Evidence generation plan for artificial intelligence (AI) technologies to aid opportunistic detection of vertebral fragility fractures: early value assessment
6 Implementation considerations
The following considerations around implementing the evidence generation process have been identified through working with system partners:
Evidence generation
-
The companies should collect and analyse outcome data carefully to ensure that important subgroups are included in the studies, such as people:
-
under 50
-
with obesity where the field of view for a diagnostic image may include more surrounding tissue
-
with cancer
-
with osteogenesis imperfecta.
-
Equalities
-
During implementation of artificial intelligence (AI) technologies, having limited access to fracture liaison services, fragility fractures and osteoporosis treatments in the NHS may drive health inequalities. This could worsen regional inequalities, particularly for people living in deprived areas.
System and implementation considerations
-
It is unknown how these AI technologies might impact the skills of the healthcare professional in detecting vertebral fragility fractures. Care should be taken to ensure that healthcare professionals are not deskilled by over-reliance on AI technologies.
-
The companies should ideally provide or support training for healthcare professionals in using the technologies.
-
The current regulatory approval for HealthVCF is due to expire in 2028, so the technology is unlikely to be available on the UK market after 2028. The preference is for evidence to be generated using HealthOST while it is used in the NHS, because this is the technology that will be more widely available in the future. Data from HealthVCF may not be generalisable to HealthOST (see section 3.13 of the guidance).
ISBN: 978-1-4731-7239-5
This page was last updated: