Click to search

Powerful spin in the conclusion of Wampold et al.'s re-analysis of placebo versus no-treatment trials despite similar results as in original review.


Hróbjartsson A1, Gøtzsche PC.
Author information
  • 1Hróbjartsson, The Nordic CochraneCentre, Rigshospitalet, Department 7112, Blegdamsvej 9, DK 2100 Copenhagen Ø,


J Clin Psychol. 2007 Apr;63(4):373-7.


Comment in

Comment on


B. E. Wampold, T. Minami, S. C. Tierney, T. W. Baskin, and K. S. Bhati (2005) re-analyzed trials included in our systematic review of randomized clinical trials that compared placebo with no treatment (A. Hróbjartsson & P. C. Gøtzsche, 2001). Based on 11 trials, B. E. Wampold et al. concluded that " ... the placebo effect is robust" (p. 850). We (2001) concluded, based on 130 trials, that "we found little evidence in general that placebos have powerful clinical effects" (p. 1599). In this commentary, we examine the reasons for this discrepancy. For trials with continuous outcomes, our analyses (82 trials) and that of B. E. Wampold et al. (5 trials) resulted in pooled standardized mean differences that were small and essentially identical: -0.28 (95% confidence interval = -0.38 to -0.19) versus -0.29 (95% confidence interval = -0.52 to -0.06). There was considerable risk of bias (e.g., reporting bias, sample-size bias). Similarly, for trials with binary outcomes, our analysis (32 trials) and that of B. E. Wampold et al. (6 trials) found no statistically significant pooled effect of placebo interventions and were essentially identical: relative risk 0.95 (95% confidence interval = 0.88-1.02) versus odds ratio 0.99 (95% confidence interval = 0.81-1.23). Thus, B. E. Wampold et al.'s conclusion was not substantiated by their data, and is best characterized as powerful spin. .


17279532 [PubMed - indexed for MEDLINE]
John Wiley & Sons, Inc.: Full text
 Citation 5 of 118 Back to results 
Standard PubMed