|
Abstract:
|
We consider evaluation of two identical pivotal trials with multiple endpoints. To obtain approval, the same endpoint must be significant in both studies. We consider the fixed sequence, weighted Holm and fallback procedures, with the intention of distinguishing which is preferable in given situations. No procedure is universally optimal. The best procedure depends on the power for the various endpoints, something that is never known with certainty. The alpha exhaustive version of the fallback procedure often has higher power for some endpoints and lower power for other endpoints compared to the weighted Holm procedure, but the differences are rarely large. Correlation among endpoints has different impact on the fixed sequence than the other procedures. The fallback procedure is rarely the best choice, but can be a compromise when the true power for the various endpoints is unclear.
|
