A0536
Title: Bayesian estimation of the survivor average causal effect for cluster-randomized crossover trials
Authors: Dane Isenberg - University of Pennsylvania (United States) [presenting]
Michael Harhay - University of Pennsylvania (United States)
Fan Li - Yale University (United States)
Nandita Mitra - University of Pennsylvania (United States)
Abstract: In cluster-randomized crossover (CRXO) trials with a binary intervention, groups of individuals are assigned to one of two sequences of alternating treatments. Since clusters act as their own control, the CRXO design is typically more statistically efficient than the usual parallel-arm cluster-randomized trial. CRXO trials are increasingly popular in critical care studies where the number of available clusters is generally limited. In trials among severely ill patients, researchers often want to assess the effect of treatments on secondary non-terminal outcomes, but there may be several patients who do not survive to have these measurements fully recorded. A causal inference framework is provided to address truncation by death in the setting of CRXO trials. The survivor average causal effect (SACE) estimand is targeted, a well-defined subgroup treatment effect represented via principal stratification. Structural and standard modeling assumptions are proposed to enable SACE identification and estimation within a Bayesian paradigm. The small-sample performance of the proposed Bayesian approach for the estimation of SACE is evaluated using CRXO trial data through a simulation study. The methods are applied to a two-period cross-sectional CRXO study examining the impact of proton pump inhibitors as compared to histamine-2 receptor blockers on certain non-mortality outcomes among adults requiring invasive mechanical ventilation.
