What happened
Skin Analytics implemented DERM, a UKCA class IIa AI device, at six NHS sites under phase IV of the AI Awards in Health and Social Care. The Department of Health and Social Care commissioned an evaluation of four of them, which Unity Insights summarized. In secondary care, DERM assessed 5,678 lesions from 4,159 referrals. The evaluators report pathway sensitivity for melanoma above 90% at every site and a secondary care sensitivity of 97.4%. A consultant dermatologist reviewed each case that DERM classified as low risk. This second read overturned 36% of potential discharges, and trust dermatologists then discharged 41% of those. Consultants found seven cancers among the 754 overturned cases, most of them at low risk. In a model of costs and benefits, the evaluators estimated that DERM saves 2,600 dermatologist appointments a year in an average trust. They estimated that it returns £1.7 for every £1 spent in secondary care. Their analysis of costs against quality of life was mostly inconclusive. Without direct baseline data, the evaluators could not fully separate DERM's effect from that of teledermatology.
Sources
-
Peer-reviewed study, by authors with no tie to the system. Context or contrary evidence.
Anderson ADG, Lo SN, Guitera P. Commentary: Real-world post-deployment performance of a novel machine learning-based digital health technology for skin lesion assessment and suggestions for post-market surveillance. Frontiers in Medicine 2024;11:1345659. doi:10.3389/fmed.2024.1345659. A peer-reviewed commentary that re-analyses the published data.
Unfortunately, the lack of transparency and failure to apply adequate scientific rigor to the current description of post-market implementation of the DERM algorithm do not support the use of AI alone as a screening tool. In particular, the reported sensitivity and missed cancer rates are potentially compromised by significant selection and attrition bias.
-
Evaluation report, by an evaluator that the makers or users commissioned.
Unity Insights. Evaluation of Deep Ensemble for the Recognition of Malignancy (DERM). Summary of an evaluation commissioned by the Department of Health and Social Care under phase IV of the AI Awards in Health and Social Care. Hosted by NIHR ARC Kent, Surrey and Sussex. 2024.
Findings suggest DERM performed accurately and safely across secondary care and community hub models, though the latter requires a larger sample size for validation. Unnecessary referrals were reduced and consultant time released due to DERM implementation.
-
Assessment report, by authors with no tie to the system.
The Unity Insights Skin Analytics evaluation report presented data from 3 NHS sites that used DERM in a post-referral setting (that is, after primary care referral to a dermatology service). This suggested that using automated DERM would lower the number of non-cancer lesions flagged for further review compared with using teledermatology alone.
-
Assessment report, by authors with no tie to the system.
Walton M, Llewellyn A, Uphoff E, Lord J, Harden M, Hodgson R, Simmonds M. Artificial Intelligence technologies for assessing skin lesions for referral on the urgent suspected cancer pathway to detect benign lesions and reduce secondary care specialist appointments: early value assessment. Health Technology Assessment 2026;30(10):1-96. doi:10.3310/GJMS0317. PubMed 41660864.
DERM might lead to around half of all patients being discharged without assessment by a dermatologist, but a small number of malignant lesions would be missed.
Details
AI system
- Workflow
-
Referrers sent photographs of lesions through existing teledermatology services. DERM assigned each lesion to a pathway, and a consultant reviewed each case that it classified as low risk.
Money
- Kind of value
-
Return on each pound spent on DERM in the secondary care model, from the evaluators' cost-benefit analysis
- Amount
-
Estimate: £1.7 returned for every £1 spent
- Recipients and payers
-
The NHS
Follow-up
- Conditions for replication
-
Referrals are appropriate and carry high-quality photographs, with a supported and trained workforce. Enough dermatologists are available for the second read, given national workforce shortages.
Entry details
- Sector
-
Healthcare.
- Purpose
-
Skin Analytics built DERM to assess photographs of skin lesions and to assign each lesion to a care pathway.
Not yet checked: model, release, provider, role, configuration, inputs and outputs, autonomy, human oversight, AI contribution, beneficiaries, reach, distribution and access, duration, every measurement question, currency and prices, period and unit, realization, gross or net, included costs, comparison, method, return and payback, ownership and control, dependencies, cross-border connections, geopolitical implications, replication attempts, availability, institutional decision, grounds, action and result, success criterion, every evidence question, observed or intended, alternative assessments, related entries, other identifiers, category mappings.
Cite this record
Peter H. Vartanian. (2026). DERM triages suspected skin cancer safely at four NHS sites (AIBD-0743). Artificial Intelligence Benefits Database, edition of 31 October 2026. https://petervartanian.xyz/aibd/records/AIBD-0743.html (archived at https://doi.org/10.7910/DVN/AADSII, version 4.0)