Evidence generation plan for artificial intelligence (AI) technologies for assessing and triaging skin lesions referred to the urgent suspected skin cancer pathway
2 Evidence gaps
This section describes the evidence gaps that need to be addressed for future committee decision making. The committee will not be able to make a positive recommendation without these being addressed.
2.1 Essential evidence for future committee decision making
How accurate DERM used in teledermatology services is at detecting cancer and non-cancer skin lesions compared with teledermatology services alone
Collecting data on the accuracy of DERM used in teledermatology services to detect cancer and non-cancer skin lesions compared with teledermatology alone is essential for determining whether it can provide assessments suitable for routine clinical settings. The data will help to determine whether DERM can effectively discharge non-urgent cases from the suspected skin cancer pathway while maintaining diagnostic accuracy for detecting high-risk lesions. This data will help evaluate the potential of DERM to enhance the diagnosis of cancer skin lesions and optimise clinical resources by reducing the burden on dermatology services. The data will also help to assess whether DERM can increase staff capacity and benefit people with non-cancer dermatological conditions. Additionally, this can help inform whether AI technologies can be used autonomously and whether this is reliable and safe.
Accuracy of DERM in people with black or brown skin
Collecting information about the accuracy of DERM to detect cancer and non-cancer skin lesions across different skin colours is vital for assessing potential biases in performance and for ensuring equitable healthcare. Skin tone should ideally be measured using spectrophotometry.
The effect of using AI technology in teledermatology services on the number of referrals for face-to-face dermatology appointments compared with established teledermatology services alone
Collecting data on the number of face-to-face dermatology referrals generated by AI technology compared with well-established teledermatology services alone is crucial for understanding the effect of AI technology on healthcare workflows. This information will help determine if AI can reduce unnecessary referrals and save dermatologist time, thereby optimising resources and improving timely access to care for people. Additional information about the number of referrals to dermatology before and after implementation of AI technology will provide further understanding of its impact.
This page was last updated: