ISPOR 2025: Exploring sessions on artificial intelligence

The ISPOR—The Professional Society for Health Economics and Outcomes Research (ISPOR) 2025 conference (May 13–16, Montreal, Canada) will feature a range of sessions exploring the growing role of artificial intelligence (AI) and generative AI in health economics and outcomes research (HEOR). Discussions will cover how AI and real-world data (RWD) are being used to support market access, improve evidence generation, and inform research in areas such as oncology, rare diseases, and health preference studies. The program will also examine the limitations and considerations involved in applying AI across different HEOR contexts.

Selected sessions

From General to HEOR-Specific: Transforming LLMs Into Reliable Research Tools

Date and time: May 14, 10:15am – 11:15am
Moderator: J Jaime Caro (Evidera)
Speakers: Apoorva Ambavane (Evidera), Baris Deniz (Aide Solutions)

This session will offer a structured examination of how large language models (LLMs) can be applied in HEOR, addressing key considerations such as data quality, reproducibility, and validation. Through a live demonstration focused on disease burden analysis, speakers will highlight the differences between basic use and advanced research design, providing practical insights for researchers and decision-makers.

Driving Evidence-Based Medicine Forward With Generative AI (GenAI)

Date and time: May 14, 11:45am – 12:15pm
Moderator: Eric Wu (Analysis Group)
Speakers: Rajeev Ayyagari (Analysis Group), Song Wang (Takeda), Guo Li (Johnson and Johnson Innovative Medicine (JJIM)), Jimmy Royer (Analysis Group)

The aim of this symposium is to explore how GenAI is transforming HEOR and RWE research by streamlining data analysis, enhancing insight generation, and advancing healthcare decision-making. Participants will examine key developments, challenges, and real-world applications of GenAI, including tools for text classification, research summarization, and multilingual data screening, highlighting new opportunities to improve research efficiency, accuracy, and patient outcomes.

Accelerating the Adoption of Generative AI in HEOR: Lessons From Early Adopters

Date and time: May 14, 1:45pm – 2:45pm
Moderator: Jag Chhatwal (Harvard Medical School/Massachusetts General Hospital)
Speakers: Turgay Ayer (Value Analytics Labs), Rachael Fleurence (National Institutes of Health), Ipek Ozer Stillman (Takeda)

This panel, convened by the ISPOR Generative AI Working Group, will explore how generative AI can advance HEOR despite current adoption challenges. Experts will highlight practical applications, regulatory considerations, and strategies to address barriers such as limited awareness and organizational hesitancy. Through case studies and actionable guidance, the session aims to help HEOR professionals integrate GenAI into research and decision-making processes.

Harnessing AI and RWE: Developing Integrated Evidence Strategies for Market Access in the IRA Era

Date and time: May 14, 3:00pm – 3:30pm
Speakers: Jacqueline Vanderpuye-Orgle (Parexel), Matthew Gordon (Parexel)

This session aims to demonstrate how integrated evidence planning (IEP), supported by real-world evidence (RWE) and AI-enabled tools, can optimize clinical development and market access – especially in light of the IRA. Through case studies, attendees will explore how RWE and AI can be embedded across the product lifecycle to enhance strategy, execution, and outcomes.

From Prompting to Policy: The Advances of Generative AI in the Last Year

Date and time: May 14, 5:00pm – 6:00pm
Moderator: Siguroli Teitsson (Bristol Myers Squibb)
Speakers: Sven L Klijn (Bristol Myers Squibb), Tim Reason (Estima Scientific), Rachael Fleurence (National Institutes of Health)

As GenAI continues to evolve, its relevance to HEOR is rapidly expanding. This session will examine recent advances in LLMs, prompting techniques, new HEOR applications, and policy developments. The discussion will explore implications for transparency, reliability, and validity, while interactive polling and open dialogue will encourage attendees to reflect on both opportunities and challenges for integrating GenAI into HEOR practice.

AI Agents and Guardrails in HEOR: The Ultimate Solution to GenAI Shortcomings or Just Another Overhyped Tool?

Date and time: May 14, 10:15am – 11:15am
Moderator: Foluso O Agboola (Institute for Clinical and Economic Review)
Speakers: Sven L Klijn (Bristol Myers Squibb); Tim Disher (Loon), Ghayath Janoudi (Loon)

As AI agents powered by LLMs gain traction in HEOR, questions around their role, risks, and oversight are coming to the forefront. This session will bring together expert perspectives to debate the value and limitations of AI-driven decision support. Panelists will explore the potential of AI agents to enhance efficiency and scalability, examine ethical and transparency concerns, and consider whether structured guardrails can offer a responsible path forward. An interactive discussion will help define practical approaches for navigating this evolving space.

Prompt Engineering: Harnessing Generative AI for HEOR

Date and time: May 15, 1:45pm – 2:45pm
Moderator: Jag Chhatwal (Harvard Medical School/Massachusetts General Hospital)
Speakers: Rachael Fleurence (National Institutes of Health), Turgay Ayer (Value Analytics Labs)

Another session from the ISPOR Generative AI Working Group will focus on the critical role of prompt engineering in optimizing generative AI for HEOR and RWE applications. Attendees will learn techniques such as zero-shot and chain of thought prompting to improve data extraction, economic modeling, and evidence synthesis. Practical demonstrations and interactive discussion will explore how thoughtful prompts can enhance accuracy, relevance, and utility while addressing challenges like contextual retention and output reliability.

How Should Artificial Intelligence (Not) Be Used in Health Preference Research?

Date and time: May 15, 5:00pm – 6:00pm
Moderator: Deborah A Marshall (University of Calgary)
Speakers: Sebastian Heidenreich (Evidera), Marco Boeri (OPEN Health), Tommi Tervonen (Kielo Research)

This issue panel will explore how AI could transform health preference research by creating efficiencies across key stages, from study design to data analysis. Panelists will discuss real-world applications of AI in developing study materials, reviewing literature, and analyzing qualitative data, including social media. The session will conclude with an interactive discussion on how AI might shape future approaches in this evolving research area.

Rare but Common: Generative AI’s Potential on Data, Evidence, and Insight Generation in Rare Diseases

Date and time: May 16, 8:00am – 9:00am
Moderator: Xiaoyan Wang (IMO health)
Speakers: Hua Xu (Yale University), Chunhua Weng (Columbia University), Jing Wang-Silvanto (Astellas Pharma)

With rare diseases posing persistent challenges for research and care, this issue panel aims to explore how GenAI can help address data gaps and improve outcomes. Speakers will examine how AI-powered tools can unlock and synthesize RWD, create rare disease-specific knowledge graphs, and generate actionable insights. Drawing on examples from academia and industry, the session will highlight opportunities to enhance trial recruitment, support decision-making, and better meet unmet patient needs.

Latest research in AI

Date and time: May 14, 1:45pm – 2:45pm
Moderator: Min-Hua Jen (Eli Lilly)

The Power and Pitfalls of AI in Health Data Analysis

Attendees will learn about the strengths and limitations of using AI in health data analysis as speakers discuss their research in this Podium Session. Research includes:

• Leveraging Real-World Data and NLP to Identify At Risk Metabolic Dysfunction Associated Steatohepatitis in the General Population – Or Shaked (Briya)
• Boosting Predictive Power: Thoughtfully Unleashing the Potential of Machine Learning In Real-World Healthcare Outcome Estimation – Achal Patel (Genentech)

AI-Assisted Literature Reviews: Requirements and Advances

Date and time: May 14, 5:00pm – 6:00pm
Moderator: Alexandre Martins (INESSS)

In this second Podium on AI-research, discussants will review their work using AI tools for literature reviews, including:

• Evaluating the Performance of Claude 3.5 Sonnet in Data Extraction Automation for Systematic Literature Reviews (SLRs) – Mir Sohail Fazeli (Evidinno Outcomes Research Inc)
• AI Tools for Literature Reviews: Are Current Guidelines Meeting the Needs of Researchers? – Grace E Fox (OPEN Health HEOR & Market Access)

Other posters of interest

• From Scarcity to Strategy: Generative AI for Rare Disease Model Conceptualization – Tushar Srivastava (ConnectHEOR)
• Using Autonomous Generative AI Agents for Data Extraction from Clinical Study Reports: Accuracy Assessment Against Canada's Drug Agency Reports – Ghayath Janoudi (Loon)
• Leveraging AI/Technology in Survey Design, Deployment, and Analysis – Jordana Schmier (OPEN Health)
• Sentence-Level Abstract Section Classification in HEOR Literature using a Biomedical Language Model – Reza Jafar (Cytel)
• Using Artificial Intelligence to Predict Patient's Preferences – Tina Cheng (Duke University)
• GenAI Goes to ISPOR: Exploratory, Descriptive Analysis of Generative AI Performance for Summarizing and Synthesizing ISPOR Research Abstracts as Sources – Cynthia D Morrow (Knowledge Resolution)
• A Review of Guidelines and Checklists for the Use of Artificial Intelligence/Machine Learning (AI/ML) in Evidence Generation: Current Landscape and Recommendations – Raju Gautam (ConnectHEOR)
• The Use of Large Language Models for Systematic Literature Review Automation: An Evaluation of Quality and Time Savings – Ryan Thaliffdeen (Gilead Sciences)
• Comparison of AI-Enhanced Tools for Automating Scientific Literature Reviews – Roman Casciano (Certara)
• Large-Language Models to Complement and Augment Literature Review: Hi! How Can I Help You? – Sarah Goring (SMG Outcomes Research)
• Leveraging Artificial Intelligence for Thematic Analysis of Qualitative Transcripts: A Feasibility Study in Insurance Payer Interviews – Alexa Klimchak (Sarepta Therapeutics)
• Emerging Trends in AI Applications: Shaping Pharmacoepidemiology and Health Technology Assessments – Aurore Bergamasco (YOLARX Consultants)
• Unlocking AI's Potential in Pricing and Reimbursement: Insights Across Global Healthcare Archetypes – Grace E Fox (OPEN Health HEOR & Market Access)
• Data Extraction in Literature Reviews Using an Artificial Intelligence Model: Prompt Development and Testing – Liz Lunn (Costello Medical Consulting)
• Using Machine Learning to Estimate Perceptions and Future Prescribing Intentions of Health Care Providers: A Novel Application of Integrated Primary and Secondary Data in Understanding HCP Decision-Making – Brittany Smith (Trinity Life Sciences)
• Key Considerations in the Use of Large Language Models for Data Extraction in Health Economics and Outcomes Research – Elise Aronitz (EVERSANA)
• GenAI for Critical Appraisal of Evidence for Systematic Literature Reviews (SLRs): A Face-Off between GenAI and Human Reviewers – Sheetal Sharma (ZS Associates)
• Integrating Artificial Intelligence (AI) and Machine Learning (ML) Techniques With Real-World Data (RWD) and Real-World Evidence (RWE) to Inform Precision Medicine: A Scoping Review – Alexandra Koumas (Axtria)
• Synthetic Patient Trajectories Using Generative AI on Electronic Health Records – Jimmy Royer (Analysis Group, Inc.)
• Large Language Models for Automatic PICO Criteria Evaluation – Lenon Mendes Pereira, (IQVIA)
• Enhancing Health Technology Assessment Accessibility: Using ChatGPT Prompts to Streamline Efficiency Frontiers, Cost-Effectiveness and Net Benefit Analyses – Bruno M Barros (Instituto Nacional de Cardiologia)
• Assessing the Effectiveness of Large Language Models in Automating Systematic Literature Reviews: Findings from Recent Studies – Sumeyye Samur (Value Analytics Labs)
• Repeatable Auto-extraction Frameworks in Clinical Systematic Literature Review: Validating a Multi-Model Human-in-the-Loop Artificial Intelligence system for Extracting Study PICOs, Location, Size, and Type –Kevin Kallmes (Nested Knowledge, Inc.)
• Performance Assessment and Validation of Real-World Response Data Generated Using a Deep Learning-Based Natural Language Processing Model Across Multiple Solid Tumors – Kelly Magee (Flatiron Health)

Coverage by The Evidence Base

The Evidence Base will be providing exclusive coverage of ISPOR 2025 (May 13–15, Montreal, Canada), including daily session highlights and in-depth features on all three plenaries. Register on The Evidence Base and follow us on LinkedIn to stay informed and up to date with the latest insights and developments from the conference.