The cost-effectiveness of an artificial intelligence software for stroke imaging and treatment decisions in English hospitals
NeuroImage Stroke · 2026 · 1-2 · 100008
DOI: 10.1016/j.ynist.2026.100008 ↗Kurz gefasst
Fachartikel in der Zeitschrift NeuroImage Stroke, 2026. Verfasst von Nichola R. Naylor und 6 weiteren Autor:innen. Die Arbeit nennt 31 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Introduction
Artificial intelligence-based imaging can support acute stroke treatment decisions, yet cost-effectiveness evidence spanning intravenous thrombolysis and mechanical thrombectomy pathways is limited. We evaluated cost-effectiveness of Brainomix 360 Stroke across National Health Service hospitals in England. Patients and Methods We developed a decision-analytic model from the National Health Service and personal social services perspective, combining a decision tree for acute stroke pathways with a long-term Markov model based on the modified Rankin Scale. The cohort comprised all acute ischaemic stroke admissions in England in one year (year = 2023, n=81,565 patients). Brainomix 360 Stroke was modelled as increasing the probability that eligible patients receive reperfusion therapy, based on a five-year real-world evaluation across 26 hospitals. Costs and quality-adjusted life years were discounted at 3.5% per year, with incremental net monetary benefit calculated at £20,000 per quality-adjusted life year.
Results
Brainomix 360 Stroke increased the number of patients receiving intravenous thrombolysis 8% and mechanical thrombectomy by 44%. The intervention was dominant, generating almost £4 million in cost savings and over 1,700 additional quality-adjusted life years, yielding an incremental net monetary benefit of over £38 million. Per AIS patient, this equates to £49 in cost savings and 0.021 additional quality-adjusted life years. Probabilistic sensitivity analysis indicated a 98.9% probability of cost-effectiveness at £20,000 per quality-adjusted life year. Results were robust across sensitivity and scenario analyses, though most sensitive to cohort age, long-term cost assumptions and treatment eligibility parameters.
Conclusion
Implementing artificial intelligence-assisted stroke imaging is highly likely to be cost-effective, driven by improved access to reperfusion therapies and better long-term functional outcomes.
Zitieren
APA
Naylor, N. R., Woodhead, Z. V., Ford, G. A., Lowe, D., Hargroves, D., Wyrtzen, J., & Harston, G. (2026). The cost-effectiveness of an artificial intelligence software for stroke imaging and treatment decisions in English hospitals. NeuroImage Stroke, 1-2, 100008. https://doi.org/10.1016/j.ynist.2026.100008
Vancouver
Naylor NR, Woodhead ZV, Ford GA, Lowe D, Hargroves D, Wyrtzen J, et al. The cost-effectiveness of an artificial intelligence software for stroke imaging and treatment decisions in English hospitals. NeuroImage Stroke. 2026;1-2:100008. doi:10.1016/j.ynist.2026.100008
BibTeX
@article{Naylor2026,
author = {Naylor, Nichola R. and Woodhead, Zoe V.J. and Ford, Gary A. and Lowe, Deborah and Hargroves, David and Wyrtzen, Jeff and Harston, George},
title = {The cost-effectiveness of an artificial intelligence software for stroke imaging and treatment decisions in English hospitals},
journal = {NeuroImage Stroke},
year = {2026},
volume = {1-2},
pages = {100008},
doi = {10.1016/j.ynist.2026.100008}
}
EUVIMED