Grabbe et al. take issue with our evidence that couples voluntary counseling and testing (CVCT) for identifiable HIV-discordant couples, although important, is likely to have limited population-level impact on the HIV epidemic . In order for CVCT to have a substantial population-level impact on the HIV epidemic through the reduction of transmission within stable identified discordant couples, all of the following conditions must apply: a substantial proportion of all transmissions in the population must occur from intra-dyad transmissions; such stable discordant couples must be identified through testing of both partners, with mutual disclosure and willingness to undertake couples counseling; and the behavior change intervention that occurs must be efficacious, at sufficient scale, cost-effective and durable. The programmatic perspective on what is actionable at large-scale level is key.
Our community-based cohort data provide key evidence regarding the first two requirements and indicate the potential for population-level impact is relatively low in the Rakai setting. We believe Rakai is a best case scenario for a potential impact of CVCT in that a very high proportion of people marry, and our intensive VCT and CVCT community outreach is probably far better than could realistically be achieved for sub-Saharan Africa as a whole. Nonetheless, our data indicate that only a minority of all new infections occur in the uninfected partner within identified stable discordant couples. The large majority of infections occurred in concordant negative couples (thus, at least one partner had to have an external source of infection), the unmarried and those married but with a partner of unknown status. Part of the reason transmissions within identified stable couples represents a minority of all infections is because transmission rates in the chronic phase of HIV infection are lower than perhaps might be expected  and that discordant relationships are prone to dissolution due to separation, divorce and death of the infected partner . We also noted in our article that viral genome studies indicate an important minority of new incident infections which occur in the HIV-negative person within a discordant partnership actually come from an outside source.
We agree with Grabbe et al. that it is important to consider the potential effects of antiretroviral therapy (ART) use on transmission in discordant couples and we have reported this elsewhere . However, due to resource constraints, ART in Rakai (as in most of Africa) is only initiated with advanced disease (in Rakai, CD4 cell count <250 cells/μl or WHO Clinical stage 4), and thus only a minority of HIV-positive partners were on HIV therapy in the period 2006–2008. None of our seroconversions occurred in the small number of discordant couples wherein the infected partner was on ART. Still, treatment might help explain why the contribution of identified discordant couples to total incident infections declined from 18.3% in the pre-ART period to 13.7% in the ART era, and providing ART to the infected partner in a discordant couple should be considered as an additional prevention modality.
We also agree with Grabbe et al. that couples counseling is not a stand-alone intervention and should complement other prevention interventions, and that it can be an excellent gateway to a variety of services. Moreover, we agree that it is important to promote outside partner reduction and condom use, within couples counseling. But, in view of the evidence that infections are so dispersed in the population, this messaging should to be provided to all couples, individuals and communities in order to influence individual behavior and larger social norms. In addition to couples and individual counseling, messaging should be reinforced through mass media and community-level communication interventions.
Conflicts of interest
1. Grabbe KL, Medley A, Bachanas P, Bock N, Marum E. The contribution of HIV-discordant relationships to new HIV infections in Rakai, Uganda.AIDS
2. Gray R, Ssempiija V, Shelton J, Serwadda D, Nalugoda F, Kagaayi J, et al. The contribution of HIV-discordant relationships to new HIV infections in Rakai, Uganda
3. Boily MC, Gaggaley RF, Wang L, Masse B, White RG, Hayes RJ, et al. Heterosexual risk of HIV-1 infection per sexual act: systematic review and meta-analysis of observational studies
. Lancet Infect Dis
4. Porter L, Hao L, Bishai D, Serwadda D, Wawer M, Lutalo T, et al. HIV status and union dissolution in sub-Saharan Africa: the case of Rakai, Uganda
5. Reynolds SJ, Makumba F, Nakigozi J, Gray RH, Wawer M, Quinn TC, et al. HIV-1 transmission among HIV-1 discordant couples before and after the introduction of antiretroviral therapy