Technical and financial support for the additional analysis of HSRC's HIV/AIDS population-based survey datasets and assistance with setting up of a Monitoring and Evaluation (M&E) unit of the South African National AIDS Council (SANAC)

STATUS: Current
PROJECT LEADER:Rehle, TMG (Prof. Thomas)
OTHER TEAM MEMBERS: Jacobs, NPP (Mr Nico), Shean, YL (Ms Yolande), Simbayi, LC (Prof. Leickness), Nyawane, CL (Ms Lebo), Wabiri, MN (Dr Njeri), Moyo, S (Dr Sizulu), Mabaso, MLH (Dr Musa), Jooste, SE (Mr Sean)
DEPARTMENT RESPONSIBLE: Social Aspects of Public Health (SAPH)

Abstract

Project Framework Narrative There are three main sub-objectives as follows; 1.1 HIV incidence estimation ??? comparison of multiple methods Methods to measure HIV incidence can be grouped into two broad categories: the epidemiological approaches and those based on laboratory methods. Triangulation of different approaches is recommended to provide well-supported HIV incidence estimates. a) Assay-derived incidence: The HIV incidence testing algorithm used in the 2012 survey applies the Limiting-Antigen Avidity Assay (LAg-Avidity EIA) in combination with additional clinical information on antiretroviral treatment exposure and HIV viral load. b) Mathematically-derived incidence: HIV incidence estimation from HIV prevalence data collected in repeated national population-based surveys. Data from the 2008 and 2012 surveys will be used to estimate national HIV incidence rates for the inter-survey period 2008-2012. c) Model-derived incidence: Besides the EPP and Spectrum software packages, ASSA 2008 and the HIV-STI Interaction model are potential models to be included in this analysis. This work could be part of a proposed model validation project by the HIV Modelling Consortium comparing age-specific model predictions to 2012 HSRC survey results (See concept note by Dr Tim Hallett in Appendix 2). 1.2 Impact of antiretroviral treatment (ART) The impact of ART on HIV/AIDS-related deaths, HIV incidence and HIV prevalence over time in South Africa will be assessed using different models already mentioned above, including the newly developed THEMBISA modeling package by Leigh Johnson, University of Cape Town. This work will be done in collaboration with T. Hallett (Imperial College) and L. Johnson (UCT). In addition, the model estimates will be compared with the empirical data on the changing HIV prevalence profile by age group and sex provided by the four national HSRC surveys conducted in 2002, 2005, 2008 and 2012. 1.3 Geo-mapping the South African HIV epidemic The very heterogeneous nature of the South African HIV epidemic warrants a more detailed analysis beyond the province level in order to design more effectively targeted prevention and care programs. Using the 2012 survey data we propose a geo-mapping exercise of the current epidemic to identify ???HIV hot spots??? at the district level and even at the level of Enumeration Area (EA) clusters. Other data sources would include the 2012 antenatal survey, 2012 HIV Communication Survey, 2012 loveLife evaluation, various RCTs including FACTS001 on microbicides and PopART on treatment as prevention as well as other specific studies carried out in selected provinces