Combining Social Network, Spatial, and Phylogenetic Approaches to Identify New HIV Infections

As nearly half of new infections are thought to be attributed to undiagnosed infections, new strategies to better identify undiagnosed infections and link them with care are critically needed. Current HIV case-finding strategies use social network (i.e. contact tracing & analyses of sex and drug use partner data to identify those at risk for infection), spatial (i.e. locate and characterize the social environment of “hot spots"), and phylogenetic (i.e. infer/characterize putative transmission links among infected individuals based on viral sequence similarities), approaches. Although combining these strategies in close to real-time could lead to a more comprehensive picture of HIV transmission and the development of better targeted and multi-level intervention approaches to identify new infections and prevent ongoing transmission to others, such integration is rare. Each approach has limitations when used alone, but they can complement and validate one another when used together. Social network data on recent sex/drug use partners provide insight into who is at risk for acquiring HIV based on the current network structure and existing relationships. Phylogenetic links are inferred between HIV-infected persons, are typically constructed retrospectively using de-identified samples, and are often not validated with sex/drug use partner data. Without data on sex and drug use partners, two genetically similar viral sequences could be due to direct transmission, indirect transmission via a common source, or a series of intermediaries. Further, because phylogenetic links only include HIV-infected individuals, they provide limited information about who might be at risk for future infection when not combined with social network and spatial data. Thus, combining phylogenetic and social network data can potentially uncover bridges between seemingly distinct networks. Finally, because individual risk behaviors are influenced by who people interact with and where people meet partners/engage in risk behaviors, analyzing social network and phylogenetic data together with spatial data can provide the social context for these interactions. The specific aims are to (1) Combine social network and spatial data from HIV infected individuals and their HIV-negative sex and drug use partners to identify novel multi-level and targeted strategies to find undiagnosed HIV infections; and (2) Overlay putative phylogenetic transmission networks with social network and spatial data and compare HIV case-findings strategies identified in Aim1 with those identified with this combined approach. To achieve the Aims, newly diagnosed HIV cases (N=50) from the Boston Medical Center HIV clinic and their peer-referred HIV negative drug and sex partners (N=75) will complete a survey to collect demographic, clinical, behavioral, social network, and spatial data (Aim 1). HIV pol sequences from routine drug resistance testing of 500 consenting BMC HIV patients will be obtained for phylogenetic analyses, to be combined with social network and spatial data (Aim 2). These aims will enable us to identify the optimal combination of social-network, spatial-, and venue-based approaches to better target HIV-case- finding strategies.