Dynamic randomization, often called minimization, assigns each new participant to the arm that best preserves balance across the prognostic factors you defined. The system simulates every possible assignment, calculates an imbalance score, then uses a biased coin to favor the arm that reduces imbalance while keeping the next allocation unpredictable. It follows the Pocock and Simon family of methods used across modern RTSM systems.
🔑 Prerequisite: the permission to configure randomization, and five decisions already made.
Your treatment arms, for example Treatment A and Treatment B.
Your allocation ratio, for example 1:1 or 2:1.
The randomization method: Static, Dynamic or Import.
The randomization type: Open label or Double blind.
Any stratification factors, for example center, gender or age range.
Turn randomization on for the study
🛠️ Step by step guide
Open your study.
Go to Study settings, open the RTSM tab, and activate Randomization.
Create and configure the randomization
🛠️ Step by step guide
Go to the RTSM tab in the study menu.
Create your randomization and give it a name. That name appears in the eCRF Builder and in trial supply management when configuring allocation.
Set up the arms: name each one and define the ratio between them, for example 1:1 or 2:1. Each arm gets a unique ID, generated automatically, usable in the eCRF.
Set the method to Dynamic, then select the distance method (Range, Variance or Maximum) and the probability.
Set the randomization type: display the arms for Open or Single Blind, hide them for Double Blind.
Define the scope and stratification: per site for independent sequences per site, or per stratum if you use stratification factors.
Add the randomization action to the eCRF
🛠️ Step by step guide
In the eCRF Builder, open the form where the action should sit.
Click Add question and choose # Randomization.
Rename the button label if you want to.
Save.
Test it before you go live
Dynamic randomization is worth testing harder than static, because the allocation depends on who came before. A single test subject tells you almost nothing.
🛠️ Step by step guide
In your draft study, generate a new test eCRF.
Click the Randomize action button.
Confirm by selecting Randomize.
Repeat with several test subjects, varying the stratification answers, so you can watch the algorithm balancing.
💡 Tip: if the configuration is not right, go back to the randomization settings, adjust and save. Saving a change resets the randomization to zero, so your tests start again from the beginning. Once you are satisfied, stop changing it and finalize the study for live deployment.