Estimation for optimal treatment regimes with survival data under semiparametric model

Abstract

In this paper, we consider a semiparametric model to find the optimal treatment regimes. A-learning type equation method is proposed to construct a doubly robust estimating equation for the parameters of interest in the optimal treatment. To overcome bias from the censoring time, we consider the inverse probability censoring weighting method in estimating equation. The resulting estimator is shown to be consistent and asymptotic normal when either the baseline effect model for covariates or the propensity score is correctly specified. Also, numerical simulations and an application with real data illustrate the proposed method.

Keywords:

• Semiparametric model
• doubly robust estimator
• A-learning estimating equations
• inverse probability weighting

Additional information

Funding

Zhou’s work was supported by the State Key Program of National Natural Science Foundation of China (71931004), the State Key Program in the Major Research Plan of National Natural Science Foundation of China (91546202).