The challenges ahead for psychedelic ‘medicine’

Abstract

With the extensive public, commercial and scientific interest from what has been widely termed the psychedelic renaissance, it is important that the scientific practices and results obtained from its implementation into medicine are put under a critical microscope. While there are numerous works on the potential benefits and applications of psychedelics as medicines, relatively little has been written about the challenges this field will face when incorporated into modern medical practice. Indeed, as a new or at least revived area of investigation, psychedelic medicine has a particular set of challenges which need to be addressed. In this viewpoint, we identify a number of these challenges. First, challenges related to the design of individual research studies are discussed, particularly focusing on current practices surrounding blinding, expectancy, the use of therapy and sources of bias. Second, the broader context of the research environment is considered, including how medical science typically establishes evidence, funding bodies and the impact of psychedelics being scheduled at odds with their risk profile. Finally, we describe challenges relating to the implementation of psychedelic therapies into modern medicine, considering the social and economic context. Alongside, we provide suggestions for what could be included into current research protocols to mitigate these challenges.

Introduction

The potential therapeutic applications of psychedelics as medicines to treat disorders such as post-traumatic stress disorder, depression and anxiety (see Reiff et al., 2020, for review) have generated wide-ranging public and clinical interest. This viewpoint highlights and addresses the particular challenges that psychedelic medicine face, to help ensure research is conducted and scrutinised with scientific rigour. The first section outlines issues associated with the internal factors of research design in psychedelic randomised controlled trials (RCTs). We have written on these issues in much greater detail elsewhere (Muthukumaraswamy et al., 2021). The second section describes challenges related to the broader context within which psychedelic research sits, including research funding, commercialisation and the effect of scheduling. The final section details challenges relating to the implementation of psychedelic therapies into medicine and gives suggestions on what could be included in current research to aid this process.

Factors related to individual research studies

Blinding/masking

The International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) guidelines nicely summarise the role of blinding in clinical trials:

Blinding or masking is intended to limit the occurrence of conscious and unconscious bias in the conduct and interpretation of a clinical trial arising from the influence which the knowledge of treatment may have on the recruitment and allocation of subjects, their subsequent care, the attitudes of subjects to the treatments, the assessment of end-points, the handling of withdrawals, the exclusion of data from analysis, and so on. (ICH Harmonised Tripartite Guideline, 1999)

In psychedelic RCTs, successful blinding is difficult to achieve and usually not maintained. We have previously contended (Muthukumaraswamy et al., 2021) that the estimation of treatment effects in all extant psychedelic clinical trials is overestimated due to unblinding effects. Via systematic review, we showed that of the six RCTs investigating classical psychedelics in mood disorders, only three reported on participant masking, with close to 100% of participants in these trials able to correctly guess when they had been administered a psychedelic. When a participant correctly believes they have received the treatment, they may show greater treatment response due to expectancy effects and, conversely, the disappointment of those who know they have received a placebo may decrease the placebo response. Therefore, it is critical that RCTs are designed such that the participants’ beliefs about whether they are in the treatment arm of the trial are approximately matched across groups (Muthukumaraswamy et al., 2021). To verify this, it is essential that blinding assessment is performed in trials and appropriate statistical techniques are applied to those assessments.

Therapeutic alliance

While it is tempting to think about psychedelic drug trials within the framework of standard drug trials, in the most prevalent models being developed (Carhart-Harris et al., 2021; Mitchell et al., 2021), psychedelic drug interventions are usually coupled with some form of psychotherapy. This adds substantial complications for controlling various treatment variables. A therapeutic alliance established between patient and practitioner can be a key predictor of treatment outcome (Zilcha-Mano et al., 2019). Whether this alliance should be considered a mediator or confound in psychedelic trials is a valid question for scientific debate. Simply increasing interaction in clinical studies can enhance the placebo response via increased exposure to known non-specific treatment effects such as emotional and cognitive care by the research team (Di Blasi et al., 2001). While this might at first be considered at risk of artificially enhancing evidence for psychedelics, any factor that increases the placebo response increases the risk of false negatives – eroding the evidence base for novel treatments. Furthermore, it is easy to envisage a situation where the therapist is aware of whether the patient has been allocated to the treatment group or not, given the patient’s immediate reaction to the drug administered. Either through conscious or sub-conscious processes, therapists might deliver differential therapy across the groups. Furthermore, the bespoke integration therapies being given to the treatment group in psychedelic RCTs might not be appropriate for the control group. These therapies might be confusing or even aversive for patients receiving a non-psychedelic placebo.

Expectancy

The placebo response refers to the observed changes in symptoms in participants who have been randomised to a placebo control group in an RCT. A key contributor to the placebo response in RCTs is expectancy – primarily response expectancy. Response expectancy refers to patients’ predictions/beliefs of their own nonvolitional response to treatment, whereas (less relevant in this context) stimulus expectancies refer to anticipation of external events (Kirsch, 2018). Given the increasing media coverage around positive therapeutics effects of psychedelics for an increasing list of indications, it is reasonable to think that expectancy effects might also be growing. There is also the question of sources of expectancy modulators in the study method. Mentioned above are the bespoke integration therapies that are likely to be at odds with the experience of any non-psychedelic inactive or active placebo. Psychedelic studies are often also or otherwise furnished with a suite of questionnaires rating experiences usually considered unique to psychedelic drugs, including magical, spiritual and mystical experiences. Such tools additionally force the consideration by the participant on whether they had such experiences and provide a means for comparison where the drug and placebo were given separately, thereby plausibly modifying a participant’s beliefs about their treatment allocation.

In principle as a pre-randomisation variable, expectancy need not be a confound. However, if post-randomisation variables such as blinding and alliance are not maintained, then expectancy/suggestibility could become confounders if a causal pathway is opened between treatment and outcome. It is perhaps debatable if expectancy could ever be held constant, and it is unclear whether there are any drugs that would serve as convincing active placebos, but it is important to remember that it is the beliefs of the participants/outcome assessors about treatment allocation that is important, not those of the investigators (Muthukumaraswamy et al., 2021).

Other biases

In addition to the methodological issues mentioned above, psychedelic studies may be affected by the usual biases including regression to the mean, sampling bias and attrition (see Muthukumaraswamy et al., 2021, for summary). Indeed, some of these, such as self-selection bias, may be overemphasised in psychedelic research. Self-selection bias is also one of several barriers to ensuring representative ethnic and demographic sampling in mental health research (Brown et al., 2014). Another potential driver of demographic participant bias is that, as with many areas of modern medical research, psychedelic scientists and therapists who identify as minorities are underrepresented in this field (George et al., 2020). Considering the role of race, ethnicity and gender in the manifestation, expression and assessment of mental health disorders (Chapman et al., 2014), this underrepresentation in research design and implementation is concerning and should be swiftly addressed. Lack of minority representation in population sampling limits the ultimate generalisability of results which may amplify any existing healthcare inequities. Minimum recruitment targets for key demographic groups can potentially be explicitly written into experimental protocols.

Other experimenter biases should also be considered; indeed, the lines between investigators and advocacy in the field of psychedelic science can be concerningly blurred. This might cause concerns for future healthcare consumers around the objectivity of the data reported from psychedelic studies.

Statistical power

Most psychedelic studies to date have tended to have relatively small sample sizes (Muthukumaraswamy et al., 2021), and this reflects several converging factors. They have primarily been conducted in academic settings where funding is often relatively limited. The heavy staff workload requirements for the collection of each data point (patient) make these very resource-intensive studies, particularly because the primary target populations are often complex treatment-resistant patients. However, to control and adjust for confounders, relatively large sample sizes are required. Similarly, if mediation analysis is to be used to control for confounding factors such as masking or alliance, then much larger sample sizes than those currently being studied will be required. Moving towards multi-site academic collaborations and the beginning of industry-funded psychedelic RCTs may help in this regard.

Factors related to the broader research context

Use of real-world/supplementary evidence

It has been previously discussed (Carhart-Harris et al., 2022) how real-world evidence can also provide evidence for therapeutic effects of interventions and that RCTs are not needed to demonstrate causation. Indeed, the most well-known example of such causal inference is the link between cigarette smoking and cancer established using the Bradford-Hill criteria (Hill, 1965). If psychedelic RCTs cannot be conducted satisfactorily given the limitations described above, then additional real-world evidence might be needed. This might include epidemiological evidence, use of surrogate and digital biomarkers, case and n = 1 studies (Carhart-Harris et al., 2022). If this approach is to be taken, then the set of criteria/evidence for and against the safety and efficacy of psychedelics should be established and a reasonable consensus decided in advance for what constitutes good evidence. Each of these types of studies has its own methodological weaknesses and convergent data would be required. That said, the double-blind RCT remains the gold standard of evidence for drug regulators world-wide and it is unclear whether these other models could be used as evidence for efficacy in this regulatory context.

Transparency

Despite the long calls for greater transparency in clinical trials, this is still a problem across the medical sciences (Hudson et al., 2016). While submission of trials to clinical trial registries is mandatory in most jurisdictions, these registries provide the minimal data set of information and key particulars such as statistical analysis plans and measurements of adverse events are missing. Furthermore, results are rarely reported when studies are completed (Casassus, 2021). Given the ‘replication’ crisis in many human science disciplines, full prospective publication of trial protocols in peer-reviewed academic journals would be preferable. Unfortunately, we have yet to see any meaningful adoption of this practice into psychedelic medicine.

Funding

To date, most psychedelic research has been funded from philanthropic sources. This has been essential to re-start psychedelic research after years of prohibition. A major problem with philanthropic funding is that the ideas implemented often do not undergo the same type of rigorous peer review and scientific assessment that grants sent to a research council would. As a result, there is potential for poorly designed trials to be funded and executed. Now that public perceptions about psychedelics as medicines have changed, the time is ripe for psychedelic research to be funded from public research councils. The recent move of the Australian government to fund $15M worth of psychedelic RCTs is a bold and innovative move that more government agencies should replicate. Recently, the field has seen the entering of commercial companies who will be funding trials to take psychedelic therapies to market. There is no inherent problem with this, but the transparent research practices mentioned earlier will be critical to developing trust with eventual consumers. The pharmaceuticalisation of mental health has a contentious history (Sadowsky, 2020), particularly when it comes to the various controversies surrounding the development and marketing of traditional antidepressants and the mistakes of the past need not be repeated.

Effects of scheduling

In most jurisdictions, psychedelics remain Schedule 1 controlled drugs and are controlled by the UN Convention on Psychotropic Substances. Nutt et al. (2013) eloquently described the situation

An important and unfortunate outcome of the controls placed on these and other psychoactive drugs is that they make research into their mechanisms of action and potential therapeutic uses – for example, in depression and post-traumatic stress disorder – difficult and in many cases almost impossible.

In our experience, the strict classification of these substances is a considerable barrier to research. In the last few years, a few jurisdictions have begun to relax some laws which is a positive sign for enabling future research.

For microdosing and other out-of-clinic or home-dosing regimens, however, the critical juncture will be the rescheduling of these drugs. Until then, having psychedelic medicines scheduled at odds with their safety profile and potential therapeutic applications will continue to impact and create biases in their evidence base. Nutt et al. (2013) identify the effects of increased time and expense which affect all research involving Schedule 1 drugs. But even once established, most countries prohibit outpatient use of psychedelics, meaning controlled trials must take place in a contrived research environment. As a result, there is a lack of evidence being gathered for the long-term effects of regular microdosing or of the more domain general effects reported in recreational and self-medicating users. This may also explain the failure of several microdosing studies to reproduce grey literature reported benefits and limit translatability of any findings. The constraint on study design has led to the development of novel approaches such as self-blinding citizen science or observing self-supply (Szigeti et al., 2021). However, with these methodologies, there are major issues yet to be addressed around the verification and validation of drug concentration and stability.

Relating also to section ‘Other biases’, perceptions of the risks around taking a Schedule 1 drug almost certainly causes self-selection and clinical referral bias both in terms of caution and perceived benefits.

Designing research for future use in standard medical practice

Ethical concerns regarding psychedelic-assisted psychotherapy

Psychedelic use involves substantial shifts in states of consciousness. This loss of control, and the often profoundly meaningful experiences patients report, could increase the risk of subtle or explicit abuses of differential power between therapist and patient already found in other areas of clinical psychology (Johnson, 2020). Consent and intercultural communication issues can become more prominent when changes to the power dynamic occur. Psychedelic therapy research design should therefore include techniques to manage these risks, such as multiple practitioners and transparent practices (Johnson, 2020), e.g. video-taping dosing sessions or supervision by other psychedelic-assisted psychotherapy experts, to ensure tested models exist when research is translated into medical practice.

Unregulated self-treatment

Unlike most other treatments for mental health disorders, psychedelics are used in traditional and underground medical practices and possess recreational properties. As such, they have been available on the black market for a long time. As the therapeutic use of psychedelics becomes more mainstream, unregulated self-treatment may become more common as the general public increasingly view this class of drugs as viable medicines. While psychedelic drugs have a relatively good safety profile (Nutt et al., 2010), increasingly mainstream unregulated self-treatment may come with an unanticipated set of harms not seen with current self-treatment. As the number of patients self-medicating increases, so too may the proportion of patients for which these treatments have known or currently poorly understood risks (e.g. history of psychosis). The lack of medical and psychiatric screening and psychotherapy support for the integration of challenging psychedelic experiences may result in a higher rate of adverse events with more serious profiles. Qualitative research models investigating the current therapeutic use of psychedelics in the community (e.g. Mason and Kuypers, 2018) could help determine the scope of these risks.

Integration into healthcare systems

Even if psychedelic medicines are to be approved by regulators for use in patients, how well could they be integrated into current healthcare delivery systems? In the country that we live in (New Zealand, NZ), most patients with depression and anxiety will never see a psychiatrist or make it into secondary care services unless they undergo acute psychiatric crisis. At best, patients will get to see their general practitioner and perhaps receive 4 hours of free talk-therapy services. Current models of psychedelic therapy are extremely resource intensive, requiring approximately 40 hours of healthcare resourcing. It is questionable ethically if a treatment paradigm established with 40 hours of supportive care can and should be used in practice with fewer hours of supportive care. Conversely, there is a real danger that resource-intensive psychedelic therapies will only be accessible to the wealthy. This will only serve to entrench the existing healthcare inequalities which exist in many countries. It is sobering to consider this given that the global burden of disease of mental disorders is 792 million persons, including 264 million people with a depressive disorder and 284 million people with an anxiety disorder (James et al., 2018). In NZ, e.g. if 2% of the population (100,000 people) were to seek psychedelic therapy at a cost of $15,000 (e.g. in Mitchell et al., 2021, there were three, 8-hour therapy sessions with two therapists to total 48 hours at $250 ph = $NZ 12,000. Add additional costs for medical oversight, facilities, drug, pharmacy, screening, follow-up and integration), this equates to $1.5 billion dollars. This is more than NZ’s entire medicine budget per annum (Ministry of Health New Zealand, 2021). As a global reference point, NZ ranks relatively high in health investment per capita (World Health Organization (WHO), 2021).

The physical infrastructure to conduct psychedelic therapy may not exist in many places, which would require significant capital investment. Considerable number of therapists would need to be trained if psychedelic medicine is to make a dent in the ocean of the global mental health pandemic. While Australia has a higher health budget than NZ and may be better placed to integrate psychedelic-assisted psychotherapy from a pure cost perspective, strains via demand on the existing workforce are well documented (Bloch-Atefi et al., 2021). We are sceptical that such resource-intensive therapy models are practical to be scaled-up given the combined pressures of appropriate infrastructure (treatment clinics), trained staff, availability of funding and, critically, patient and health system acceptability. These issues expose a need for more research into lower cost alternatives to the resource-heavy therapy model, potentially including online resources designed by therapists instead of in-person sessions, or hybrid models of both online and in-person therapy sessions, and even investigations of the efficacy and safety of treatment without therapy. However, any such approach adopted should be evaluated in clinical trials to assess both safety and efficacy of that specific therapy regimen as opposed to adapting approved treatment regimens after regulatory approval.

There are further issues/costs in terms of delivery to consider – despite the long-standing use of psychedelics outside of formal healthcare, to meet regulatory standards their integration into modern care as medicinal products will occur with the patenting of novel formulations and delivery systems. Initial approval will only be for specific indications and be inaccessible or involve navigating off-label prescribing for otherwise promising indications.

Finally, the acceptability of patients taking an intense ‘trip’ as part of therapy has not been well explored for patients, caregivers, family and usual healthcare providers.

Incorporating frameworks from traditional medical models

While the above argument focuses on the way forward for integration of psychedelics into a modern medicine framework, it must also be considered what the impact is on traditional practices being swept up in the wave of mainstream adaptation. Psychedelic medicines, particularly psilocybin and ayahuasca, have long histories of tradition and ritual, with sacred and deeply cultural significance. In these contexts, the hypothetico-deductive model underlying gold standard RCTs and evidence-based medicine can be seen as reductionist. The clash between traditional use of psychedelics and the aspiration for psychedelic medicines to be offered and approved in modern clinical practice is well documented (e.g. George et al., 2020; Tupper, 2008). Maintaining the heritage of psychedelic use requires ethnographic focused research and knowledge sharing led by Indigenous peoples. How modern medicine shaped by economic constraints and its philosophies can authentically build on, and benefit from millennia of evolving psychedelic knowledge – without reciprocally undermining and appropriating its heritage and indigenous guardians – will always be an important and ongoing issue.

Summary

It is understandable that the first wave of pioneering psychedelic trials currently being reported may not meet the many criteria we lay out here for scientific rigour. Science is a process of continual improvement, and it is reasonable to expect psychedelic study implementation to improve and it would be disturbing if it did not in the coming years. There are also significant concerns as to how psychedelic medicine could be implemented into healthcare systems. So, while psychedelics have shown early promise in the treatment of mental health disorders (see Reiff et al., 2020), with all the excitement around the potential of psychedelic medicine, it is important to remain objective and realistic about the difficult path that lies ahead given the pandemic it aims to treat.