Psychedelic Trials Are Effectively Open-Label. The FDA Has to Decide What to Do With That.

Picture the moment a Phase 2 participant opens their eyes ninety minutes into a psilocybin session and the room is breathing. They know. Their therapist, sitting across from them watching their face reorganize into something between wonder and grief, knows too. Whatever the randomization envelope says, the blind is gone — and it was gone before the first questionnaire was scored. The trial has already become, in every functionally meaningful sense, an open-label experiment with paperwork that says otherwise.

This is not an edge case. A systematic review published in PubMed covering 112 randomized clinical trials of psychedelics as psychiatric interventions found that psilocybin, LSD, and ayahuasca studies frequently reported blinding failure rates exceeding 90% among participants and raters. Only 29.5% of those 112 trials — 33 studies — even evaluated blinding integrity. And 57.1% acknowledged blinding as a limitation in their methods section, which is academic shorthand for “we know this is broken and we are moving on.” A recent Lancet Psychiatry editorial described the entire literature as effectively open-label. That is not a fringe position. That is the field describing itself.

The striking efficacy signals keep coming anyway. COMPASS Pathways received FDA Breakthrough Therapy designation for its psilocybin therapy COMP360 for treatment-resistant depression in October 2018. The signal was real enough for the agency to act. But Breakthrough designation is granted on early evidence of substantial improvement over existing therapy — not on evidence that the blind held. Those are two entirely different evidentiary questions, and the field has been answering only one of them.

The Mechanism Behind the Broken Blind

To understand why this matters statistically, you have to understand what blinding actually does in a trial. The randomized controlled trial derives its power from one assumption: that neither the participant nor the assessor knows which arm they are in, and therefore any psychological response to being treated — hope, attention, expectation of relief — distributes equally across active and placebo groups. When that assumption holds, the difference in outcome between arms measures drug effect alone. When it fails, the difference measures drug effect plus the psychological response to knowing you received the active agent. The two are confounded, and you cannot algebraically separate them after the fact.

In most trials, blinding leaks slowly — a side effect here, a taste difference there. In psychedelic trials, it does not leak. It ruptures. A participant who receives 25mg of psilocybin in a monitored clinical session experiences a complete alteration of their perceptual and cognitive state for four to six hours. A participant who receives a 100mg niacin active placebo — one of the more common comparators — experiences flushing and tingling. The experiential difference is not subtle. There is no statistical adjustment that can compensate for participants correctly identifying their assignment 90% of the time, because the expectancy mechanism is not random noise — it is systematic, directional, and correlated with outcome.

Here is where it becomes a measurement problem with specific consequences. Expectancy effects in psychedelic research have been documented through assessments run alongside escitalopram comparison work. A trial published in Psychological Medicine examining psilocybin versus escitalopram for depression used the Beliefs About Treatment scale to assess pre-treatment expectancy and suggestibility. What it found: pre-treatment expectancy scores were predictive of outcome — meaning some portion of the measured antidepressant effect was being generated by what participants believed before the drug touched their receptor. When the blind is broken and that belief is confirmed in real time, the expectancy signal amplifies. The effect size you report at week six is not a clean pharmacological readout. It is a composite.

Call this the Expectancy Inflation Principle: in any trial where functional unblinding exceeds 80%, the primary efficacy endpoint absorbs a directional bias whose magnitude cannot be estimated from the trial data alone because the unblinding and the treatment response are measured on the same subjects, at the same time, using the same self-report instruments. Every point on a Hamilton Depression Rating Scale or a MADRS that a functionally unblinded participant completes is potentially inflated — not by fraud, not by bias in the pejorative sense, but by the basic psychology of knowing you got the drug that you hoped would work.

This is not a problem unique to psychedelics — it appears in any trial using pharmacologically active placebos or drugs with distinctive phenomenological signatures. But psychedelics represent the extreme end of a spectrum. The FDA’s 2019 guidance on placebos and blinding in randomized controlled cancer trials acknowledges that blinding failure increases risk of bias and instructs sponsors to assess blinding integrity and report the results. That guidance was written for oncology. The agency has issued no psychedelic-specific blinding framework. The gap between where the science is and where the regulatory infrastructure is could not be wider.

The Second Example That Confirms the Pattern

MDMA-assisted therapy for PTSD offers a parallel case — and a sharper regulatory wound. The FDA rejected MAPS’s NDA for MDMA-assisted therapy in August 2024, citing functional unblinding among its evidentiary concerns. Participants in the MDMA arms of the Phase 3 MAPP1 and MAPP2 trials reported subjective drug effects that were unmistakable, and the blinding assessment data confirmed near-total identification. The agency’s complete response letter asked for an additional Phase 3 trial. What that means in practice: a literature built over decades, a patient population with limited treatment options, and a regulatory agency that looked at the evidence base and concluded it could not read the effect size with confidence. The blinding failure was not a side note in the FDA’s reasoning. It was load-bearing.

The structural parallel to COMPASS’s psilocybin program is uncomfortable. Both compounds produce unmistakable subjective experiences. Both have efficacy literatures generating headlines. And both now face the same underlying evidentiary question that no amount of p-value reporting resolves: how much of the signal is drug, and how much is the participant knowing they received the drug in a highly supportive therapeutic context designed to maximize their expectancy of healing?

The Dissent Worth Hearing

Balázs Szigeti, PhD, a clinical data scientist at UCSF’s Translational Psychedelic Research Program and one of the field’s more rigorous internal critics, has argued that the framing of blinding failure as a fatal flaw may itself be too simple. His position is that the expectancy effect may be a genuine component of the therapeutic mechanism — that for psychedelic-assisted therapy, the drug’s ability to induce a state of heightened suggestibility and openness is not noise contaminating the pharmacological signal, but part of what the therapy actually does. Under that framing, trying to strip expectancy out of the outcome measure might be scientifically incoherent, like measuring the analgesic effect of physical therapy while controlling for the patient’s belief that movement helps. The argument has real force. But it also carries a regulatory consequence that the field tends to understate: if expectancy is part of the mechanism, then the product being approved cannot be described as a drug. It is a drug-plus-context system, and the FDA’s approval architecture was not designed for that.

What the FDA Is Actually Being Asked to Evaluate

The lesson embedded in this mechanism is broader than psychedelics. Every time the agency receives an NDA or BLA built on a blinding paradigm that functionally failed, it is being asked to evaluate a hypothesis — “this molecule produces therapeutic benefit” — using evidence that conflates the molecule with the belief that the molecule was received. The 112-trial meta-analysis makes clear that the field has known about this problem for years: 57.1% of studies cited it as a limitation, which means sponsors, IRBs, and statistical reviewers all signed off on trial designs they themselves described as methodologically compromised. That is not a failure of one company. It is a failure of the evidentiary standard that the field collectively accepted and regulators collectively allowed.

The FDA now faces a version of this at scale. Whatever decision it makes on COMP360 or the next psilocybin NDA will implicitly set the standard for how much functional unblinding an approval-grade evidence package can contain. There is no guidance document that answers that question. There is no precedent from an approved compound with a comparable blinding failure rate. The agency will have to make a judgment call about how to read an efficacy estimate that was, by the field’s own admission, built on a broken blind — and that judgment call will define psychedelic regulatory science for the next decade.

The data for these compounds may ultimately prove strong enough to survive the scrutiny. But the scrutiny has to happen first, and right now, the FDA is being handed effect sizes without a validated method for decomposing them. That is the actual problem on the desk in Silver Spring — not whether psilocybin works, but whether anyone in this field has built the evidentiary tools to prove it in a way the agency can defend.

References

1. PubMed — Systematic review of 112 RCTs: blinding integrity in psychedelic psychiatric trials
2. COMPASS Pathways — “COMPASS Pathways receives FDA Breakthrough Therapy designation for psilocybin therapy for treatment-resistant depression” (October 2018)
3. Psychological Medicine — “Assessing expectancy and suggestibility in a trial of escitalopram v. psilocybin for depression”
4. FDA — “Placebos and Blinding in Randomized Controlled Cancer Clinical Trials for Drug and Biological Products: Guidance for Industry” (October 2019)
5. HCPLive — Balázs Szigeti, PhD, on psychedelic therapy efficacy and blinding methodology