Development of a Provider-Level Measure of Pharmacotherapy for Opioid Use Disorder

Introduction: Opioid use disorder (OUD) has continued to have devastating effects in the United States. In 2019, opioids were involved in 49,912 overdose deaths in the United States. While medications for opioid use disorder (MOUD) are known to be an effective treatment for OUD and have been shown to contribute to significant reductions in opioid use, overdose, and other risky behaviors, MOUD is an underused treatment strategy (Degenhardt et al., 2010; Sordo et al., 2017; Volkow et al., 2019). In this study, we adapt an existing health plan-level measure of the percentage of patients with OUD who are receiving MOUD to the provider level and show how the measure can be used to improve provider-level quality of care.

Methods: Data for this study included Medicaid claims for calendar year 2014 from 32 states from the Medicaid Analytic Extract files. We identified 9,398 unique providers with 10 or more beneficiaries between the ages of 18 and 64 with a diagnosis of OUD for inclusion in the analyses. These unique providers included 5,344 individual clinicians and 4,054 facilities or agencies. We conducted tests of reliability to assess signal- to-noise and to ensure the measure could be used to make meaningful comparisons between providers. Convergent validity was established by determining whether the provider-level measure was correlated with a conceptually similar measure.

Results: The mean provider score was 38.4%, indicating that for the average provider 38.5% of their patients with an OUD received a medication to treat OUD within 30 days of visiting them for OUD treatment (43.9% for individual clinicians and 31.0% for facilities). The range of the provider score was 0% to 100%. Among the bottom 25% of providers, the percentage of patients with OUD receiving MOUD was 10% and 9%, for clinicians and facilities, respectively. Among the top 75% of providers, the percentage of patients with OUD receiving MOUD was 79% and 45%, for clinicians and facilities, respectively. The measure was determined to be highly reliable in that it was able to meaningful distinguish lower and higher performing providers. The measure also demonstrated convergent validity as indicated by the fact that providers with more patients receiving MOUD also has fewer patients with hospitalizations or emergency department visits for substance use or overdose.

Conclusion: This study describes the successful adaptation of an existing health plan-level MOUD measure for Medicaid beneficiaries to the provider-level. Our findings indicate that measurement at the provider level is feasible, in that there are sufficient numbers of qualifying interactions to report at that level. Further, our analyses show that the provider-level measure is valid and reliable. Finally, the measure demonstrates clinical usefulness as the variation in performance allows considerable room for improvement at the provider level.

References: Degenhardt L, Bucello C, Mathers B, et al.: Mortality among regular or dependent users of heroin and other opioids: a systematic review and meta‐analysis of cohort studies. Addiction 2011; 106:32-51
Sordo L, Barrio G, Bravo MJ, et al.: Mortality risk during and after opioid substitution treatment: systematic review and meta-analysis of cohort studies. bmj 2017; 357
Volkow ND, Jones EB, Einstein EB, et al.: Prevention and treatment of opioid misuse and addiction: a review. JAMA psychiatry 2019; 76:208-216

Learning Objective: Participants will learn about aspects ofthe development of a quality measure that assesses meaningful differences between quality of care among providers.