Risk factors for human immunodeficiency virus seroconversion among out-of-treatment drug injectors in high and low seroprevalence cities

From 1988 to 1991, 6,882 drug injectors in 15 US cities were interviewed and had serum samples collected. The interviews and samples were analyzed for determination of significant predictors of human immunodeficiency virus (HIV) seroconversion in the 10 low seroprevalence cities and the five high seroprevalence cities. The unit of analysis was the period of observation between consecutive paired interviews/blood samples. In Cox proportional hazards regression, significant predictors of seroconversion in the low seroprevalence cities were: not being in drug treatment, injecting in outdoor settings or abandoned buildings, using crack cocaine weekly or more frequently, engaging in woman-to-woman sex, being of non-Latino race/ethnicity, and city seroprevalence. Predictors in high seroprevalence cities were: injecting with potentially infected syringes, not being in drug treatment, and having a sex partner who injected drugs. These findings suggest that HIV may be concentrated in sociobehavioral pockets of infection in low seroprevalence cities. For reducing HIV transmission, these results suggest: 1) in low seroprevalence cities, localized monitoring to detect specific emerging sociobehavioral pockets of infection, and quick implementation of appropriate targeted interventions if necessary; 2) in high seroprevalence cities, relatively more emphasis on locality-wide outreach and syringeexchange projects to reduce risky behavior; and 3) in both types of cities, considerable expansion of drug treatment programs.