Debunking the placebo effect in depression: The effect of patient and investigator expectation on escitalopram efficacy

In approximately half of the major depressive disorder (MDD) antidepressant trials published in the last decade, 30% or more of the patients assigned to the placebo arm showed clinically significant improvements. Attempts to reduce the placebo effect in a variety of ways have proven mostly unsuccessful. The aim of this study was to determine whether trial design has an effect on the efficacy outcome in a mock placebo versus escitalopram treatment of adult outpatients with MDD. An 8-week study was designed to evaluate the placebo effect on the response to fixed doses of escitalopram (10 and 20 mg/day) in patients with MDD. The variables affecting placebo response evaluated were as follows: patient expectation, rater expectation, three different outcome measures and the number of visits during the study. Investigators were blinded to the inclusion and exclusion criteria. Forty patients were randomized to receive what they and their treating physicians conceived of as double-blind treatment. The mean age of the patients in the group was 45.1 years, 19 women (47.5%) and 21 men. The mean change from baseline to week 8 in the Montgomery-Åsberg Depression Rating Scale total score was -13.7 for participants with 'monthly' visits and -12.9 for the 'fortnightly' group (P=0.75). In each group, 14/16 responders and their physicians thought that they were receiving active treatment. Of 22 nonresponsive patients, 17 thought that they had been receiving placebo. The pharmacological effect of escitalopram observed in the present study is almost identical to that observed in open-label studies, even when patients and clinicians are misled by the study design, placebo presence or raters' blindability.