Bookmarks
REAL-DERM: Evaluation of the AI software Deep Ensemble for Recognition of Malignancy (DERM) versus standard care in the NHS skin cancer pathway: a comparative observational cohort study using real-world data
Safe People
Organisation name
University Hospital Southampton
Applicant name(s)
Funders/ Sponsors
Safe Projects
Project ID
SDE_WXS_PROJ_121
Lay summary
The need for this registry has arisen following the NICE Early Value Assessment supporting the provisional use of autonomous AI tools, such as DERM AI, for skin cancer diagnosis while further evidence is gathered. To evaluate these technologies safely and robustly in real-world NHS settings, linked and high-quality pathway data are required. At present, key information is often spread across referral, tele-dermatology, appointment, procedure and pathology systems. The registry will provide the data infrastructure needed to understand how urgent suspected skin cancer pathways operate and how new technologies or service models affect care.
Public benefit statement
For patients, the registry will provide a real-world view of the fairness of these AI tools as currently deployed within NHS tele-dermatology services. For clinicians and services, it will support pathway optimisation and reduce avoidable system failures and resulting harm. For the NHS, it will provide a robust, real-world evidence base to inform future commissioning around these autonomous AI systems and their impacts on service planning and policy decisions. This database may identify opportunities to reduce unnecessary face-to-face demand, improve prioritisation, maintain or improve cancer detection, support faster diagnosis, and understand the operational and economic implications of AI-supported pathways.
Other approval committees
Latest approval date
28/09/2026
Safe Data
Dataset(s) name
REAL-DERM
Safe Setting
Access type
TRE
Safe Outputs
Link to research outputs