Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Mindstretcher: blind leading blind

Placebo and acute pain
The answer

Continually challenging what constitutes good evidence is a good thing. It keeps us on our toes, it reminds us of our foibles, and we may have missed something. Given that the keystone of good evidence is the randomised, double blind trial, a challenge that claims them to be improperly conducted and open to bias is a serious matter.

This thought provoking assertion [1] comes from people with backgrounds in behavioural science and mathematical psychology. Hardly everyday knowledge, this, so it deserves our attention.


The argument goes something like this. Suppose doctors or nurses involved in a trial see that some patients are doing well, or badly. Even though the trial is randomised and properly blinded, their beliefs and expectations about the results of the trial are consequently influenced. Their beliefs could influence others concerned with the trial, or the patients. Subtle and even unconscious cues imparted to professionals or patients could influence professionals or patients in making assessments, or interpreting the results.

These unconscious feedbacks could also occur between patients in some circumstances, and true blinding could be lost. Since lack of blinding is known to impart bias, then the trial could become unblinded, even partially. The result would then be a trial in which the quality we seek from randomised double blind trials would be lacking.

For example, if we were comparing an analgesic and placebo in acute pain, a good analgesic might influence a nurse to encourage patients, while a poor analgesic might influence the same nurse to discourage them. A consequence of this would be for placebo responses to vary with the quality of the analgesic under test. Trials of good analgesics would have higher placebo responses than poor analgesics. It has been said that placebo is 55% as effective as the analgesic used [2].

Placebo and acute pain

Two common misconceptions are that a fixed fraction (one third) of the population responds to placebo, and that the extent of the placebo reaction is also a fixed fraction (again about one third of the maximum possible [3]). As Wall points out, these ideas stem from a misreading of Beecher's work of forty years ago [4].

When placebo controls are used in randomised, parallel-group, double-blind, trials in similar clinical settings measuring the same outcome over the same period they would be expected to give about the same overall placebo response. They do not, as Beecher's data showed [3].

The answer

It comes in two parts. Firstly, an apparent correlation between placebo effect and extent of response to analgesic results from using mean values of highly skewed data. When medians are used, the relationship disappears [5]. The second argument is one of size. In a comparison of about 12,000 patients given placebo in acute pain studies, 18% of patients had at least half pain relief over 4-6 hours. In over 50 meta-analyses of identical trials only when there were about 1,000 patients was this accurately measured [6]. With a smaller sample the variability was huge.


In one clinical area we can be reasonably sure that we are safe, and that leakage of blinding is unnecessary. But it is an important consideration that needs thinking about.


  1. MD Kirk-Smith, DD Stretch. Evidence-based medicine and randomized double-blind clinical trials: a study of flawed implementation. Journal of Evaluation in Clinical Practice 2001 7: 119-123.
  2. FJ Evans. The placebo response in pain reduction. In: Bonica JJ, ed. Advances in Neurology, Vol 4. New York: Raven Press; 1974:289-296.
  3. HK Beecher. The powerful placebo. JAMA 1955 159:1602-6.
  4. PD Wall. The placebo effect: an unpopular topic. Pain 1992 51:1-3.
  5. HJ McQuay et al. Variation in the placebo effect in randomised controlled trials of analgesics: all is as blind as it seems. Pain 1996 64:331-335.
  6. RA Moore. Understanding clinical trials: what have we learned from systematic reviews? Proceedings of the 9th World Congress on Pain, Progress in Pain Research & Management, Vol 16, Eds M Devor, MC Rowbotham, Z Weisenfeld-Hallin. IASP Press, Seattle, 2000, 757-770.
previous story in this issue