Artificial Intelligence Integration in Health Technology Assessments: A Review of Global Policies and Practices
Author(s)
Eon Ting, MSc, MBA1, Matthew Badin, MSc, MBA1, Nishu Gaind, MBA2, Kimberly Hofer, BSc2, Mir-Masoud Pourrahmat, MSc2, Luka Ivkovic, MSc2, Thomas Haugli-Stephens, PhD, MPhil3, Johanna Jacob, PhLic Economics4, Mir Sohail Fazeli, MD, PhD2;
1AstraZeneca, Mississauga, ON, Canada, 2Evidinno Outcomes Research Inc, Vancouver, BC, Canada, 3AstraZeneca AS, Oslo, Norway, 4AstraZeneca AB, Stockholm, Sweden
1AstraZeneca, Mississauga, ON, Canada, 2Evidinno Outcomes Research Inc, Vancouver, BC, Canada, 3AstraZeneca AS, Oslo, Norway, 4AstraZeneca AB, Stockholm, Sweden
Presentation Documents
OBJECTIVES: Artificial intelligence (AI) is revolutionizing healthcare, including drug development, clinical decision-making, and health technology assessments (HTAs). Although its application in HTA is still emerging, AI holds promise for enhancing evidence generation, dossier development, and review quality and efficiency. This study examines the current landscape of AI use and acceptance by HTA agencies globally.
METHODS: A comprehensive review of guidance documents, policy statements, and opinions on AI use published by HTA agencies across Canada, Europe, and Asia-Pacific was conducted. A supplementary search of Embase, bibliographies of previous reviews, and gray literature was completed on Dec-11-2024. Focus was on HTA agencies from 14 countries (England, Australia, Canada, France, Germany, Italy, Spain, Scotland, Sweden, Denmark, Finland, Norway, Japan, and Singapore) and EUnetHTA.
RESULTS: Few agencies reported guidance on the use of AI or machine learning in HTA submissions. Among those providing guidance, National Institute for Health and Care Excellence (NICE), Institute for Quality and Efficiency in Health Care (IQWiG), and EUnetHTA have outlined AI use in evidence generation, including use of machine learning tools for priority screening. NICE’s position statement emphasizes justifying AI use, outlining assumptions (e.g., PALISADE checklist), ensuring ethical and regulatory compliance, and prioritizing explainability, human oversight, and transparency. Innovative applications include NICE’s HTA Lab which explores generative AI in economic modelling, while Canada’s Drug Agency (CDA-AMC) has developed an evaluation instrument for AI search tools to monitor and assess evolving technologies. Quebec’s National Institute for Excellence in Health and Social Services (INESSS) has created a GPT-4-based publication screening tool to assist with internal literature reviews.
CONCLUSIONS: The findings highlight the evolving yet uneven integration and acceptance of AI into HTA submissions, with limited guidance from most HTA agencies. Increased collaboration among HTA bodies, industry and academia can clarify acceptable HTA submission methods, enhance existing methods, and facilitate sharing of best practices among stakeholders.
METHODS: A comprehensive review of guidance documents, policy statements, and opinions on AI use published by HTA agencies across Canada, Europe, and Asia-Pacific was conducted. A supplementary search of Embase, bibliographies of previous reviews, and gray literature was completed on Dec-11-2024. Focus was on HTA agencies from 14 countries (England, Australia, Canada, France, Germany, Italy, Spain, Scotland, Sweden, Denmark, Finland, Norway, Japan, and Singapore) and EUnetHTA.
RESULTS: Few agencies reported guidance on the use of AI or machine learning in HTA submissions. Among those providing guidance, National Institute for Health and Care Excellence (NICE), Institute for Quality and Efficiency in Health Care (IQWiG), and EUnetHTA have outlined AI use in evidence generation, including use of machine learning tools for priority screening. NICE’s position statement emphasizes justifying AI use, outlining assumptions (e.g., PALISADE checklist), ensuring ethical and regulatory compliance, and prioritizing explainability, human oversight, and transparency. Innovative applications include NICE’s HTA Lab which explores generative AI in economic modelling, while Canada’s Drug Agency (CDA-AMC) has developed an evaluation instrument for AI search tools to monitor and assess evolving technologies. Quebec’s National Institute for Excellence in Health and Social Services (INESSS) has created a GPT-4-based publication screening tool to assist with internal literature reviews.
CONCLUSIONS: The findings highlight the evolving yet uneven integration and acceptance of AI into HTA submissions, with limited guidance from most HTA agencies. Increased collaboration among HTA bodies, industry and academia can clarify acceptable HTA submission methods, enhance existing methods, and facilitate sharing of best practices among stakeholders.
Conference/Value in Health Info
2025-05, ISPOR 2025, Montréal, Quebec, CA
Value in Health, Volume 28, Issue S1
Code
MSR109
Topic
Methodological & Statistical Research
Topic Subcategory
Artificial Intelligence, Machine Learning, Predictive Analytics
Disease
No Additional Disease & Conditions/Specialized Treatment Areas