1: \begin{abstract}

2: Clinical studies often encounter truncation by death, which may render certain outcomes undefined. Statistical analysis based solely on observed survivors may lead to biased results because the characteristics of survivors may differ between treatment groups. Under the principal stratification framework, a meaningful causal parameter called the survivor average causal effect can be defined in the always-survivors subpopulation. This causal parameter may not be identifiable in observational studies, where treatment assignment and the survival or outcome processes are confounded by unmeasured factors. In this paper, we propose a novel method for handling unmeasured confounding when outcomes are truncated by death. First, a novel method is proposed to identify the heterogeneous survivor average causal effect based on a substitutional variable. Second, under additional assumptions, the survivor average causal effect on the overall population is identified. Furthermore, we consider estimation and inference for the conditional survivor average causal effect based on parametric and nonparametric methods with asymptotic properties. Good finite-sample properties are demonstrated through simulations and sensitivity analyses. The proposed method is applied to investigate the effects of allogeneic stem cell transplantation on leukemia relapse.

3: 

4: {\bf Keywords:} Causal inference; principal stratification; substitutional variable; truncation by death; unmeasured confounding.

5: \end{abstract}

6: