Synthetic Opioids Should Be Excluded When Examining the Association Between Cannabis Laws and Opioid Overdose Mortality

We read with great interest the recent work in PNAS by Shover et al. revisiting the relationship between medical cannabis laws and opioid overdose mortality. They expand the 2014 Bachhuber et al. by adding 7 recent years of mortality data to cover a longer study period 1999-2017. They suggest that the reduction in opioid overdose mortality following medical cannabis laws became statistically indistinguishable from 0 when data in 2013-2016 are added and reversed after data in 2017 are added. The authors interpret the sensitivity of the results to the study endpoint to suggest that the association is spurious. We agree that the associations found in ecological studies should not be interpreted as causal relationships. However, we argue that the sensitivity of the results could be likely attributable to the recent spike in overdose mortality related to synthetic opioids.