ABSTRACT
OBJECTIVE: To assess the relationship between experiencing homelessness and assisting injection drug use (IDU) initiation among people who inject drugs (PWID) in Tijuana, Mexico and Vancouver, Canada. METHODS: We used self-reported questionnaire data collected semi-annually on PWID from Tijuana (n = 703) and Vancouver (n = 1551) between 2014 and 2017. Within each setting, the effect of recent (i.e., past six months) homelessness on recent provision of injection initiation assistance (i.e., helping anybody inject for the first time in the past six months) was estimated using inverse-probability-of-treatment (IPT)-weighted estimation of a marginal structural model. RESULTS: Across follow-up, the prevalence of recent homelessness at a given visit ranged from 11.6%-16.5% among Tijuana-based participants and 9.4%-18.9% among Vancouver-based participants; the prevalence of recent provision of injection initiation at a given follow-up visit was lower, ranging from 3.3%-5.4% in Tijuana and 2.5%-4.1% in Vancouver. Based on the IPT-weighted estimates, recent homelessness was associated with 66% greater odds among Tijuana-based PWID (Adjusted Odds Ratio [AOR] = 1.66; 95% CI: 1.01-2.73) and 47% greater odds among Vancouver-based PWID (AOR = 1.47, 95% CI: 1.02-2.13) of providing injection initiation assistance over the same six-month period. CONCLUSION: We found that recently experiencing homelessness was associated with an increased likelihood of PWID reporting IDU initiation assistance over time in both Tijuana and Vancouver. Addressing homelessness may decrease the initiation of IDU via multiple pathways.