The placebo response in depression clinical trials is a major contributing factor for failure to establish the efficacy of novel and repurposed treatments. However, it is not clear as to what the placebo response in treatment-resistant depression (TRD) patients is or whether it differs across treatment modalities. Our objective was to conduct a systematic review and meta-analysis of the magnitude of the placebo response in TRD patients across different treatment modalities and its possible moderators.

Searches were conducted on MEDLINE and PsychInfo from inception to January 24, 2020. Only studies that recruited TRD patients and randomization to a placebo (or sham) arm in a pharmacotherapy, brain stimulation, or psychotherapy study were included (PROSPERO 2020 CRD42020190465). The primary outcome was the Hedges’ g for the reported depression scale using a random-effects model. Secondary outcomes included moderators assessed via meta-regression and response and remission rate. Heterogeneity was evaluated using the Egger's Test and a funnel plot. Cochrane Risk of Bias Tool was used to estimate risks.

46 studies met our inclusion criteria involving a total of 3083 participants (mean (SD) age: 45.7 (6.2); female: 52.4%). The pooled placebo effect for all modalities was large (N = 3083, g = 1.08 ,95% CI [0.95-1.20)I 2 = 0.1). The placebo effect in studies of specific treatment modalities did not significantly differ: oral medications g = 1.14 (95%CI:0.99-1.29); parenteral medications g = 1.32 (95%CI:0.59-2.04); ayahuasca g = 0.47 (95%CI:-0.28-1.17); rTMS g = 0.93 (95%CI:0.63-1.23); tDCS g = 1.32 (95%CI:0.52-2.11); invasive brain stimulation g = 1.06 (95%CI:0.64-1.47). There were no psychotherapy trials that met our eligibility criteria. Similarly, response and remission rates were comparable across modalities. Heterogeneity was large. Two variables predicted a lager placebo effect: open-label prospective design (B:0.32, 95%CI: 0.05-0.58; p:0.02) and sponsoring by a pharmaceutical or medical device company (B:0.39, 95%CI:0.13-0.65, p:0.004)). No risk of publication bias was found.

The overall placebo effect in TRD studies was large (g = 1.08) and did not differ among treatment modalities. A better understanding of the placebo response in TRD will require: standardizing the definition of TRD, head-to-head comparisons of treatment modalities, an assessment of patient expectations and experiences, and standardized reporting of outcomes.