Abstract:
|
Simon’s single-arm two-stage design is one of the most popular designs for phase II clinical trials aiming to test the efficacy of a new regimen or treatment. Based on this frequentist design, both Frequentists and Bayesians have been developing new single-arm two-stage designs. Recently, a hybrid approach appeared as a combination of the Bayesian and Frequentist (BF) frameworks for this type of design, where the frequentist type I, type II errors, and other quantities are redefined in a Bayesian way. But there has not been a widely accepted calculation method to derive all necessary quantities under the BF framework. We proposed new design algorithms under the BF framework and extended the 2-stage design to K-stage where K can be greater than 2. Based on the new algorithms, two single-arm multi-stage (SAMS) designs are proposed: the two-stage SAMS design and the extended K-stage SAMS design. The two-stage SAMS design reduces sample sizes needed compared with other two-stage designs; the K-stage SAMS design provides a more flexible design that can potentially reduce the sample sizes further.
|