Every psilocybin efficacy number this desk has covered so far has come from a clinical trial: controlled conditions, selected patients, standardized protocols, structured psychological support built around the dosing session. A new study in The Lancet Regional Health Europe, alongside a commentary from Vassilis Martiadis published in the same journal examining whether trial results actually translate to clinical practice, is one of the first to report psilocybin outcomes generated outside that structure. The results are real, they are positive, and they are considerably more modest than the numbers that built psilocybin’s reputation as a breakthrough treatment.
Where this data actually comes from
Switzerland maintains a legal framework, distinct from anywhere else this desk has covered, permitting limited medical use of psilocybin for treatment-resistant depression outside a formal clinical trial. Researchers at the University Hospital of Psychiatry Zurich retrospectively reviewed the records of 19 patients who received one to four psilocybin dosing sessions under that exemption. This is not a trial with a protocol designed in advance. It is a look at what happened when psychiatrists treated real patients, with real comorbidities and real concurrent medications, using clinical judgment rather than a standardized research design.
What the numbers actually show
Depression scores fell meaningfully: a mean reduction of roughly 11 points on the MADRS scale, from an average starting score near 31 to about 20 within 42 days of the final session, with no serious or sustained adverse events documented in the medical record. Response and remission rates landed at approximately 33 percent and 22 percent. Those are real, clinically meaningful improvements for a treatment-resistant population. They are also substantially lower than the numbers that established psilocybin’s reputation. The original 2021 Johns Hopkins trial reported 71 percent response and 54 percent remission. Compass Pathways’ own Phase 2b trial in treatment-resistant depression, the largest of its generation, started at roughly 37 percent response and declined to 20 percent by week 12. The Zurich cohort’s real-world results sit close to that Phase 2b endpoint, and land in the same broad zone as STAR*D’s first-step remission rates of 28 to 33 percent, the largest real-world antidepressant effectiveness study ever conducted, and far below the 54 to 71 percent the early psilocybin trials reported.
Why this comparison matters more than the number itself
The pattern here is not unique to psilocybin. Pivotal trials for essentially every psychiatric drug tend to report larger effects than real-world use eventually confirms, because trial populations are selected, monitored, and supported in ways ordinary clinical practice cannot fully replicate. What makes this data point notable is that it is one of the first times anyone has measured where on that familiar curve psilocybin lands, rather than assuming it would prove exceptional. The early answer is that it looks a great deal like other antidepressants once trial-level selection and support are removed, not like the dramatic outlier its most-cited early results suggested.
What the study cannot tell you
This is a single retrospective chart review of 19 patients at one hospital, with no control group, real heterogeneity in how many sessions patients received and what other treatments they were using concurrently, and no long-term follow-up beyond 42 days after the final dose. It cannot establish causation the way a randomized trial can, and a sample this size cannot rule out that a different patient population, a different protocol, or a different country’s implementation would produce meaningfully different results. The study’s own authors are explicit about these limits and call directly for larger real-world studies, the appropriate and modest claim their data actually supports.
Why Martiadis’s framing is the useful addition
The commentary accompanying this data, from Martiadis, engages directly with the question this desk has asked about nearly every psychedelic trial covered this year: does a result generated under tightly controlled research conditions actually hold once psilocybin reaches ordinary clinical practice. That question matters enormously for how regulators and payers eventually treat this drug class, since a coverage or approval decision built on trial-level effect sizes could set expectations that real-world use does not meet, precisely the dynamic this data now offers early direct evidence of.
The caveats
Switzerland’s regulatory framework is unusual and does not generalize directly to how psilocybin might eventually be used in the United States or elsewhere, where any approved product would move through a different regulatory and reimbursement structure entirely, with its own effect on how the drug is actually administered and supported in practice. A single 19-patient cohort is a genuinely small evidence base to draw firm conclusions from, and the comparison to STAR*D, while informative, is not a formal head-to-head study; both figures come from different populations, different eras, and different measurement approaches. This data should be read as an early, honest signal worth taking seriously, not as a settled verdict on how well psilocybin performs outside a trial.
The frame
This desk has spent the year reading psychedelic trial data carefully, flagging where topline numbers rest on post hoc analyses, unblinding risk, or narrow evidentiary bases. Here, for once, the caution runs the other direction: real-world data, generated with no funding and no company involved, suggesting the field’s headline efficacy numbers may not be what patients actually experience once psilocybin leaves the tightly controlled conditions that produced them. Neither the celebrated early trial results nor this modest real-world signal should be treated as the final word. But an industry that has spent years citing 50 and 70 percent response rates now has an early, honest look at what happens when psilocybin is used the way most drugs actually get used, in ordinary clinics, by ordinary psychiatrists, on patients a trial’s inclusion criteria would often have excluded. The number that comes back looks a lot less like a breakthrough and a lot more like psychiatry as usual.