HIV Serostatus Disclosure, Condoms, and HIV/STD Prevention

Abstract

To the Editor: In a recent study, Rietmeijer et al. found that 36.3% of 1400 men who have sex with men who reported having one or more “occasional” sex partners in the previous 4 months “always” engaged in human immunodeficiency virus (HIV)-serostatus discussions with potential partners before sex.1 Open and honest discussion of HIV status could have important public health benefits by increasing condom use within serodiscordant partnerships. However, as the authors note, for seroconcordant couples, serostatus discussions could decrease the likelihood of condom use, facilitating the transmission of non-HIV sexually transmitted diseases (STDs) and, within seroconcordant-positive relationships, increasing the risk of superinfection with a different, possibly drug-resistant strain of HIV. We conducted a model-based analysis of the impact of HIV-positive serostatus disclosure on the probability of HIV transmission within serodiscordant relationships.2 The main analysis assumed that between 25% and 75% of uninfected prospective sex partners would decline to have intercourse after learning their HIV-positive partner's serostatus, that disclosure would increase the likelihood of condom use from 72.7% to 77.6%,3 and that condoms are 90% effective at preventing HIV transmission.4 Under these rather conservative assumptions, disclosure by HIV-positive persons would reduce the per-act probability of HIV transmission during sexual encounters with uninfected partners by between 34.6% and 78.2%, depending on the proportion of partners who declined intercourse (25% or 75%, respectively). Moreover, a 78.3%–92.8% reduction in HIV transmission risk would be obtained if condoms were used 100% of the time in serodiscordant relationships. This same model can be applied to the risk of STD transmission within HIV-concordant relationships. Again assuming that condoms are used 72.7% of the time absent serostatus disclosure, if HIV-concordant couples decreased their frequency of condom use to an average of 37.1% after mutual serostatus disclosure, the risk of STD transmission within STD-discordant partnerships would increase by 92.8% (this is the same as the reduction in HIV-transmission risk obtained by increasing condom use to 100% within HIV-discordant couples). We are not aware of any empirical estimates of potential decreases in condom use after serostatus discussions between HIV-negative partners. Regardless, whether potential increases in STD transmission within HIV-concordant/STD-discordant partnerships would be sufficient, from a public health standpoint, to offset decreases in HIV transmission in HIV-discordant relationships depends on the relative proportions of HIV and STD-discordant relationships in a particular population and on the relative value placed on preventing HIV and STD transmission. In light of the substantial uncertainty that surrounds these issues, we strongly second the call of Rietmeijer et al. for additional research on the effectiveness and potential limitations of interventions to increase HIV-serostatus discussions.1 It might be helpful, however, to distinguish between interventions to increase serostatus disclosure by HIV-positive persons and generic efforts to increase serostatus discussions by all persons, irrespective of serostatus. A better-targeted approach potentially could reduce HIV transmission without substantially increasing STD incidence.