Observing the unobserved confounding through its effects: toward randomized trial-like estimates from real-world survival data
arXiv:2604.12137v1 Announce Type: cross
Abstract: Background: Randomized controlled trials (RCTs) are costly, time-consuming, and often infeasible, while treatment-effect estimation from observational data is limited by unobserved confounding.
Metho…