Placebo and Sham Trials

Abstract

The traditional view of a randomised trial is the double- or triple-blind placebo drug trial in health care research. The basic outline of this study design is as follows. Participants are randomly allocated to receive an active or placebo drug. A placebo looks and tastes the same as an active drug. The use of placebo achieves several important theoretical goals. First, it ‘blinds’ or masks the participant, researchers and the clinicians to the group allocation. This reduces the chance of a Hawthorne effect differentially affecting one of the trial arms, which would bias the results. It avoids the resentful demoralisation of the control group, as they will not be aware they are taking an inactive treatment. Differential attrition bias may be less likely, as any attrition will be evenly spread across both groups (unless one treatment has side-effects). It also addresses performance bias, as participants are less likely to differentially seek alternative treatments to compensate for their allocation to the control arm. Finally, subversion bias is less likely, as concealment of allocation is more easily attainable than in a non-placebo trial.