Surrogate endpoints increasingly underpin regulatory approvals, accounting for around 60% of new FDA drug and biologic approvals over the past two decades. Yet robust validation is often lacking, which leaves HTA agencies and payers deciding under considerable uncertainty. In this webinar, Oriana Ciani (Bocconi University) and Dawn Lee (University of Exeter), co-authors of the ISPOR Good Practices Task Force Report (Value in Health, 2026), will present its recommendations. They will cover the three levels of evidence for surrogacy (biological plausibility, individual-level association and trial-level association) and the statistical methods used to assess them. These include multivariate meta-analysis that accounts for within-study correlation, Bayesian approaches that borrow information across treatment classes and indications, and the role of real-world data. The second part will address how surrogate relationships should be built into health economic models, how to assess plausibility, and how to characterise parameter and structural uncertainty through probabilistic, scenario, threshold and value of information analyses. Discussion will follow. The target audience is anyone interested in medical statistics and health care decision making.
 
This presentation will be preceded by the Royal Statistical Society Medical Section AGM 2026.