Therapeutic potential of psychedelic drugs: navigating high hopes, strong claims, weak evidence, and big money

Abstract

Therapeutic claims about many psychedelic drugs (e.g., ayahuasca, mescaline, ibogaine, N, N-Dimethyltryptamine) have not been evaluated in any studies of even modest rigor. Claims about the medical benefits of other psychedelic drugs are reliant on a limited number of studies in which research design weaknesses (e.g., lack of successful blinding in the presence of significant expectancy effects) may account for some or all apparent benefit. The science of psychedelic drugs is strengthening however, making it easier to differentiate some promising findings amidst the hype that suffuses this research area. Ketamine for major depressive disorder is the most studied psychedelic drug indication. It is probably non-inferior to electro-convulsive therapy (ECT) and superior to quetiapine for treatment-resistant depression. S-ketamine is FDA-approved as a supplement to traditional anti-depressant medication on the basis of positive trial results, but has shown no or minimal benefit in other clinical trials. Ketamine has initial evidence of benefit in treatment of alcohol use disorder and cocaine disorder, but has not proven effective for treating post-traumatic stress disorder. Even for conditions for which ketamine has some evidence of benefit, its risk of adverse side effects (e.g., addiction and cystitis) should be borne in mind by clinicians. Other therapeutic signals from psychedelic drug research which merit rigorous replication studies include 3,4-Methyl​enedioxy​methamphetamine (MDMA) for post-traumatic stress disorder (PTSD), and psilocybin for depression, end of life dysphoria, and alcohol use disorder. The precise mechanisms through which psychedelic drugs can produce benefit and harm are not fully understood. Rigorous research conducted at equipoise is the best path forward for evaluating the therapeutic potential and mechanisms of psychedelic drugs. Policies governing the clinical use of these drugs should be evidence-informed and prioritize protection of public health over the profit motive.

Introduction

The past decade has witnessed an explosion of scientific, clinical, media, and public interest in the potential therapeutic effects of drugs like psilocybin, ayahuasca, and lysergic acid diethylamide (LSD). These drugs have been hailed as transformational cures for depression, addictions, trauma syndromes and many other disorders, and also promoted as methods to enhance creativity, increase empathy, expand consciousness and even prevent war (Cunliffe, 2023; Psychedelic Frontier, 2013). Given the range and number of claims made about psychedelic drugs, it would be rational for a reader to expect that the authors of an Annual Review would have great difficulty selecting which of the many rigorous demonstrations of effectiveness to highlight. However, to prefigure one of our main conclusions, review of the literature led us instead to paraphrase Winston Churchill: “Never has so much been claimed, by so many, with the support of so few solid scientific studies”. This does not obviate the need for this review, for two reasons.

First, the fact that so many studies in this area are of limited quality means neither that all are poor quality, nor that no substantive conclusions can be drawn. The late Roland Griffiths, perhaps the most respected scientist to conduct research in this area, spent the last year of his life warning about the bursting of the “hype bubble” of psychedelic medicine (e.g. Griffiths, 2023; Yaden, Potash, & Griffiths, 2022). His caution is salutatory. That said, some gold nuggets amidst the pyrites merit attention. Second, the psychedelic medicine field provides an important opportunity to illustrate more broadly-applicable principles of how scientific evidence can be misrepresented and/or misinterpreted. Understanding these realities, which we highlight in a series of “warnings to the wise” (see text boxes), can help scholars to better evaluate the current state of evidence in psychedelic medicine as well as in other areas where extra-scientific forces are strong and scientific evidence is weak.

To put our own cards on the table, if even a fraction of the claims being made for psychedelic medicine are true (see box 1), we would be delighted as scientists, clinicians, and human beings. We write therefore not in a spirit of negativism but with knowledge that hopes and reality often part company in this world, and one of science’s most critical roles is to help differentiate the two. In their award-winning 1970 book LSD and Alcoholism, Ludwig, Levine, and Stark described their years of work to develop a “hypnodelic alcoholism treatment” combining psychotherapy, hypnosis and LSD. At the launch of a 4-year randomized clinical trial of hypnodelic treatment, versus LSD only, versus no therapy, they noted that “Because of our prior work with LSD as a therapeutic agent, our stake in the outcome of this study was far from impartial and detached (p.5–6).” Yet when the results were in, the authors’ candor was impressive: “It would have been professionally gratifying to report significant results for the hypnodelic treatment technique…unfortunately, our conscious wishes have yielded to the overwhelming evidence that none of the LSD treatment procedures produced any greater benefit than could be realized by the milieu (no therapy) condition”. Those scholars’ commitment to using rigorous science to evaluate beliefs about medications as well as to change their own beliefs in response to evidence is the animating spirit of this review.

Word to the Wise #1: Resist the temptation to hype results.

Commercial sellers of putative psychedelic cures often make expansive claims of benefit. A similar phenomenon exists in the psychedelic research literature, even beyond the usual level of overenthusiasm to which emerging scientific fields are sometimes prone. In a recent review, van Elk and Fried (2023) documented a number of egregious examples: “Abbar et al. (2022) found in a randomized controlled trial comparing ketamine against placebo that there was no persistent benefit of ketamine over placebo at the exit timepoint of the trial in week 6, but concluded in the abstract that “ketamine [..] has persistent benefits for acute care in suicidal patients”. Ionescu et al. (2016) found in an open label ketamine study that only 2 of 14 patients show sustained improvement at 3 month-follow-up (which may well be due to the placebo effect or other factors), but the title of the paper reads “Rapid and Sustained Reductions in Current Suicidal Ideation”. Palhano-Fontes et al. (2019) concluded in their ayahuasca study (n=14 treatment, n=15 placebo) that “blindness was adequately preserved”, when all participants in the treatment group said they believed they had received ayahuasca, but less than half of participants in the placebo group said so. Daws et al. (2022) compared two treatment arms, including one using psilocybin-assisted-psychotherapy, against each other, concluding that one treatment outperformed the other despite the lack of a statistically significant interaction term between the treatments.”

Poorly substantiated claims about the benefits of psychedelic drugs are also prevalent in mass media outlets. Although some blame for this must go to journalists, Sumner and colleagues (2014) documented that a leading source of hype in media accounts of health-related scientific findings is the university press releases which researchers are typically involved in writing. There have also been instances of scientists writing their own mass media pieces that make therapeutic claims far beyond what their own scientific articles stated (Hall & Humphreys, 2022).

The desire to hype can have many roots, but we needn’t speculate about motives to suggest strongly that careful research carefully described is the best way forward both for science and for public health.

Definitions and Scope

Because the term “psychedelic” is as much a cultural as scientific term, this class of drugs has fuzzy boundaries. It typically includes entirely synthetic drugs (e.g., LSD, ketamine) as well as some that are found in plants (e.g., ayahuasca, mescaline, ibogaine). Clinical trials usually evaluate synthetic forms using standard dosing of active ingredients (e.g., N, N-Dimethyltryptamine instead of lickable toads and lab-produced psilocybin instead of “magic” mushrooms). Other than MDMA, the drugs referred to as psychedelic have the subjective effect of causing perceptual changes and hallucinations which can be quite vivid and prolonged. Other than ketamine and perhaps MDMA, psychedelic drugs do not seem to have significant addictive liability. Their chemical structure is varied and includes tryptamines like psilocybin, ergolines like LSD, and phenylethylamines like mescaline.

Psychedelic drugs can also be defined by their affinity and agonism in the brain. Numerous studies using radioligands demonstrate that psychedelic drugs have affinity for the 5-HT_2A receptor (Aghjanian & Foote, 1968; McClure-Begley & Roth, 2022). Animal research (e.g., in a two-lever paradigm) indicates that several psychedelic drugs fully substitute for one another, and studies that employ serotonergic antagonists (e.g., Barrett et al., 2018) show that the 5-HT_2A receptor is a common factor in the interoceptive properties across multiple psychedelics (McClure-Begley & Roth, 2022). This picture, however, is incomplete as serotonergic psychedelics also show high affinity for a broad range of other receptors, and it is unclear whether these other targets are relevant for the purported therapeutic effects of psychedelic drugs (Sard et al., 2005). A further complexity is that there are non-psychedelic drugs that also have affinity for the 5-HT_2A receptor.

This review addresses the potential of psychedelic drugs to treat diagnosable diseases. The claims that these drugs will expand consciousness, augment creativity, advance social compassion, prevent war, and the like will not be evaluated here.

At this writing, there is insufficient high-quality scientific evidence to review the potential therapeutic benefits and risks of ayahuasca, mescaline, ibogaine, and N, N-Dimethyltryptamine. If phenylcyclohexyl piperidine (also known as PCP or “angel dust”) is considered a psychedelic drug, we are not aware of any therapeutic claims made about it so it too is out of scope for this review. We therefore confine our evidentiary review to psychedelic drugs for which at least some provisional assessments can be made: ketamine, MDMA, psilocybin, and LSD.

Finally, unlike many reviews in this area, we will not revisit the historical, political, artistic, and cultural history of psychedelic drugs, a story that has been well-related elsewhere (Barber, 2018; Dyck, 2008; Pollan, 2019). We would however offer a caution about the potential biasing effect of these drugs’ provenance. (See box 2).

Warning to the wise #2: Aim to be right rather than on the “right side”.

Science at its best is about finding out what is true, but sometimes becomes about being on the “right side”. Two related processes that produce this dynamic are ingroup bias (i.e., evaluating evidence based on whether it is favorable to people like oneself) and badging bias (i.e., expressing a view of the evidence designed to express one’s values in a social context, Perlmutter, Campbell, & MacCoun, 2024). Such “team loyalty” effects can emerge in psychedelic research because the history of these medications in the US is usually cast as a story of left-wing rebels in tie-dye clothes battling establishment kill-joys in grey flannel suits. In the 1950s, when academia was in many respects politically conservative and culturally straight-laced, this could well have produced an underappreciation of evidence suggesting benefits of psychedelic medications. But within academia today (certainly including psychology and psychiatry departments), the political culture is both more left-wing politically than the overall culture and increasingly uniformly so (Gross, 2013). In such an environment, “believing in” psychedelics can become something between an identity marker and an obligation. A combination of cognitive bias and desire for approval can thus lead academics to hold psychedelic medicine studies to lower standards of evidence than they would a treatment with a different political coding (e.g., the Christian Evangelical addiction program Teen Challenge).

In reality of course, the politics of who invented or advocated for a medication are irrelevant to whether it works. Methadone, the best evidenced treatment for opioid use disorder, was developed as a synthetic opioid by the Nazis during World War II. The left-wing history and coding of psychedelic drugs is irrelevant to whether they are effective.

Minimal research quality standards imposed in this review

Psychedelic drugs have been the subject of at least some enthusiasm and optimism among clinical researchers for over 75 years. Over that time, a large number of papers have been published on these drugs. Overwhelmingly these have been case studies, observational studies, single group treatment evaluations, very small sample clinical trials, and in some cases ecological correlational studies. Many psychedelic reviews exhaustively recount all such studies, but we avoid this potentially misleading approach because it dilutes the impact of the important findings from the most rigorously designed studies (see box 3).

Warning to the wise #3: Resist the seductive persuasiveness of a long list of weak studies.

Human beings sometimes come to conclusions by attending to peripheral information that is simple to gather, for example finding a proposition more believable when the number of weak arguments supporting it exceed the number of strong arguments against it (Petty & Cacioppo, 1984). The influence of quantity versus quality of information is inbuilt into some research summary strategies, for example “box score” reviews in which a hypothesis is deemed correct if numerically more studies confirm than reject it (Finney, 2020). To the extent people update prior beliefs in a Bayesian fashion, even the release of higher-quality studies has limited power to overturn incorrect assessments of the state of the science built on many earlier, low quality studies (MacCoun & Paletz, 2009).

This bias is a particular challenge for assessing the effects of psychedelic drugs because an unusually large proportion of evaluation research studies in this area have extremely weak methodology. The most prudent way to read the literature, therefore, is to focus attention on the subset of studies that have a reasonable prospect of generating a valid result rather than on the much longer list of studies that are subject to multiple serious internal and external validity threats.

A key challenge of this review is that if we restrict discussion to research that meets high methodological standards we will only have a handful of studies to discuss for some psychedelic drugs (i.e., all of them other than ketamine). To illustrate this point, consider that Kisely et al.’s (2023) well-conducted meta-analysis of international clinical literature on psilocybin and MDMA screened 837 studies but identified only 14 randomized trials. And only one study on that list met the modest methodological standard of enrolling at least 30 participants in total; none met the equally modest standard of demonstrating that blinding of research staff and subjects had been successful.

In an effort to balance competing needs for rigor and sufficient content to review, we set only two modest standards for inclusion in the section of this review that draws conclusions about the risks and benefits of psychedelic drug treatments. First, given our focus on therapeutic effects, we limit discussion to randomized controlled trials of people with diagnosed disorders. Second, we only discuss studies that enrolled more than 15 participants in each condition. Open-label trials and trials where blinding was never assessed or was assessed and failed will be included, but with appropriate cautions.

Finally, because there are many meta-analyses in the field, it is worth noting that contrary to what is widely believed, meta-analysis does not rectify the problems of tiny research samples. When a meta-analysis is reliant on very small trials, effect sizes tend to be inflated (Beets et al, 2023).

Safety and Effectiveness of Psychedelic Drugs as Treatments for Different Conditions

Ketamine as an augmentation for antidepressant medication for treatment-resistant depression

Ketamine is by far the most studied psychedelic medication and depression (particularly when prior treatments have proved ineffective) is by far its most commonly studied indication. In 2019, the Food and Drug Administration judged the safety and efficacy evidence sufficient to approve S-ketamine (racemic ketamine is an equal mixture of two mirror image molecules, R-ketamine and S-ketamine) as a supplement to anti-depressant medication for treatment-resistant depression. It would supererogatory to recapitulate here every study that informed FDA’s decision. We instead give readers a sense of the literature by highlighting key clinical trials. Readers interested in a more exhaustive discussion are referred to Nikolin and colleagues’ (2023) recent review of all 49 adequately-powered randomized clinical trials (total n = 3299) of racemic ketamine and S-ketamine for depression.

As an organizational note, some reviews attempt to divide trials into those that evaluate ketamine as a single treatment versus placebo and those trials that evaluate ketamine combined with anti-depressant medication. We do not adopt that approach here because, practically speaking, people who enroll in clinical trials for depression are almost always taking anti-depressant medication (see, e.g., Loo et al., 2023). Because attempting to remove research participants from their antidepressant medication in ketamine studies would be unethical, even “comparisons of ketamine treatment versus placebo” are actually comparisons of ketamine + anti-depressant medications with placebo + anti-depressant medication.

The first notable randomized study in this area was a Phase 2 trial conducted by Daly and colleagues (2018), who assigned individuals with treatment resistant depression to placebo (n=33) or different doses of nasal spray S-ketamine (n=34). In a comparison of placebo with all dose levels combined, Montgomery-Åsberg Depression Rating Scores (MADRS; Montgomery & Åsberg, 1979) participants receiving S-ketamine improved significantly more at day 8 and 15. In a larger subsequent discontinuation trial by the same team (Daly et al., 2019), 297 adults with stably remitted depression who were taking anti-depressant medication and nasal S-ketamine, were randomly assigned to continued S-ketamine or a placebo spray. Relapse rates as measured by the MADRS were significantly higher in the placebo group.

Ochs-Ross and colleagues (2020) conducted a double-blind study in which 138 elderly patients with treatment resistant depression without psychotic features were randomized to receive either (1) flexibly dosed S-ketamine nasal spray in combination with a new oral antidepressant or (2) placebo spray in a combination with a new oral antidepressant. Investigators, site personnel, data managers and statistical analysts were blind as to study treatment. A bitter compound was added to the placebo spray in the hopes of making patients blind to condition, but the study reported no assessment of whether this strategy was successful. Outcomes were assessed at 28 days by blinded raters using the MADRS. No significant differences were found between conditions on total change in MADRS score, rates of response (≥50% reduction from baseline in the MADRS total score) or rates of remission (MADRS ≤12).

Another negative trial (N=202) was conducted in Japan by Takahisi and colleagues (2021). Treatment-resistant depression was initially treated with a new anti-depressant, with non-responders randomized to supplemental nasal S-ketamine or placebo. The primary outcome of MADRS score improvement at day 28 did not differ significantly between conditions.

In a randomized trial conducted in 39 outpatient referral centers, Popova et al (2019) compared the efficacy and safety of switching patients from an ineffective antidepressant to a newly initiated antidepressant combined with flexibly dosed S-ketamine nasal spray versus placebo spray containing a bittering agent (though again, the effectiveness of this blinding effort was not assessed). A total of 227 patients were randomized to treatment and 197 completed the 28-day treatment phase. Improvement in clinician-assessed MADRS score was significantly greater in the S-ketamine plus antidepressant group than the antidepressant plus placebo group at day 28. The proportion of patients with a >50% decrease in MADRS by day 2 and day 28 did not differ significantly between conditions. Discontinuation rates because of adverse events were almost 8 times higher in the S-ketamine (7%) than placebo control group (0.9%).

Canuso et al (2018) compared standard-of-care treatment plus intranasal S-ketamine (84mg) or placebo in their capacity to produce a rapid reduction in symptoms in 68 depressed patients who were at imminent risk of suicide. There was significantly greater decrease in MADRS in the S-ketamine group compared with the placebo group at 4 hours and at ∼24 hours, but not at day 25. There was also greater improvement in the S-ketamine group on the MADRS suicidal item score at 4 hours, but not at 24 hours or day 25. Clinician global judgments of suicide risk were not statistically different at any time point.

The often-transitory nature of S-ketamine’s benefits were also evident in Chen and colleagues’ (2023) randomized trial with 252 patients with treatment-resistant depression. MADRS were significantly lower 24 hours after receipt of antidepressants plus S-ketamine versus antidepressants with placebo, but no such difference was evident at 28 day follow-up.

The KADS Study (Loo et al., 2023) differed in multiple respects from the above trials in studying racemic ketamine administered by subcutaneous injection compared with an active placebo (midazolam). After finding no benefits to a fixed dose of ketamine in an initial cohort of 68 individuals, the team ran a second trial (N= 106) with a protocol allowing increased dosing in the event of non-response. In this trial, racemic ketamine produced significantly higher rates of remission (19.6% vs. 2.0% for midazolam placebo) as measured by the MADRS at 4 week follow-up.

As mentioned, adding S-ketamine to anti-depressant medication is an FDA-approved treatment, and clearly some trials have reported positive results. At the same time, a number of trials have reported no benefit and others have found small and transitory benefits. When this is considered in light of S-ketamine’s addictive liability and adverse events profile, its clinical utility appears limited. The KADS study raises the possibility that racemic ketamine might have more effectiveness, as does some other data (see Nikolin et al, 2023), but further research is needed to take this hypothesis beyond the realm of conjecture.

More generally, a major problem with all the above trials is that ketamine’s characteristic and vivid dissociative effects are hard to disguise, even with active placebos like midalozam. This may mean that estimates of the benefits of ketamine and S-ketamine were inflated by expectancy effects.

To wit, Lii et al (2023) conducted a small (N=40), randomized, placebo-controlled trial (N=40) to assess the antidepressant efficacy of intravenous racemic ketamine given to depressed individuals undergoing surgery under anesthesia, when patients would not be able to know their treatment condition (the anesthetist and clinical assessor were also blinded). MADRS ratings showed that participants in each condition had similar and moderate levels of depression at baseline. At 1, 2 and 3 days after the ketamine or placebo infusion, there was no statistically significant difference between the two groups on MADRS. Only 36.8% of participants correctly guessed their condition, which is significantly less than chance. Individuals whose depression abated were significantly more likely to believe that they received ketamine whether they did or not, suggesting that expectancies can significantly alter the findings of trials in this area. This cleverly designed study merits replication in a larger sample, and raises concerns that unblinded trials of ketamine are significantly biased.

Ketamine for depression compared to other treatments with no placebo condition

Anand et al (2023) conducted an open-label, randomized trial to test the non-inferiority of racemic ketamine to ECT in outpatients with treatment-resistant major depression and no psychotic symptoms. A total of 403 patients were randomized, but 38 withdrew before the initiation of treatment, leaving 195 patients received twice-weekly ketamine infusions and 170 receiving thrice-weekly ECT over 3 weeks. Just over half (55.4%) of patients in the ketamine group and 41.2% in the ECT group benefitted on the primary outcome of a 50% or more decrease in self-reported depressive symptomology. Improvement in quality-of-life was similar in the two treatment groups. Anand et al concluded that ketamine was noninferior to ECT on the primary outcome of depression response.

S-ketamine has abeen compared to extended-release quetiapine, which although developed as an atypical antipsychotic medication, is approved by the Food and Drug Administration (FDA) as an augmentation treatment in treatment-resistant depression. In an open-label, single-blind multicenter clinical trial of patients taking traditional antidepressant medication, 336 patients were randomly assigned to receive flexibly dosed S-ketamine nasal spray and 340 were assigned to receive extended-release quetiapine (Reif et al., 2023). More patients in the S-ketamine group than in the quetiapine group achieved remission at week 8 as measured by a score of 10 or less on the MADRS (91 of 336 patients [27.1%] vs. 60 of 340 patients [17.6%]). More of the patients who received s-ketamine stayed in remission through week 32 after being in remission at week 8, namely, 73 of 336 patients [21.7%] vs. 48 of 340 patients [14.1%]). Over 32 weeks of follow-up, the percentage of patients in remission and the change in the MADRS from baseline was greater in those given the S-ketamine nasal spray.

Notably, almost twice as many patients discontinued quetiapine than S-ketamine (137 [40.3%] vs. 78 patients [23.2%]) either because of the adverse effects of quetiapine or because the treatment lacked perceived efficacy. The percentage of patients with remission increased over time in both groups. Remission at week 32 occurred in 49.1% of patients in the S-ketamine group compared with 32.9% of those in the quetiapine group (55.0% and 37.0% of patients). MADRS also decreased over time in both groups but the S-ketamine group showed a greater reduction than the quetiapine group at each time point.

Drawing strong conclusions from two trials would not be prudent. That said, the above results are encouraging regarding ketamine being a potential alternative to other treatments for depression. This is particularly true of Reif and colleagues’ (2023) trial of quetiapine and ketamine, one of very few studies in the field to have a longer-term follow-up assessment.

Ketamine for substance use disorders

Initial evidence from Russia that racemic ketamine may be effective in sustaining abstinence from alcohol (Ezquerra-Romano et al., 2018) led Grabski et al (2022) to conduct a clinical trial with 96 severe alcohol use disorder patients. Participants were randomly assigned to one of four conditions: 1) three weekly ketamine infusions (0.8 mg/kg i.v. over 40 minutes) plus psychotherapy, 2) three saline infusions plus psychological therapy, 3) three ketamine infusions plus alcohol education, or 4) three saline infusions plus alcohol education. Ketamine was well tolerated and produced no serious adverse events. Intent-to-treat analysis found significantly more (10.1%) days abstinent from alcohol in the ketamine than the placebo group at 6-month follow-up. However, relapse rates were comparable between the ketamine and saline (placebo) groups. The authors acknowledged that blinding may not have been successful, and suggested that future studies should include an active control.

A small randomized trial (N=40) adopted such an approach by using midazolam as an active placebo compared to a single ketamine infusion in individuals receiving 5 weeks of motivational enhancement therapy for alcohol use disorder (Dakwar et al., 2020). Those who received the ketamine infusion at week 2 had higher rates of abstinence for the remaining 3 weeks of treatment than did those receiving midazolam. Unfortunately, neither the success of blinding nor longer-term outcomes were assessed.

The same team conducted a very similar small trial for individuals with cocaine use disorder (N=55) receiving behavioral treatment in an inpatient setting followed by mindfulness-based relapse prevention on an outpatient basis for 4 weeks (Dakwar et al., 2019). The rate of urinanalysis-confirmed abstinence in the final two weeks of the trial was markedly higher (48.2%) in the ketamine condition versus the midazolam placebo condition (10.7%). As in the team’s alcohol trial, blinding effectiveness and longer-term outcome were not assessed.

Despite their limitations, these trials send a positive therapeutic signal. The evidence in this area seems more than sufficient to justify investment in large clinical trials of ketamine’s effect on alcohol and cocaine use disorder.

Ketamine for post-traumatic stress disorder

Some early, small sample clinical trials reported large benefits of ketamine in patients with post-traumatic stress disorder (PTSD, see Borgogna et al., 2024, for a review). However, the only randomized trial with an acceptable sample size (N=158) found no differences in self-reported or clinician assessed PTSD symptoms in patients who received intravenous infusions of low-dose ketamine, high-dose ketamine, or placebo (Abdallah et al., 2022).

MDMA for post-traumatic stress disorder

Two clinical trials of MDMA for PTSD have been conducted by a consortium assembled and funded by the Multidisciplinary Association for Psychedelic Studies (MAPS), an advocacy organization. The first enrolled individuals with severe PTSD (Mitchell et al., 2021) and the second enrolled individuals with moderate to severe PTSD (Mitchell et al., 2023).

The first trial randomized participants to three sessions of MDMA dosing (80–180mg per session) separated by 4 weeks versus placebo dosing. All participants received 3 hours of manualized therapy prior to the first dose and three weekly 90 minute therapy sessions after each dosing session. The primary outcome was change in the Clinician Administered PTSD Scale for DSM-5 (CAPS-5, Weathers et al., 2018) between baseline and 18 weeks. The mean difference in change in CAPS-5 scores was significantly larger in the MDMA than placebo group.

Although Mitchell and colleagues (2021) describe this trial as double-blind, the study did not formally test for blinding. The published paper argued that blinding was successful because “at least 10% had inaccurately guessed their treatment arm”. But this logically implies that up to 90% of participants could correctly guess their assigned arm. As multiple scholars noted (Burke & Blumberger, 2021; Flameling et al, 2023) this raises significant concern that expectancy biases affected the results through placebo and nocebo effects.

Mitchell and colleagues (2023) second trial had a similar design, randomizing individuals to psychotherapy either with MDMA (n=53) or placebo (n=51). Severe adverse events were more common in the MDMA (9.4% of participants) versus placebo (3.9%) condition. At 18 weeks, both groups experienced large decreases in symptoms as measured by the CAPS-5, with this improvement being significantly larger in the MDMA group. This trial improved on the original study by conducting a formal assessment of blinding, which showed that 75% of individuals in the placebo condition and 94.2% of those in the MDMA condition correctly guessed what they had received. Expectancy effects may thus account for some or all of the difference between groups at follow-up assessment. Heigthening this concern, the Institute for Clinical and Economic Review (Mustafa et al., 2024) recently argued in a quite critical report that because MAPS is embedded within the community of psychedelic drug enthusiasts, enrollment in its studies may have been skewed toward individuals with strong faith in the value of MDMA and a desire to convince others to share their enthusiasm.

Based mainly on these trials, MAPS has applied to the FDA for approval of MDMA-assisted psychotherapy as a treatment for PTSD (Adams, 2023). As the results of both trials were positive, and adequate blinding is not required for approval of medications under FDA rules, it seems likely that this treatment will be approved for clinical use in the near future.

Psilocybin for major depressive disorder

Goodwin et al. (2022) conducted a phase 2, multisite, double-blind, randomized trial of psilocybin for treatment-resistant depression in 10 countries. Participants were randomized to receive a single dose of psilocybin 25mg (n=79), 10mg (n=75), or 1mg (n=79; hoped to produce a sub-perceptual placebo equivalent effect). All participants met with a psychotherapist for 3 sessions prior to, and 2 sessions following psilocybin administration and were followed for 12 weeks after administration. MADRS improved in all groups at 3 weeks after drug administration, but only the 25mg dose was superior to placebo equivalent. Worryingly, suicidality increased from baseline to week 3 occurred in 11 (14%) participants in the 25-mg group, 13 (17%) in the 10-mg group, and 7 (9%) in the 1-mg group. Three participants reported suicidal behavior between weeks 3 and 12, all of whom received the 25mg dose.

A similar though smaller clinical trial randomized depressed individuals to either two doses of psilocybin 25 mg (n=30) three weeks apart and daily placebo pills versus psilocybin 1 mg/day (considered sub-perceptual placebo dose) plus escitalopram 10–20 mg/day (n=29) during 6 weeks of treatment (Carhart-Harris et al., 2021). All participants received 3 hours of preparatory psychotherapy, two in-person integration psychotherapy sessions, and 6 additional integration psychotherapy phone calls. Unlike in Goodwin et al.’s trial, there were no increase in suicidality or other serious adverse events, but neither did the treatment groups differ on the primary outcome of depressive symptoms at the end of treatment.

Employing a different active placebo (100mg of niacin), Raison et al. (2023), conducted a phase 2, multisite, randomized trial of single dose psilocybin 25mg for major depression. All participants received 6 to 8 hours of preparatory psychological support and 4 hours of therapy in the 43 days following dosing. Participants randomized to psilocybin (n=51) experienced significantly greater reductions in the primary outcome of change in MADRS at day 43 than those randomized to placebo (n=53). Adverse events were more than twice as common in the psilocybin than placebo condition, but there were no reports of suicidal behavior during the follow-up period.

Von Rotz and colleagues (2023) filled a gap in the literature by examining the potential effects of a moderate dose of psilocybin (~15mg for a 70kg person) as most other clinical trials compare higher doses to subperceptual low-dose psilocybin. Participants (N=52) were randomized to psychological support with psilocybin or placebo. At 14 days, participants receiving moderate dose psilocybin experienced significantly greater decreases in MADRS than did the placebo group. No serious adverse events were reported.

In summary, most (though not all) trials in this area report better outcomes for depressed individuals receiving psilocybin in combination with psychological support than those receiving such support in combination with an inactive or active placebo. Some studies suggest a risk of adverse events, up to and including suicidality, from psilocybin, suggesting a need for careful prescribing and close monitoring of patients. None of the trials in this area assessed participants for effectiveness of blinding or expectancies, so again, the reported results may be significantly biased.

Psilocybin for end of life anxiety and depression

Palliative care represents an intriguing and potentially impactful setting in which to use psychedelic medications. The best randomized trial in this area compared psychotherapy combined with high dose psilocybin (22 to 33 mg/70 kg) to sub-perceptual-dose psilocybin (1 or 3 mg/70 kg; intended as a placebo dose) in 51 participants with life-threatening cancer and a diagnosis of anxiety and/or mood disorder (Griffiths et al., 2016). All participants received both doses of psilocybin in random order a mean of 38 days apart. Participants also received a mean of 7.9 hours of preparatory psychotherapy prior to the first session, a mean of 3.4 hours integrative psychotherapy following the first session, and mean of 2.4 hours integrative psychotherapy following the second session. Approximately 5 weeks after high-dose sessions, participants experienced improvements in clinician-rated depression and anxiety compared with 5 weeks following sub-perceptual-dose psilocybin. In planned pooled analysis across dose conditions, 71% of participants experienced remission of depressive symptoms and 63% experienced remission of anxiety symptoms 6 months following their last dose. Investigators assessed for integrity of psychotherapy blinding by asking therapists to rate the dose magnitude following each dosing session. Therapists incorrectly guessed the psilocybin dosing assignment in 13% of high doses sessions and 12% of low-dose sessions, which means blinding was largely unsuccessful.

Psilocybin for alcohol use disorder

The first adequately-powered clinical trial of psilocybin for substance use disorder evaluated psilocybin-assisted psychotherapy (n=49) versus diphenhydramine (n=46) in participants with DSM-IV alcohol dependence who reported at least 4 heavy drinking days in the prior 30 days (Bogenschutz et al., 2022). Participants were randomized to receive two medication sessions of psilocybin (25 mg/kg for first session; 25 mg/kg to 40 mg/kg in second session) or diphenhydramine (50mg in first session; 50mg-100mg in second session), administered four weeks apart. All participants received 12-weeks of manualized psychotherapy including motivational interviewing and cognitive behavioral therapy. The percent of heavy drinking days decreased from 56 days in the 12 weeks before screening to 10 days at 32 weeks in the psilocybin group and from 49 days in the 12 weeks before screening to 24 days at 32 weeks in the diphenhydramine group (mean difference 13.9%). There were no serious adverse events in the psilocybin group. Blinding failed: both participants and therapists correctly identified the treatment assignment group for more than 90% of dosing sessions. Expectancy effects may therefore partly or fully account for the apparent results.

LSD for alcohol use disorder

In the late 1950s, when psychiatric research standards were substantially less rigorous than today, clinicians at Saskatchewan Hospital reported that LSD greatly improved the outcome of severe alcohol use disorder patients (Barber, 2018;Dyck, 2008;Mangini, 1998). These claims were greeted with skepticism because they came from uncontrolled studies involving small patient samples, with outcomes assessed by having psychotherapists (some of whom were enthusiastic users of LSD themselves) gather data on their own patients at 6-month posttreatment follow-up (Grinspoon, 1981).

Subsequently, a small randomized trial conducted at the Addiction Research Foundation in Toronto in the early 1960s compared the effects on heavy drinking of an 800 mcg dose of LSD with ephedrine and treatment as usual (Barber, 2018; Mangini, 1998). In this study (and similar small sample trials) the investigators reported substantial improvements in patients for all treatment conditions but no better outcomes for patients given LSD (Mangini, 1998). As mentioned in the introduction of this review, despite their great enthusiasm for LSD-based treatment of alcohol use disorder, Ludwig and colleagues’ (1970) well-designed clinical trial found no benefits at any follow-up window of LSD alone or in combination with a psychotherapy specifically designed to work synergistically with it. A later meta-analysis of the five randomized controlled trials of LSD conducted between the late 1960s and 1970 (with a median of 44 patients per trial) showed better short-term alcohol outcomes in individuals receiving LSD-based treatment, but these dissipated by 12 month follow-up (Krebs, 2012).

Summary of Effectiveness Evidence for Psychedelic Medicines

Because this review was limited to studies with at least a few quality indicators (albeit rather modest ones) it was not able to identify many studies for psychedelic drugs other than ketamine. Notably, there is at present no convincing evidence one way or the other on whether the therapeutic claims made for ayahuasca, mescaline, ibogaine, and N, N-Dimethyltryptamine have merit.

Many of the studies that were identified have serious threats to validity such as insufficient demonstration of blinding, incomplete reporting, and insufficient controls for expectancy effects. Follow-ups are generally short-term, sometimes extremely so (e.g., 24 hours or less). The conclusions drawn in this foregoing discussion of available trials should thus all be viewed as tentative.

To recapitulate the major themes, ketamine (whether racemic or S-ketamine) is the most extensively evaluated psychedelic drug, and individual trials suggest it is non-inferior to ECT and superior to the atypical antipsychotic quetiapine for treatment-resistant depression. Although approved by FDA as safe and effective as an augmentation of anti-depressant medication for treatment-resistant depression, S-ketamine’s empirical record is uneven, with some positive trials as well as multiple trials that find no or minimal (i.e., small or transient) benefit. Given that it has a high rate of adverse events and is addictive, its clinical utility appears limited. The study by Lii et al suggests that expectancy effects are significant in evaluations of ketamine’s effect on depression, which could both bias individuals who knew they received ketamine to overreport benefit and those who desired it and knew they did not receive it to register their disappointment in their self-reported depression score. That Lii et al’s negative trial is the only one in this area to have shown that blinding was effective, it is reasonable to worry that all the positive trials of ketamine are positively biased by expectancy effects.

There is no compelling evidence that ketamine is beneficial for PSTD. For other drug-indication pairings, there is a positive signal. Ketamine may have promise in the treatment of substance use disorders (specifically alcohol use disorder and cocaine use disorder), as might MDMA for treatment of PTSD. Psilocybin research has generated positive initial signals for depression, end of life dysphoria, and alcohol use disorder. Again, due to the absence of successful blinding and the likely presence of positive and negative expectancy effects, none of these signals should be taken as definitive without replication in methodologically stronger studies. At the same time, all of them certainly warrant the investment in such attempted replications.

Finally, in terms of LSD for alcohol use disorder, although assessing research conducted by the dominant (very low) scientific standards of 60 years ago is challenging, our judgement is there is no compelling evidence of enduring benefit. LSD should be a lower research priority than psychedelic drugs with a stronger therapeutic signal for substance use disorders.

Mechanisms of action by which psychedelic drugs could generate benefit and harm

The review has highlighted the many reasons why conclusions about the therapeutic effects of psychedelic drugs must be tentative. Presuming however there is in fact “a there there”, attention naturally turns to what mechanisms could explain how psychedelic drugs can benefit or damage those who take them. Theory and findings on effect mediators are even more embryonic than the efficacy assessment literature, but four questions are currently hubs of important discussion and debate.

1. What are the effects of psychedelic drugs in the brain?

The principal molecular mechanism by which psychedelic drugs elicit their subjective acute effects is the 5-HT_2A receptor (see recent reviews by Jaster and González-Maeso, 2023 and Gumpper and Roth, 2024). This may simultaneously be a true and also insufficiently complex explanation. Off-target actions of psychedelics have also been established, perhaps most consistently for the 5-HT_2B receptor on which all known psychedelics act as agonists (Ray et al, 2019).

Many scholars argue that the effect of psychedelics on psychopathology is similar to that of traditional antidepressants (Moda-Sava et al., 2019): rapid and sustained alterations of synaptic plasticity (that is, augmented dendritic spine formation, Jones et al., 2009; Shao et al, 2021). Again, however, some complexities arise. Many groups have drawn links between 5-HT_2A activation by psychedelics and psilocin (the psychoactive product of ingested psilocybin) with brain-derived neurotropic factor (BDNF) release, which in turn activates TrkB receptors (Cameron et al., 2023; Moliner et al., 2023). This suggests an alternative hypothesis that the potential antidepressant effects of classical psychedelics may be mediated independently of 5-HT_2A activation.

Not all psychedelics are 5-HT_2A receptor agonists and not all 5-HT_2A receptor agonists have psychedelic effects. A number of novel 5-HT_2A receptor agonists have been developed which appear to have therapeutic effects in animal models but do not induce the animal analogue indicators (e.g., a head twitch response) of human psychedelic experience (Cao et al., 2022; Kaplan et al., 2022). This opens the possibility of developing therapeutics that produce the benefits of psychedelic drugs in the absence of their subjective effects.

The activation of 5-HT_2A receptors by psychedelic drugs indirectly activates glutamate networks and downstream increases in BDNF, which in turn supports growth and maintenance of neurons and enhances neuroplasticity. This may be at least a partial mechanism by which psychedelics could induce antidepressant effects. BDNF levels are reduced in major depressive disorder, and interventions for depression increase BDNF (Lee et al., 2007). Ayahuasca, psilocybin and LSD have been shown to increase BDNF levels (de Almeida et al, 2019).

2. Do psychedelic drugs account for change themselves or do they create a context that increases patients’ ability to resolve their disorder?

A behavioral catalyst model proposes that psychedelic drugs are insufficient to cause a therapeutic effect, but can facilitate a more traditional psychotherapeutic process (Pokorny et al, 2016). A similar view is that psychedelic drugs accelerate the natural course of recovery by overcoming obstacles that would otherwise impede psychotherapy (Mithoefer, Grob, and Brewerton, 2016). Any corresponding increase in neuroplasticity would therefore reflect experience-dependent learned adaptive change rather than drug effects independent of experience.

For example, psychedelics may function within psychotherapy by temporarily weakening the certainty of previously held beliefs (e.g., that depression is due to enduring character flaws and will never go away), thereby allowing the individual to revise them. Studies supporting this hypothesis find that psychedelic drugs temporarily increase cognitive flexibility and reinforcement learning in both human and animal model experiments (e.g., Doss et al., 2021).

3. Are the vivid subjective effects of psychedelic drugs essential to their therapeutic effect?

Researchers vigorously debate whether ecstatic or hallucinogenic experiences induced by psychedelic drugs are therapeutic or are epiphenomena (see e.g., Olson 2020; Yaden & Griffiths, 2020). Some experimental data suggests that the subjective experiences of individuals using psychedelics does not necessarily have a causal relationship with any therapeutic outcomes. For example, Moda-Sava and colleagues (2019) showed that when the dendritic spine growth induced by ketamine administration was photablated in mice, the antidepressant effect of ketamine was eliminated, establishing a causal role for spinogenesis in the therapeutic effect. Parallel studies have not been conducted with serontonergic psychedelics.

4. Are any therapeutic effects of psychedelic drugs setting-specific?

Clinical researchers devote significant attention to establishing a particular environment (the “setting” in “set and setting”) in which psychedelic drugs are administered to patients. The majority of trials have been performed in environments where participants are lying down on a couch, and sensory stimulation is limited or highly controlled (Johnson et al., 2008). This is very different from studies of the acute and persistent effects of psychedelics in other animals where relatively little attention is paid to environmental factors, despite well-known context dependent effects of other drugs (e.g., opioids) in both humans and rodents (Badiani et al., 2000).

Unravelling the potential interactions between the effects of drug and setting would require very well-powered clinical trials that randomly assign distinct settings and therapeutic procedures to patients across different doses of the same psychedelic drug. We are not aware of any trials of this sort to date.

Policy Implications

The FDA has granted psilocybin and MDMA “breakthrough therapy” status and also provided guidance to researchers studying psychedelics on how to conduct studies that meet FDA standards. As approval applications are submitted, FDA should proceed with proposed psychedelic medications like any other, i.e., if rigorous research demonstrates safety and efficacy, new psychedelics should be approved. This of course presumes that clinical trials are of high-quality, e.g., use reasonable sample sizes, credible comparison conditions, verified blinding of researchers and participants, and analyze pre-registered outcomes.

Some advocates have argued, most notably in Australia but also in the U.S. and Europe, that psychedelic drugs should be made available for compassionate use in advance of their (presumed) approval by regulatory agencies. Similar arguments have been made in the past for many drugs (recent examples include ivermectin for COVID-19 and cabotegravir + rilpivirine for HIV) with varying effects. The strongest case to do this for psychedelic drugs is for psychiatric distress in terminal illness, where there is encouraging evidence available and any potential risks of long-term harm are less relevant.

Beyond that extreme case, the FDA and indeed everyone else should be cognizant of the history of “medical cannabis” as an example of how medical language and putative health goals can be misused for other purposes. The concept of medical cannabis was created by some activists with the explicit purpose of being a route to recreational legalization (Hall & Lynskey, 2020). It also received generous financial backing from the for-profit cannabis industry. As a result, use of cannabis for putatively medical purposes rapidly outpaced the still limited scientific evidence for its safety or efficacy.

Policymakers, scientists, clinicians, and activists for therapeutic psychedelic drugs should also be realistic that once these drugs are prescribed, there will be extensive off-label prescribing for other conditions, as well as extensive prescribing without any supportive psychological interventions. Medical psychedelic practice after FDA approval will more closely resemble the medical cannabis industry than traditional, well-regulated health care, both for financial (e.g., rapidly writing prescriptions to as many patients as possible is lucrative) and pragmatic reasons (e.g., many healthy individuals want to use these drugs and will falsely claim psychiatric disorders to get them). To cite a relevant example, FDA’s 2019 approval of S-ketamine for the narrow indication of anti-depressant medication supplementation in treatment-resistant depression led to a rapid promulgation of more ketamine providers than would be needed for such a narrow indication, and FDA has been issuing warnings ever since (FDA, 2022; 2023) about the widespread, unevidenced (but highly profitable) promotion and use of ketamine for a broad range of disorders.

To be clear, voters who wish to legalize psychedelic drugs for non-medical purposes, have every right in a free society to endorse that policy, as of course do elected officials who hold the same view. But cloaking such goals in medicalization is bad for public health and for trust in medicine (Shover & Humphreys, 2019). For example, based on weak ecological correlational studies (that not incidentally did not even stand up to replication on their own terms, Shover et al., 2019), the cannabis industry aggressively promoted medical cannabis as the solution to the opioid crisis (Humphreys et al, 2022). Some individual providers even encouraged people addicted to opioids to stop taking FDA-approved medication and use cannabis instead (Humphreys & Saitz, 2019). It would be impossible to know definitely how many people were harmed by such false claims and abuse of the trust that medicine commands, but it is surely not zero. Rather than repeat this experience with psychedelic drugs, dishonest efforts at medicalization and any distorting influence of for-profit industry should be vigorously resisted (see box 5).

Word to the Wise #5: Beware the power of loot and clout to distort science and regulation.

The influence of opioid manufacturers and their owners was a significant factor in the emergence of the opioid crisis which is still ravaging North America today. They callously misled the public, distorted the conduct and interpretation of scientific studies, subverted medical practice, and corrupted some regulators, elected officials, and academic institutions, reaping enormous profits in the process (Humphreys et al, 2022).

The deep-pocketed companies and investors in the emerging psychedelic drug industry are rapidly accruing loot and clout in the arenae of science and policy. Companies obviously have a right to manufacture and sell FDA-approved medications and to conduct and fund research. But all ethical people interested in psychedelic medicine should be wary of how big money can influence the results and interpretation of research studies, the education of students, and the decisions of drug regulators in ways that advance private profit more than public health. This is particularly the case for psychedelics that are surely (e.g., ketamine) or possibly (e.g., MDMA) addictive, but it applies across the whole class of drugs.

Australia as a case study in regulatory failure

In 2023 Australia’s drug regulator, the Therapeutic Goods Administration (TGA), decided to permit the prescription of psilocybin and MDMA as unapproved drugs under the Approved Prescriber provisions of the Special Access Scheme (SAS). This Scheme allows psilocybin to be used to treat intractable depression and also permits MDMA to be used to treat PTSD. Prescribers must be psychiatrists who provide these treatments under the ethical oversight of either the Australian New Zealand College of Psychiatrists or a Hospital Ethics committee.

The SAS of the Australian Therapeutic Goods Act was originally introduced during the HIV/AIDS epidemic to allow the prescription of unapproved medicines by any medical practitioner to a named patient after obtaining the TGA’s approval. The SAS also allowed “Authorised Prescribers”, i.e., physicians recognised by the TGA as experts in the treatment of certain classes of disorder (e.g. psychiatric disorders, neurological diseases) to prescribe unapproved medicines to a class of patients. These prescribers were required to notify the TGA after doing so and to provide bi-annual reports on the numbers of patients they treated and any adverse effects. The SAS and Authorized Prescriber provisions were designed to allow patients with serious illnesses to get early access to drugs that were undergoing clinical trials or approved in other countries but not available in Australia.

The TGA decision on psychedelics was contrary to the advice of TGA’s own expert panel and was also opposed by the Royal Australian and New Zealand College of Psychiatrists and the Australian Psychological Society. These bodies were overruled after investigators promoting psychedelics visited Australia (funded by Mind Medicine Australia) and gave public lectures and talks to government officials advocating the therapeutic use of psychedelics in advance of their formal approval as medicines. This campaign received generally uncritical and positive media attention. Mind Medicine Australia also conducted an intensive lobbying campaign using former politicians on their advisory board to target state and Federal ministers and policymakers. Mind Medicine Australia had previously made several unsuccessful applications to have psychedelic drugs rescheduled for unapproved medical use. It has operated a training program for psychedelic therapists (which costs over USD$7000 to enroll for and attend) in anticipation of obtaining TGA approval for psychedelics (Mind Medicine Australia, 2024).

The campaign for approval of psychedelic drugs followed the model used to secure the legalization of medical cannabis as an unapproved therapeutic good in Australia in 2016. In the case of medical cannabis, the SAS initially limited prescribing to specialized physicians and pharmaceutical cannabis products and it restricted the medical conditions for which cannabis could be prescribed. Once that foot was in the door, these restrictions were progressively loosened in response to lobbying. The SAS now allows the prescription of more than 250 different cannabis products. Family medical practitioners (and in some states nurse practitioners) can prescribe these products for any condition if they believe it may be therapeutic. For profit medical cannabis centers have proliferated to provide patients with prescriptions for medical cannabis and to facilitate their access to the drug. For profit treatment providers will likely place similar pressure on the TGA to liberalize regulations governing access to unapproved psychedelic drugs (Graham et al., 2023).

The U.S. experience of medical cannabis indicates that promises to carefully regulate a drug are often forgotten once a state allows a for profit industry to exist (Humphreys & Shover, 2020). Jurisdictions that allow provision of psychedelic drugs have a public health responsibility to rigorous evaluate the safety and outcomes of people receiving these drugs and associated services, and to adjust regulation as needed in response.

An additional concern is that early and ready access to unapproved psychedelic drugs may make it harder to complete clinical trials of these drugs. The high costs of psychedelic assisted psychotherapy are also likely to produce demands for public subsidies to offset the costs of providing the treatment, as has happened with medical cannabis.

Conclusion: How can science increase the likelihood that psychedelic drugs are a net benefit to medicine and public health?

Science is only one input into public health policy (Humphreys & Piot, 2012). That said, research evidence is a highly important input and we therefore close with some suggestions on how it can help facilitate the use of psychedelic drugs doing more good than harm to patients 10, 20, or more years from now.

Research on psychedelic medicine has improved methodologically in recent years. For example, randomized trials are more common, small pilot studies are beginning to progress to larger phase 2 and phase 3 trials, and blinding success is assessed at least occasionally (even if this has so far generally proved that blinding has failed, Nayak et al, 2023). That said, more strides are needed to develop a more credible experimental research base on these drugs’ therapeutic potential as well as their risks. Extended discussion of areas for improvement have been offered by multiple teams (e.g., Aday et al, 2022; Nayak et al., 2023; van Elk and Fried, 2023). Here we highlight the three areas we consider highest priority.

First, the field needs to develop better methods of blinding conditions. Relatedly, it needs to assess participant expectations and blinding success/failure in every trial, report the results of those tests, and take them seriously in the discussion of results. There are good reasons to assume that even studies described as “blinded” in the psychedelic literature are not blind in reality. Difficulties in blinding are well-understood by psychotherapy researchers. But the problem is much more severe in psychedelic drug research. As disappointing as this may be to psychotherapy gurus, most patients in clinical trials are not able to differentiate between for example Gestalt, Rogerian, and Adlerian therapy, if they know what they are at all. But it is hard to imagine a person not knowing the difference between experiencing and not experiencing the vivid subjective effects that psychedelic drugs can produce. In combination with the high expectations generated by the hype around psychedelic drugs, blinding failures are probably more destructive to valid inferences than they are for other psychiatric medications for which expectations are less inflated (e.g., another anti-depressant medication for someone who has already had poor outcome on other anti-depressant medications). Concerns about blinding are exacerbated by most evaluations of the effects of these drugs using an inert placebo (Nayak et al, 2023).

Lii and colleagues’ trial of depressed individuals receiving ketamine during surgery (described earlier) should command high attention in the future. This study employed a more effective blinding strategy than any in the ketamine literature and found that reduction in depressive symptoms was not reduced by ketamine, yet subjects believed that it was. Indeed, a number of individuals in the control condition hailed the life-transforming power of ketamine as a successful treatment for their depression (Bai, 2023), which ought to raise worries about the validity of this literature more generally. Testing psychedelic drugs under more effective blinds is thus critical to the future credibility of the field.

Second, in addition to recruiting larger samples, clinical trials should enroll participants who are more representative of the patients who will receive these treatments if they are FDA approved. Currently, about half of patients in trials are experienced with psychedelic drugs (Nayak et al, 2023) against about 10% of the general population (Krebs and Johansen, 2013). In addition to making blinding success even less plausible than in naive populations, it suggests that research may be built on an enthusiastic base of trial participants and thus overstate real-world benefits, because enthusiasts may over-report positive benefits when they get the drug and reflect their disappointment in reporting an outcome when they don’t. Further, the general population of patients may experience higher rates of adverse events than do individuals who have prior experience with psychedelic drugs and choose to enroll in a trial where they may take one again.

More generally, like many other areas of clinical research less use of eligibility criteria would increase the likelihood that results apply in practice (Humphreys, 2023). Careful reporting of the operationalization of the exclusion criteria as well as the number of individuals excluded by each should be universal in randomized clinical trials.

Third, to take a lesson from other pharmaceutical research areas, longer follow-ups are required in clinical trials. Short-term outcomes, both positive and negative, often change over in the weeks and months after the typical trial ends. Regulators, prescribers, and patients have a right to know about longer-term effects of psychedelic drugs, and scientists should take that obligation seriously.

Conducting higher-quality research will require higher research budgets. That is largely a responsibility of government, which multiple agencies are now, to their credit, taking seriously. Quality research also requires the involvement of researchers who want to evaluate rather than evangelize, and who are prepared to creatively correct the prevalent design problems noted in this review. The challenges of doing so are many, but the possibility of helping people with serious illnesses more than justifies the endeavour.

Word to the Wise #4: Understandable desperation for cures can affect judgements of treatment effectiveness and risk.

Many of the conditions which psychedelic drugs hope to cure, including post-traumatic stress disorder and treatment-resistant major depression, cause enormous suffering to those who experience them and their families. Intense desire for a cure is understandable, compassionate, and laudable. That said, “wishful thinking” can sometimes bias judgements in an overoptimistic direction (Krizan and Windschilt, 2009; MacCoun, 1998). All consumers of psychedelic research should recognize the tendency to see proof of benefit in ambiguous data when the humanitarian stakes are high.

The well-documented phenomenon of risk seeking in the domain of losses (i.e., if some loss is certain, willingness to risk even greater loss increases, Kahneman & Tversky, 1984) may lead patients suffering from serious disorders to underestimate the risks of psychedelic drugs. Rettig and colleagues (2007) documented how desperation led to faith in the powers of high-dose chemotherapy with autologous bone marrow transplantation to cure late-stage breast cancer. Patients and their families did not want to hear that there was negligible evidence supporting this treatment, and policymakers and scientists who knew better were not keen to deliver this bad news. As a result, oncologists offered this costly, grueling, and ineffective treatment to already suffering cancer patients, many of whom died in agony. The admirable desire to ease suffering and to give hope may contribute to hyping the benefits of psychedelic medications. Without criticizing the impulse, we note that while diseases can be cruel, offering false hope of a cure can be even crueler. The most ethical and compassionate thing we can say when we have no evidence that a treatment works is just that.