Is South Africa on track to meet NSP targets?

On World AIDS Day 2018 we assess how South Africa is faring against 10 key targets set in the National Strategic Plan (NSP) on HIV, TB and STIs 2017 – 2022. The current picture is mixed, with areas of impressive progress, such as HIV testing,  offset by some serious red flags, such as retention in care.

Of the 10 targets, we have assessed two as “Appears target will be reached”, two as “Target is within reach”, four as “Reaching the target will be difficult”, and two as “no sufficiently reliable figures”.

In future articles we will assess progress against other important issues in the NSP not covered here such as stigma, availability of data, and accountability in the public healthcare system. Accountability for the development of implementation plans and actual implementation is an ongoing problem with the NSP – although there are now signs that it may be addressed. The South African National AIDS Council (SANAC) reports that it is currently “developing an NSP Accountability framework and scorecard for accountability to achieve the goals of the NSP 2017-2022 for HIV, TB and STIs”.

“The accountability framework will assist in ensuring commitment to set priorities by individual role players in the current and subsequent NSPs,” SANAC explained in response to questions from Spotlight. “The Framework will nationally determine who should be accountable, for what, why and how. The development of the accountability framework will require extensive consultations to ensure ownership in the national response.”

SANAC also indicated that its mid-term review of the NSP will be published by August 2019. The mid-term review is expected to include up-to-date estimates for all the indicators covered in this article.

1.Reduce new HIV infections to less than 100 000 by 2022 (or 88 000 by 2020)

According to the Thembisa model (a sophisticated mathematical model of HIV in South Africa) there were around 275 000 new HIV infections in South Africa in 2017. This number is higher than the 231 000 estimated in the South African National HIV Prevalence, Incidence, Behaviour and Communication Survey 2017 (the HSRC survey).

Either way, given the current trajectory it seems highly unlikely that South Africa will get new infections down to below 100 000 by 2022. The Thembisa model projects that new infections will only drop to around 198 000  by 2022 (with the low end of the 95% confidence interval around this estimate at 184 000).

According to SANAC, the target for New infections has now changed to 88 000 by 2020, an even more ambitious, and arguably more unreachable, target.  “The change/adjustment has been made in line with  the global UN Prevention road map to accelerate HIV prevention strategies and reach national and global goals to end the threat of AIDS by 2030,” SANAC explained in a response to Spotlight. “The NSP Steering Committee agreed that the change in the targets according to the Prevention Revolution will be incorporated into the NSP M&E Plan (rather than in the NSP programme narrative).  The NSP M&E Plan will be available once approved by all SANAC structures.”

Verdict: Reaching the target will be difficult

2. Reduce annual new HIV infections in women and girls aged 15 to 24 to 30 000 by 2022

The targets section of the NSP lists the 30 000 target, while earlier in the NSP reference is made to bringing the infection rate in young women and girls down to less than 800 new infections per week (which works out to around 41 000 per year). Either way, it is widely agreed that reducing new infections in young women and girls is of critical importance.

According to the Thembisa model there was around 81 000 new infections in women and girls aged 15 to 24 in 2017 (this works out to just over 1 500 per week, rather than the more widely quoted 2 000 per week.). Thembisa projects annual infections in this group only dropping to around 58 000 by 2022 – almost double the target.

Verdict: Reaching the target will be difficult

3. The rate of mother-to-child transmission of HIV at 18 months must be below 2% by 2022

According to the Thembisa model the mother-to-child transmission (MTCT) rate was 4.81% in 2017. The model projects this dropping to around 3.5% by 2022. This number however includes transmission that occurs after 18 months, and the 18 month number might thus be very slightly lower. While there are other sources for MTCT rates, we prefer quoting the Thembisa estimates since the model factors in MTCT cases where the mother is not aware of her status – likely the majority of MTCT cases.

For the MTCT rate to come down further will require that a higher percentage of HIV-positive women of child-bearing age take antiretroviral therapy and achieve viral suppression. It will also require that we help more breastfeeding women who are not living with HIV to stay HIV negative – since women who become HIV-positive in the months they are breastfeeding account for a large proportion of onward transmission to children.

Verdict: Reaching the target will be difficult

Implement the 90-90-90 strategy for HIV

The 90-90-90 targets first proposed by UNAIDS were adopted in South Africa’s current NSP. There was previously some uncertainty regarding whether the three targets (described below) are to be met by 2020 (as per UNAIDS) or by 2022 (the end date of the NSP). Our understanding is that the agreed target date for South Africa is 2020.

4. 90% of all people living with HIV know their HIV status

The Thembisa model estimates that South Africa has already reached the first 90, with an estimated 90% of people living with HIV knowing their status in 2017. By contrast, the HSRC survey suggests that South Africa is just below the target, with around 85% of people aged 15-64 in South Africa who are living with HIV knowing their HIV status.

While South Africa appears to have done well on the first 90, maintaining a high score on the first 90 will require sustained testing efforts given that over 200 000 people are newly infected every year (and obviously do not know their status until testing). Thembisa projects that South Africa will reach around 92.4% by 2020.

Verdict: Appears target will be reached

5. 90% of all people with diagnosed HIV infection receive sustained antiretroviral therapy

According to the Thembisa model, South Africa is doing much poorer on the second 90, with only around 61.9% of people diagnosed with HIV receiving antiretroviral therapy in 2017. The HSRC survey paints a slightly rosier picture, estimating that around 71% of people aged 15-64 with diagnosed HIV infection are receiving sustained antiretroviral therapy. According to the Thembisa model projections South Africa will reach only around 70.4% by 2020.

South Africa’s poor performance on the second 90 is arguably the biggest red flag in the set of indicators examined in this article.

Verdict: Reaching the target will be difficult

6. 90% of all people receiving antiretroviral therapy are virally suppressed

In terms of the third 90, the Thembisa model estimates that around 77.8% of HIV positive people who were receiving treatment in 2017 were virally suppressed (the virus was so successfully suppressed in their bodies that it could not be detected with standard tests). Again the HSRC survey paints a more positive picture than Thembisa, estimating that 86% of people aged 15-64 receiving antiretroviral therapy have viral suppression.

Even though these figures are relatively high and may well go up with the introduction of dolutegravir-based antiretroviral therapy in South Africa, it is by no means certain that it will rise to 90 in the coming years and then be maintained above 90 (Thembisa projects around 86% by 2020).

Verdict: Target is within reach

A related indicator that does not directly form one of the three 90s, is the percentage of all people living with HIV who are virally suppressed (not just those on treatment as per the third 90). The Thembisa model estimates that 43.3% of all people living with HIV in South Africa were virally suppressed in 2017 (the implicit target based on the 90-90-90 target is 73%). Compelling scientific evidence shows that people who are virally suppressed do not transmit HIV.

Implement the 90-90-90 strategy for TB

As with HIV, TB also has a set of 90-90-90 targets that have been integrated into the NSP. Given that TB figures are harder to find than HIV figures, Spotlight twice wrote to the Department of Health to request the most recent estimates they have for these indicators. While the department did acknowledge our request, we did not receive any figures by the time of publishing. SANAC also failed to provide Spotlight with estimates regarding the 90-90-90 targets for TB.

Below we use the targets as they are presented in the final NSP. It is notable however that the 90-90-90 targets for TB has been taken to mean different things in different contexts. A recent Health Systems Trust publication describes the targets as “90% of people with TB will be screened, 90% will be initiated on  treatment, and 90% will successfully complete treatment” – which is not the same as the targets from the NSP used below. Where we found targets unclear we have attempted to clarify them with reference to the WHO’s version of the 90-90-90 targets for TB.

7. Find 90% of all TB cases and place them on appropriate treatment

The WHO’s version of this target is formulated as follows: “Reach at least 90% of all people with TB and place   all of them on appropriate therapy: first-line, second-line and preventive therapy, as required.”

In our understanding the indicator that most closely tracks whether we are finding people with TB is the case detection rate. According to figures available from the WHO’s TB data portal South Africa’s case detection rate was 68% in 2017. But the WHO also has the case detection rate at 68% for each of the last six years – something which does not inspire much confidence in this number. Neither the Department of Health or SANAC provided figures for this indicator despite various requests from Spotlight.

Recent testing campaigns in mining and other high-burden areas have been an important step in the right direction. Indications are that we are most likely still failing to diagnose many thousands of TB cases – one recent study found that most people with active TB who attend clinics in the Eastern Cape were not properly screened and tested. In addition to improving the quality of screening that should be happening at primary healthcare facilities, rapid and widespread implementation of the increased contact tracing and active-case-finding efforts described in the NSP will be critical if this target is to be met.

Verdict: No sufficiently reliable figures

8. Treat at least 90% of those diagnosed with DS TB and 75% of those with DR TB.

We understand “treat” here to refer to successful treatment as per the WHO version of this target. The WHO states it as follows: “Achieve at least 90% treatment success for all   people diagnosed with TB through affordable   treatment services, adherence to complete and   correct treatment, and social support.”

According to WHO figures around 82% of people with drug-sensitive TB and 55% of people with drug-resistant (MDR or Rifampicin Resistant) TB in South Africa were successfully treated in 2017. Neither the Department of Health or SANAC provided figures for this indicator despite various requests from Spotlight.

While reaching the 90% treatment success target for DS TB is within reach, reaching the target cannot be taken for granted given the many challenges facing the public healthcare system in South Africa. The introduction of better and less toxic treatments for drug-resistant forms of TB should help push the DR TB treatment success numbers up in the coming years.

Verdict: Target is within reach

9. Find at least 90% of the TB cases in key populations (the most vulnerable including people living with HIV with low CD4 counts, under-served, at-risk) and place them on appropriate treatment. Successfully

We have not been able to find reliable figures for this indicator. Neither the Department of Health or SANAC provided figures for this indicator despite various requests from Spotlight.

Verdict: No sufficiently reliable figures

10. Reduce TB incidence by at least 30%, from 834/100,000 population in 2015 to less than 584/100,000 by 2022

According to the 2018 World Health Organization (WHO) World TB Report, TB incidence in South Africa was 567 per 100 000 in 2017. At first glance it thus seems that we have already achieved the NSP target of less than 584 per 100 000. The reality is however more complicated. In 2018 the WHO made major adjustments to its TB estimates for South Africa. This included recalculating the estimate for 2015 as 759 per 100 000. A 30% reduction of this figure would set a target for 2022 of 531 per 100 000. The estimate of 567 per 100 000 is still above this level, but by 2022 South Africa will be well below this adjusted target if the downward trend suggested by the WHO’s figures  continues.

That said, there is significant uncertainty regarding TB incidence in South Africa and the picture might well change substantially as new information becomes available in the coming years. While the 567 per 100 000 figure may be the figure most often quoted, the WHO estimates that the real figure could be anywhere between 406 and 754 per 100 000.

Verdict: Appears target will be reached


  • Regarding HIV, the good news is that South Africa is doing well when it comes to HIV testing and where people living with HIV are on treatment the treatment is generally working and saving lives.
  • The bad news is that many people who are living with HIV are not on treatment – either never having started or having quit. Arguably the biggest challenge facing the public healthcare system today is to support many more people living with HIV to start and stay on treatment. Doing this in an often dysfunctional healthcare system will not be easy.
  • Regarding TB, the good news this year is that South Africa’s TB epidemic appears not to be quite as large as previously thought. TB rates appear to be coming down slowly, although it is hard to say with any certainty exactly how fast – chances are not fast enough.
  • The bad news is that even in 2018 many people in South Africa with TB do not get diagnosed. Again the available figures cannot be trusted, but even if the real figure is only half of the estimated 90 000 to 100 000, we are still facing a very serious situation.
  • Either way, it is clear that in addition to gathering better data on TB, as is being done with South Africa’s first TB prevalence survey, South Africa also needs better epidemiological models of TB along the lines of what we have for HIV.

Leave a Reply

Your email address will not be published. Required fields are marked *