In his landmark paper on the power of the placebo, Beecher found that the number of patients who responded to a placebo varied between 15 and 53% (Beecher 1955). Other investigators examining such various diseases as headaches, low back pain, and angina have even reported response rates higher than 50%. The oft-cited statement that the response rate to placebo is 30% likely derives from the average of Beecher's original observations.
These figures, however, represent the average of many individual placebo responses and do not indicate how each member of the group responds. One might imagine all members of the group responding equally well or in contrast, some members responding extremely well, and other members not responding at all, with a group response average of 30%. Levine et al. demonstrated this concept in a study of pain following tooth extraction. When given placebo medication, he found that 39% of the patients had some response to the placebo while 61% had no response at all (Levine et al. 1979). Thus, he was able to categorize individual patients as "placebo responders" or "placebo non-responders."
Predicting which individuals would respond to placebo administration becomes important, but this information is difficult to identify. Various studies have determined that intelligence or susceptibility poorly predicts the response to placebo. Furthermore, gender has been shown to be a poor predictor of placebo response, and there have been varied results in attempting to link personality traits with placebo response. In addition, people who respond to placebo in one setting may not respond in another setting (Oken 2008, Harrington 1997). However, adherence to a placebo regimen has been shown to be predictive of high placebo response (Horwitzet al. 1990).
Was this article helpful?