Trends in prescription opioid use and dose trajectories before opioid use disorder or overdose in US adults from 2006 to 2016: A cross-sectional study

BACKGROUND: With governments’ increasing efforts to curb opioid prescription use and limit dose below the Centers for Disease Control and Prevention (CDC)-recommended threshold of 90 morphine milligram equivalents per day, little is known about prescription opioid patterns preceding opioid use disorder (OUD) or overdose. This study aimed to determine prescribed opioid fills and dose trajectories in the year before an incident OUD or overdose diagnosis using a 2005–2016 commercial healthcare database. METHODS AND FINDINGS: This cross-sectional study identified individuals aged 18 to 64 years with incident OUD or overdose in the United States. We measured the prevalence of opioid prescription fills and trajectories of opioid morphine equivalent dose (MED) prescribed during the 12-month period before the diagnosis. Of 227,038 adults with incident OUD or overdose, 33.1% were aged 18 to 30 years, 52.9% were males, and 85.0% were metropolitan residents. Half (50.5%) of the patients had a diagnosis of chronic pain, 32.7% had depression, and 20.3% had anxiety. Overall, 79,747 (35.1%) patients filled no opioid prescription in the 12 months before OUD or overdose diagnosis, with the proportion significantly increasing between 2006 and 2016 (adjusted prevalence ratio, 1.86; 95% CI 1.79–1.93; P < 0.001). Patients without (versus with) prescribed opioids tended to be younger males and metropolitan and Northeast US residents. Of 145,609 patients who filled opioid prescriptions, 5 distinct prescribed daily dose trajectories preceding diagnosis emerged: consistent low dose (<3 mg MED, 34.6%), consistent moderate dose (20 mg MED, 27.3%), consistent high dose (150 mg MED, 15.0%), escalating dose (from <3 to 20 mg MED, 13.7%), and de-escalating dose (from 20 to <3mg MED, 9.4%). Overall, over two-thirds of patients with OUD or overdose with prescription opioids were prescribed a mean daily dose below 90 mg MED before diagnosis. Major limitations include the limited generalizability of the study findings and lack of information on out-of-pocket drug spending, race/ethnicity, and socioeconomic status of participants, which prevents analyses addressing these characteristics. CONCLUSIONS: In this study, we found that absence of opioid prescription fills in the year before incident OUD or overdose diagnosis was prevalent, and the majority of the patients received prescription opioid doses below the risk threshold of 90 mg MED. An increasing proportion of high-risk patients could be missed by current programs solely based on opioid prescribing and dispensing information in this new era of limited access to prescription opioids.