| Introduction. Although efforts to prevent HIV transmission among injection drug users (IDUs) are important, the ultimate harm reduction for IDUs is long-term injection cessation. Previous research on injection cessation, however, has investigated short-term stops (i.e., 6 months). The main objective of this dissertation is to characterize longitudinal patterns of injection behaviors, with a specific focus on long-term injection cessation, among injection drug users (IDUs) using data from the ALIVE (AIDS Link to the IntraVenous Experience) study, a community-based cohort of IDUs in Baltimore, Maryland.;Methods. ALIVE is a prospective cohort of IDUs with semi-annual follow-up since 1988. Latent class growth analysis among 1,716 participants was conducted to determine the longitudinal trajectories of injection drug use through 2008. Discrete-time survival analysis among 1,697 participants was employed to determine the incidence and effects of residential relocation, neighborhood deprivation, and non-injection drug use on long-term injection cessation. Discrete-time survival analysis was also conducted among 673 former IDUs to determine the incidence and predictors of relapse. Finally, we conducted a qualitative study with 20 former injectors.;Results. Multiple patterns of injection drug use were identified, including long-term injection cessation (incidence rate: 3.78 per 100 person-semesters). Residential relocation increased the likelihood of long-term injection cessation, and a dose-response relationship was observed between quartiles of neighborhood deprivation at cessation initiation and long-term injection cessation. Non-injection drug use decreased the likelihood of long-term injection cessation and subsequently increased the risk of injection relapse. Non-injection drug use following injection cessation was commonly reported qualitatively. The motivators of non-injection drug use among former IDUs were varied and the social context was important to the maintenance of long-term injection cessation.;Conclusion. A substantial proportion of IDUs made established changes in their injection behaviors over a 3-year period and beyond. The findings suggest that programs for IDUs should be targeted to sub-groups of IDUs over extended timeframes to be most effective. Programs should focus not only on relapse and HIV prevention, but on supporting former IDUs throughout their cessation efforts, with particular attention given to non-injection drug use. A consideration of the social and neighborhood context is essential to this support. |
