Assessing the Value: Health Technology Assessment and the Governance of AI in Stroke Care
Digital tools are reshaping the acute stroke pathway – but whether they truly enter national health systems depends on something less visible than the technology itself: a rigorous, multidimensional Health Technology Assessment.
The digital transformation of stroke care is well underway. Deep-learning software, advanced neuroimaging, and telemedicine now support clinical decisions across the hyperacute, pre-hospital, intra-hospital, and post-acute phases of a time-critical disease. Many forms of “smart Tech and unregulated wearables are being incorporated in different shap eand form into Stroke care in the border sense (we cover some of these here). Yet innovation adopted without a structured evaluation of its wider consequences bares some risk.. This is where Health Technology Assessment (HTA) becomes decisive: it weighs not only clinical efficacy and safety, but the economic, organisational, ethical, and legal impact of a technology before it is embedded in routine care, also allowing policy makers to distinguish a commercial hit from a clinical tool [2, 6].
The tools under assessment
AI-driven platforms such as Brainomix apply deep learning to advanced neuroimaging, transforming how CT and MR scans are read in acute stroke [4]. These systems can automatically identify large-vessel occlusions (LVO), quantify the ischemic core and salvageable penumbra through perfusion maps, and compute the ASPECTS score — reducing inter-observer variability and accelerating the decision between intravenous thrombolysis and mechanical thrombectomy [4].
Telemedicine applied to stroke, known as ‘Telestroke’, extends specialist expertise across geography. It allows clinicians at Hub centres such as Siena’s ‘Le Scotte’ Hospital to remotely assess patients admitted to Spoke centres [3].
Integrating these technologies into National Health Systems cannot proceed on clinical promise alone. HTA subjects each innovation to a multidimensional appraisal built on several pillars.
Economic and sustainability analysis.
Using cost-effectiveness (CEA) and cost-utility (CUA) analyses, HTA establishes whether a technology delivers genuine value for the resources it consumes — the difference between an investment that strengthens a health system and one that quietly drains it [2, 4]. The evaluation also serves as a safeguard for end users (patients and their families) to ensure ethical standards regarding their sensitive data and their wellbeing is protected from innovations that serve shareholders more than the public interest.
Organisational and training impact.
Software is only as effective as the teams operating it. HTA makes explicit the need to train medical and nursing staff and to guarantee the interoperability of hospital information systems, so that a tool validated in a trial actually performs in a real Hub-and-Spoke network [2, 4].
Ethical and legal dimensions.
As AI becomes indispensable to treating time-dependent conditions like stroke, it also raises questions HTA is designed to address: medical liability in the event of a diagnostic error by the software, algorithmic bias, and data privacy — all now framed within the European Regulation on Artificial Intelligence (the AI Act) [2, 6].
From evidence to policy
Technological innovation offers extraordinary opportunities to improve prognosis and quality of life for stroke patients. But introducing these technologies without centralised governance risks creating inequities in access to care and financial inefficiencies across regions and centres. Health Technology Assessment confirms itself as an indispensable politico-scientific instrument — the mechanism that translates clinical evidence into health policy that is sustainable and of real value [2, 6].
Author(s): Martin Vitiello, Lorenzo De Stefano, Siena University Hospital, “Santa Maria alle Scotte
References:
[1] World Heart Federation / World Stroke Organization (2024–2026). Reports & Global Fact Sheets. Updated global epidemiological data on the clinical, economic, and social impact of stroke and the importance of timely intervention.
[2] European Network for Health Technology Assessment (EUnetHTA). The HTA Core Model for Rapid Relative Effectiveness Assessment. Methodological guidelines for standardising multidimensional health technology evaluation.
[3] Bladin, C. F., & Cadilhac, D. A. (2024). Effectiveness of Telemedicine in Stroke Systems of Care: A Systematic Review. Journal of Stroke and Cerebrovascular Diseases.
[4] Kunz, W. G., et al. (2025). Artificial Intelligence in Acute Stroke Imaging: A Health Technology Assessment of Automated Software Packages. Stroke Journal.
[6] World Health Organization (WHO) / Agenzia Nazionale per i Servizi Sanitari Regionali (AGENAS). Programmatic documents on the Governance of Digital Health Technologies. Guidelines on integrating HTA processes into health systems for the acquisition of complex medical devices, AI, and telemedicine services.
Keywords:
#Stroke #HTA #AIinhealth #Telestroke #EUAIact