Trajectories of Prescribed Medical Cannabis Filled Over Time

Background: Medical cannabis (MC) use is rising with limited clinical data to support products and dosing for specific conditions. This study relies on observational data to examine MC purchases across time, and to assess dosing trajectories for different conditions.

Methods: A retrospective study of MC patients of dispensaries located in New York (NY). This study relies on secondary analysis of point-of-sale (POS) invoice data from 16,727 unique patients with 79,885 purchases between 2016-2019. Group-based-trajectory modeling (GBTM) was used to identify clusters of MC patients following similar progressions in potency utilization (e.g. THC and CBD) over time. Multinomial logit models were estimated to identify group membership based on patient level characteristics and qualifying medical conditions.

Results: Six distinct trajectory groups were identified. Four of the groups compromised (75%), (39.9%), (8.2%), and (8.5%) of the population and purchased a steady dosage (ranging from low to high) of THC over time. The fifth (14.9%) and sixth group (10.1%) demonstrated MC patients who gradually increased their THC dosage across time. Patient characteristics and qualifying medical conditions for MC use were strong predictors of group membership. Men, older individuals, and those with a qualifying pain condition were more likely to be part of a group that consumed higher doses of THC across time, compared to other reference groups. This study identified distinctive trajectories of monthly THC and CBD potency levels purchased, and factors associated with these trajectories.

Conclusion: Examining MC purchasing patterns over time may help understand whether MC treatment works, subgroups of MC patients, and risk factors. This study pioneer’s analysis of POS data, which could help guide policy decisions to effectively monitor MC use, aid in the design of future MC programs and target prevention efforts. This study provides a strong foundation upon which that research can build by utilizing new-technologically advanced sales data sets.