Graphs that tell the story of HIV in South Africa’s provinces

By Marcus Low and Sean MacDonell

The most recent outputs of the Thembisa mathematical model (version 4.2) of HIV in South Africa not only allow us to see the burden of HIV at the country level, but also allow for provincial comparisons. Through several tables, graphs, and maps, we illustrate how the HIV epidemic differs in each of the country’s nine provinces.

1. Which provinces have the most people living with HIV?

The table below shows which provinces have the largest number of people living with HIV. The ongoing increases in these numbers are both positive and negative. On the negative side, it reflects the fact that the rate of new infections remains high; on the positive side, it reflects the reality that people with HIV are not dying at the rates they used to since antiretroviral therapy is keeping millions of people alive. 

KwaZulu-Natal is clearly at the centre of South Africa’s HIV epidemic. The province has more people living with HIV (over two million) than North West, Limpopo, Western Cape, Free State, and Northern Cape combined. In interpreting these numbers one should of course keep in mind that KwaZulu-Natal is South Africa’s second most populous province behind Gauteng – which partly explains the high absolute numbers, but not entirely.

2. What percentage of the population is living with HIV in each province?

To contextualise the number of people living with HIV in a province in terms of the size of provincial populations we can look at the percentage of a province’s population who are HIV positive (also known as prevalence). In the same table above we can see that KwaZulu-Natal still tops the list with an incredible 18.23% of the population living with HIV.  While Free State ranked eighth for the absolute number of people living with HIV in the province, it ranks third in terms of prevalence at 14.62%. Despite the large absolute number of people living with HIV in Gauteng (nearly two million), the province ranks only fifth in terms of prevalence at 13.05%. Here we can see how prevalence has changed over time in South Africa’s four most populous provinces: Eastern Cape (EC), Gauteng (GT), KwaZulu-Natal (KZ) and Western Cape (WC). While the prevalence of HIV continues to rise in the Eastern Cape, KwaZulu-Natal, and Western Cape, we can see that it has now begun to fall (albeit slightly) in Gauteng. Prevalence continues to rise in most provinces for the same reasons that the absolute number of people living with HIV is rising, although prevalence is also impacted by changes in the province’s population.We can also conceptualize prevalence using the map above. One interesting trend is that provinces in the east of the country tend to have higher HIV prevalence than provinces in the west. (See the column on prevalence in the table for exact numbers.)

3. How many new cases of HIV are there each year in each province?The graph above shows the number of people newly infected with HIV each year in the four most populous provinces. Here we can see that KwaZulu-Natal has dramatically reduced the number of new HIV infections, now having fewer people newly infected with HIV than in Gauteng. While the number of new HIV infections continues to come down in the Western Cape, a much flatter line suggests a slower rate of decline than in the other three most populous provinces. 

4. The 90-90-90 targets

The 90–90–90 targets are a set of global goals established by the United Nations Programme on AIDS and HIV. By 2020, the goal is that “90% of people living with HIV will know their HIV status, 90% of those who know their HIV-positive status will be accessing treatment, and 90% of people on treatment will have suppressed viral loads.”

These targets provide a good measure of how well different provinces are performing in key areas such as the provision and promotion of HIV testing and helping people who test positive to start treatment and to stay on treatment.

The table above displays 90-90-90 targets for each province. This allows us to examine how each province is either achieving or falling behind each target of 90%. The fourth column is the product of the three target percentages; in other words, it is the percentage of all people living with HIV in the province who are both accessing antiretroviral treatment and have suppressed viral loads. As this fourth column represents 90% of 90% of 90%, a province achieving all three targets would have at least 72.9% in the fourth column. In other words, at least 72.9% of people living with HIV in a province would be accessing ART and would have viral loads below 1000 if they are meeting all targets. 

In the table above we can also see that the Northern Cape is closest to achieving this target percentage as 63.66% of people living with HIV are both accessing ART and have suppressed viral loads.

It is clear from the table that the greatest barrier to achieving the 90-90-90 targets is ensuring that people with diagnosed HIV are taking antiretroviral treatment. Only in the Northern Cape is the percentage of people with diagnosed HIV who are taking treatment over 80%. While North West comes close to achieving the first and third targets, only 57% of people with diagnosed HIV in that province are on treatment.

The dynamics between the three 90s can be interesting. For example, when a province suddenly initiates a lot more people onto treatment, we may at first see a reduction in the percentage of those with suppressed viral loads since many of the new patients will still have unsuppressed viral loads. Either way, ensuring that those who know their status receive the treatment and care they need is a priority in order for South Africa to achieve the 90-90-90 targets. The graph above shows the percentages of people with HIV who are both on treatment and virally suppressed for the four most populous provinces from the year 2001 until 2018. KwaZulu-Natal and Gauteng have consistently high rates, while there has been a slight drop-off in the Western Cape and Eastern Cape during the last few years.

Note: The graphs in this article were produced using RStudio and the ggplot2 package. Graphs are exclusively based on publicly available Thembisa model outputs. Spotlight takes full responsibility for any errors or misrepresentations there may be in the graphs.