This application requests funding to evaluate two community-based AIDS prevention models to assist low income African-American women in reducing activities which confer HIV infection risk and evaluate their outcome against a social support self-help model. Minority women in urban areas characterized by poverty, drug use, high pregnancy rates, and high STD prevalence represent a population at high HIV infection risk, yet there are few controlled evaluations of AIDS risk reduction interventions for this population. The proposed study will recruit 240 women from three primary health care clinics serving low income residents in Jackson, Mississippi. 80 women will be randomly assigned to participate in a 6-session group social support and self-help intervention (SS), 80 women will participate in a 6-session social support plus modeling intervention based on social learning theory (SS+M), and 80 women will be randomly assigned to a social support and modeling plus behavior rehearsal condition (SS+M+BR) grounded in traditional behavior theory. Training will be followed by booster sessions 3 months and 1 year after the intervention to enhance maintenance. Before intervention, three months after training, and at 12-month followup, all participants will complete measures of AIDS risk knowledge; perceptions of social norms; estimates of personal vulnerability, self efficacy, and response efficacy beliefs; self -reported sexual and substance use behavior, health locus of control, condom application and syringe cleaning skill, behavioral skill in role-play simulations involving sexual negotiation with a partner, and report the frequency with which she has discussed AIDS precautions with others in her social network. As corroborative measures, STD data will be collected from medical records before and l2 months after the intervention. Multivariate data analysis will compare the two experimental groups and social support group to evaluate the interventions. It is hypothesized that all participants will obtain higher knowledge scores and social support scores following their respective interventions. Participants in the SS+M+BR groups are predicted to be lower in risk behavior, more behaviorally skilled in role play enactments, more skilled in condom application and syringe cleaning skills, endorse more internal health locus of control beliefs, and evidence lower STD rates after intervention . This research will provide direct comparison of three widely proposed intervention models and yield information that will be practical and feasible in urban health care settings which serve urban poor African American women.
