The Direct Assignment Option as a Modular Design Component: An Example for the Setting of Two Predefined Subgroups

Figure 1

Distributions of observed treatment differences in the M− subgroup at interim analysis, across 500 simulated trials, using control response rate uniformly distributed over and response rate ratio for treated versus control arms uniformly distributed over . Sample sizes in the first stage, , 16, and 48. Proportion of trials where observed treatment difference exceeds in absolute value is 67%, 32.8%, and 13%, respectively.