University of California Los AngelesUniversity of North Carolina at Chapel HillAlexander BreskinApproval & CommercializationYesOral PresentationR&D
Fusion designs and estimators for treatment effects.
Fusion designs and estimators for treatment effects
Challenge
When trials compare new treatments to standards of care rather than to placebo, direct randomized comparisons are often unavailable. No formal data fusion estimator had been developed for this bridged treatment comparison problem.
Solution
Target RWE researchers and UNC collaborators developed and published a formal data fusion estimator for bridged treatment effects in Statistics in Medicine, combining data from two studies with shared arms using inverse probability weighting.