In 2018, high-profile campaigns on behalf of children with treatment-resistant epilepsy led to the legalisation of medical cannabis in the UK.
The change in law permitted specialist consultants to prescribe cannabis-based products for medicinal use (CBPMs) for any condition if in the best interests of the patient.
Without a prescription, the possession and use of cannabis (a Class B drug) remains illegal in the UK.
As most of these products are unlicensed, meaning they lack market authorisation from the Medicines and Healthcare Products Regulatory Agency (MHRA), they’re rarely prescribed through the UK’s National Health Service (NHS). What regulators didn’t foresee was that a private market would emerge to fill this gap.
More than 30 specialist cannabis clinics are now registered with the Care Quality Commission in England, prescribing CBPMs to an estimated 80,000 patients for everything from chronic pain to anxiety and ADHD.
Data suggests almost half (42 per cent) are prescribed medical cannabis for psychiatric conditions, such as anxiety, depression, post-traumatic stress disorder (PTSD) and OCD, similar to trends in Australia and the US.
But a new review published in Lancet Psychiatry examined over 50 randomised control trials (RCTs) and found ‘no evidence’ that cannabinoids are beneficial in treating anxiety, PTSD, substance-use disorders, ADHD, bipolar disorder, psychotic disorder or anorexia.
While there was some evidence for their efficacy in cannabis-use disorder, insomnia, Tourette’s syndrome and autism spectrum disorder, this was considered ‘low quality’.
The study comes as the Advisory Council for the Misuse of Drugs (ACMD) is conducting a review into the prescribing of medical cannabis in the UK, including whether there have been any ‘unintended consequences’ of the law change.
Former chair of the ACMD Prof Owen Bowden-Jones said the findings give the “clearest indication yet” that the benefits of medical cannabis “may have been overstated for many conditions” and that these products “should not be offered for the many mental illnesses for which no benefit was found.”
“We need to reduce the barriers to enable more high-quality research to further examine the effects of cannabis products,” he said.
The review, which concludes that the routine use of cannabinoids in mental health conditions is ‘rarely justified’, raises important questions. Perhaps most importantly: why is cannabis prescribed if there is so little evidence for its efficacy?
Treatment options
“The absence of evidence isn’t evidence of absence,” says Dr Niraj Singh, a consultant psychiatrist in the UK, who has been prescribing medical cannabis for over six years.
“Tens of thousands of patients would testify that it benefits them across an array of symptoms, and the vast majority use it responsibly. In my experience, it has had some incredible outcomes, helping them get to a place where they’re able to lead a happy, fulfilling life.”
Prescribers report that many patients arrive at medical cannabis clinics having “exhausted every conventional option” or are unable to access other mental health support. As of January 2026, more than 1.5 million adults were in contact with NHS mental health services, while 8.7 million people in England were prescribed antidepressants in 2023/24, which are thought to work in around three-quarters of cases.
In a survey conducted by the medical cannabis advocacy group United Patients Alliance, one respondent with anxiety, depression and PTSD said they felt “seen and supported” after receiving treatment for their condition with a medicine that “actually works”, and “without the harmful side effects of previous prescriptions”.
“Often people hit a ceiling in terms of their treatment options and, for many of them, medicinal cannabis has worked,” Singh adds.
This is supported by real-world evidence, published in peer-reviewed journals, linking cannabis to improvements in symptoms and quality of life in conditions such as PTSD, OCD and insomnia. However, observational studies were not included in the review as they’re ‘more likely to have potential biases’ and ‘cannot establish a causal relationship’.
While more robust clinical trials are needed, Prof David Nutt, former chair of the ACMD and founder of the independent charity Drug Science, disagrees with the idea that only RCTs provide adequate data on a medicine’s efficacy.
He’s not alone. In his 2008 lecture for the Royal College of Physicians, former head of the MHRA and the National Institute for Health and Care Excellence (NICE), Sir Michael Rawlins, suggested that the placebo-controlled RCT “is not the gold standard of evidence that most regulators and prescribers think it is”. He recommended gathering real-world evidence, which can offer “better clinical data and greater statistical power.”
“Placebo-controlled trials are an intensive and therefore expensive way of gathering clinical data,” says Nutt. “Such clinical trials are conducted on highly selected patient populations, restricting generalisability to clinical practice.”
As cannabis contains hundreds of active compounds, and can vary significantly in dose and formulation, researchers often overlook the challenges of conducting double-blind placebo-controlled trials in these products, adds Prof Mike Barnes, chair of the Medical Cannabis Clinicians Society.
“The study shows a significant lack of knowledge of the cannabis plant and a misinterpretation of the fact that you can apply pharmaceutical approaches to a botanical compound, as well as a lack of understanding of the nuances of mental health in relation to cannabis,” he says. “You can’t take a blanket approach to prescribing in mental health.”
Clinical oversight
Medical cannabis can cause side effects like increased anxiety and paranoia, and isn’t recommended for patients with a history or family history of psychosis.
According to findings published in BMJ Mental Health, those who use cannabis to ‘self-medicate’ not only use it more frequently, but also consume higher levels of tetrahydrocannabinol (or THC, the principal psychoactive compound in cannabis), which is associated with greater levels of paranoia.
“Cannabis is not an entirely side-effect-free substance,” says study co-author Dr Marta Di Forti, professor of drug use, genetics and psychosis at King’s College London, who runs a clinic for psychosis patients in London.
“People with a pre-existing mental health condition, such as anxiety, depression and even more so paranoia, if exposed to products with a THC content of 10 per cent or above, are more likely to experience a worsening of their baseline condition and to become dependent.”
A patient under Di Forti’s care became unwell after being prescribed cannabis containing 19 per cent THC, she says, despite a known history of psychosis. She is also aware of cases where patients prescribed cannabis for chronic pain have been hospitalised with psychotic symptoms. But this, like so much of our understanding of medical cannabis, is anecdotal.
“I think there are good reasons to prescribe this as a medicine,” she adds. “But only in indications where there’s clear and established evidence, and where there’s adequate monitoring – which isn’t happening.”
The Medical Cannabis Clinicians Society’s best practice guidelines recommend that any prescription of more than 60g per month, and for products containing more than 25 per cent THC, should be reviewed by a peer panel. Like any controlled medicine, prescribing of CBPMs requires “careful clinical oversight and robust governance” and isn’t appropriate in every circumstance. In complex cases, particularly where there’s a significant mental health history, this means “thorough assessment, clear clinical reasoning, careful risk evaluation and ongoing monitoring”.
In Singh’s experience, side effects are the “exception rather than the norm”, but he also shares concerns about the “escalating” availability of products containing high levels of THC.
“There are definitely checks and balances needed,” he said. “It has to be tailored, carefully titrated [to deliver the correct THC concentration] and medically monitored regularly.”
Prescribers insist that there are robust clinical oversight processes in place and say they have never felt “pressured” to prescribe. To be eligible for medical cannabis, patients must have received at least two prior treatments and undergone at least one assessment by a psychiatrist, followed by a multi-disciplinary team review.
However, there are concerns that clinics could do more to support prescribers with appropriate training, and that they have a responsibility to help generate the research to support their claims.
“The industry has not been rigorous enough in collecting and analysing patient outcomes,” says Barnes. “Clinics have a moral obligation to collect their own data and liaise as best they can with other clinics to do so collaboratively.”
The evidence gap
While everyone agrees there’s an urgent need to build the evidence base, beyond this, the debate reaches a stalemate. Some say it works, some say it doesn’t and very few are conducting the studies to find out for sure.
“The current system for conducting clinical research has failed medical cannabis,” says Nutt.
“In 2018, the Department of Health promised to conduct efficacy trials in children with epilepsy, yet it has done nothing so far. This failure is mirrored by the fact that pharmaceutical companies aren’t working in this space, as patenting plant medicines isn’t feasible.”
This can’t be rectified by “more calls for research”, he adds, but by placing more value on the real-world data and lived experiences that support the use of cannabis in these conditions.
In the meantime, patients fear being forced back to the illegal market where they won’t have access to medical supervision or regulated products – and which most agree would be far more harmful.
Dismissing medical cannabis based on “incomplete evidence” doesn’t just “misrepresent the science”, but causes “direct harm” to patients who rely on it, says United Patients Alliance.
“Real-world evidence studies, patient-reported outcomes and research into treatment-resistant populations are urgently needed,” they add. “We’re not asking anyone to ignore the science, we’re asking that the science catches up with patients.”
Read more:
