Incidence is the most sensitive indicator of epidemiological trends, but it is hard to estimate for HIV. Epidemic surveillance therefore largely relies on population-level prevalence surveys, antenatal clinic surveys and various forms of routine data. A range of methods are used to estimate HIV incidence, but no single approach attains the levels of precision one might hope for with the data available. To help address this, we developed an approach that implements an optimal weighting of a biomarker-based incidence estimator and an incidence estimator based on estimates of age-specific prevalence and excess mortality.
Intimate partner violence (IPV) is a worldwide epidemic. Physical and sexual violence are the most well-studied forms of IPV from an epidemiological point of view. Both of these have serious implications for other aspects of physical and mental health. In particular, women who have experienced IPV are more likely to be HIV-positive. In a modelling study we investigated some closely related questions about HIV and IPV in South Africa.
: In 1998, the US CDC announced the publication of a simple laboratory technique that would allow the differentiation of ‘recent’ from ‘longstanding’ HIV infections. Detection of recent infections with a laboratory test allows one to estimate HIV incidence using a cross-sectional survey. We believed the quest to be able to measure HIV incidence and monitor the epidemic in real time was over. How wrong we were! Eighteen years later we published a paper highlighting the incredible progress that has been made, but also outlined the many difficulties that remain.
The use of laboratory assays to identify recent infections among samples collected in cross-sectional surveys provides a potentially powerful way to measure HIV incidence. Ongoing evaluations of candidate laboratory assays have highlighted improvements and shortcomings of individual assays in correctly identifying and classifying people as recently infected. We tested two candidate assays, Sedia Limiting Antigen (LAg) and BioRad avidity assays (BioRad) against samples from a prospective cohort study in which incidence was measured directly. BioRad and LAg avidity assays have false recent rates that are up to 8 times lower than those for the BED assay, providing the potential for improved estimates of HIV incidence.
A new formal ‘R Package’ to support incidence estimation is available on the Comprehensive R Archive Network (CRAN). This is the canonical way that the R community distributes stable packages to share functionality, and it is the heart and soul of the R coding environment. The new release through CRAN will make a substantial range of functionalities around incidence survey design and survey data analysis seamlessly and flexibly available to any skilled R programmer/analyst.
The remarkable expansion in access to ART globally since 2004 has transformed HIV from a life-threatening into a chronic illness. Improved survival as a result of ART has starkly highlighted the lack of preparedness amongst health systems to deal with the complex needs of children living with HIV as they grow older and enter adolescence. While the drive to increase coverage to ART needs to continue, there is also an urgent need for policymakers and healthcare providers to focus beyond the goal of prolonging survival and to concentrate ensuring that adolescents living with HIV achieve an optimum quality of life.
The most recent South African National HIV Prevalence, Incidence and Behaviour Survey conducted in 2012 incorporated new tools for generating greater information about the current state of the HIV epidemic from the national household survey. The results from the direct HIV incidence estimates from the multi-assay recent infection algorithm from the survey and estimates from more conventional modelling approaches to estimating incidence were found to generate fairly similar results. The 2012 household survey data also provided an opportunity to externally validate model projections over time.
In populations where most subjects know their HIV status, population-based prevalence HIV estimates can be heavily biased due to high rates of non-response to HIV testing. Inverse probability weighting could potentially be used to correct for non-response to HIV testing in order to derive sub-national level HIV statistics, especially where the data at these levels are sparse. Its usefulness can be enhanced by incorporating antenatal clinics’ HIV data, often the only source of HIV prevalence.
UNAIDS has reported that the prevalence of people infected with HIV but who are not on ART, the incidence of HIV, and AIDS related mortality are falling. The Health Metrics Institute recently made their own, semi-independent, assessment of the trends in each of these indicators and reached similar conclusions with small differences arising from the use of somewhat different assumptions. Both analyses suggest that the world is on track to end AIDS by 2030, but this will depend on continued expansion of treatment at about the present rate together with supportive prevention efforts in Sub-Saharan Africa. Unfortunately, the data on which these analyses are based is weak in almost all places and better data on patient monitoring, follow-up and support, including drug procurement, supply and delivery, and better routing surveillance are needed.
The recent HSRC household survey reports that the HIV prevalence among adolescents and young people is declining. Although the decline is important, the focus needs to be on the fact that the reported HIV prevalence levels are still very high, together with alarmingly high levels of HIV incidence. Prevention methods have demonstrated effectiveness in reducing the risk of HIV acquisition among many of the most-at-risk populations. More research is needed, however, into how HIV is spread among the adolescent population and how to decrease this spread.