Development of new measures of cannabis market exposure: a case study in Colorado

As efforts are underway to assess the impact of cannabis policy reforms, little methodological work has been completed assessing meaningful metrics of the presence and size of local cannabis markets. This paper attempts to move the science forward in this area by examining information in Colorado on dispensary licenses, sales tax receipts, and total amount produced and distributed within a county to test the extent to which different measures of the existence and maturity of the cannabis market are correlated with various measures of cannabis use and cannabis-related harm.

Study Design: We geocode medical and recreational cannabis dispensary licensing information throughout the state and link this to county-level rates of cannabis-involved poison calls and emergency department (ED) visits using data from the Rocky Mountain Poison and Drug Center and the Colorado Hospital sociation over the period 2010 -2017. We exploit within state and temporal variation in the number of medical and recreational cannabis dispensaries operating as well as different indicators of the volume of sales associated with these dispensaries (through sales receipts and production measures) to assess the extent to which particular measures of the market are more or less correlated with cannabis use and cannabis-involved harm.

Principal Findings: The number of medical dispensaries did not change from 2014 (when the recreational dispensaries could first open) to 2018 while the number of recreational dispensaries increased 162% over the same time period. Preliminary results suggest that cannabis poison center calls, which rose 107% over the same time period, are strongly correlated with not just outlet density but also measures reflecting volume sold. The simple presence of a store was not sufficient to generate statistically significant associations.

Preliminary conclusions: Exposure to cannabis markets is indeed correlated with cannabis use and harm, although measures that do not account for the density within the market (in stores or volume of goods sold) are not sufficient to identify associations. This study is the first to develop and test measures of local cannabis marijuana exposure. These measures will be useful in future studies of the health effects of changing marijuana policies on health and other outcomes, including opioid use.