Despite available strategies to prevent sex risk behaviours in Puerto Rico, heterosexual transmission of HIV continues to increase. Since 1990, heterosexual contact has been the fastest growing infection risk category among the island's general population, and the primary transmission route for women and children. To understand change in sex risk behaviours and factors related to change, 911 drug injectors and 359 crack smokers were recruited from the San Juan metropolitan area following a stratified cluster design. This study comprised a total of 1,004 (79.1%) drug users who were assessed at follow-up. Abstinence from sex behaviour increased from 54.6% to 61.1% (p < 0.01), use of condoms during vaginal sex also increased from 26.4% to 36.9% (p < 0.01). In multivariate analysis, significant predictors of abstinence were gender, injection drug use, HIV seropositivity and not having a steady partner. Predictors of using condoms during vaginal sex were HIV seropositivity, STD diagnosis and participation in an HIV preventive programme. These findings indicate that additional HIV preventive efforts are needed to reduce sex risk behaviours among drug users who have a steady sex partner, as well as among drug users who are HIV-negative.
PIP: Factors associated with changes in HIV risk behaviors were investigated in a prospective study of 911 injecting drug users and 359 crack smokers recruited on the basis of a stratified cluster design in metropolitan San Juan, Puerto Rico, in 1992-93. The findings presented in this paper are based on the 1004 drug users (79.1%) who were available for a follow-up interview 6 months after the baseline interview. Respondents were randomly assigned to receive a standard HIV prevention intervention (pretest counseling, optional HIV testing, and provision of HIV results) or the standard intervention plus a site-specific enhanced intervention entailing access to condoms and needle hygiene equipment, development of a personal risk reduction plan and training in the skills required to implement it, and ongoing contact with outreach workers. Overall, abstinence from sexual behavior increased from 54.6% at baseline to 61.1% at follow-up (p 0.01) while condom use during vaginal sex rose from 26.4% to 36.9% (p 0.01). In multivariate analysis, significant predictors of abstinence were male gender, injecting drug use, a positive HIV test at baseline, and not having a steady partner. Significant predictors of condom use were HIV seropositivity and a sexually transmitted disease diagnosis. Participation in the enhanced compared to the standard HIV prevention program was associated with an odds ratio of 1.29 (95% confidence interval, 0.95-1.74) for abstinence and 1.97 (95% confidence interval, 1.18-3.29) for condom use in vaginal sex.