New HIV estimates shows too many people are still not on treatment
South Africa has made great strides in scaling up HIV testing and increasing HIV suppression in patients receiving antiretroviral treatment (ART), but is still struggling to reach ART coverage targets and is falling short of HIV prevention goals. This is according to new findings from the Thembisa model of HIV in South Africa presented today at the 9th South African AIDS Conference held in Durban.
Overall, the model estimates that in 2018 there were 7.39 million people in South Africa living with HIV – equivalent to 12.9% of the population”, said a press release from the University of Cape Town. Around 4.57 million of these were receiving ART.
Around 80 000 people are estimated to have died of AIDS in South Africa from mid-2017 to mid-2018.
Progress against 90-90-90
The model painted a mixed picture regarding South Africa’s progress against UNAIDS and National Strategic Plan (NSP) targets of diagnosing at least 90% of people with HIV, treating at least 90% of those diagnosed with HIV, and achieving suppression of the virus in 90% of those on treatment (the so-called 90-90-90 targets).
According to the model around 90.5% of people living with HIV in South Africa have received a positive HIV diagnosis, 68.4% of those diagnosed are receiving treatment, and 88.4% of those treated had viral suppression (according to the UNAIDS definition of <1000 RNA copies/ml – <400 RNA copies/ml is also used some times).
Of all people living with HIV in South Africa (including undiagnosed cases and people not on treatment), only around 54.7% are virally suppressed. This figure is important since people with viral suppression do not transmit HIV.
Still too many new infections
Although South Africa has been making progress in reducing the rate of new infections, the statement expressed concern that the progress may be too slow. “Over the period from mid-2017 to mid-2018, Thembisa estimates there were 249 000 new HIV infections in South Africa – a 36% reduction on the 388 000 new infections that occurred between mid-2010 and mid-2011. However, the UNAIDS target is to achieve a 75% reduction in the annual number of new infections between 2010 and 2020. To meet this target, South Africa would need to reduce its annual number of new infections to less than 100 000 within the next year, which appears unlikely given the slow pace of decline to date.”
Dr Leigh Johnson, lead developer of the Thembisa model, identified two factors that were hampering progress towards the HIV incidence reduction target: “Firstly, ART coverage is lower in South Africa than in other southern African countries, and we need to do better in linking HIV-diagnosed individuals to care and retaining them in care. Secondly, there is increasingly strong evidence of reductions in condom use, relative to the levels we would expect in the context of high levels of HIV diagnosis.”
The model estimates that adolescent girls and young women (ages 15-24) account for 31% of all sexually-acquired HIV infections. It is estimated that around 22% of women and girls in this group used condoms the last time they had sex. This percentage appears to have been falling steadily from a high of around 30% in 2005-2007.
Although the rate at which new infections is coming down is too slow, the statement points out some success stories. “KwaZulu-Natal, the province with the most severe HIV epidemic, has succeeded in reducing its total annual number of new infections by 49% over the period from 2010-11 to 2017-18. Thembisa also estimates that the annual number of new infections in children declined from 29 000 in 2010-11 to 13 000 in 2017-18, a reduction of 55%.”
Apart from a variety of other sources, the latest version of the Thembisa model also takes into account data from the 2016 Demographic and Health Survey. Work on the model is funded by UNAIDS and from 2017 Thembisa has been the source on which official UNAIDS estimates for South Africa are based.
Spotlight will in the coming weeks be publishing more in-depth analysis of the new Thembisa estimates.