Agent-based network model predicts strong benefits to youth-centered HIV treatment-as-prevention efforts

Abstract

______________________________________________________________________________________We used an agent-based network model to examine the effect of targeting different risk groups with unsuppressed HIV viral load for linkage or re-linkage to HIV-related treatment services in a heterosexual population with annual testing. Our model identifies prevention strategies that can reduce incidence to negligible levels (i.e., less than 0.1 infections per 100 person-years) 20 years after a targeted Treatment-asPrevention (TasP) campaign. The model assumes that most (default 95%) of the population is reachable (i.e., could, in principle, be linked to effective care) and a modest (default 5% per year) probability of a treated person dropping out of care. Under random allocation or CD4-based targeting, the default version of our model predicts that the TasP campaign would need to suppress viral replication in ~80% of infected people to halt the epidemic. Under age-based strategies, by contrast, this percentage drops to 50% to 60% (for strategies targeting those <30 and <25, respectively). Age-based targeting did not need to be highly exclusive to yield significant benefits; e.g. the scenario that targeted those <25 years old saw ~80% of suppressed individuals fall outside the target group. This advantage to youth-based targeting remained in sensitivity analyses in which key age-related risk factors were eliminated one by one. As testing rates increase in response to UNAIDS 90-90-90 goals, we suggest that efforts to link all young people to effective care could be an effective long-term method for ending the HIV epidemic. Linking greater numbers of young people to effective care will be critical for developing countries in which a demographic "youth bulge" is starting to increase the number of young people at risk for HIV infection.