Introduction
Digital twins are dynamic digital representations that combine patient-specific multidimensional data with simulation and, increasingly, AI-based prediction1,2
In healthcare, they are being explored for disease modelling, treatment optimisation and synthetic or in silico comparator generation, particularly where conventional comparative evidence is difficult to obtain2,3,7
Despite rapid development, their role and acceptance within health technology assessment (HTA) remain unclear4,8
Objectives
- To assess whether AI-enabled digital-twin comparator evidence has been used in formal HTA appraisal and identify the agency response
- To describe emerging HTA positions and explore stakeholder views on acceptability, methodological requirements and barriers
Conclusions
Digital twins are beginning to intersect with HTA through methodological guidance, early HTA and post-launch economic evaluation, 9–11 but formal appraisal use was not identified.
This suggests a gap between growing clinical and methodological interest in digital twins and demonstrated acceptance of digital-twin-generated comparative evidence in formal HTA. Existing experience with external and synthetic control arms provides relevant methodological precedent but should not be interpreted as direct precedent for AI-generated patients. 1–8
Survey findings indicate that acceptance of AI-generated comparative evidence may be conditional on robust validation, transparency, reproducibility and appropriate characterisation of uncertainty. Respondents generally envisaged a limited or supportive role rather than use as primary evidence, although findings should be interpreted cautiously given the small sample (n=5) 8
Further research and HTA experience are needed to establish when AI-enabled digital-twin comparators can credibly reduce uncertainty in comparative effectiveness and what evidentiary standards agencies will require