对英国国家医疗服务体系人工智能实验室的混合方法形成性评估。

A mixed methods formative evaluation of the United Kingdom National Health Service Artificial Intelligence Lab.

作者信息

Cresswell Kathrin, Williams Robin, Dungey Sheena, Anderson Stuart, Bernabeu Miguel O, Mozaffar Hajar, Yang Xiao, Sai Varun, Bea Sara, Eason Sally

机构信息

Usher Institute, University of Edinburgh, Edinburgh, UK.

Institute for the Study of Science, Technology and Innovation, University of Edinburgh, Edinburgh, UK.

出版信息

NPJ Digit Med. 2025 Jul 17;8(1):448. doi: 10.1038/s41746-025-01805-w.

DOI:10.1038/s41746-025-01805-w

PMID:40676147

原文链接:https://pmc.ncbi.nlm.nih.gov/articles/PMC12271305/

Abstract

Internationally, health systems are investing in Artificial Intelligence (AI) to improve safety, quality, and efficiency, yet many efforts remain localised and do not progress beyond early development stages. In 2019, National Health Service (NHS) England and the Department of Health and Social Care launched the AI Lab to accelerate safe AI adoption. We conducted a mixed-methods evaluation of the AI Lab, analysing 1021 documents and 85 stakeholder interviews. The AI Lab made important contributions to national AI policy, regulation, and capability building, and positioned the United Kingdom as a global leader in AI deployment for health. Despite progress, implementation and scaling were hindered by shifting objectives, limited capacity, and systemic misalignment with service needs. Some AI technologies demonstrated high return on investment and improved clinical processes. Lessons from the AI Lab highlight critical socio-organisational factors, gaps in scaling support, and the need for sustained coordination to realise the long-term benefits of AI in health and social care systems.

摘要

在国际上，卫生系统正在投资于人工智能（AI）以提高安全性、质量和效率，但许多努力仍局限于局部，尚未超越早期开发阶段。2019年，英国国家医疗服务体系（NHS）以及卫生和社会保健部启动了人工智能实验室，以加速安全采用人工智能。我们对该人工智能实验室进行了混合方法评估，分析了1021份文件并与85位利益相关者进行了访谈。该人工智能实验室为国家人工智能政策、监管和能力建设做出了重要贡献，并使英国成为人工智能在医疗领域部署的全球领导者。尽管取得了进展，但目标的不断变化、能力有限以及与服务需求的系统性不匹配阻碍了实施和推广。一些人工智能技术显示出高投资回报率并改善了临床流程。人工智能实验室的经验教训凸显了关键的社会组织因素、扩大支持方面的差距，以及为实现人工智能在卫生和社会保健系统中的长期效益而进行持续协调的必要性。