Abstract:
|
Subgroup analyses in early phase trials are frequently conducted, both in a prospectively planned but also in a post-hoc manner. In particular the latter situation poses challenges for decision makers in their decision whether to proceed a compound to the next phase of development or not. This presentation reviews and compares recent statistical tools in exploratory subgroup analyses to deal with the challenges posed by evaluating multiple models and subgroups, for example in terms of estimates and uncertainty intervals for the treatment effect of a selected subgroup.
|