Effect of information on reported adverse events in a placebo-controlled trial

Although placebo controls are a standard measure in clinical trials the mechanisms underlying placebo effects are still not fully understood. We hypothesised that information about the likelihood of receiving placebo might influence the perception of adverse effects in volunteers participating in a clinical trial.

Healthy subjects received either nifedipine 20 mg or placebo in an adaptive two-stage crossover study. Sixty subjects were randomised to a group given either correct (50% chance) or misleading (100% chance) information about the likelihood of receiving the active drug. A sum of the severity scores from visual analogue scales over all individual adverse effects was defined as the primary endpoint.

The analysis revealed no difference in the primary endpoint between the two groups. This lack of difference may in part be attributable to a conditioning effect as on the first study day higher symptom scores were reported by the participants than on the second study day. Furthermore, the day effect seemed to arise mainly when the first day treatment was the placebo. For the placebo the day effect was clearly significant (p = 0.012), with higher scores on the first day. A further explorative finding in patients given placebo was a tendency for higher scores in the group with the misleading information (p = 0.08). Nothing of that sort was found in the analysis for active treatment. The day effect collapsed and the factor information did not show any tendency of being a potential influence.

In the present study we did not find a statistically significant effect of misleading information on reported adverse events. The large treatment and day effects observed made it difficult to detect a potential small information effect. However, this study excluded a strong and relevant effect of information on the frequency and severity of reported adverse events.