Sample size and screening size trade-off in the presence of subgroups with different expected treatment effects.

When designing clinical trials, sometimes we may expect a larger treatment effect in one group while others exhibit an attenuated effect. In these settings there can be a trade-off between a smaller average treatment effect with broader enrollment criteria and a larger effect with restricted criteria but longer enrollment duration. Identification of subgroups will often use a clinical decision rule, for example, biomarker cutoff, but may be imprecise, that is, with sensitivity and specificity not simultaneously 100%. We evaluate the impact of including attenuated subgroups on design operating characteristics and illustrate scenarios where overall trial duration may be shorter by not being restrictive.