A feasibility analysis of implementing interventions for discordant couples in 14 African countries: implications for epidemic control.

We find interventions targeting serodiscordant couples (SDC) may not be feasible in countries where HIV prevalence is less than 5%, because only 3-19/1000 individuals are HIV-positive/negative and in SDC. Interventions may be feasible in countries where prevalence is greater than 10%, because 34-48/1000 individuals are HIV-positive/negative and in SDC. We calculated that 20-27% of all HIV-positive individuals, but less than 6% of all HIV-negative individuals, are in SDC. Consequently, targeting HIV-positive partners could significantly reduce transmission, whereas targeting HIV-negative partners may have little impact.