Recent incarceration and risk of first-time injection initiation assistance: A prospective cohort study of persons who inject drugs

Background: Given the prevalence and harms of incarceration among persons who inject drugs (PWID) and their role in injection drug use initiation, we aimed to investigate whether recent incarceration influences the like-lihood PWID assist others in their first-ever injection. Methods: Prospective cohort study of PWID in Vancouver, Canada who had their PReventing Injecting by Modifying Existing Responses (PRIMER) baseline visit between December 2014 and May 2017, reported never providing injection initiation assistance previously, and had >= 1 follow-up visit. The primary outcome, provision of injection initiation assistance, was defined via self-report as helping anybody inject for the first time in the past six months. The primary exposure was recent incarceration, i.e., self-report of being jailed, imprisoned or detained in the past six months. Participants were assessed biannually until November 2017, drop-out, or their first report of the primary outcome. Results: 1,199 PWID (62.1% male; mean (SD) age, 44.4 (12.3) years) were included in our study. Across 4,171 follow-up visits, 67 participants (5.6%) reported providing injection initiation assistance. The proportion of participants reporting recent incarceration varied between 2.4% to 5.1% per follow-up visit. Based on a multivariable discrete-time proportional hazards regression analysis, recent incarceration was associated with an increased risk of providing injection initiation assistance during the same six-month period (adjusted hazard ratio, 2.64; 95% CI, 1.19 to 5.86). Conclusions: The observed association between recent incarceration and risk of providing injection initiation assistance suggests that incarceration could be contributing to the expansion of injection drug use practices within vulnerable populations over time.