbers and biopsy rates is relevant to interpreting the claim. A reported reduction cannot, on its own, describe the overall effect on workload or establish whether patient outcomes changed.

Clinical oversight

The clinician check of every image is a central feature of the pathway described by BBC News. It is important not to read the reported capacity and biopsy figures as results from a system operating without human involvement.

However, it is not yet clear how clinicians and the AI system interact within the assessment process, how differing assessments are handled, or how responsibility for subsequent pathway decisions is organised. The BBC News report, as summarised here, does not provide enough detail to characterise those arrangements further.

Questions for further reporting

The Bradford experience reported by BBC News may be useful as an early account of service change, rather than a settled measure of effectiveness. Further detail could clarify the duration of the observation period, the consistency of the reported changes and their relationship to other aspects of care.

For surgeons examining similar innovations, relevant unanswered questions include whether specialist referral volumes changed, whether the reported capacity increase persisted, and how patient outcomes were assessed. Those questions remain separate from the two headline measures reported by BBC News. Neither increased session activity nor fewer reported unnecessary biopsies, taken alone, establishes the overall performance of the pathway.

This article is an overview for information only, and readers should consult the original sources and appropriate professional guidance.