Why Did the Placebo Benchmark Change the Estimated Psilocybin Effect?
A network meta-analysis can compare treatments studied in different trials, but its estimates depend on how the trials are connected. Hsu and colleagues examined whether using a different placebo benchmark changed the estimated effect of psilocybin on depressive symptoms.
Why distinguish the placebo groups?
The researchers pooled randomized trials involving adults with depressive symptoms and used change on the 17-item Hamilton Depression Rating Scale as the primary outcome. They distinguished placebo responses in psychedelic trials from those in escitalopram trials. Placebo response was lower in the psychedelic-trial group. The authors considered unsuccessful blinding a possible contributor because noticeable effects may influence participants’ expectations about which treatment they received.
What changed in the model?
The standardized mean difference for high-dose psilocybin was 0.88 when the model used the placebo benchmark from psychedelic trials and 0.31 when it used the benchmark from antidepressant trials. These are comparative estimates from the analysis. The change illustrates how much the estimated effect depended on the reference group.
What the comparison supports
Differences across trial populations and settings matter for comparisons assembled through a network. The alternate benchmark is a sensitivity analysis of those differences. It does not identify how much of the observed difference came from participants guessing their assignment.
Source
Hsu TW, Tsai CK, Kao YC, et al. Comparative oral monotherapy of psilocybin, lysergic acid diethylamide, 3,4-methylenedioxymethamphetamine, ayahuasca, and escitalopram for depressive symptoms: systematic review and Bayesian network meta-analysis. BMJ. 2024;386:e078607. DOI 10.1136/bmj-2023-078607.



