Medical cannabis authorization did not significantly reduce opioid prescribing in a university pain clinic

Average opioid dose at the final time point was 33.4 mg/day OME overall. Medical cannabis authorization predicted a non-significant decrease of 14.25 mg/day OME. No significant difference between cannabis-authorized (38.51 mg/day) and non-authorized patients (32.60 mg/day). Long-term opioid users had significantly higher OME (85.34 mg/day, 63 mg/day higher than the rest, p<0.0001).

Longitudinal retrospective cohort analysis of electronic health records from a university-based pain clinic, July 2016 to August 2019. Longitudinal multilevel modeling with maximum likelihood estimation compared opioid milligram equivalents over time between patients with and without medical cannabis authorization.

Claims that medical cannabis reduces opioid use are central to many advocacy and policy arguments. This real-world clinical data from a pain specialty setting shows no statistically significant opioid-sparing effect, tempering expectations while not ruling out a modest benefit.

The null finding joins a growing body of clinical studies showing that population-level claims about cannabis replacing opioids are not consistently supported by individual-level data. The dominant predictor of high opioid use was having a long-term opioid use diagnosis, not cannabis access, suggesting that the opioid problem is driven by factors that cannabis alone cannot address.

Retrospective design with no randomization. Patients who sought cannabis authorization may differ from those who did not. Cannabis use was captured by authorization, not actual consumption. The 3-year study period may be too short. A university pain clinic population may not represent all chronic pain patients.