OBJECTIVES: Determine a generic set of criteria to quantify the holistic value of AI interventions in UK NHS radiology practice.

METHODS: Local adoption pathways for health technology in the NHS currently prioritise cost savings demonstrated with budget impact modelling. This risks missing the potential of technologies to address other areas of unmet need e.g. addressing workforce burnout. The MCDA approach allows for the quantification of monetary and non-monetary aspects of novel technologies. This is opposed to traditional economic evaluations that focus on direct financial and quality of life impacts.

RESULTS: The MCDA approach has received a degree of acceptance for the health technology assessment of medicines and therefore can also be considered for the evaluation of AI-driven medical devices. To our knowledge, no previous MCDA has been conducted of AI radiology interventions. Relevant value frameworks (for provider facing and radiology) point to value drivers that can be categorised under: patient outcomes, health equalities and integration with current service. Data gathered from NHS clinical stakeholders suggest that the value of technologies in reducing environmental impact, increasing alignment with national guidance, legal cases and workforce wellbeing and training should also be quantified when assessing considering technologies for adoption.

CONCLUSIONS: Outcomes beyond the financial implications must be considered during quantitative assessments that support reimbursement decisions of AI-assisted devices. Future research should aim to perform an MCDA of an AI-assisted radiology intervention using the categories identified above.