Blinding (medicine)

In medicine, blinding is used in clinical trialsof treatment or other interventions to increase the reliability of the result, by reducing the potential for bias, especially observer bias. A blind trial implies that one or more of the humans involved in a trial do not know what intervention, or lack of intervention, has been given.

There are three levels of blinding: single, double and triple blind. In a single blind trial, the subject of the trial does not know whether he is being given the intervention or a control, often a placebo. Single blind trials are used when, for practical reasons, it is impossible for the clinician not to know what intervention has been given. Likewise, if the subject knows what intervention he has been given but the clinician does not, this is also known as a single blind trial.

In double-blind trials, neither the subject of the intervention nor the clinician knows which intervention has been given.

Triple-blind trials are those where the statistician interpreting the results does not know which intervention has been given; as well as the subject and clinician. Sometimes triple-blinded is used to mean when a second clinician is also blinded, such as a radiologistor pathologistwho interprets the results of medical investigations.

Of course, somebody, somewhere must know which intervention has been given otherwise it would be impossible to interpret the results.

Effective blinding can be difficult to achieve where the treatment is notably effective and is being compared to placebo, or where the treatment is very distinctive in taste or the incidental effects. (Consider a trial of coronary artery bypass surgery, where a general anaesthetic might be ruled ethical, but a thoracotomy incision hopefully would not.) A good trial will forsee these potential problems to ensure blinding is as effective as possible.

Randomisationis an essential part of clinical trials and integral to the process of blinding. Medical ethicsdemands that both the clinician and the patient (if capable) can withdraw from the trial at any time. A clinician may hold, consciously or unconsciously, the view that the intervention (or control) is not beneficial to a particular group of people. Effective randomisation minimises the risk that the clinician will introduce bias by excluding a particular group of patients, as the clinician will not know whether they are to be given the intervention or control.

Evidence-based medicinepractitioners prefer blinded randomised controlled trials(RCTs), where that is a possible experimental design. These are high on the hierarchy of evidence; only a meta analysisof several well designed RCTs is considered more reliable.

This article is licensed under the GNU Free Documentation License.
It uses material from the http://en.wikipedia.org/wiki/Blinding+%28medicine%29 Wikipedia article Blinding (medicine).