The National Institute for Health and Care Excellence (NICE) has issued Early Value Assessment (EVA) guidance on the use of artificial intelligence (AI) technologies for assessing and triaging skin lesions referred to the urgent suspected skin cancer pathway.
Introduction
The guidance evaluates the Deep Ensemble for Recognition of Malignancy (DERM), an AI-based tool designed to assist in the assessment and triage of skin lesions in adults referred to the urgent suspected skin cancer pathway.
Conditional Use of DERM in the NHS
NICE recommends that DERM can be used within NHS teledermatology services during a three-year evidence generation period, provided that:
- Evidence outlined in the evidence generation plan is being collected.
- DERM has obtained appropriate regulatory approvals, including NHS England’s Digital Technology Assessment Criteria (DTAC) approval.
To mitigate potential risks of missed or delayed cancer diagnoses, especially in individuals with black or brown skin, additional safety measures such as healthcare professional reviews and regular performance monitoring are advised.
Evidence Generation Requirements
During the evidence generation period, further data is needed on:
- The accuracy of DERM in detecting cancerous and non-cancerous skin lesions, particularly in people with black or brown skin.
- The impact of DERM on the number of benign lesions redirected to non-urgent pathways and the reduction in face-to-face dermatology appointments.
- The proportion of skin lesions referred from primary care eligible for assessment by DERM.
The evidence generation plan provides detailed information on prioritised evidence gaps, ongoing studies, and potential real-world data sources.
Technology Overview
DERM is an AI-based skin lesion analysis tool intended for use in screening, triage, and assessment of suspected skin cancer lesions in adults. It operates within teledermatology services, analysing dermoscopic images captured via smartphone to classify lesions as benign, pre-cancerous, or cancerous.
The cost per assessment using DERM is £30, with additional charges for virtual reviews and second-read dermatologists. Discounts are available when biopsy results are shared with the company.
Conclusion
NICE’s Early Value Assessment of DERM represents a cautious approach to integrating AI technologies into the NHS. The conditional recommendation allows for the collection of real-world evidence to determine the tool’s effectiveness and safety, particularly concerning diverse skin types.
Citation:
National Institute for Health and Care Excellence. (2025, May 1). Artificial intelligence (AI) technologies for assessing and triaging skin lesions referred to the urgent suspected skin cancer pathway: early value assessment. NICE.