NHS Trust Clears Inpatients to Continue Cannabis Prescriptions

For a UK patient prescribed medical cannabis, a hospital admission has carried an uncomfortable risk: ward staff unsure what to do with a legally prescribed medicine, and a real chance it is confiscated or simply withheld. Devon Partnership NHS Trust has moved to close that gap on its own wards, approving what patient advocates describe as the first formal NHS policy in the country for continuing prescribed cannabis-based medicines during an inpatient stay.

The trust — a mental health and learning disability provider — ratified the procedure, known as CD21, through its Medicines Safety Committee in May 2026. It is the first NHS trust in England, and according to PatientsCann UK, the patient organisation that helped develop it, the first in the UK, to spell out how staff should verify, store and administer these products since medical cannabis was legalised in November 2018.

What the policy actually does

The central move is a distinction that had been tripping clinicians up: continuing an existing prescription is not the same as starting one. Under CD21, transcribing a patient’s authenticated cannabis prescription onto the inpatient record counts as a direction to administer a medicine the patient already takes — not a new prescription that ward staff, who are not specialist cannabis prescribers and are often unfamiliar with these medicines, would be in no position to write. The trust does not dispense the product or pay for it; patients keep sourcing and funding supplies through their private clinic.

CD21 then sorts cannabis products into categories and handles each differently. Licensed medicines — the anti-sickness drug nabilone, the multiple sclerosis spray Sativex and the epilepsy solution Epidyolex — follow standard controlled-drug procedures.

Unlicensed products from private clinics, the most common category for UK patients, have to arrive in their original pharmacy packaging with a legible dispensing label showing the patient’s name, the product, the dose and its THC and CBD content; a commercial “cannabis card” does not count as proof of a legal prescription. Once a pharmacist confirms the prescription is genuine, the medicine is treated like any other Schedule 2 controlled drug, with locked storage, dosing records and oversight by the trust’s controlled-drugs officer. Smoking remains banned on NHS premises, so inhaled flower must be taken through a dry-herb vaporiser or heated nebuliser.

Where the evidence stands

Because Devon Partnership treats mental illness, the policy had to take a position on a genuinely unsettled clinical question: does cannabis use blunt psychiatric treatment? CD21’s answer is that it should be weighed in treatment planning but should not, on its own, bar a patient from psychological therapy — and it grounds that stance in two peer-reviewed studies rather than in assumption.

The first, a 2022 study in the Journal of Psychiatric Research, tracked 253 adults through cognitive behavioural therapy for anxiety and related disorders. Frequent users — twice a week or more — did fare worse on average than non-users. But they still recorded a statistically significant and large reduction in anxiety symptoms over the course of treatment, and the severity of their cannabis-related problems did not predict how they responded. It was an observational study rather than a controlled trial, so it can flag an association without establishing that cannabis caused the weaker response.

The second, a 2024 individual-patient meta-analysis in the Journal of Anxiety Disorders, pooled four randomised trials of treatment for post-traumatic stress disorder occurring alongside substance use. Trauma-focused therapy came out ahead for cutting PTSD severity whether or not patients reported recent cannabis use, and the cannabis and non-cannabis groups attended similar numbers of sessions and improved to a similar degree. The NIH-funded analysis concludes the priority should be removing barriers to effective treatment, not screening cannabis users out of it — the same logic CD21 adopts.

Neither study settles the matter. Both examine recreational-pattern use rather than prescribed dosing, the samples are modest, and the wider evidence for cannabis-based medicines is still thin: NICE endorses them for only a narrow set of uses and advises against them for chronic pain, citing insufficient evidence. What CD21 reflects is less a claim that cannabis helps than a clinical judgment that abruptly stopping a medicine a patient depends on carries a risk of its own.

A model other trusts may copy

Continuity of treatment is a principle UK guidance already recognises. NICE says NHS patients already established on these products for chronic pain before its 2019 guidance should be able to keep taking them until they and their clinician agree to stop, and GMC guidance treats maintaining a patient’s supply during an unexpected gap as a legitimate reason to prescribe. Yet legality in principle since 2018 has not translated into consistent practice, and most trusts still have no procedure at all. Access also remains lopsided: very few patients obtain cannabis on the NHS, while the majority pay privately, and CD21 does not change that — a limit its drafters acknowledge no single trust can fix.

Other health systems have taken narrower routes to the same problem; Delaware, for one, has legislated hospital cannabis access only for terminally ill patients. PatientsCann UK says it has now written to NHS bodies, regional commissioners and controlled-drugs officers across all four UK nations, offering CD21 as a template and urging them to adopt equivalent procedures. Whether other trusts follow will decide if a policy governing one provider’s wards becomes the standard the group is pushing for.

References:

^{1. Ouellette, Mélise J.; Rowa, Karen; Cameron, Duncan H.; Elcock, Ashleigh; Soreni, Noam; Pawluk, Elizabeth J.; and McCabe, Randi E., “Does cannabis use impact cognitive behavioural therapy outcomes for anxiety and related disorders? A preliminary examination” (2022). Journal of Psychiatric Research, 156, 690–697. https://www.sciencedirect.com/science/article/abs/pii/S0022395622006057
2. Hill, Melanie L.; Kline, Alexander C.; Saraiya, Tanya C.; Gette, Jordan; Ruglass, Lesia M.; Norman, Sonya B.; Back, Sudie E.; Saavedra, Lissette M.; Hien, Denise A.; and Morgan-López, Antonio A., “Cannabis use and trauma-focused treatment for co-occurring posttraumatic stress disorder and substance use disorders: A meta-analysis of individual patient data” (2024). Journal of Anxiety Disorders, 102, 102827. https://www.sciencedirect.com/science/article/abs/pii/S0887618524000033}