Effectiveness of HIV Interventions Among Crack Users

SUMMARY This study is one of the first to concentrate upon chronic crack users in determining the effectiveness of HIV interventions in changing risk behaviors. A population of 185 crack users were randomized into two arms of an experimental design. One received a standard intervention, designed by the National Institute of Drug Abuse and a committee of grantees, and the other arm received a locally designed innovative intervention. Equivalent characteristics in both groups confirmed the effectiveness of randomization. Base line and follow-up measures are compared for both groups indicating significant behavioral changes for each of the two groups. Also, significant differences are noted between the standard and innovative interventions for a specific subgroup. The statistical tests of group differences employing composite variables included a multiplicative interaction term in the regression model. This study suggests the increasing need to develop, implement and evaluate intervention models targeted to specific subgroups and specific behaviors in order to find the most appropriate and cost-effective prevention models.