Abstract
BACKGROUND: HIV prevention intervention efficacy is often assessed in the short term. Thus, we conducted a long-term (mean 4.4 years) follow-up of a woman-focused HIV intervention for African American crack smokers, for which we had previously observed beneficial short-term gains. METHODS: 455 out-of-treatment African American women in central North Carolina participated in a randomized field experiment and were followed up to determine sustainability of intervention effects across three conditions: the woman-focused intervention, a modified NIDA intervention, and a delayed-treatment control condition. We compared these groups in terms of HIV risk behavior at short-term follow-up (STFU; 3-6 months) and long-term follow-up (LTFU; average 4 years). RESULTS: The analyses revealed two distinct groups at STFU: women who either eliminated or greatly reduced their risk behaviors (low-risk class) and women who retained high levels of risk across multiple risk domains (high-risk class). At STFU, women in the woman-focused intervention were more likely to be in the low HIV risk group than the women in control conditions, but this effect was not statistically significant at LTFU. However, low-risk participants at STFU were less likely to be retained at LTFU, and this retention rate was lowest among women in the woman-focused intervention. CONCLUSIONS: Short-term intervention effects were not observed over 4 years later, possibly due to differential retention across conditions. The retention of the highest risk women presents an opportunity to extend intervention effects through booster sessions for these women.