A sample of 582 injecting drug users were interviewed as part of an evaluation of an AIDS prevention programme for drug users. This paper examines the biographic and predispositional determinants of five HIV preventive behaviours--equipment sharing (not receiving and not giving) and and condom use (with regular partners, casual partners and sex clients). A two-stage sequential approach was adopted for a logistic regression analysis. Initially, to model each of the five preventive behaviours, biographical and drug use variables were entered. In a second set of models, behavioural predisposition factors were included. Age, drug use and prison experience correlate with variables in both models, although not consistently in the same direction. While a predisposition to reject sharing correlates with safer rejecting and condom use, the predisposition to safer sex only correlates with condom use. Needle exchange programmes that only target the individual would seem to be inadequate. To enhance targeted interventions changes in public and agency policy that create a social environment conductive to behaviour change are required.
PIP: Interviews were conducted with 582 intravenous (IV) drug users aged 16-55 (85.9% of whom were male) living in Toronto, Canada, so researchers could examine the biographic and predispositional determinants of HIV preventive behaviors: equipment sharing (not receiving and not giving) and condom use (with regular partners, casual partners, and sex clients). This study was part of an evaluation of one of Canada's first HIV prevention programs for IV drug users, which included education, counseling, needle and syringe exchange, and an environmental campaign to Keep the Needles Off the Street. Cocaine was the most commonly abused drug (95.4%) in the last six months. Condom use was highest with sex clients (53.1%), followed by regular partner (43.1%) and casual partner (30.8%). The researchers had adopted a two-stage sequential approach for a logistic regression analysis. They first applied biographical and drug use variables to model each of the preventive behaviors. They then included the behavioral predisposition factors in a second set of models. Younger IV drug users were more likely to practice safer sex and safer needle use than older IV drug users. IV drug users who had been in prison were more likely to use condoms with sex clients and not to give used equipment to others. Among all substance types, only heroin and barbiturate use were linked to sexual risk. The predisposition to reject equipment sharing related to safe injecting and condom use, while the predisposition to accept safer sex only related to condom use, suggesting that the mechanisms for behavior change associated with safer sex and equipment sharing may vary. Needle exchange programs exist for IV drug users and focus only on the individual. These findings suggest that these programs appear to be ineffectual. Improvement of targeting and prevention strategies, including creation of a social environment conducive to change, is needed.