INTRODUCTION: Medication adherence may be associated with improved healthcare outcomes. Understanding risk factors of poor adherence to buprenorphine medication assisted treatment (BMAT) among patients with opioid use disorders (OUD) is key to improving management of this population.
OBJECTIVE: Identify factors associated with BMAT adherence one year following BMAT initiation.
METHODS: OUD patients initiating BMAT were selected from the 2008-2014 MarketScan Commercial and Medicaid databases (earliest claim = index). Patients were followed 6 months pre-index and 12 months post-index. Logistic regression models were used to identify patient characteristics predictive of adherence, defined as post-index proportion of days covered (PDC) ≥80%.
RESULTS: 16,085 Commercial and 5,688 Medicaid patients were analyzed, with 37.1% and 41.3%, respectively, classified as adherent. Among Commercial patients, factors that significantly lowered the likelihood of being adherent (all p <0.05) were BMAT average daily dose <12 mg (odds ratio [OR]=0.58), being a dependent/spouse of the primary insurance subscriber (OR=0.61), and pre-index alcohol abuse disorder (OR=0.82), non-opioid substance abuse disorder (OR=0.76), depression (OR=0.92), or bipolar disorder (OR=0.74). In Medicaid patients, factors predictive of decreased likelihood of adherence were BMAT average daily dose <12 mg (OR=0.38), access restrictions to buprenorphine treatment (OR=0.80), pre-index use of narcotic medication (OR=0.82), and pre-index diagnosis of non-opioid substance abuse disorder (OR=0.79) or bipolar disorder (OR=0.86).
CONCLISIONS: Adherence to BMAT is associated with dose, insurance subscriber status, and comorbidities including alcohol use disorder, pain, depression, and bipolar disorder. BMAT dose should be carefully selected for patients at risk of non-adherence.
