#199 Missing “High” Quality Evidence: Medical Cannabinoids for Pain?
Reading Tools for Practice Article can earn you MainPro+ Credits
Join NowAlready a CFPCLearn Member? Log in
- Any chronic pain: Systematic review of systematic reviews.1
- Pain reduction (15 Randomized Controlled Trials (RCTs), 1,985 patients): 39% versus 30%, Number Needed to Treat (NNT)=11.
- Larger (>150 patients) and longer (9-15 weeks) RCTs: No effect.
- Mean pain improvement ~0.5 (0-10 scale, not clinically meaningful).2
- Pain reduction (15 Randomized Controlled Trials (RCTs), 1,985 patients): 39% versus 30%, Number Needed to Treat (NNT)=11.
- Neuropathic pain:
- Inhaled MC (five RCTs, 178 patients):3 NNT=6.
- Any MC (15 RCTs, 1,619 patients):4 NNT=14.
- Cancer pain (six RCTs):5 Pain reduction not statistically significant.
- HIV neuropathy, smoked MC (two RCTs, 89 patients):6 NNT=4.
- Multiple sclerosis pain (seven RCTs, 298 patients):7 Mean pain improvement over placebo ~0.8 (0-10 scale, borderline clinically insignificant).
- Acute pain (seven RCTs): One positive, one negative, and five equivalent to placebo.8
- Versus medications: Cannabinoids no better with more adverse events (versus low-dose amitriptyline)9 or inferior with similar adverse events (versus dihydrocodeine).10
- No difference in Quality of Life.2,4,5
- Very sparse evidence for back pain, fibromyalgia, or osteoarthritis.11-13
- Issues:
- Cannabinoids generally adjunctive to other pain treatments.1,2
- Quality often poor: Of 28 RCTs, two low risk of bias and 16 high risk.2
- When assessed, unblinding common, likely exaggerating effectiveness.6,14
- For inhaled marijuana, data on pain is very sparse and poor:1
- Only five RCTs with 189 patients followed 6 hours to 12 days.
- Represents <1% of the total patient-years studied of MC for pain.
- Prescribing guidance available through the College of Family Physicians of Canada15 and multiple reliable sources,16-19 including international sites (example20).
- Health Canada provides clinician21 and patient information.22








DATA DOESNOT SUGGEST GREAT BENEFIT IN TREATING PAIN WITH MEDICAL CANNABINOIDS
despite low quality evidence NNT is reasonably low for the right patient population, perhaps that is the item that is hardest to define.