Structural Cumulative Survival Models for Robust Estimation of Treatment Effects Accounting for Treatment Switching in Randomized Experiments
share
We propose an instrumental variable estimator to estimate the treatment causal effect when treatment switching is present in a randomized experiment, under a structural cumulative survival model. Our estimator is robust to violation of the exclusion restriction, a commonly adopted assumption for IV methods that is untestable and usually subject to questioning in practice, especially in an open-label randomized trial. We derive large-sample properties of our proposed estimator, along with inferential tools. We apply the estimator to evaluate the treatment effect of Nucleoside Reverse Transcriptase Inhibitors on a safety outcome in the Optimized Treatment That Includes or Omits NRTIs trial.
READ FULL TEXT