DEVELOPING A PROOF-OF-CONCEPT WORKFLOW COMBINING AI-DRIVEN HTA COMMITTEE SIMULATION AND AI-GENERATED PAYER QUESTIONNAIRES FOR PROACTIVE PRICING AND ACCESS PLANNING
Author(s)
Shrutya Bhalla, MBA1, Ghayath Janoudi, PhD, MD2, Stefan Walzer, MA, PhD3.
1Director, Mindrium Global Ltd., Toronto, ON, Canada, 2Loon, Ottawa, ON, Canada, 3MArS Market Access & Pricing Strategy GmbH, Weil am Rhein, Germany.
1Director, Mindrium Global Ltd., Toronto, ON, Canada, 2Loon, Ottawa, ON, Canada, 3MArS Market Access & Pricing Strategy GmbH, Weil am Rhein, Germany.
OBJECTIVES: Health technology assessment (HTA) recommendations are shaped by clinical, economic, and evidentiary considerations that influence pricing and access decisions. This research aims to demonstrate a proof-of-concept workflow integrating AI-driven HTA committee simulation with AI-generated payer research questionnaires to support more proactive pricing and access planning in public payer settings.
METHODS: A hypothetical therapy profile in SMA (Spinal Muscular Atrophy) was constructed for using standard market access inputs, including target population, comparator landscape, clinical evidence, health economic evidence, and pricing scenarios. An AI-based multi-persona HTA committee simulator using persona-conditioned prompts and Monte Carlo sampling was used to simulate reimbursement deliberations reflecting the perspectives of UK, German, and Canadian HTA decision makers. From these simulations, probabilistic predictions of the price expectations and ICER/QALYs, reimbursement outcomes, together with decision patterns and uncertainty across clinical, economic, and evidence domains, were generated. These outputs were translated into structured topic areas and passed to a Large Language Model, to generate a targeted payer research questionnaire. The workflow and the research questionnaire were then externally validated with a AI-based payer research platform and the three payers within the scope countries using a structured rubric assessing relevance, clarity, and practical implementability.
RESULTS: he workflow was able to produce structured, indication-specific payer questionnaires linked to simulated HTA concerns and reimbursement scenarios and could be adapted as strategic scenarios changed. The resulting questionnaires focused payer research on reimbursement-relevant uncertainties across clinical, economic, and evidence domains.
CONCLUSIONS: This proof-of-concept suggests that integrating AI-based HTA simulation with AI-generated payer questionnaires may provide a practical methodology for more proactive, strategy-aligned payer research design in UK, German, and Canadian settings. Further work is needed to compare simulation-guided questionnaires with traditional guides in live payer research projects and to assess their impact on pricing and access decisions.
METHODS: A hypothetical therapy profile in SMA (Spinal Muscular Atrophy) was constructed for using standard market access inputs, including target population, comparator landscape, clinical evidence, health economic evidence, and pricing scenarios. An AI-based multi-persona HTA committee simulator using persona-conditioned prompts and Monte Carlo sampling was used to simulate reimbursement deliberations reflecting the perspectives of UK, German, and Canadian HTA decision makers. From these simulations, probabilistic predictions of the price expectations and ICER/QALYs, reimbursement outcomes, together with decision patterns and uncertainty across clinical, economic, and evidence domains, were generated. These outputs were translated into structured topic areas and passed to a Large Language Model, to generate a targeted payer research questionnaire. The workflow and the research questionnaire were then externally validated with a AI-based payer research platform and the three payers within the scope countries using a structured rubric assessing relevance, clarity, and practical implementability.
RESULTS: he workflow was able to produce structured, indication-specific payer questionnaires linked to simulated HTA concerns and reimbursement scenarios and could be adapted as strategic scenarios changed. The resulting questionnaires focused payer research on reimbursement-relevant uncertainties across clinical, economic, and evidence domains.
CONCLUSIONS: This proof-of-concept suggests that integrating AI-based HTA simulation with AI-generated payer questionnaires may provide a practical methodology for more proactive, strategy-aligned payer research design in UK, German, and Canadian settings. Further work is needed to compare simulation-guided questionnaires with traditional guides in live payer research projects and to assess their impact on pricing and access decisions.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
MSR269
Topic
Health Technology Assessment, Methodological & Statistical Research, Study Approaches
Disease
Rare & Orphan Diseases