Primary Submission Category: Regression Discontinuity Designs
Nonparametric Regression Discontinuity Designs with Survival Outcomes
Authors: Maximilian Schuessler, Erik Sverdrup, Robert Tibsirani, Stefan Wager,
Presenting Author: Maximilian Schuessler*
Quasi-experimental evaluations are critical for generating real-world causal evidence and complementing insights from randomized trials. The regression discontinuity design (RDD) is a quasi-experimental framework for estimating causal effects when treatment assignment depends on a running variable crossing a threshold. Such threshold-based rules are ubiquitous in healthcare, education, policy and beyond. However, standard RDD estimators rely on complete outcome data, an assumption often violated in time-to-event analyses such as in healthcare where censoring arises from loss to follow-up. To address this issue, we propose a nonparametric approach that leverages doubly robust censoring corrections and can be paired with existing RDD estimators. Our approach can handle multiple survival endpoints, long follow-up times, and covariate-dependent variation in survival and censoring. We discuss the relevance of our approach across biomedical applications and demonstrate its usefulness through simulations and the PLCO Cancer Screening Trial for prostate cancer, a US-based phase III clinical trial with right-censored survival outcomes. Our results show that our approach is more robust to misspecification and yields higher precision than censoring corrections based on inverse probability of censoring weighting. We have also developed an open-source software package rdsurvival that enables estimation with existing RDD approaches in the R language.