TO SYSTEMATICALLY EVALUATE THE APPLICATION OF ARTIFICIAL INTELLIGENCE IN COST-EFFECTIVENESS MODELING, AND ASSESS ITS IMPACT ON MODEL EFFICIENCY, ACCURACY, AND DECISION-MAKING IN HEALTH TECHNOLOGY ASSESSMENT (HTA).
Author(s)
Rohini Yadav, M.Pharm, Sagar Rai, PharmD, Venkata Sharmila Uribandi, M.Pharm, Richa Goyal, M.Pharm, Ph.D.
Dual Sphere Consulting, 427 Hougang Avenue 6, #01-34, Singapore.
Dual Sphere Consulting, 427 Hougang Avenue 6, #01-34, Singapore.
OBJECTIVES: While there is an increased interest in utilizing Artificial Intelligence (AI) technology to efficiently calculate cost-effectiveness, the methodology is still fragmented across many different disease areas, geographic locations, and modeling approaches. This review aimed to summarise the use of AI in cost-effectiveness models in HTA (Health Technology Assessment) and the effect of AI on the efficiency and accuracy of the models and the quality of the evidence produced for health economic decisions.
METHODS: Cost-effectiveness analyses or HTA studies where AI was used (e.g., machine learning, generative AI, and computer-assisted detection systems) were searched using a systematic search within key electronic databases (like PubMed) using specific keywords related to AI, such as "AI", "AI in cost-effectiveness", "AI in HTA". All searches were performed independently and conducted in accordance with the PRISMA guidelines to ensure comprehensive data extraction, assessment of relevant study characteristics, methodological quality, modelling frameworks used (e.g., Markov models and individualized estimated treatment effect models), and outcome measures related to the efficiency and accuracy of decision-making.
RESULTS: The reviewed evidence demonstrates that when integrated into cost-effectiveness models, AI-assisted tools can enhance screening accuracy, improve individualized treatment effect estimation, and reduce per-unit screening costs compared to standard care pathways. Applications of AI across colorectal cancer screening, breast cancer detection, and community-based fundus disease programs have demonstrated favourable cost-effectiveness profiles. However, significant barriers remain, particularly in Central and Eastern European countries, where data infrastructure, regulatory frameworks, and institutional capacity constrain AI adoption in HTA.
CONCLUSIONS: AI has considerable potential to enhance CEA within the framework of HTA. However, its effective implementation requires standardized methodologies for conducting AI-based CEAs, equitable access to AI technologies across healthcare systems, and robust policy guidance for integration into decision-making processes.
METHODS: Cost-effectiveness analyses or HTA studies where AI was used (e.g., machine learning, generative AI, and computer-assisted detection systems) were searched using a systematic search within key electronic databases (like PubMed) using specific keywords related to AI, such as "AI", "AI in cost-effectiveness", "AI in HTA". All searches were performed independently and conducted in accordance with the PRISMA guidelines to ensure comprehensive data extraction, assessment of relevant study characteristics, methodological quality, modelling frameworks used (e.g., Markov models and individualized estimated treatment effect models), and outcome measures related to the efficiency and accuracy of decision-making.
RESULTS: The reviewed evidence demonstrates that when integrated into cost-effectiveness models, AI-assisted tools can enhance screening accuracy, improve individualized treatment effect estimation, and reduce per-unit screening costs compared to standard care pathways. Applications of AI across colorectal cancer screening, breast cancer detection, and community-based fundus disease programs have demonstrated favourable cost-effectiveness profiles. However, significant barriers remain, particularly in Central and Eastern European countries, where data infrastructure, regulatory frameworks, and institutional capacity constrain AI adoption in HTA.
CONCLUSIONS: AI has considerable potential to enhance CEA within the framework of HTA. However, its effective implementation requires standardized methodologies for conducting AI-based CEAs, equitable access to AI technologies across healthcare systems, and robust policy guidance for integration into decision-making processes.
Conference/Value in Health Info
2026-09, ISPOR Asia Pacific 2026, Bangkok, Thailand
Value in Health, Volume 55, Issue S1
Code
HTA42
Topic
Health Technology Assessment
Topic Subcategory
Systems & Structure
Disease
No Additional Disease & Conditions/Specialized Treatment Areas