Evidence generation plan for artificial intelligence technologies to help detect fractures on X-rays in urgent care

5 Minimum evidence standards

The selection of the artificial intelligence (AI) technologies evaluated for this assessment was based on a limited evidence base gathered from company submissions, the external assessment group's review of the available literature and committee discussions.

The technologies were primarily assessed on diagnostic accuracy. Initial findings indicated a potential benefit in reducing missed fractures. However, this evidence had significant limitations such as risk of bias, small sample sizes and heterogeneity in study designs. Additionally, the evidence lacked consistency across different subgroups, such as children or people with conditions that affect bone health.

It is important to note that there was an absence of studies done in the NHS. This contributed to the decision for further data collection aimed at supporting the implementation of these technologies across urgent care settings in the NHS.

To have a deeper understanding of the full range of benefits these technologies can offer and to support their implementation, additional data is needed on diagnostic accuracy, efficacy in specific subgroups, clinical and service outcomes, and economic impact.