Marijuana didn’t feature in my textbooks when I started Medical School nearly 20 years ago. Then again, I didn’t study the world’s oldest surviving medical text, the 2nd century BC Shen-nung Pen-tshao Ching. This championed the herb in its treatment of “rheumatism, female weakness, absent-mindedness, and malaria”. Meanwhile, the Chinese surgeon, Hua Tuo, (c. 140-208) was the first to use a popular mix of cannabis and wine as an anaesthetic.

Had I been a doctor in 1870, vials of marijuana, opium, and alcohol would happily have sat in my briefcase. However, within a decade, hemp would fall out of favour - it was insoluble, unpredictable, and had a short shelf-life. Soon, it would be banned entirely. Now, turn to 2014. Could medical marijuana be making a come back?

Charlotte Figi is a seven-year-old girl with an intractable form of epilepsy called Dravet syndrome. From the age of two onwards, she has had dozens of fits each day, despite being on a cocktail of heavy-duty benzodiazepines and barbituates. Eventually her desperate parents sourced a canniboid derivative to add to her existing anti-seizure drugs. This reduced her seizure frequency from nearly 50 convulsive seizures per day to two to three per month. Her other anti-seizure drugs have been weaned and 20 months later, she is still responding to a few drops of cannabis oil twice a day in her food. This treatment has now aptly been named “Charlotte’s Web”.

Charlotte’s story has stuck with me. I have a few patients just like her: teenagers suffering from dozens of seizures each day, on a cocktail of six drugs which make a small difference. Their desperate parents are always looking for the next new cure. So should I be prescribing Charlotte’s Web?

How to make that decision? First, do no harm. Marijuana is not entirely safe. Its use in adolescence is associated with lower IQ, likely linked to impaired activity and neural connectivity in brain regions that mediate alertness, self-conscious awareness, learning, inhibition, multi-tasking, and memory.

As a neurologist, I am all too aware of shortcomings in the medications I prescribe. Yes, they can be wonderfully effective. But I cannot ignore the adverse effects that some patients experience with painkillers and anti-seizure drugs: sedation, confusion, dizziness, and dependence. Also, here’s the thing: they don’t always work.

Solid evidence for marijuana is also not easy to find. Few controlled studies comprehensively compare it to existing treatments. One important reason is that cannabis is a Schedule 1 drug, along with LSD and ecstasy. This means means that it is deemed to have no therapeutic value, cannot be lawfully possessed or prescribed, and requires a Home Office licence for research. The Liberal Democrats have call for a reform of our cannabis policy, but a call from Norman Baker, Minister of State for Crime Prevention, to review its medicinal use was quickly rejected earlier this year.

In short, I won’t be writing prescriptions for medical marijuana in 2014. Well, not yet anyway. It’ll take solid evidence-based studies before I think about doing so. That means working further with drugs like Sativex (which is not Schedule 1) or reconsidering the Schedule 1 status for cannabinoids under tight regulation. Exposing my often very unwell patients to the risks of psychosis, dependence, and cognitive impairment is not something that I can bear to do. And certainly not if the benefits of marijuana over existing treatments are uncertain.

Those patients I mentioned – the teenagers who come to me with refractory epilepsy, and dozens of seizures each day – maybe one day they’ll be sitting across from me in clinic while I write a prescription for Charlotte’s Web. Just imagine if they could live normal lives again.

But first give me those decent trials, and decent evidence. Lobby the Government (and NICE) for change. And then I’ll be the first one to make a decision that serves my patients best.

Medical marijuana: An overview

There is already solid trial evidence for the use of marijuana in pain syndromes; one randomised study in the US recruited patients with HIV who were suffering pain. They were allocated to smoke cannabis cigarettes or placebo cigarettes three times daily for five days. The first puff of cannabis smoke reduced pain by a median of 72 per cent, compared to 15 per cent with the placebo. These effects were sustained over the course of the trial. The problem was that the extent of benefit only seemed to lie in the same range as existing tablets for pain, including opiates.

Furthermore, research has linked early adolescent cannabis use with increased risk of later life psychosis. Not to forget dose-related effects in users of all ages: dizziness, sedation, dry mouth, muscle weakness, low blood pressure, palpitations, and anxiety. Addiction is a risk in about 9% (versus 30% for tobacco).

The only way to get medical marijuana in the UK is if you have MS and are prescribed Sativex. This is a sublingual spray, licensed for spasticity treatment in multiple sclerosis (MS) when other treatments have failed. Up to 35 per cent of patients report improved spasticity outcome measures. However just this week, the National Institute for Health and Clinical Excellence (Nice) ruled that Sativex is not cost- effective. Patients with MS will now only receive his drug if their local commissioning group chooses to ignore the Nice guidelines.

Current research is evaluating the effectiveness of marijuana in cancer, epilepsy, post-traumatic stress disorder, and schizophrenia, but randomised clinical trials comparing marijuana and placebo are few and far between. The only way to be prescribed medical marijuana in the UK is if you have multiple sclerosis (MS), although its availability has just been limited by the National Institute for Health and Clinical Excellence.