Abstract:
|
The Signature Design of Freidlin and Simon splits the trial data into two stages. The data from the first stage is used to identify a potential signature for sensitive patients. On average, this design reduces the chance of false rejection over a standard design. However, it is not clear whether the result is driven by the effect in the sensitive subgroup. Modified approaches are proposed in the talk to better differentiate the full population and sensitive subgroup effects.
|