Estimand

The ICH E9 (R1) addendum, introduced in 2019, implements statistical thinking in clinical trial by providing a structured framework to define estimands, enhancing collaboration between clinicians and statisticians, and facilitating regulator-sponsor interactions. This chapter explicates the practical application of the estimand framework in Clinical Outcome Assessments (COA), with a focus on addressing intercurrent events that could impact the interpretability of clinical trials. While ICH E9 (R1) lists common intercurrent events, it does not exhaustively cover events such as noncompliance and decisions by patients or physicians, which are also pivotal. The chapter delves into the nuances of choosing appropriate strategies for such events, covering the hypothetical and principal stratum strategies, particularly relevant to pediatric and rare diseases. It also examines the application of estimands in Bayesian approaches and the intricacies of population-level summaries. By discussing the determination of intercurrent events, defining treatment effects, and the selection of suitable estimand strategies, this chapter advances a comprehensive understanding of the estimand framework’s application in clinical trial design and analysis, ultimately aiming to refine the estimation of treatment effects and facilitate regulatory approval processes.