Use of condoms and other contraceptive methods by HIV-infected and at-risk women.

Knowledge of factors associated with incremental frequency of condom use will aid in the design of interventions to prevent HIV and other sexually transmitted diseases. In this cross-sectional analysis 604 HIV-infected and 322 at-risk uninfected women enrolled in the multi-site HIV Epidemiologic Research study of HIV disease progression and underwent standardized interviews. Information on demographic characteristics contraceptive use drug use and sexual behaviors was examined by using a polytomous (multinomial) logistic regression model to determine predictors of always sometimes and never use of condoms. HIV infection was the strongest independent predictor of always use of condoms (relative odds ratio [ROR] = 6.21 95% confidence interval [CI] 4.04-9.56) when compared with never users. Compared with women with multiple male sex partners women reporting one male sex partner in the previous 6 months were one-third as likely to report always condom use compared with never (ROR = 0.31 95% CI 0.91-0.51) and one-fifth as likely to report sometimes use of condoms compared with never (ROR = 0.22 95% CI 0.14-0.35). Similarly use of non-barrier coitally independent contraception (i.e. surgical sterilization hormonal or IUD was independently associated with never use compared with always and sometimes use of condoms. The authors study suggests that specific risk behaviors including non-barrier contraceptives and sexual and substance abuse behavior should be the focus of interventions to promote condom use. Efforts to promote condom use may need to be tailored to women according to HIV status. (authors)