Randomization
Randomization helps you avoid potential bias in the selection and allocation of subjects arising from the predictability of treatment assignments. From a statistical perspective, randomization provides unbiased estimates of error variance and independence of errors.
Randomization configuration and management in IWRS depends on the randomization method you select upon configuring cohorts of your study. There are two available methods of randomization in IWRS:
-
Permuted Block (PB): this randomization method involves the preparation of a randomization list before initiating the randomization process for the study. Randomization of subjects is first performed into blocks according to preset stratification factors and then into arms according to specific ratios.
An example of when this method is most convenient is the study with a large group of patients screened and their data collected before the randomization phase starts. This way, you efficiently randomize the subjects into groups (arms) and immediately achieve balance.
-
Dynamic: this randomization method adjusts the allocation probabilities based on the current distribution of participants. Thus, when each patient comes to the study, the system assesses their compliance with established factors, evaluates the existing patient groups and site imbalance, and allocates the new patient to the most imbalanced arm. Unlike permuted block randomization, dynamic randomization does not involve the predefined randomization list and offers more adaptivity to achieve arm balance at all times.
An example of when this method is most convenient is the study for which you cannot form the groups in advance and want to achieve arm balance at all stages of the randomization phase. It allows patients to enter the study individually considering the current state of groups and treatments.
Another scenario for dynamic randomization can be a study that uses multiple randomization stratification factors but a small sample size. In such a case, dynamic randomization can better ensure the balance of subjects' distribution by strata.