Download files
Abstract
Aims:
This thesis examined non-adherence with methadone (MET), buprenorphine (BPN) and buprenorphine-naloxone (BNX), and evaluated the abuse liability of BNX in Australia. Key practices of interest included: diversion of opioid substitution therapy (OST) medications to illicit markets; injection of medication intended for oral/sublingual administration; and non-adherence with the supervised dosing requirements of OST programs.
Methods:
Data from Australian post-marketing surveillance studies of BNX (2006-2008) were utilised, including: sales data; surveys of people who inject drugs (PWID) and OST clients; a census of authorised OST prescribers; and needle-syringe program data. The levels of injecting among PWID were adjusted for background availability of medication, and non-adherence among OST clients (i.e. removal of supervised doses, diversion of prescribed medication, etc) was quantified relative to the total number of daily doses dispensed. Latent Class Analysis was used to characterise groups of OST clients based on patterns of reporting stability/adherence indicators (derived from Australian OST guidelines). A linear regression model examined predictors of unsupervised dosing.
Findings:
Adjusting for background availability, BNX was less commonly and less frequently injected among PWID and OST clients, and diverted in smaller quantities, than BPN (at similar levels to that for MET). The most common motivation for using diverted BNX among out-of-treatment PWID was self-treatment of withdrawal. In 2008, the street price of BNX was equivalent to that for BPN. Two latent classes of OST clients were identified: a smaller non-adherent group (33%) with higher probabilities of diversion, injection and removal of supervised doses, and a larger adherent group (67%). Unsupervised dosing was associated with demographic and treatment characteristics, but not class membership. Prescribers perceptions were that the majority of patients adhered with OST, possibly underestimating how frequently diversion and injection occurs.
Conclusions:
Agonist-antagonist formulations, such as BNX, may reduce injection, but clearly will not deter it completely. Multiple measures are needed: clinical risk assessment, careful patient selection for unsupervised dosing and ongoing monitoring. The OST program faces challenges: evaluating the differing harms associated with different opioid formulations; agreeing whether there is an acceptable level of diversion and injection; accurately identifying stable patients; and expanding the provision of unsupervised dosing.