Abstract
INTRODUCTION: The period post-residential treatment for opioid use is associated with heightened risk of overdose. The current study aims to: (i) describe characteristics of men and women attending residential treatment for primary opioid use; (ii) identify classes of clients based on primary opioid and other substance/s of concern and (iii) identify relationships between substance use profile and sociodemographic risk factors for opioid overdose, including differences by gender. METHODS: Data from 2994 clients (29.6% women) attending residential treatment for opioid use in the non-government sector in New South Wales, Australia, were included in the analysis. Descriptive and chi-square statistics analysed demographic, clinical, substance use and service use characteristics of participants. Participants were grouped using latent class analysis based on their primary opioid of concern and other substance/s of concern. Multinomial logistic regression examined the relationship between latent classes and sociodemographic risk factors for overdose, including interactions with gender. RESULTS: Men and women presented to residential treatment with different demographic, clinical, substance use and service use characteristics. A five-class model of polysubstance use was identified: heroin + lower polysubstance use (52.3%), heroin + polysubstance use (22.2%), pharmaceutical + lower polysubstance use (10.1%), pharmaceutical + polysubstance use (6.7%) and OAT + polysubstance use (8.7%). There were some associations between sociodemographic risk factors and class membership, though limited interactions between sociodemographic risk and gender. DISCUSSION AND CONCLUSIONS: Effective overdose prevention and harm reduction strategies during and post-residential treatment need to consider individuals' complex and unique accumulation of risk.