Download files
Abstract
Heroin dependence is a chronic condition associated with significant health and social harms. The most effective treatment for heroin dependence is opioid substitution treatment (OST), in which long-acting opioid medications such as methadone or buprenorphine are prescribed with the goal of reducing heroin use and associated harms. Internationally, OST is rarely available in prisons, despite the high proportion of heroin users among prisoners. Furthermore, limited research attention has been given to examining how prison-based OST can reduce the harms of heroin dependence.
This thesis reports on two systematic literature reviews and four data linkage studies on the effects of prison-based and post-release OST. The first systematic review found that there is good evidence that prison OST reduces heroin use and needle and syringe sharing among prison inmates. The second review found that the evidence relating to the effects of prison OST on post-release outcomes is inconsistent and has limitations. As such, four data linkage studies were undertaken to assess incarceration, offending and mortality outcomes for a cohort of 375 male heroin users recruited in prisons in New South Wales (NSW), Australia, in 1996-7. Data were linked for the nearly ten-year period 1 June 1997 – 31 December 2006.
The first data linkage study assessed whether the baseline data for the cohort could be linked to other databases with sufficient sensitivity and specificity to obtain reliable and valid results regarding episodes of OST. Results showed that maximum sensitivity and specificity were achieved when participants’ aliases were included as identifiers during the linkage process, and that enrolment in OST during the observation period had been reliably ascertained by linkage.
The second data linkage study demonstrated that exposure to OST while in prison did not in itself reduce risk of re-incarceration; rather, it was continuation of treatment as the individual returned to the community that reduced the risk of returning to prison. Among participants who remained in OST post-release, risk of re-incarceration was, on average, 80% that of participants not in OST. The third study, assessing re-offending, did not find a relationship between OST exposure and criminal convictions; however, there were indications of bias in the analysis as a result of informative censoring.
The fourth data linkage study analysed mortality outcomes for the cohort. Participant mortality was six times that seen in the age-, sex- and calendar-adjusted NSW population, but was moderated while in OST and while in prison. Although mortality was elevated in the 28 days immediately after release from prison in comparison to all other time at liberty, this difference was not statistically significant; a larger sample size may have resulted in a significant finding in this regard.
Although OST has been studied extensively, few studies have employed data linkage to examine long-term treatment outcomes, particularly in relation to treatment participation while in prison. The evidence presented in this thesis provides support for the provision of OST in prisons, and for programs that facilitate prisoners’ access to post-release OST. Integration of prisoner healthcare into public health systems may assist in improving continuity of OST as well as general standards of care. Future research should explore how the duration of pre-release treatment affects post-release outcomes and how OST can be combined with therapeutic approaches that address other risk factors for offending. Further follow-ups of the cohort would provide insights into the course and consequences of heroin use in Australia.