All Times EDT
Keywords: ITT, causal inference, principal stratum, Tripartite Approach, direct treatment effect, ICH-E9(R1)
The recently released ICH-E9 addendum on estimands and sensitivity analysis, ICH-E9(R1), has opened the door to considerations beyond ITT for clinical trials, including pivotal trials, for the primary estimand of a treatment effect. One approach that has gained some interest and attention is the Tripartite Approach (Akacha, Bretz and Ruberg, 2017). In the Tripartite Approach, one estimand is the effect of treatment in the principal stratum of patients who can adhere to their study medication. Qu et al. (2020) have addressed the estimation of this estimand with two estimators using causal inference approaches. Those estimators are consistent for what they have called the Adherer Average Causal Effect (AdACE).
ITT is an estimand related to the initiation of treatment and considers intercurrent events irrelevant to the estimation, including cessation of study treatment and the use of rescue medication. The Tripartite Approach is aimed at the direct treatment effect; that is, the effect of treatment when taken as prescribed. Arguments have been made that the Adherer Average Causal Effect (AdACE) may be more important, relevant, and understandable to patients and physicians. Furthermore, the impact of these estimands, estimators and estimates on product labeling will be considered. In the Roundtable Discussion, the participants will be challenged to think about different disease states or trial designs in order to compare and contrast ITT and the Tripartite Approach with a guarantee of insights and learnings for all!