🤖 AI Summary
This study addresses the estimation of state occupation probabilities in multistate processes when right censoring and baseline exposure coexist. Under a coarsening-at-random assumption, the proposed approach abandons the conventional Markov assumption and accommodates time-varying confounding. It introduces an augmented inverse probability weighting (AIPW) semiparametric estimator grounded in causal inference and missing data methodology. The method enjoys double robustness and statistical efficiency, ensuring reliable performance even under complex coarsening mechanisms. Both theoretical analysis and simulation studies demonstrate that the estimator yields robust and efficient inference for state occupation probabilities in non-Markov multistate processes.
📝 Abstract
We derive augmented inverse probability weighted estimators for occupation probabilities of multistate models under two levels of coarsening; right-censoring and baseline exposure. The key exchangeability assumption for identification is coarsening at random, while allowing for time-varying confounders, but not requiring Markov properties. Using existing techniques from causal inference and missing data literature, the derived estimators have highly desirable robustness and efficiency properties. These properties are demonstrated through both theoretical results, and a simulation study.