|
Abstract:
|
Interim assessment of accumulating data from ongoing clinical trials is performed for various reasons, including stopping for futility. Several methods have been developed to assess futility, including conditional power and predictive power calculations, but these are focused on one univariate outcome variable only. However, some clinical trials may need to consider two correlated endpoints for futility assessment at the interim analysis. Such situations are: 1) trials with co-primary endpoints that are correlated, 2) futility assessment at the interim analysis based on a surrogate marker (or biomarker) which is a secondary endpoint. In this talk, we consider extensions of the existing methods of conditional power and predictive power for futility assessment to situations where two correlated endpoints should be considered for futility, and apply to a clinical trial example.
|
