Covariate-adaptive biased coin randomization for master protocols with multiple interventions and biomarker-stratified allocation.

In a multi-arm trial with predefined subgroups for each intervention to target, it is often desirable to enrich assignment to an intervention by enrolling more biomarker-positive participants to the intervention. We describe how to implement a biased coin design to achieve desired allocation ratios among interventions and between the number of biomarker-positive and biomarker-negative participants assigned to each intervention. We illustrate the proposed method with the randomization algorithm implemented in the Precision Interventions for Severe and/or Exacerbation-prone Asthma (PrecISE) trial.