Online Program

Return to main conference page

All Times EDT

Wednesday, September 22
Wed, Sep 22, 2:15 PM - 3:30 PM
Virtual
Statistical Design and Considerations in the Evaluation of Subsequent Therapy in Time-to-Event Analysis

Estimand and Estimation in Time-to-Event Analyses (303540)

*Yongming Qu, Eli Lilly and Company 

ICH E9(R1) Addendum has provided a framework of estimands and analyses, including how to handle intercurrent events. Many discussions on the guidance have occurred. Examples shared in literature mostly focus on continuous variables. In this presentation, we will use the ICH E9(R1) guidance to define various estimands for time-to-event variables using the potential outcome introduced from causal inference according to the study objectives. Then, we also discuss a generic imputation method that can be used to impute the missing data either due to censoring or because of using a hypothetical strategy. The performance of the imputation will be evaluated through simulation studies. The method is also illustrated using an example from a cardiovascular study.