Medical Marijuana

medical marijuana20 states have laws authorizing the use of medical marijuana, but what is the evidence for the medical effectiveness of marijuana?

An exhaustive review of the topic just published in the Journal of the American Medical Association finds that, overall, the evidence for marijuana’s effectiveness is quite poor.

And while many of my younger patients tell me that the problem is a profound bias on the part of medical researchers against medical marijuana, this review suggests that there is a significant bias for medical marijuana amongst at least some researchers.

The authors of this study performed a meta-analysis of 79 trials that involved 6462 patients.

Of these 79 trials, four were felt to be well-designed and unlikely to be biased, and 55 of the studies were felt to be at high risk, generally because the authors did not completely report their results, and thus might have been selectively reporting favorable data in their publication.

In keeping with the overall medical literature, where it has been noted for a long time that published articles tend to have more positive findings (tend to show that the studied treatment was effective) than studies that are not published, most of these studies reported positive findings. But even more so than in typical clinical trials, the written summaries of the research in these articles tended to describe benefits of treatment even though the results were not statistically significant.

Four studies were judged to be at low risk of bias, whereas 55 studies were judged to be at high risk, mostly because of incomplete reporting of results.

Another concern was the adequacy of blinding. Double-blinded studies are ones where neither the researcher nor the patient is supposed to know whether the subject is taking the active ingredient. When studying a medication that has psychotropic effects it is particularly important to pay attention to blinding since patients and researchers are more likely to be able to gas whether the substance was an active or placebo agent because of the secondary effects  (in this case the subjective high). Although most studies were reported as double-blind, their blinding methods were considered inadequate in many cases.

Journal Watch reviewer Thomas L. Schwenk concludes –

The investigators found moderate-quality evidence that cannabinoid use might benefit patients with chronic pain or muscle spasticity; they found low-quality evidence that marijuana use prevents nausea and vomiting secondary to cancer chemotherapy, leads to weight gain in patients with HIV, promotes uninterrupted sleep, or lessens tic severity in Tourette syndrome. No evidence of benefit was shown for several other conditions, including depression, anxiety disorder, psychosis, hepatitis C infection, Crohn disease, Parkinson disease, and glaucoma; at least some of these are indications for which marijuana use is legal in some states. Cannabinoids were significantly more likely than placebo to be associated with dizziness, nausea and vomiting, sleepiness, disorientation, confusion, and hallucinations.


Whiting PF et al. Cannabinoids for medical use: A systematic review and meta-analysis. JAMA 2015 Jun 23/30; 313:2456. (

Vandrey R et al. Cannabinoid dose and label accuracy in edible medical cannabis products. JAMA 2015 Jun 23/30; 313:2491. (

D’Souza DC and Ranganathan M.Medical marijuana: Is the cart before the horse? JAMA 2015 Jun 23/30; 313:2431. (