Chronic opioid use in a universal health care setting following the implementation of risk communication

Audrey Beliveau, Pharmacy, Universite De Montreal, Montreal, Canada
Anne-Marie Castilloux, Pharmacy, Universite De Montreal, Montreal, Canada
Iris Pujade, Clinical Pharmacology, Universite De Bordeaux, Bordeaux, France
Yola Moride, Pharmacy, Universite de Montreal, Montreal, Canada (yola.moride@umontreal.ca)

BACKGROUND :  To address the opioid crisis, risk communication targeted to health care providers and/or patients have been implemented in Canada, with a universal health care system.  Chronic opioid use in non-cancer patients is a well-known predictor of opioid-related harms. Patterns of chronic opioid use and associated factors following the implementation of risk communication remains poorly examined.

OBJECTIVES : i) To estimate the incidence of chronic opioid use; ii) To describe patterns of opioid acquisition; iii) To compare the characteristics of chronic and non-chronic opioid users.

METHODS: A cohort study was conducted using the Quebec administrative claims (RAMQ) databases. We selected a random sample of non-cancer outpatients aged >18 initiating an opioid (index date) between 01 January 2014 and 31 December 2016. Patients were followed for 12 months after index date. The guideline definition for chronic opioid use adopted was 90 consecutive or 120 cumulative days with an active opioid prescription during the 12-month follow-up. Demographic characteristics, medical history, comorbidity, opioid treatments, number of prescribers and of pharmacies were compared in chronic and non-chronic opioid users.

RESULTS: The cohort included 84,816 opioid-initiators, of whom 2,742 (3.23%) transitioned to chronic use (mean [SD] duration of use was 239.8 [86.9] days vs 10.8 [15.1] days in non-chronic users). Chronic users were more likely (61.2%) than non-chronic users (55.5%) to be female. Over a third (37.3%) of chronic users were >75 years of age vs 17.2% in non-chronic users.  In the year prior, 41.7%  chronic users had a diagnosis associated with pain (arthritis, spinal stenosis, fibromyalgia, neuropathy, etc.) vs 21.6% in non-chronic users.  One-fifth (20.7%) of chronic users had a psychiatric comorbidity vs 10.7% in non-chronic users. Of chronic users, 21.7% started opioid therapy with a long-acting opioid vs 1.0% of non-chronic users and transdermal fentanyl at initial prescription was more frequent in chronic (5.4%) than non-chronic (0.2%) users. Almost a quarter of chronic users (26.1%) had an initial opioid prescription >30 days vs 2.7% in non-chronic users. Doctor shopping was common in chronic users (31.9% had > 3 prescribers and 5.8% had > 3 pharmacies). 

CONCLUSION: Female sex, age >75, chronic pain, mental health issues, initial prescription of >1 month and long-acting opioids were more common in chronic opioid users. Despite communication and policies in place to prevent doctor shopping, multiple prescribers and dispensing pharmacies are still occurring.  A better understanding of these factors may inform risk communications targeted to high-risk opioid users.   

---

Abstract Reference & Short Personal Biography of Presenting Author

Professor Moride has a strong track record of methodological expertise in pharmacoepidemiology and therapeutic risk management. She is Full Professor at Université de Montréal, Research Professor at Rutgers University, and President of YOLARX Consultants. She serves on the Steering Group of the European Network of Centres of Pharmacoepidemiology and Pharmacovigilance (ENCePP) at the European Medicines Agency (EMA).  She is Past President of the International Society for Pharmacoepidemiology (ISPE) and Past Vice-President of the International Society of Pharmacovigilance (ISoP).  She was a member of CIOMS Working Group IX on «Practical Approaches for Risk Minimisation for Medicinal Products»