Restricted Randomized Experiment
An Restricted Randomized Experiment is a two-phase randomization experiment that first applies a restricted randomization procedure.
- Context:
- It can range from being a Stratified Randomized Experiment to being an Stratified Non-Randomized Experiment.
- It can range from being a Subject-level Restricted Randomized Experiment to being an Group-level Restricted Randomized Experiment.
- Example(s):
- a comparative experiment that first separates members into male/female blocks.
- …
- Counter-Example(s):
- See: Sampling.
References
2013
- http://en.wikipedia.org/wiki/Randomized_controlled_trial#Restricted_randomization
- To balance group sizes in smaller RCTs, some form of "restricted" randomization is recommended. The major types of restricted randomization used in RCTs are:
- Permuted-block randomization or blocked randomization: a "block size" and "allocation ratio" (number of subjects in one group versus the other group) are specified, and subjects are allocated randomly within each block. For example, a block size of 6 and an allocation ratio of 2:1 would lead to random assignment of 4 subjects to one group and 2 to the other. This type of randomization can be combined with "stratified randomization", for example by center in a multicenter trial, to "ensure good balance of participant characteristics in each group." A special case of permuted-block randomization is random allocation, in which the entire sample is treated as one block. The major disadvantage of permuted-block randomization is that even if the block sizes are large and randomly varied, the procedure can lead to selection bias. Another disadvantage is that "proper" analysis of data from permuted-block-randomized RCTs requires stratification by blocks.
- Adaptive biased-coin randomization methods (of which urn randomization is the most widely known type): In these relatively uncommon methods, the probability of being assigned to a group decreases if the group is overrepresented and increases if the group is underrepresented. The methods are thought to be less affected by selection bias than permuted-block randomization.
- To balance group sizes in smaller RCTs, some form of "restricted" randomization is recommended. The major types of restricted randomization used in RCTs are:
1999
- (Kernan et al., 1999) ⇒ Walter N Kernan, Catherine M Viscoli, Robert W Makuch, Lawrence M Brass, and Ralph I Horwitz. (1999). “Stratified Randomization for Clinical Trials." Elsevier. doi:10.1016/S0895-4356(98)00138-3
- QUOTE: Simple randomization can fail in two important ways. First, it can fail to assign the compared treatments to equal numbers of patients. To avoid this chance imbalance, investigators commonly randomize patients in permuted blocks [8]. A block usually comprises four or six randomly ordered treatment assignments. Within each block, equal numbers of patients receive each treatment. For example, if two treatments, A and B, are being compared, the permuted blocks of four treatment assignments would be AABB, ABAB, BBAA, BABA, ABBA, and BAAB. As patients enter the trial, they receive the next treatment in the current block. Patients are randomized within block after block until the study is complete. With permuted blocks of size four, the number of patients assigned to each treatment within a block can never differ by more than two.
Second, simple randomization can fail if it creates treated groups of patients that are unbalanced for critical features that are known or suspected to affect prognosis. Significant imbalance is most likely to happen in small trials where chance may result in sicker patients being in one treatment arm than in another. It may also occur for large trials at the time of interim analyses when the full sample size has not been obtained. Finally, it may occur for trials with prolonged recruitment periods, if changes occur over time in prognostic characteristics of enrolled patients.
To illustrate the chance that simple (unstratified) randomization may lead to treatment groups that are unbalanced with respect to a prognostic factor, consider a trial of two therapies in a disease with an important prognostic factor that is present in 15% of patients. The chance that the two treatment groups will differ by more than 10% for the proportion of patients with the prognostic factor is 33% for a trial of 30 patients, 24% for a trial of 50 patients, 10% for a trial of 100 patients, 3% for a trial of 200 patients, and 0.3% for a trial of 400 patients (Table 1). The chance of an imbalance is greater when prognostic factors are present in 30% of patients than when they are present in 15%.
- QUOTE: Simple randomization can fail in two important ways. First, it can fail to assign the compared treatments to equal numbers of patients. To avoid this chance imbalance, investigators commonly randomize patients in permuted blocks [8]. A block usually comprises four or six randomly ordered treatment assignments. Within each block, equal numbers of patients receive each treatment. For example, if two treatments, A and B, are being compared, the permuted blocks of four treatment assignments would be AABB, ABAB, BBAA, BABA, ABBA, and BAAB. As patients enter the trial, they receive the next treatment in the current block. Patients are randomized within block after block until the study is complete. With permuted blocks of size four, the number of patients assigned to each treatment within a block can never differ by more than two.