Real-world longitudinal treatment trajectories following opioid use disorder diagnosis among commercially insured beneficiaries in the United States.

INTRODUCTION: Early initiation and continuity of treatment are critical for the effective management of opioid use disorder (OUD). Despite the availability of effective medications for OUD (MOUD), many individuals never initiate treatment, and those who do often discontinue early. A better understanding of real-world treatment patterns is needed to inform strategies for improved long-term engagement in care. This study uses a novel technique to characterize and visualize real-world treatment trajectories following an OUD diagnosis among commercially insured individuals.

METHODS: Retrospective, longitudinal descriptive study utilizing claims data from the IBM MarketScan® (now MerativeTM) commercial and Medicare Supplemental and Coordination of Benefits Databases, from 2010 to 2019. The study included beneficiaries aged 18-85 years with an OUD diagnosis, requiring continuous insurance enrollment for at least 6 months before and 3 months after the first date of OUD diagnosis (index date). Beneficiaries with any prescription or medical service claims for MOUD or with OUD remission codes during the 6-month baseline period were excluded. Initial OUD treatment was assessed in the 3-month period after the index date (treatment assignment period). Beneficiaries were categorized into 8 groups based on treatment receipt during the treatment assignment period: behavioral therapy (BH) only, buprenorphine (BUP), buprenorphine with behavioral therapy (BUP + BH), naltrexone (NTX), naltrexone with behavioral therapy (NTX + BH), methadone with or without behavioral therapy (MTD+/-BH), no treatment, and all other treatment combinations. Monthly treatment assignments, including disenrollment, were summarized into trajectories during the subsequent 9-month follow-up period. Sankey diagrams were used to visualize monthly transitions between treatment states and quantify the number of beneficiaries in each trajectory.

RESULTS: The study identified 339,224 beneficiaries with an OUD diagnosis and 22,902 distinct treatment trajectories. The most common treatment assignment period category was BH only, with 75.5 % of beneficiaries, while 14.1 % of beneficiaries were in a category that included any MOUD. Treatment continuity was highest among those starting with MTD+/-BH, and lowest among those NTX with or without BH. Overall, high disenrollment rates were observed regardless of treatment assignment, especially early in the follow-up period.

CONCLUSIONS: Findings highlight a significant discordance between real-world treatment patterns and evidence-based guidelines for OUD treatment.