An estimator for treatment comparisons among survivors in randomized trials

In clinical trials of advanced-stage disease it is often of interest to perform treatment comparisons for endpoints which are defined only for survivors. Examples include time on ventilation in ventilation studies, change in quality of life in health-related quality-of-life studies, and duration of response to therapy in therapeutic trials. Randomized treatment comparisons for these endpoints cannot be performed because the outcomes are only defined in the nonrandomly selected subgroup of survivors. We propose a new evaluation of the survivor average causal effect (SACE) for treatment comparisons of this nature. We provide an estimator of SACE in the presence of no unmeasured confounders, a nontestable assumption, which identifies SACE. We also outline a sensitivity analysis for exploring robustness of conclusions to deviations from this assumption. We apply this method to duration of ventilation in a clinical trial of acute respiratory distress syndrome.