ABSTRACT
When there are multiple promising new therapies in a disease setting, it may not be feasible to test all of them against the standard treatment in a definitive phase III trial. The sample sizes required for a phase III study with more than three arms could be prohibitive.1 In addition, the analysis can be highly complex and prone to errors due to the large number of possible comparisons in a multi-arm study. An alternative strategy is to screen the new therapies first in a phase II setting and choose one to test against a standard treatment in a simple two-arm phase III trial. Selection designs can be used in such circumstances.
