Developing a regionally representative risk assessment tool to identify men at highest risk of HIV acquisition in sub-Saharan Africa

Univ Of North Carolina Chapel Hill, Chapel Hill NC

Investigators

Abstract

PROJECT SUMMARY/ABSTRACT Background: In sub-Saharan Africa (SSA), HIV prevention has lagged behind UNAIDS targets. Approximately 730,000 adults are infected each year, with men accounting for more than one third of these new infections. Biomedical HIV prevention strategies, such as oral and injectable pre-exposure prophylaxis (PrEP), are promising measures for reducing HIV acquisition. However, providing PrEP to all men is infeasible and unnecessary. Risk stratification approaches are needed to identify men at highest risk of HIV acquisition and in greatest need of PrEP. Although multiple risk assessment tools have been developed for women, few tools have been developed for men. Those that have been developed have important limitations with respect to internal and external validity. The goal of this proposal is to develop a risk assessment tool for identifying men at highest risk for HIV acquisition using rigorous analytic methods in a regionally representative dataset. Methods and Specific Aims: In Aim 1, we will use nationally representative data from 15 SSA countries collected from 2015-2019 as part of the Population-based HIV Impact Assessments (PHIAs), the Nigeria AIDS Indicator and Impact Surveys (NAIIS), and the South African National HIV Prevalence, Incidence and Behaviour Survey (SABSSM). These 15 surveys represent 249,000 new annual male infections in a population of 122 million susceptible men. These surveys all have a set of biomarkers (limiting-antigen avidity enzyme immunoassay, HIV-1 viral load, and antiretroviral drug detection) indicative of recent HIV-1 infection. We will apply two machine-learning methods to develop risk assessment tools to identify men aged 15-49 at greatest risk of HIV-1. We will calculate the area under the receiver operating characteristic curve (AUC), as well as sensitivity and specificity at select thresholds. In Aim 2, we will externally validate the tools in ten second-round national surveys conducted in SSA from 2020-2022 and select a final tool. Using the final tool, we will estimate the AUC and the sensitivity and specificity at meaningful thresholds. At each threshold, we will calculate the number of men at increased risk of HIV-1 infection, the number needed to treat with PrEP to avert one new infection, and the estimated number of infections averted. Team: This is a proposal led by multiple Principal Investigators (mPIs)—one with epidemiologic expertise in HIV prevention in SSA and the other with extensive statistical experience using complex survey data. The team includes co-investigators with expertise in machine learning and biomedical HIV prevention. They are supported by a technical advisory panel of regional HIV prevention specialists; the principal investigators of the PHIA, NAIIS, and SABSSM surveys; and an expert biostatistician in the field of HIV research. Impact: We will use rigorous methods to develop a regional tool to identify men at highest risk of HIV acquisition. This research is crucial for a targeted HIV response and ultimately ending the HIV epidemic in SSA.