Measurement Approaches to Estimating Methadone Continuity in Opioid Use Disorder Care

Abstract

Background: Long-term treatment with medications for opioid use disorder (OUD), including methadone, is lifesaving. There has been little examination of how to measure methadone continuity in claims data. Objectives: To develop an approach for measuring methadone continuity in claims data, and compare estimates of methadone versus buprenorphine continuity. Research Design: Observational cohort study using de-identified commercial claims from OptumLabs Data Warehouse (January 1, 2017–June 30, 2021). Subjects: Individuals diagnosed with OUD, ≥1 methadone or buprenorphine claim and ≥180 days continuous enrollment (N=29,633). Measures: OUD medication continuity: months with any use, days of continuous use, and proportion of days covered. Results: 5.4% (N=1607) of the study cohort had any methadone use. Ninety-seven percent of methadone claims (N=160,537) were from procedure codes specifically used in opioid treatment programs. Place of service and primary diagnosis codes indicated that several methadone procedure codes were not used in outpatient OUD care. Methadone billing patterns indicated that estimating days-supply based solely on dates of service and/or procedure codes would yield inaccurate continuity results and that an approach incorporating the time between service dates was more appropriate. Among those using methadone, mean [s.d.] months with any use, days of continuous use, and proportion of days covered were 4.8 [1.8] months, 79.7 [73.4] days, and 0.64 [0.36]. For buprenorphine, the corresponding continuity estimates were 4.6 [1.9], 80.7 [70.0], and 0.73 [0.35]. Conclusions: Estimating methadone continuity in claims data requires a different approach than that for medications largely delivered by prescription fills, highlighting the importance of consistency and transparency in measuring methadone continuity across studies.