Background
ICH E9 (R1) definition of estimands
-
Treatment condition, and as appropriate, the alternative treatment condition
-
Population of patients targeted (e.g. patients with type 2 diabetes)
-
Variable (or endpoint) obtained for each patient to address the clinical question (e.g. change from baseline to week 26 in HbA1C)
-
Population-level summary for the variable, providing a basis for comparison between treatment conditions (e.g. difference in means in change from baseline to week 26 in HbA1C)
Strategy | Relevant values for patients with the intercurrent event |
---|---|
Treatment policy | Actual values of the variable regardless of whether the intercurrent event has occurred |
Composite | Modified definition of the variable incorporating the intercurrent event |
Hypothetical | Values the variable would have taken in the hypothetical scenario envisaged in which the intercurrent event would not occur |
While on-treatment/at-risk | Values of the variable up to the time of the intercurrent event |
Principal stratum | Restrict population to patients who would not experience the intercurrent event |