On Two-Stage Adaptive Seamless Design with Count Data from Different Study Durations under Weibull Distribution

In clinical development, a two-stage design combining two separate studies (e.g., a phase II dose finding study and a phase III confirmatory study) into a single trial is commonly considered. The purpose of a two-stage design is not only to reduce lead time between the two studies, but also to evaluate the treatment effect in a more efficient way. In practice, one of the difficulties in utilizing a two-stage design is that the study endpoints at different stages may be different. For example, a biomarker (or the same study endpoint with different duration) may be considered at the first stage, while a regular study endpoint is used at the second stage. As per the studies the case where both study endpoints are continuous variables with certain correlation structure. In this paper, our attention is on the case where the study endpoints are count data which are obtained at the two stages with different time intervals. Statistical procedure for combining data observed from the two different stages are proposed. Furthermore, results on hypotheses testing and sample size calculation are derived for the comparison of two treatments based on data observed from a two-stage design.