CONSISTENT CONDOM USE IN RELATIONSHIPS BETWEEN SEROPOSITIVE INJECTINGDRUG-USERS AND SEX PARTNERS WHO DO NOT INJECT DRUGS

Objectives: To study how condom use in injecting drug users' (IDU) rel ationships differs according to whether they are HIV-infected, and to whether their sex partner is an IDU. Design and methods: A total of 31 7 street-recruited IDU were HIV-antibody tested and interviewed about 421 relationships with particular sex partners. Results: Condoms were consistently (100%) used in sex between partners (during the previous 30 days) in 33% of these relationships, and their use was significantl y more frequent in relationships of seropositive IDU and in relationsh ips with non-IDU partners. In relationships between seropositive IDU a nd non-IDU, consistent condom use was reported to be high (68%); this remained unchanged under multivariate controls. Conclusions: Self-repo rted condom use by IDU in New York, with its relatively mature epidemi c, appears to be concentrated where it may most reduce the spread of H IV to non-IDU heterosexuals, i.e., in relationships between infected I DU and non-IDU partners. Differential condom use by serostatus and by partners' drug injection should be incorporated into mathematical mode ls of the HIV epidemic. Causes of the high level of condom use in this subset of relationships may include drug injector altruism and pressu re by sex partners; prevention programs should develop ways to use bot h of these factors to motivate increased condom use.