The role of large connected components in sexual networks for HIV spreading in Sub-Saharan Africa #

Francesco Di Lauro

13:50 Tuesday in 2Q50/51.

Part of the Modelling and inference in public health session.

Abstract #

The HIV epidemic in sub-Saharian Africa is historically characterised by high-levels of prevalence and incidence. Thanks to a global effort to reach UNAIDS 90-90-90 target worldwide and to the scaling-up of HIV treatment, as well as youth-focused protection interventions, the incidence has finally declined over the past ten years, albeit non-consistently across different sex and age-groups. A few questions need to be addressed in order to tailor setting-specific interventions aimed at reducing the burden of HIV: are there unobserved groups of individuals who are responsible for most of the transmissions? Is incidence declining across all groups? In this setting, modelling is a valuable tool to assess the potential impact of actionable measures. Recently, an individual-based model calibrated to (anonymised) age-and-sex stratified data, named PopART-IBM, has been developed to in the context of a HPTN-071 (PopART) study community in Zambia. The outputs of the model include the full transmission and partnership networks. In this work, we explore these and show that the sexual partnership network exhibits a strongly connected component (giant component), embracing the majourity of the population, in each of the communities under investigation. We will analyse the demographic characteristics of people in the connected component, and discuss what are the implications of this topology for policy-makers. Lastly, we will present early results from an improved analysis and modelling of the available data.