Graphs that paint the picture of HIV in SA: Part 2

News24

19 hours ago

Eight million people living with HIV. Just over 6 million on treatment.
Behind these big numbers lurk a universe of fascinating epidemiological dynamics.
In this special briefing, Spotlight editor Marcus Low unpacks what we know about the state of HIV in South Africa. This is part 2 of 3.
In Part 1 of this Spotlight special briefing, we looked at some of the big picture dynamics of HIV in South Africa. Most of the news there was good. In Part 2, however, we look at why making progress is getting harder and consider some key vulnerabilities of our HIV programme.
The final 10% or 20% will be the hardest
Spotlight
The above graph shows the number of people starting treatment by year. As you can see, around a decade ago, this number was often higher than 500 000. Since 2021, it has been below 300 000.
Government could no doubt do a better job at helping people start and stay on treatment, but there are also good epidemiological reasons for the slow-down.
In short, those who were sicker or more eager to start treatment - and for whom taking treatment is relatively easy - are mostly already on treatment.
Those who are not on treatment are likely less motivated to take treatment or may have challenges in their lives that make it hard for them to start and stay on treatment.
Convincing and supporting this latter group to start and stay on treatment is quite simply harder than it is for the prior group, but it is where the focus will have to be if we are to get our numbers higher.
It is one reason why many are so concerned about recent aid cuts that have disproportionately impacted HIV services for marginalised groups like sex workers, people who use drugs, and men who have sex with men.
These groups of people often feel unwelcome at public sector clinics.
One key thing to keep in mind when looking at HIV numbers is that people cycle in and out of HIV treatment a lot.
A misleading impression can, for example, be created if one looks at the number of people starting or restarting treatment without subtracting the number of people who have stopped taking treatment over the same period. It is for this reason that it is best to focus mainly on the total number of people on treatment and how that number changes from year to year (you can see a graph of this in Part 1 of this series).
Finally, another important nuance in the numbers is how much damage has been done to people's immune systems by the time they start treatment – the earlier you start, the better your long-term prognosis. It is concerning that, of the 241 000 people who started treatment in 2023/2024, 54 000 had CD4 counts below 200 cells per cubic millimetre, an indication that their immune systems had already been hit hard.
Bigger programme, bigger vulnerabilities?
Much of how an epidemic grows or wanes depends on how many people are infectious at a specific point in time. Generally, the more infectious people there are in a population, the more other people will get infected. It follows that, if you can't cure people, the next best thing from an epidemiological control perspective, is to make people non-infectious. (See our recent Special Briefing on the search for an HIV cure.)
Fortunately, apart from keeping people healthy, consistently taking antiretroviral medicines also make the vast majority of people living with HIV non-infectious. Thus, while all eight million people living with HIV in South Africa could potentially become infectious, antiretroviral treatment means that only around two million of them actually are.
Spotlight
It is very unlikely that we will drop from 6.2 million people on treatment to 5.2, or even 5.5, but even just dropping by 100 000 or 200 000, will mean 100 000 or 200 000 more people become infectious. That would give our HIV epidemic a blast of oxygen it wouldn't have had otherwise.
There are real vulnerabilities here. According to preliminary modelling that the Department of Health commissioned in response to US funding cuts, failure to replace the HIV services that were funded by the US in South Africa could result in an additional 150 000 to 295 000 additional infections by 2028. It will take a year or two before we see the actual impact of the aid cuts in data sources such as Thembisa, but, as reported by Daily Maverick, there have been some early indications from the country's public lab service that important numbers are down.
*Check back tomorrow for Part 3 of this series. You can also find the complete version of this #InTheSpotlight special briefing as a single page on the Spotlight website.
Note: All of the above graphs are based on outputs from version 4.8 of the Thembisa model published in March 2025. We thank the Thembisa team for sharing their outputs so freely. Graphs were produced by Spotlight using the R package ggplot2. You are free to reuse and republish the graphs. For ease of use, you can download them as a Microsoft PowerPoint slide deck.
Technical note: The Thembisa model outputs include both stock and flow variables. This is why we have at some places written 2024 (for stock variables) and 2023/2024 (for flow variables). 2024 should be read as mid-2024. 2023/2024 should be read as the period from mid-2023 to mid-2024.
Show Comments ()