Prescription opioid characteristics at initiation for non-cancer pain and risk of treated opioid use disorder: A population-based study

Background: Long-term prescription opioid use has been associated with adverse health outcomes, including opioid use disorder (OUD). We examined a population of opioid na?ve individuals who initiated prescription opioids for non-cancer pain and investigated the associations between opioid prescription characteristics at initiation and time to treated OUD. Methods: We conducted a retrospective population-based cohort study in Ontario, Canada among opioid na?ve individuals aged 15 years and older dispensed an opioid for non-cancer pain between 2013 and 2016. We used the Narcotic Monitoring System to abstract opioid dispensing data. A multivariable Cox regression model was used to examine the association between average daily dose and time to treated OUD. Results: We identified 1,607,659 opioid-na?ve individuals who initiated a prescription opioid within the study period. The incidence of treated OUD within the study period was 86 cases per 100,000 person-years. Compared to an average daily dose of <20 morphine milligrams equivalent (MME), higher average daily doses at initiation were associated with greater hazard of treated OUD, 20?50 MME (HR 1.11, 95% CI: 1.02, 1.21), >50-90 MME (HR 1.29, 95% CI: 1.16, 1.44), >90-150 MME (HR 1.29, 95% CI: 1.06, 1.56), >150?200 MME (HR 2.49, 95% CI: 1.54, 4.03) and >200 MME (HR 4.15, 95% CI: 2.89, 5.97). Long-acting formulations and days? supply ?11 days were also associated with greater hazard of treated OUD. Conclusion: Prescription opioid characteristics at initiation are associated with risk of treated OUD, identifying potentially important and modifiable risk factors among people initiating opioids for non-cancer pain.