Frequent Needle Exchange Use and HIV Incidence in Vancouver, Canada

Purpose Opposition to needle exchange programs has been fueled by a Vancouver study showing an association between frequent program use and elevated rates of human immunodeficiency virus (HIV) infection among injection drug users. Methods We evaluated possible explanations for the observed association between elevated HIV rates and frequent needle exchange attendance using a prospective observational cohort study of injection drug users in Vancouver, BC, Canada. HIV incidence rates were examined using stratified Kaplan-Meier methods and Cox proportional hazards regression. Results Between May 1996 and December 2004, 1035 individuals were recruited. At 48 months after recruitment, the cumulative HIV incidence rate was 18.1% among those reporting daily needle exchange use at baseline, compared with 10.7% among those who did not report this behavior (P<.001). However, comparing HIV incidence among daily versus nondaily exchange users, while stratifying the cohort into those who did (23.2% vs 16.8%; P = .157) and did not (11.4% vs 9.0%; P = .232) report daily cocaine injection at baseline, the association between daily exchange use and HIV incidence was no longer significant. In an adjusted Cox model, daily exchange use was not associated with the time to HIV seroconversion (relative hazard = 1.41 [95% confidence interval, 0.95-2.09]). Conclusions Differential HIV incidence rates between frequent and nonfrequent needle exchange attendees can be explained by the higher risk profile of daily attendees. Causal factors, including the high rates of cocaine injection and other local injection drug user characteristics, explain the Vancouver HIV outbreak.