Insurance-related barriers to medications for opioid use disorder in private and Medicaid plans

PROJECT SUMMARY In 2020, a record 71,000 opioid overdose deaths occurred in the U.S. The number of such deaths can be reduced by buprenorphine, methadone, and extended-release naltrexone, the three FDA-approved medications for opioid use disorder (MOUD). Despite this, many patients with opioid use disorder (OUD) never initiate treatment with MOUD, and those that do often drop out of treatment, increasing their risk of death. Although the uninsured are over-represented among patients with OUD, privately insured and Medicaid patients represent three-quarters of non-elderly Americans with OUD, suggesting that mitigating barriers to MOUD use in these patients is an especially important goal. Among the many such barriers, insurance-related barriers are an appealing target for intervention, as insurance benefit design is directly under the control of policymakers, insurers, and employers. However, the effects of insurance-related barriers on MOUD use in privately insured and Medicaid patients have not been rigorously studied. Consequently, it is unclear which barriers should be targeted in policy interventions and how these interventions should be designed. In this proposal, we will provide actionable, policy-relevant information on the effect of cost-sharing for MOUD in privately insured patients and on the effect of removing prior authorization requirements for MOUD in Medicaid patients. To do so, we will analyze 2017-2023 data from several state-of-the-art national databases, including prescription dispensing data from 92% of U.S. pharmacies, a pharmacy transactions database capturing 63% of U.S. prescription volume, commercial claims from 55 million Americans, and 100% Medicaid claims from all states. Analyses will focus on patients aged 15-64 years, an age group accounting for 96% of U.S. opioid overdose deaths. In Aim 1, we will evaluate the association between cost-sharing and the risk of MOUD prescription non-dispensing in privately insured patients, both overall and among key patient subgroups, such as those initiating MOUD therapy. Additionally, we will use the commercial claims database to evaluate the association between cost-sharing for MOUD and opioid overdose risk in privately insured patients. In Aim 2, we will use difference-in-differences models, national pharmacy databases, and Medicaid claims to evaluate whether state policies eliminating prior authorization requirements for MOUD in Medicaid plans were associated with changes in MOUD dispensing, retention in MOUD therapy, and opioid overdose risk in Medicaid patients. Findings from this proposal will inform and spur efforts to optimize insurance benefit design for MOUD in private and Medicaid plans, potentially leading to the removal of insurance-related barriers that may be contributing to rising opioid-related morbidity and mortality.