ABSTRACT
In clinical research, primary clinical endpoints for evaluation of the treatment effect of the compound under study could be discrete variables, for example, clinical response (e.g., complete response, partial response, and stable disease), survival in cancer trials, and the presence of adverse events in clinical trials. For evaluation of treatment effect based on discrete clinical endpoint, the proportions of events that have occurred between treatment groups are often compared. Under a given study design, statistical tests for specific hypotheses such as equality or equivalence/non-inferiority can be carried out based on the large sample theory in a similar manner as continuous responses discussed in Chapter 3. In this chapter, our primary focus will be placed on comparing proportions between treatment groups with binary responses.
