Abstract:
|
In a pivotal registration clinical trial, key secondary endpoints may be formally tested in order for additional claims or useful information from these endpoints to be included in the product label. This talk considers multiplicity adjustment to control the familywise error rate (FWER) testing both primary and key secondary endpoints when designing a Phase III confirmatory oncology clinical trial with three treatment arms. A multistage stepwise parallel gatekeeping procedure with truncated Hochberg test is utilized to control the overall FWER, where families of hypotheses for the primary and secondary endpoints are tested in a hierarchical order and each family serves as a parallel gatekeeper to testing subsequent families. The impact of choosing multiplicity adjustment strategies on sample size and statistical power for both primary and secondary endpoints is examined and evaluated through statistical simulations.
|