The Impact of Medical Cannabis Legalization on Opioid Prescribing in the Medicare Population

• In Event: Policies related to Opioid Use and Health Outcomes

Abstract

Prior research finds that opioid prescribing significantly decreases following the implementation of medical cannabis laws (MCLs) and recreational cannabis laws (RCLs) in the Medicaid and Medicare populations, though some recent studies show mixed findings. This study investigates discrepancies in existing literature by examining opioid prescribing behaviors in the Medicare population from 2008-2021, focusing on non-cancer beneficiaries with chronic pain. 

We use retrospective cohort analysis to assess the impact of MCLs on Medicare Part D opioid prescription fills from 2008-2021. We quantify the impact of dispensary openings on Medicare Part D opioid prescription fills using difference-in-difference models accounting for staggered policy adoption. We observe all Medicare Part D prescription claims for beneficiaries in our sample from 2008-2021, comparing opioid utilization in states with open medical cannabis dispensaries to states without open medical cannabis dispensaries. We draw a random sample of 5.4 million beneficiaries from the Medicare Master Beneficiary Summary File, and we identify approximately 650,000 enrollees with non-cancer chronic pain based on the Medicare Chronic Conditions Warehouse algorithms. We examine all opioid prescription claims in Medicare Part D for these enrollees from 2008-2021, aggregating outcomes at the state-quarter level (N=2,646). 

The primary measure of MCL implementation is whether medical cannabis dispensaries are open in each state and quarter. We estimate the impact of medical cannabis dispensaries on the following outcomes: (a) total morphine milligram equivalents (MMEs), (b) number of patients with non-cancer pain, (c) rate of non-cancer pain patients with an opioid prescription, (d) rate of opioid prescriptions per non-cancer pain patient with an opioid prescription, and (e) rate of MMEs per opioid prescription. 

We find statistically significant reductions in MMEs per prescription (8.9%), but significant increases in enrollees prescribed opioids (10.0%) and opioid prescriptions per patient (2.9%). We find a 21.1% reduction in overall MMEs, but this finding is not statistically significant. Our findings may help reconcile the discrepancy in existing literature which provides evidence of overall decreases in MMEs prescribed following MCL, but possible increases in the number of opioid prescriptions per patient with non-cancer pain. 

Author

• Victoria Elizabeth Bethel, University of Georgia
  Presenting Author