Randomization in clinical trials Edit

There are two processes involved in randomizing patients to different interventions. First is choosing a randomization procedure to generate a random and unpredictable sequence of allocations. This may be a simple random assignment of patients to any of the groups at equal probabilities, or may be complex and adaptive. A second and more practical issue is allocation concealment. which refers to the stringent precautions taken to ensure that the group assignment of patients are not revealed to the study investigators prior to definitively allocating them to their respective groups.

Randomization procedures Edit

There are a couple of statistical issues to consider in generating the randomization sequences [1].

Balance. since most statistical tests are most powerful when the groups being compared have equal sizes, it is desirable for the randomization procedure to generate similarly-sized groups. Selection bias . depending on the amount of structure in the randomization procedure, investigators may be able to infer the next group assignment by guessing which of the groups has been assigned the least up to that point. This breaks allocation concealment (see below) and can lead to bias in the selection of patients for enrollement in the study. Accidental bias. if important covariates that are related to the outcome are ignored in the statistical analysis, estimates arising from that analysis may be biased. The potential magnitude of that bias, if any, will depend on the randomization procedure.

Complete randomization Edit

In this commonly used and intuitive procedure, each patient is effectively randomly assigned to any one of the groups. It is simple and optimal in the sense of robustness against both selection and accidental biases. However, its main drawback is the possibility of imbalances between the groups. In practice, imbalance is only a concern for small sample sizes (n < 200).

Permuted block randomization Edit

In this form of restricted randomization, blocks of k patients are created such that balance is enforced within each block. For instance, let E stand for experimental group and C for control group, then a block of k = 4 patients may be assigned to one of EECC. ECEC. ECCE. CEEC. CECE. and CCEE. with equal probabilities of 1/6 each. Note that there are equal numbers of patients assigned to the experiment and the control group in each block.

Permuted block randomization has several advantages. In addition to promoting group balance at the end of the trial, it also promotes periodic balance in the sense that sequential patients are distributed equally between groups. This is particularly important because clinical trials enroll patients sequentially, such that there may be systematic differences between patients entering at different times during the study.

Unfortunately, by enforcing within-block balance, permuted block randomization is particularly susceptible to selection bias. That is, since toward the end of each block the investigators know the group with the least assignment up to that point must be assigned proportionally more of the remainder, predicting future group assignment becomes progressively easier. The remedy for this bias is to blind investigator from group assignments and from the randomization procedure itself.

Strictly speaking, permuted block randomization should be followed by statistical analysis that takes the blocking into account. However, for small block sizes this may become infeasible. In practice it is recommended that intra-block correlation be examined as a part of the statistical analysis.

A special case of permuted block randomization is random allocation. in which the entire sample is treated as one block.

Covariate-adaptive randomization Edit

Outcome-adaptive randomization Edit

For a randomized trial in human subjects to be ethical, the investigator must believe before the trial begins that all treatments under consideration are equally desirable. At the end of the trial, one treatment may be selected as superior if a statistically significant difference was discovered. Between the beginning and end of the trial is an ethical grey zone. As patients are treated, evidence may accumulate that one treatment is superior, and yet patients are still randomized equally between all treatments until the trial ends.

Outcome-adaptive randomization is a variation on traditional randomization designed to address the ethical issue raised above. Randomization probabilities are adjusted continuously throughout the trial in response to the data. The probability of a treatment being assigned increases as the probability of that treatment being superior increases. The statistical advantages of randomization are retained, while on average more patients are assigned to superior treatments.

Allocation concealment Edit

In practice, in taking care of individual patients, clinical investigators often find it difficult to maintain impartiality. Stories abound of investigators holding up sealed envelopes to lights or ransacking offices to determine group assignments in order to dictate the assignment of their next patient [2]. This introduces selection bias and confounders and distorts the results of the study. Breaking allocation concealment in randomized controlled trials is that much more problematic because in principle the randomization should have minimized such biases.

Some standard methods of ensuring allocation concealment include:

Sequentially-Numbered, Opaque, Sealed Envelopes (SNOSE) Sequentially-numbered containers Pharmacy controlled Central randomization

Great care for allocation concealment must go into the clinical trial protocol and reported in detail in the publication. Recent studies have found that not only do most publications not report their concealment procedure, most of the publications that do not report also have unclear concealment procedures in the protocols [3] .