Comparison of statistical and operational properties of subject randomization procedures for large multicenter clinical trial treating medical emergencies

Large multicenter acute stroke trials demand a randomization procedure with a high level of treatment allocation randomness, an effective control on overall and within-site imbalances, and a minimized time delay of study treatment caused by the randomization procedure. Driven by the randomization algorithm design of A Study of the Efficacy and Safety of Activase (Alteplase) in Patients with Mild Stroke (PRISMS) (NCT02072226), this paper compares operational and statistical properties of different randomization algorithms in local, central, and step-forward randomization settings. Results show that the step-forward randomization with block urn design provides better performances over others. If the concern on the potential time delay is not serious and a central randomization system is available, the minimization method with an imbalance control threshold and a biased coin probability could be a better choice.