ABSTRACT
In clinical research, clinical trials are usually conducted for evaluation of the efficacy and safety of a test drug as compared to a placebo control or an active control agent (e.g., a standard therapy) in terms of mean responses of some primary study endpoints. The objectives of the intended clinical trials usually include (i) the evaluation of the effect, (ii) the demonstration of therapeutic equivalence/non-inferiority, and (iii) the establishment of superiority. For evaluation of the effect within a given treatment, the null hypothesis of interest is to test whether there is a significant difference in mean response between preand post-treatment or mean change from baseline to endpoint. We refer to this testing problem as a one-sample problem. For establishment of the efficacy and safety of the test drug, a typical approach is to test whether there is a difference between the test drug and the placebo control and then evaluate the chance of correctly detecting a clinically meaningful difference if such a difference truly exists. Thus, it is of interest to first test the null hypothesis of equality and then evaluate the power under the alternative hypothesis to determine whether the evidence is substantial for regulatory approval. For demonstration of therapeutic equivalence/non-inferiority and/or superiority as compared to an active control or standard therapy, it is of interest to test hypotheses for equivalence/non-inferiority and/or superiority as described in Chapter 1. In this chapter, under a valid design (e.g., a parallel design or a crossover design), methods for sample size calculation are provided to achieve a desired power of statistical tests for appropriate hypotheses.
