🌍 Public RSVP

Augmenting randomized trials with failure-time outcomes using external data

abstract: Randomized trials provide credible evidence about treatment effects but are often limited by costly recruitment and follow-up. I present an extension of Hybrid Augmented Inverse Probability Weighting (H-AIPW) for trials with failure-time outcomes, including survival and other time-to-event endpoints. The approach uses randomization-aware estimators that exploit trial randomization to remain consistent and asymptotically normal, even when external data are misaligned with the trial. The construction allows flexible nuisance estimation using modern machine learning and AI methods, enabling external data to support richer representations of the covariate–outcome relationship. Simulations and an application to stroke-trial data show precision gains over trial-only estimators without introducing asymptotic bias or compromising confidence interval coverage. bio: Issa Dahabreh is Associate Professor of Epidemiology and Biostatistics at the Harvard T.H. Chan School of Public Health. His research develops methods for causal inference, evidence synthesis, and the use of trial and external data to improve the design and analysis of studies informing clinical and public health decisions.

Where

OAT, Room 11, ETH AI Center, Zurich

Organizer

ETH AI Center+ Follow

Source

Original listing on curated ↗

Who's going

~60 expected

Registration

🌍 Open to all

“Interested” saves it to your list and sharpens your recommendations. “Going” marks your spot and opens the organizer’s page so you can grab the actual ticket.