Issue #014Clinical Science7 min readOctober 30, 2026

NHS psilocybin trial: 12 of 30 participants in remission at week 3 after one 25 mg dose, vs 1 on placebo

A 60-person NHS feasibility trial found large short-term effects, and near-complete functional unblinding. Both results matter for how to read it.

Rotem Petranker
Rotem Petranker
PhD, Clinician and Psychedelic Researcher
The Rose Hill Review
014

THE DISPATCH

Most of the psilocybin depression data this newsletter has covered comes from industry-sponsored trials run in specialist research units. On August 6, 2026, Nature Medicine published a smaller trial asking a different question: can psilocybin therapy for treatment-resistant depression be delivered inside the UK's National Health Service at all?¹ The King's College London team, led by James Rucker, randomized 60 participants to a single 25 mg dose of psilocybin or a matched placebo, both with psychological support. The trial was designed to test feasibility, not to prove efficacy. That being said, its efficacy numbers are large, but they come with serious limitations (see below).

In this issue:

→ Who was enrolled, what was measured, and what the week 3 and week 6 results show

→ Why nearly every participant could tell which arm they were in, and what that does to the effect size

→ What moving the trial from a research facility into a community clinic partway through tells us about delivery


THE SCIENCE

The trial (NCT04959253) was run at a single NHS site, South London and Maudsley NHS Foundation Trust, which co-sponsored it with King's College London. Study visits took place first at King's College Hospital's Clinical Research Facility and later at a community mental health research setting.¹ Participants were adults aged 25 to 80 with major depressive disorder that had not responded adequately to at least two antidepressants, or to one antidepressant plus one course of psychotherapy, and a HAM-D-17 score of 14 or higher. Participants had been ill for 20 years on average.¹ Of 75 people screened, 62 were eligible and 60 were randomized: 30 to psilocybin and 30 to placebo.¹ Retention was high. 59 of 60 participants completed MADRS assessments at every follow-up visit, and the one withdrawal was in the placebo arm, for personal reasons.¹ The psilocybin was COMP360, a synthetic formulation supplied by Compass Pathways, which had no role in the trial's design, conduct, or analysis. The trial itself was funded by the UK's National Institute for Health and Care Research.¹

The primary outcomes were feasibility measures: recruitment, retention, and an estimate of how much MADRS scores vary, which future trials need in order to calculate sample sizes. The paper does not specify the time point for this estimate; its effect sizes are standardized against the pooled baseline standard deviation.¹ The efficacy results are secondary and should be considered exploratory. At week 3, MADRS scores in the psilocybin arm had fallen 10.41 points more than in the placebo arm (95% CI, −14.86 to −5.95; Cohen's d = −1.70). At week 6, the between-group difference was 12.92 points (95% CI, −17.37 to −8.47).¹ Response, defined as a reduction of at least 50 percent in MADRS score, was reached by 13 of 30 psilocybin participants at week 3 and 15 of 30 at week 6, compared with 1 participant on placebo at both time points. Remission, defined as a MADRS score of 10 or below, was reached by 12 of 30 psilocybin participants and 1 placebo participant at both week 3 and week 6.¹

One of the limitations in this study is its unsuccessful blinding. Blinding was assessed only in a subsample added partway through the trial: of the 23 participants asked, all 13 in the psilocybin arm and 7 of 10 in the placebo arm guessed their allocation correctly.¹ When participants know which arm they are in, expectancy can inflate the apparent drug effect in one arm and deflate it in the other (the "know-cebo effect").² The authors call functional unblinding an obvious limitation and note that psilocybin's subjective effects may be part of how it works, which makes a fully blind comparison hard to design. For this reason, a Cohen's d of 1.70 from 60 participants should be read as a provisional upper estimate, not as the expected effect size in routine care.

On safety, there were 164 non-serious adverse events in the psilocybin arm and 123 in the placebo arm, most mild and resolved by the end of the trial.¹ There were four serious adverse events in total: three in participants who received psilocybin and one in the placebo arm. Three of the four occurred during the trial's later open-label phase, and none was judged related to psilocybin.¹ Safety scales for suicidality and mania showed no meaningful change across the trial. With 30 participants per arm, the absence of drug-related serious events does not rule out rare or delayed events.


THE STORY

One detail in the methods matters more for patients than the effect size. Partway through recruitment, dosing moved from King's College Hospital's Clinical Research Facility to a community mental health setting.¹ The authors noted no new logistical or safety concerns after the move, which they present as informal commentary rather than a formal finding.¹ That does not show that psilocybin therapy is ready for routine NHS care. It does show that the procedure, from preparation through a supervised dosing day to integration sessions on days 1, 7, 21, and 42, could run outside a dedicated research unit in this sample. For a health system deciding whether it could ever offer this treatment, that operational evidence is the main contribution of a feasibility trial, and it is the part most likely to be missed when a headline reports only the remission figures.

Rose Hill Life Sciences is a psychedelic research organization, specializing in the production and research of Psilocybe cubensis operating at the intersection of science and therapeutic integration.


ON OUR RADAR

① FDA Sets Out How It Expects Trials to Handle Functional Unblinding

A Special Report in the New England Journal of Medicine, published September 10, 2026, describes FDA's evolving framework for psychedelic drug development, including trial designs that account for functional unblinding, such as active placebo conditions and measures of expectancy.³ ⁴ The NHS trial above shows why this matters: without these design choices, even well-run trials produce effect sizes that are difficult to interpret.

② A Second Phase 3 Anxiety Result for an LSD Formulation

On September 14, 2026, Definium Therapeutics (formerly MindMed) reported topline results from Panorama, a Phase 3 trial of a single dose of DT120 (lysergide) in generalized anxiety disorder. 245 participants were randomized to 100 µg (n = 96), 50 µg (n = 52), or placebo (n = 97). At week 12, HAM-A scores fell 5.1 points more with 100 µg than with placebo (p < 0.0001; Cohen's d = 0.64).⁵ The company included the 50 µg arm partly to reduce functional unblinding. These are company-reported topline data and have not been peer reviewed.

③ What Happens at 12 Months: A Useful Comparison From Sweden

A Karolinska Institutet trial in JAMA Network Open (May 2026) randomized 35 adults with major depression to one 25 mg psilocybin dose (n = 17) or niacin as an active placebo (n = 18). Psilocybin led on MADRS at day 8 (difference 7.27 points; P = .01) and through day 42, but at 12 months the between-group difference was no longer significant (P = .20).⁶ Six weeks of follow-up, as in the NHS trial, cannot yet show how long an effect lasts.


YOU ASKED

Q: "An effect size of 1.70 is far larger than anything I see with antidepressants. Should I take it at face value?"

No, and the authors don't ask you to. Three features of this trial push the estimate upward. First, it is small: with 30 participants per arm, effect sizes are imprecise, and the 95% confidence interval for the week 3 difference runs from about 6 to 15 MADRS points. Second, functional unblinding was close to complete, so expectancy may have contributed to the gap between arms. Third, the placebo arm improved very little, which widens any between-group difference. None of this means the effect is absent. The response and remission counts favor psilocybin at every time point, and the direction of the result matches earlier randomized trials, although the largest trial in treatment-resistant depression to date found more modest effect sizes.¹ What it means is that this trial was designed to tell future researchers how to size and run the next trial, and the most accurate description of its efficacy result is "large and provisional."


This week's takeaway: in a 60-person NHS feasibility trial, 12 of 30 participants who received a single 25 mg dose of psilocybin met MADRS remission criteria at weeks 3 and 6, compared with 1 on placebo. Functional unblinding was near complete and follow-up lasted six weeks, so the effect size is provisional, while the feasibility result, including delivery in a community setting, is the trial's firmest finding.

A question worth sitting with: if psilocybin's subjective effects are part of how it works, what would a genuinely blind comparison even look like, and would we trust its answer more?

Read the study directly: Nature Medicine: Psilocybin-assisted therapy for treatment-resistant depression in a public healthcare setting

If a colleague has asked you whether psilocybin could ever work in a public health system, this trial is the most direct evidence so far. Forward it to them.

— Rotem Petranker

Rotem Petranker

PhD, Clinician and Psychedelic Researcher

Advancing the development of novel psychedelic-based therapeutics


FOOTNOTES

¹ Rucker, J.J., et al. "Psilocybin-assisted therapy for treatment-resistant major depressive disorder in a public healthcare setting: a randomized controlled trial." Nature Medicine, Vol. 32, published online August 6, 2026. NCT04959253. https://www.nature.com/articles/s41591-026-04541-0

² Williams, Z.J., Barnett, H., Szigeti, B. "Psychedelic Therapy vs Antidepressants for the Treatment of Depression Under Equal Unblinding Conditions: A Systematic Review and Meta-Analysis." JAMA Psychiatry, 2026;83(5):461-468. DOI: 10.1001/jamapsychiatry.2025.4809. https://jamanetwork.com/journals/jamapsychiatry/article-abstract/2846479

³ "FDA's New Framework for Psychedelic Drugs." New England Journal of Medicine, Vol. 395, No. 10, September 10, 2026. https://www.nejm.org/doi/full/10.1056/NEJMsb2609286

⁴ Clinical Trial Vanguard. "What the FDA's New Psychedelic Framework Actually Means for the Clinic." September 2026. https://www.clinicaltrialvanguard.com/article/psych-pulse/what-the-fdas-new-psychedelic-framework-actually-means-for-the-clinic/

⁵ Definium Therapeutics. "Definium Therapeutics Announces Positive Topline Results from Phase 3 Panorama Study of DT120 ODT in Generalized Anxiety Disorder." September 14, 2026. https://ir.definiumtx.com/news-events/press-releases/detail/251/definium-therapeutics-announces-positive-topline-results-from-phase-3-panorama-study-of-dt120-odt-in-generalized-anxiety-disorder

⁶ Yngwe, H., et al. "Short-Term and Late-Term Effects of Psilocybin on Symptoms in Major Depression: A Randomized Clinical Trial." JAMA Network Open, May 15, 2026. https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2849099


The Rose Hill Review

Rotem Petranker
Rotem Petranker
PhD, Clinician and Psychedelic Researcher

Advancing the development of novel psychedelic-based therapeutics.

Enjoy this issue?

Get this analysis every week, in your inbox.

Clinical science, veteran data, and honest coverage of psychedelic medicine, every Tuesday.

Browse all issues