Josh D. Woolley, MD, PhD, professor of psychiatry and behavioral sciences, University of California, San Francisco.

Josh D. Woolley, MD, PhD

More than 90% of participants in high-dose 

 correctly identify their treatment, compared with roughly 60% in antidepressant trials. In a recent Viewpoint published online in 

, the authors argue active placebos such as niacin and diphenhydramine never plausibly matched psilocybin's intensity, and even phenomenologically richer candidates, including tetrahydrocannabinol, salvinorin A, and dextromethorphan, fail to reproduce its full experiential breadth as single agents.¹ Some investigators have concluded effective masking of high-dose psychedelics is not achievable and have shifted toward comparative efficacy trial designs instead.¹

Josh D. Woolley, MD, PhD, corresponding author of the Viewpoint and a member of the Department of Psychiatry and Behavioral Sciences at the University of California, San Francisco, proposes an alternative called UMBRAA (unidentifiable multidrug blinding with reduced, authorized awareness). The framework combines active placebos built from multiple drug classes with a consent model called authorized incomplete disclosure, in which participants agree in advance not to be told certain design details, such as the number of treatment arms.² Woolley argues regulators have held active placebos to an outpatient tolerability standard irrelevant to a single supervised psilocybin session, which routinely requires 8-hour monitored dosing days with 2 providers present.¹

In the following Q&A, Woolley discusses why current active placebos fail, how UMBRAA's multidrug comparator addresses critiques comparing it to induced delirium, and what he sees as the framework's realistic next steps.

Woolley on UMBRAA's Multidrug Approach to Psilocybin Trial Masking

Why has masking failed so badly, and why has the field been slow to address it?

 Why it's failed is pretty clear. The doses we use in these trials are high doses, and people have big experiences: visual changes, intense experiences of bliss, unity, [and] feeling one with the universe. In most trials, two-thirds of people say it's among the top 5 most meaningful experiences of their life, compared with sitting there looking at the walls, nothing happening, or even worse in trials using niacin, where, as we say in the paper, either you meet God or you get itchy.

Having blinded assessors helps, but the patient isn't blinded; the patient knows, and probably the treatment they get is quite different. If the preparation is preparing people for a high-dose experience, that actually makes people, when they know they didn't get it, probably disappointed.

Masking failure…hasn't really been addressed in pharmacology or pharmaceutical trials more broadly. For some things, it might be possible to address this with active placebos, but active placebos aren't without their own challenges. Inactive placebos are obviously cleaner; give someone niacin and now you have side effects from the placebo you have to deal with, and how are you sure your active placebo doesn't make people less depressed or more depressed?

There's been a real reluctance to use active placebos in general, and with psychedelics, because the experience is so big, people have felt it was impossible and haven't really tried.

Why do regulators hold placebos to an outpatient tolerability standard when a psilocybin session isn't outpatient tolerable?

 People say psilocybin is quite tolerable and the safety is pretty high, which is true; the clinical trials haven't had many adverse outcomes. But that's in the setting of a very intensive delivery model.

People come in; they have all this preparation and screening, and then they're there all day with 2 providers in the room, vital signs monitored, and an MD on call. That's not outpatient; that's a procedure. I'm just saying it's tolerable given all of that.

The point we're making is that expecting some drug you could give as an outpatient to match that is like comparing apples and oranges. It's not a fair comparison because we're not talking about chronic dosing, which is what FDA approval typically requires.

We've gotten pushback when we've said we want to give doses that have been studied at higher levels in single-dose sessions in healthy people, and we're confident it's at least as tolerable as psilocybin, and we've been told we can't do that.

Can you walk me through the UMBRAA framework?

 The umbra is the area of totality of the shadow during an eclipse; a penumbra is the area of partial shadow. We picked that acronym partly because of the idea of blinding.

Our proposal…is a combination of 2 arms, because this is a really hard problem. One arm is active placebos: we argue we should be using them, and a single agent as an active placebo for high-dose psilocybin might work.

And do those drugs have therapeutic efficacy? Dextromethorphan is one example. Some healthy-volunteer studies show people can confuse it with psilocybin, but it's ketamine-like, with an NMDA mechanism, and it's being used as an antidepressant. Not at a single high dose, but still, no one knows, and we can't say it doesn't have antidepressant effects, so that's a problem.

What we suggest is to think more broadly. Maybe we can't find it in a single drug, but maybe we can make it happen by combining drugs… [to] confuse people, because that's the goal.

The second element is… the idea of incomplete disclosure. The ethics of the regular placebo-controlled trial haven't always been obvious, because you're a clinician giving someone an inactive treatment on purpose and not telling them whether it's happened, an incomplete disclosure. You tell people, look, you can participate, you might get the active treatment, or you might get inactive, you have a 50-50 chance, and I can't tell you which you're getting.

We argue, in this article and in one from 2025, psychedelics and psychedelic-like treatments may require a more aggressive version of that.² Just as you don't tell people exactly what they'll get in a double-blind trial, we argue there are other elements of the trial design you can ethically not disclose, like the number of treatment arms, a 2-arm trial versus a 3-arm trial. If you tell people you're withholding that information, they can consent; it's called authorized incomplete disclosure.

Critics say a multidrug comparator sounds more like inducing delirium than a legitimate control. How do you respond?

 I'd ask what they mean by delirium. Delirium is a nonspecific syndrome, and the psychedelic state is different from a drug-induced delirium in some ways; people lose track of time and have cognitive difficulties, among other things. If you separate out the physiological parts, delirium is often caused by infection or metabolic abnormalities, which are quite risky on their own, apart from being intoxicated and having something that alters how you process the world.

If you want to call that second part delirium, I'd call it intoxication, altered sensory processing, and other things, and I think that's what psilocybin does. We have to figure out how to do something similar enough that people get confused.

Yes, it is an altered state, and yes, it's a complex pharmacology, but I see that as a strength, or a requirement, to achieve effective masking. If there were something simpler, I'd happily use it, but there doesn't seem to be.

Why do you think the shift toward comparative efficacy designs instead of masking is the wrong call?

 I wouldn't necessarily call it the wrong call. These are interesting, different approaches. Comparative efficacy can work, but it's tricky, because in an unblinded situation, like a trial comparing an SSRI with psilocybin, masking still fails.

We're also not sure our approach is going to work; it may not be possible; you can throw everything at it, and people may still tell. That's possible, but no one's really tried. I think when people say masking doesn't matter or that it's impossible, so it doesn't matter anyway, that's 1 point of view, but I don't agree with it.

How worried are you that a comparator like dextromethorphan could confound the outcome you're measuring?

 Finding any active placebo is challenging. They add complexity, they can add their own side effects, and they may add their own therapeutic outcomes, whether antidepressant or pro-depressant. We don't usually run studies to prove something isn't an antidepressant; that's not something you typically do, and it would actually be quite hard to do.

Dextromethorphan has this challenge in particular because it's already being explored as an antidepressant. That doesn't mean you couldn't still use it as a comparator; people have talked about comparing ketamine to psilocybin, where you might be able to get participants to confuse the 2. There are challenges around time course, but it's possible; ketamine is clearly an antidepressant, but if you believe the literature, a single dose should have an effect over days, versus weeks to months with psilocybin.

Even though ketamine is an antidepressant, it's a well-characterized one, and you could still see a differential outcome despite that effect. Having antidepressant effects isn't necessarily bad, but an unknown antidepressant effect is a problem, and we don't really know what a single high dose of dextromethorphan does.

What's the realistic next step for UMBRAA?

 We and others are beginning to do studies with incomplete disclosure. Trying this out and seeing how well it works…is 1 step.

The other is these drug combinations. People have talked about combinations like THC and dextromethorphan, and there's been a little work, but not much.

I don't want to give up on masking before we've really given it a good try.

Watch parts 1, 2, 3, and 4 of our interviews with Woolley here: 

UMBRAA Framework Aims to Fix Blinding Failures in Psychedelic Trials

UMBRAA Framework Targets Blinding Failures in Psilocybin Trials

Defending Multidrug Masking in Psilocybin Trials, With Josh Woolley, MD, PhD

Confounding Risk in Psilocybin Trial Comparators, With Josh Woolley, MD, PhD

Editor’s note: A reported disclosure for Woolley includes Eli Lilly and Company.

Yang KH, Heifets BD, Krystal A, et al. The case for more aggressive masking in psychedelic trials. 

Matvey M, Kelley DP, Bradley ER, Chiong W, O'Donovan A, Woolley J. Modifying informed consent to help address functional unmasking in psychedelic clinical trials. 

Articles

High-dose psilocybin trials show profound functional unmasking, with most participants correctly inferring assignment due to intense, distinctive subjective effects that standard active placebos cannot approximate. A proposed solution, UMBRAA, pairs multidrug active-placebo combinations with authorized incomplete disclosure to reduce expectancy-driven bias while maintaining ethical consent. The approach reframes psilocybin administration as a highly supervised procedure, challenging regulators’ outpatient tolerability expectations for comparators. Key concerns include comparator-induced therapeutic effects (eg, dextromethorphan) and feasibility, motivating early studies testing incomplete disclosure and candidate drug combinations.

Josh Woolley, MD, PhD, explains why current active placebos fail to mask high-dose psilocybin trials and outlines a framework proposed to fix it.

Q&A: Woolley on UMBRAA's Multidrug Approach to Psilocybin Trial Masking

 Viewpoint, Josh D. Woolley, MD, PhD, of the University of California, San Francisco, and colleagues proposed the UMBRAA blinding framework for high-dose 

, which uses a multidrug comparator to mask the placebo.¹

Active placebos have historically been rare in psychiatric trials because they add complexity, carry their own adverse events, and may produce therapeutic effects of their own. Dextromethorphan (DXM), 1 candidate comparator, illustrates this risk directly, since its N-methyl-d-aspartate (NMDA) antagonism at dissociative doses has drawn interest as a standalone antidepressant.

“Even though it's an antidepressant, it's a well-characterized antidepressant, and you could still have a differential outcome, even though it has an antidepressant effect,” Woolley told 

 “Having antidepressant effects is not necessarily bad, but an unknown antidepressant effect is problematic.”

DXM and Ketamine Illustrate the Confounding Comparator Problem

Woolley said researchers do not typically design studies to prove a drug is not an antidepressant, since disproving a therapeutic effect is difficult to demonstrate. He pointed to ketamine as a further example: investigators have proposed comparing ketamine with psilocybin because participants may confuse the 2 experiences despite differing time courses, with ketamine's effects reported within days and psilocybin's extending across weeks to months.

Woolley contrasted DXM with agents such as zolpidem and niacin, neither of which is considered an antidepressant. He said if a combination of drugs not believed to have antidepressant properties produced a delirium-like state matching high-dose psilocybin's outcomes, the finding would suggest the treatment effect is not pharmacologically specific to psilocybin. Woolley called this scenario a genuinely radical result forcing investigators to rethink how psilocybin works.

UMBRAA's Next Steps: Incomplete Disclosure and Drug Combinations

Why is DXM considered a risky comparator for psilocybin trials?


Its NMDA receptor antagonism at dissociative doses has drawn interest as a standalone antidepressant, so a single high dose could confound trial outcomes with an unknown therapeutic effect.

Is authorized incomplete disclosure a new concept?


No. Woolley traces its lineage to early Johns Hopkins University psilocybin studies in patients with cancer, where participants were told a dose range but not the exact number of trial arms.

What are the next steps for testing UMBRAA?


Woolley said investigators are beginning to test incomplete disclosure directly and plan to evaluate specific multidrug combinations, such as THC and DXM, in healthy volunteers or clinical trials.

Woolley said authorized incomplete disclosure is not new to psychedelic research, citing early studies from Johns Hopkins University in the early 2010s testing psilocybin in patients with cancer. Those studies told participants they would receive psilocybin from within a stated dose range without disclosing the exact number of trial arms, an early form of incomplete disclosure.² Woolley said he and other investigators are now beginning to test the approach directly to determine whether it improves masking success.

A second priority is testing specific multidrug combinations, since limited empirical work has been done despite years of informal discussion. Woolley pointed to tetrahydrocannabinol (THC) and DXM as 1 combination with only preliminary testing. He said this work could proceed in healthy volunteers or within active clinical trials designed to improve outcomes for participants.

Woolley's comments reflect the Viewpoint's central argument: current active placebos do not adequately mask the effects of high-dose psilocybin, leading some investigators to consider comparative efficacy designs instead.¹ Woolley said UMBRAA remains untested and that its authors have not endorsed a single drug combination for all investigators to use.

"I don't want to give up on masking before we've really given it a good try," Woolley said. "Scientists and patients should advocate for that, because they don't want something that waste[s] their time and money [and] puts them at risk without being pretty sure it works."

Watch parts 1, 2, and 3 of the interview with Woolley here: 

Videos

Woolley discusses why an active placebo's own therapeutic effects complicate psilocybin trial blinding and outlines next steps for testing UMBRAA.

Confounding Risk in Psilocybin Trial Comparators, With Josh Woolley, MD, PhD 

 Viewpoint, Josh D. Woolley, MD, PhD, of the University of California, San Francisco, and colleagues proposed the UMBRAA blinding framework for 

UMBRAA (unidentifiable multidrug blinding with reduced, authorized awareness) proposes combining agents from distinct mechanistic classes for disclosure to address unmasking rates exceeding 90% in psilocybin trials. The framework includes a consent model called authorized incomplete disclosure.

, Woolley addressed concerns the comparator resembles induced delirium rather than a controlled placebo condition.

“I would call it intoxication [with] alerted sensory processing,” Woolley said. “I think that's what psilocybin does.”

Why the Multidrug Comparator Isn't Delirium

Woolley noted delirium is a nonspecific syndrome often caused by infection or metabolic abnormality, conditions carrying physiological risk independent of any drug effect. He separated this risk from the intoxication and altered sensory processing the comparator itself produces.

Woolley said reproducing a similar altered state through a different pharmacology, if the approach succeeds, would force investigators to rethink assumptions about the mechanisms behind psilocybin's effects. He described the pharmacological complexity of the multidrug comparator as a requirement for effective masking rather than a design flaw.

“Yes, it's a complex pharmacology, but I see that as [a] strength or required to achieve the goal of effective masking,” he said. “If there was something simpler, I’d happily use that, but it doesn't seem like there is.”

Comparative Efficacy Trials and IRB Feasibility

Woolley says the multidrug comparator produces intoxication and altered sensory processing comparable to psilocybin, not delirium.

He argues comparative efficacy designs still tend to unmask.

He acknowledges the approach is untested and may not succeed.

Woolley defended pursuing more aggressive masking instead of a shift to comparative efficacy trial designs. He said the field has not ruled it out through trial and error.

Some investigators have already moved toward comparative efficacy designs, arguing effective masking of high-dose psychedelics is not achievable. Woolley said this alternative, which compares psilocybin directly against another active treatment such as an SSRI, carries its own blinding problem, since masking still tends to fail and participants often learn which arm they are in.

He pointed to psychotherapy research on waitlist controls, where participants assigned to wait for treatment sometimes improve less than people who never enrolled in a study at all, a finding some investigators attribute to participants knowing they are not expected to improve. Woolley said this dynamic also appears in placebo response rates: sugar pills perform better in SSRI trials than in psilocybin trials because participants assigned to SSRI placebo remain uncertain of their assignment, and psilocybin trial participants who receive placebo typically know it.

“We're not sure that our approach is going to work,” Woolley said. “It may not be possible. You may throw everything at it, and people can still tell. That is possible, but no one's tried, so I think we're all just trying to figure out how to get reliable information.”

Watch parts 1 and 2 of the interview with Woolley here: 

Woolley separates intoxication from delirium and defends aggressive masking instead of a shift to comparative efficacy trial designs.

A Viewpoint published online recently in 

 proposes a new blinding framework called UMBRAA (unidentifiable multidrug blinding with reduced, authorized awareness) to address years of failed masking in high-dose 

Correct treatment identification in psilocybin trials now exceeds 90%, even against active placebos selected to produce noticeable effects, far above the roughly 60% rate seen in antidepressant trials.¹ Many investigators have concluded effective masking of high-dose psychedelics is not achievable and have called for a shift toward comparative efficacy designs instead.¹ Josh D. Woolley, MD, PhD, corresponding author of the Viewpoint and a member of the Department of Psychiatry and Behavioral Sciences at the University of California, San Francisco, argues the field has abandoned the problem before sufficiently aggressive approaches were tried.

"The reason we do masking is because people have ideas about how the treatment will help them, which is totally okay," Woolley said. "But if they can tell what they get, expectancy is differentially activated.”

Why Active Placebos Fail to Mask Psilocybin's Effects

A 25 mg dose of psilocybin produces perceptual distortions, ego dissolution, and pronounced emotional swings lasting 4 to 6 hours.¹ Early comparators such as niacin and diphenhydramine never plausibly mimicked this profile, and more recent proposals, including tetrahydrocannabinol (THC), salvinorin A, and dextromethorphan (DXM), still fail to reproduce psilocybin's full experiential breadth as single agents.¹ Woolley and colleagues note regulatory agencies have held active placebos to an outpatient tolerability threshold irrelevant to a single supervised administration, since a 25 mg psilocybin dose routinely requires 8-hour supervised sessions, intensive monitoring, and postdosing driving restrictions.¹

DXM illustrates a second problem beyond its incomplete resemblance to psilocybin. At dissociative doses, its N-methyl-d-aspartate (NMDA) antagonism can produce rapid-acting antidepressant effects similar to ketamine, so investigators cannot rule out an antidepressant effect from a single high dose in depression trials. Woolley and colleagues argue expecting a mild comparator to mask this experience is not a cautious choice but a methodological mismatch that all but guarantees unmasking.

UMBRAA Combines Multidrug Comparators with Authorized Disclosure


UMBRAA stands for unidentifiable multidrug blinding with reduced, authorized awareness, a proposed framework combining multidrug active placebos with authorized incomplete disclosure to improve blinding in high-dose psilocybin trials.

Why do single-agent active placebos fail to mask psilocybin trials?


No single comparator, including THC, salvinorin A, or DXM, reproduces the full experiential breadth of a 25 mg psilocybin dose, and each carries a recognizable pharmacological fingerprint experienced participants can identify.

What is authorized incomplete disclosure?


It is a consent model in which participants agree in advance not to be told certain non-safety-related trial design details, such as the number of treatment arms, to reduce belief-driven expectancy effects.

UMBRAA proposes combining agents from distinct mechanistic classes, potentially a stimulant, sedative, anticholinergic, and cannabinoid, chosen to avoid psilocybin's primary 5-hydroxytryptamine receptor 2A (5-HT2A) mechanism and to produce an intense, unfamiliar profile.¹ Woolley and colleagues cite a 19th-century precedent in which combining hashish with strong coffee produced classically psychedelic experiences, including visions and ego dissolution, effects neither substance reliably produces alone.¹ Combining classes can also extend the duration and offset dose-limiting adverse effects of individual agents; stimulants, for example, can counteract sedative-induced sleep and augment perceptual effects, and hallucinations occur in up to 4.7% of zolpidem-treated individuals, rising to nearly 8% when stimulants are added.¹

The second element pairs this multidrug approach with what Woolley and colleagues term authorized incomplete disclosure, an extension of standard blinding in which participants consent in advance to not being told certain design details, such as the number of treatment arms.

"We argue there are some other elements of the trial design that you can ethically not disclose, for example, the number of treatment arms in the study," Woolley said. "If you tell people you're withholding that information, they can consent."

Watch part 1 of the interview with Woolley here: 

Josh Woolley, MD, PhD, says active placebos fail to mask high-dose psilocybin and propose combining drug classes with authorized incomplete disclosure.