The evidence bar impacting UK medical cannabis access

pharmafile | November 6, 2026 | Feature | |  Cannabinoids, Chronic Diseases, Legislation, NICE, Pharmacy, chronic pain, drug approval 

By Mike Morgan-Giles

Private prescribing of cannabis-based medicines in the UK has more than doubled in a year. Getting the same treatment on the NHS remains close to impossible — and much of the explanation lies in a mismatch between how these medicines are made and how they’re expected to prove themselves.

A handful of cannabis-derived medicines — Sativex (nabiximols), Epidyolex (a purified CBD extract) and the synthetic Nabilone — are licensed, pharmaceutical-grade products: single or narrowly defined formulations that went through conventional randomised controlled trials (RCTs) and hold a marketing authorisation, much like any other prescription medicine.

The vast majority of what’s actually prescribed in the UK, however, is botanical: whole-plant flower or full-spectrum oils containing hundreds of interacting compounds — cannabinoids, terpenes, flavonoids — whose combined action is often described as the “entourage effect”. NICE and other Health Technology Assessment bodies apply essentially the same RCT-first standard to both categories, and critics argue that’s a poor fit for the botanical side specifically: these products vary by cultivar and growing conditions, resist the standardisation an RCT demands, and are difficult to blind or placebo-control in trial design.

The alternative on the table is formalising real-world evidence (RWE) — patient registries, observational data, patient-reported outcomes — as a legitimate evidence base for botanical CBPMs, rather than a fallback used only when trial data is unavailable. Health Canada’s RWE reporting guidance and Australia’s Special Access Scheme, which has tracked outcomes across nearly its entire medical cannabis patient population since 2016, are frequently cited as models the UK could adapt.

A regime built for one-off cases

New CQC data shows the scale of that growth. The regulator’s Safer Management of Controlled Drugs: Annual Update 2025 recorded a 2.3x increase in CBPMs prescribed in the year to 31 March 2025, with 39 providers now prescribing and 63% of repeat prescriptions issued by pharmacists rather than doctors — reflecting how far multidisciplinary prescribing models have developed since 2018. A separate FOI analysis by Business of Cannabis, using NHSBSA data, sharpens the picture further: pharmacist independent prescribers accounted for 955,471 of the 1.54 million CBPM items dispensed between January 2019 and May 2025 — 61.9% of the total, against 35.8% from doctors.

Almost all of that activity sits outside the NHS. Most unlicensed CBPMs are supplied through the MHRA’s “Specials” regime. Procurement is restricted to doctors on the GMC Specialist Register, special importers, registered pharmacies and licensed wholesalers, and the “special clinical need” justifying use explicitly excludes reasons of cost or convenience. Because Specials sit outside standard NHS commissioning, Trusts must fund them from existing unlicensed-medicines budgets rather than a dedicated pathway — discouraging prescribing even where a clinician judges a patient would benefit.

That access gap persists despite economic modelling suggesting it may be self-defeating. A 2024 analysis commissioned by the CIC and Drug Science, built by the York Health Economics Consortium and published in Expert Review of Pharmacoeconomics & Outcomes Research, modelled NHS cannabis prescribing against standard chronic pain treatments — analgesics, physiotherapy, CBT — over a one-year horizon using real-world data from Project Twenty21. It estimated potential NHS savings of close to £4 billion a year if cannabis-based medicines were prescribed for chronic pain, alongside reduced sickness absence and higher patient earnings, and concluded the intervention would meet NICE’s own cost-effectiveness threshold. The gap between that finding and the almost total absence of NHS prescribing for pain is, for advocates, a clear illustration of how far reimbursement decisions still lag behind the RWE base already available.

Much of this is not new territory. NHS England’s own 2019 review of prescribing barriers recommended exploring alternative study designs and RWE generation. NICE issued a clarification in 2021 confirming that the absence of population-wide recommendations in its guideline (NG144) should not itself prevent clinicians prescribing unlicensed CBPMs where clinically appropriate.

Proposals put to policymakers over the past year include: formal integration of RWE into NICE and MHRA decision-making; extending prescribing initiation beyond the GMC Specialist Register to trained GPs and non-medical prescribers for stable patients; reformed commissioning so unlicensed-medicine cost rules don’t structurally block CBPM funding; and regulatory clarity on driving, housing, employment and possession.

A government-commissioned review of the 2018 legislation, with a call for evidence that closed in October 2025, has repeatedly slipped past earlier expected publication dates. In its Autumn 2026 update to ministers, dated 21 September, the ACMD confirmed its CBPMs working group report is “concluding” and now expected to publish by the end of the year.

The CQC’s latest annual controlled drugs report, published July 2026, shows the private prescribing market continuing to grow at pace while the regulatory framework it operates under remains unchanged.

Pic: Celadon

Mike Morgan-Giles is chief executive of the Cannabis Industry Council