Blinding vs. Double Blind in Statistics: Definition, Examples

RCTs > Blinding in Statistics

What is Blinding in Statistics?

In medical trials, the term blinding, or double-blind, usually refers to the practice of keeping patients in the dark as to whether they receive a placebo or not. It can also refer to allocation concealment, which is used to avoid selection bias.

Double-blind vs. Blinding in Statistics

Blinding, or double-blinding, is when a patient does not know what treatment they receive. They could be getting either a placebo or the real drug. Blinding also refers to the practice of keeping the name of the treatment hidden. For example, patients might know they are in a trial for arthritis, but they will not know the name of the brand name in the trial. In “double-blind” trials, the clinical team— people involved in the patient’s management, the people collecting samples from the patient and the people analyzing the data also do not have any knowledge of whether the patient gets the placebo or not.

blinding in statistics
Placebos can ensure blinding in statistics.

the Importance of Blinding in Statistics

Blinding is especially important in subjective trials to avoid skewed results. For example, blinding is appropriate for pain relief studies. If a patient knows they are receiving a “real” drug, they will be more likely to report pain relief than those patients receiving a placebo. Blinding is less important in trials when there is a more objective criteria at stake, such as avoiding a patient’s death (for example, in cancer trials).

Blinding in Statistics: Double Dummy Method

The double dummy method studies a pair of medicines. For example, a combination drug for HIV treatment. Double dummy trials have two placebo drugs and two active drugs. The patients may be given a double placebo, or one placebo and one active drug of each type.

Blinding in Statistics: Allocation Concealment

Allocation concealment is the practice of not revealing to a patient their treatment allocation until the trial enrollment, to avoid selection bias. In other words, a clinician might ensure that certain patients (perhaps those in more pain) receive the “real” drug. The practice also keeps patients and the trial team unaware of upcoming assignments. Although the practice frequently annoys clinicians and their patients (see this NCBI article) allocation concealment ensures random distribution of treatments.

References

Everitt, B. S.; Skrondal, A. (2010), The Cambridge Dictionary of Statistics, Cambridge University Press.
Kotz, S.; et al., eds. (2006), Encyclopedia of Statistical Sciences, Wiley.


Comments? Need to post a correction? Please Contact Us.