The association of Medicare Part D prior authorization for buprenorphine–naloxone with adherence to opioid use disorder treatment guidelines in the United States

Abstract

Aims To assess differences in the quality of opioid use disorder (OUD) treatment received by Medicare beneficiaries enrolled in health plans that used prior authorization (PA) for buprenorphine–naloxone compared with those enrolled in plans that did not use PA. Design, Setting and Participants Cross‐sectional observational study, United States. Continuously enrolled beneficiaries (71 294) with an OUD who filled at least one prescription for buprenorphine–naloxone between March 2012 and July 2017. Measurements Percentage of patients tested for hepatis B, hepatis C, HIV and liver functioning; percentage of patients with urine drug screens and number of urine drug screens; continuous use of buprenorphine–naloxone for at least 180 days; co‐use of benzodiazepines; number of outpatient visits with and without an OUD diagnosis. Findings PA was significantly associated with a lower likelihood of testing for hepatitis B [−3.5, 95% confidence interval (CI) = −4.4, −2.7] and C (−5.9, 95% CI = −6.9, −4.9), but the findings were inconclusive as to whether or not there was a difference in HIV (−1.1, 95% CI = −2.5, 0.4) or liver function testing (1.3, 95% CI = −0.1, 2.7). PA was associated with a lower likelihood of urine drug screening (−25.5, 95% CI = −26.8, −24.1) and with fewer drug screens (−2.5, 95% CI = −3.0, −2.1). Findings were inconclusive as to whether or not there was a difference in continuous use of buprenorphine–naloxone (0.3, 95% CI = −1.2, 1.8). PA was associated with fewer outpatient visits (−2.1, 95% CI = −3.0, −1.2) and fewer outpatient visits with an OUD diagnosis (−1.7, 95% CI = −2.1, −1.3). PA was associated with a lower likelihood of filling benzodiazepine prescriptions before and after buprenorphine–naloxone induction (−28.9, 95% CI = −29.6, −28.3) but a greater likelihood of only using benzodiazepines after buprenorphine–naloxone induction (10.6, 95% CI = 9.3, 11.8). Conclusions US Medicare patients subject to prior authorization for buprenorphine–naloxone are not more likely to receive high‐quality treatment for opioid use disorder than patients not subject to prior authorization.