What are the rules for randomization designs?

You can create blinded studies with or without open-label extensions. You can also create open-label studies, as long as randomization occurs when the open-label period starts. A study can have one or more randomization events.

Rules for randomization designs

• You must drag randomization designs to visits in order (drag the first randomization design first, the second randomization design second, and so on).
• The first randomization design must randomize subjects (that is, you cannot start with an open-label period that doesn't involve randomization).
• The first randomization design in a study must have No selected for the Re-Randomization field.
• Only one randomization design in the study can randomize into cohorts.
• If a randomization design randomizes subjects into cohorts, it must be the first randomization design in the study.
• The second and later randomization designs must have Yes selected for the Re-Randomization field.
• If you are stratifying subjects:
  • You can use only one stratified randomization design in a study.
  • A stratified randomization design can be used only as the first randomization design.
• A randomization design must be assigned to a required and scheduled visit.
• A randomization design can't be assigned to the study completion or withdrawal visits.

Options for a second randomization event

If you randomize a second time, you have the following options for the second randomization event:

• Randomize all subjects again.
• Map subjects to new treatment arms, with or without randomization.

  For example, consider a study with 3 treatment arms: 5 mg, 10 mg, and placebo. Subjects in the 5 mg and 10 mg arms remain in their treatment arms throughout the study, and subjects in the placebo arm are randomized to determine whether they move into either the 5 mg or 10 mg arms.

  More variations are possible. For instance, the study can add or drop treatment arms, and subjects can be mapped or randomized into a new treatment arm.

Tip:

If you're doing a crossover study where all subjects swap treatment arms without being randomized, you don't need to create a second randomization design. Instead, you can simply update the dispensation schedule so that subjects receive the appropriate treatment for each visit.