The effectiveness of condoms in reducing heterosexual transmission of HIV.

CONTEXT: It is not established whether the condom is as effective at preventing heterosexual transmission of HIV as it is for preventing conception. An overall estimate of condom effectiveness for HIV prevention is needed.
METHODS: Information on condom usage and HIV serology was obtained from 25 published studies of serodiscordant heterosexual couples. Condom usage was classified as always (in 100% of acts of intercourse), sometimes (1-99%, 0-99% or 1-100%) or never (0%). Studies were stratified by design, direction of transmission and condom usage group. Condom efficacy was calculated from the HIV transmission rates for always-users and never-users.
RESULTS: For always-users, 12 cohort samples yielded a consistent HIV incidence of 0.9 per 100 person-years (95% confidence interval, 0.4-1.8). For 11 cohort samples of never-users, incidence was estimated at 6.8 per 100 person-years (95% confidence interval, 4.4-10.1) for male-to-female transmission, 5.9 per 100 (95% confidence interval, 1.5-15.1) for female-to-male transmission and 6.7 per 100 (95% confidence interval, 4.5-9.6) in samples that specified the direction of transmission. Generally, the condom's effectiveness at preventing HIV transmission is estimated to be 87%, but it may be as low as 60% or as high as 96%.
CONCLUSIONS: Consistent use of condoms provides protection from HIV. The level of protection approximates 87%, with a range depending upon the incidence among condom nonusers. Thus, the condom's efficacy at reducing heterosexual transmission may be comparable to or slightly lower than its effectiveness at preventing pregnancy