How many SAs have had Covid-19? And why does no one know?

While the latest Covid-19 statistics will tell you that just over 1.4 million South Africans have tested positive, the real number is far higher. In September last year – when the official number was closer to 650,000 – BizNews spoke to Ryan Noach. The Discovery Health CEO and his team of experts estimated the number closer to 13 million. So, how many South African’s have actually had Covid-19? When the IRR’s Gabriel Crouse posed that question to two prominent medical professionals, both gave him the same answer – ‘that is the million-dollar question’. Below, Crouse explores the reality of Covid-19 in SA. This article was first published on the Daily Friend. – Jarryd Neves

The million-dollar question: How many have had Covid-19?

By Gabriel Crouse*

How many people have Covid in South Africa? How many have had it? What difference does it make and why does no one seem to know? These are a few of the questions that have been overlooked in fashionable conversation since the second surge, but answers are forthcoming.

How many people have or have had Covid-19 in South Africa? I put this question to Professor Willem Hanekom, director of the Africa Health Research Institute, and to Dr Marvin Hsiao of UCT, in separate interviews. Both gave the same answer. ‘That is the million-dollar question’.

‘As at August 24,’ Discovery Health estimated, as many as ‘15 million’ South Africans had been infected with and recovered from Covid-19.

In October, Professor Shabir Madhi, Dean of Wits Faculty of Health Services, was reported to believe that ‘the coronavirus had stimulated a level of immunity in approximately 12 to 15 million people’.

Emeritus Professor Jonny Myers, occupational and public health specialist at the Centre for Environmental and Occupational Health Research at UCT, argued that 60% of the Western Cape population may have been infected by mid-October, and that the rest of the country would not be far behind.

How Many Recovered to Date?

On 5 January, I reached out to Discovery Health to find their updated estimate. On 12 January, Emile Stipp of Discovery Health said the estimate had been reached and they were ‘triangulating this with lots of other data: testing rates, positivity rates, trends by age etc. I promise we will let you know as soon as we have a firm view…Hard to put a time frame on this, but it is a matter of days.’

On 18 January, Nthabiseng Chapeshamo, Senior Reputation Manager at Discovery Health, said ‘the data is finalised and undergoing internal reviews’.

By 26 January, the ‘internal reviews’ were still not over. We await Discovery Health’s update on the ‘million-dollar question’.

In the meantime, using Discovery Health’s spring 2020 formula, I made a calculation myself. Discovery assumed that 90% of excess deaths were caused by Covid-19, and that the Infection Fatality Ratio in South Africa is 0.3%. If both were true, that would mean roughly 35 million South Africans have been infected with SARS-CoV-2 to date.

This would roughly line up with Discovery’s earlier estimate. On 24 August there were 13 159 confirmed deaths, whereas on 24 January there were 40 874 confirmed deaths, an increase by a factor of 3.1 in confirmed deaths. Apply that factor to their estimate for infections of 12 million (their headline August figure) and you get 36 million today, or 15 million (their August upper bound) to 45 million infected today.

To be sure, Discovery Health’s formula leaves much to be desired. By my calculations (which will be published in a separate piece) anywhere between 25% and 75% of the population in South Africa could have been infected to date.

The million-dollar question gapes.

An Answer Will Come

The Daily Friend can report that a seroprevalence study is already under way in South Africa, meaning a study that tests random people for antibodies to see how many have recovered from Covid-19.

Professor Adrian Puren, acting executive director of the National Institute of Communicable Diseases (NICD) and head of the Centre for HIV and STI, explained as follows.

‘The HSRC [Human Sciences Research Council] is leading a consortium to test for prevalence of antibodies to SARS CoV-2’ which will be done in two ‘cross-sectional’ surveys one of which is ‘currently under way’, while the other ‘is planned for this year but the timing is yet to be decided’.

Furthermore, they will ‘recruit 19 000 participants for each survey’, which should provide statistically significant results.

To explain the significance of this, Prof Puren said: ‘It will provide critical information on the extent of exposure of the South Africa population to the virus, which age groups, genders, geographic spread. The data if sufficiently robust can be used in mathematical models for projecting needs/planning for future surges.’

Why does it matter? Herd protection

There are many reasons why the million-dollar question matters. Here is one.

‘Herd immunity’ can be a misleading term because it confuses two things. On the one hand there is a situation in which so many people are immune that the virus withers to extinction. On the other hand, there is the downward pressure on viral spread that strengthens every single time a person becomes less vulnerable to infection.

The first definition is binary. Either you have ‘herd immunity’ in that sense or you don’t. But the second is scalar; it comes in degrees.

Going by the ancient tradition of using different terms for different things I recommend calling the second scalar ‘herd protection’, which comes in degrees, not absolutes.

The point about herd protection is that it increases from the very first infection and recovery from SARS-CoV-2. It also occurs from the very first vaccination.

This is totally uncontroversial, and has been well known to scientists for about as long as epidemiology has been an institutionalised topic of analysis. Its implications are profound.

Perhaps most telling is the following scenario. If the average number of persons a Covid-19 carrier strikes with sufficient viral load to infect is 1.35, and recovery provides near-perfect protection, then viral spread will increase exponentially until about 25% of the population has been infected, at which point ‘herd protection’ will be strong enough to slow the virus down.

Those numbers are a pretty tight fit with South African data in early spring last year when the country opened up, but viral spread continued to slow down.

‘What has happened in SA today, the only way to explain it,’ Professor Mahdi said in mid-October, ‘the only plausible way to explain it, is that some sort of herd immunity has been reached when combined with the use of non-pharmaceutical interventions…’

Dr Hsiao at the end of October added to this theory, too, by saying, ‘when we analyzed the data it became clear, this immunity within the population (linked to) the big surge in infections is probably the main reason why we’ve seen the decrease of numbers of infected’.

The Sky News headline in which that interview appeared was Coronavirus: South Africa’s COVID lockdown may have created ‘herd immunity’.

If you listened closely you could hear the National Coronavirus Command Council shudder.

If 50% or 60% of the population has been infected to date then that fact is again almost certainly the predominant reason for flattening the second surge. But what happened in between?


Professor Darren Martin, a viral evolutionary biologist at UCT, told me that ‘perhaps the worst messaging failure of the entire pandemic’ is that scientists acted as if we needed to flatten the curve to buy time so that a vaccine could come along and save us all.

‘We needed to flatten the curve so that by the time the vaccine comes the virus hasn’t been able to spread around so much and evolve so much that the vaccine is significantly less effective, giving the virus yet another chance to mutate.’

The paper that identified the now famously evolved 501Y.V2 ‘South African strain’ of SARS-CoV-2 proposed two hypotheses for why South Africa discovered possibly the world’s most advanced form of SARS-CoV-2 mutations, beyond our exemplary genomic sequencing efforts. One is raised, only to be severely limited, to do with HIV. The only other hypothesis left standing is this:

Whilst we have yet to characterise how the mutations (particularly those in the RBM) affect antigenicity, it is plausible that high levels of population immunity could have driven the selection of [501Y.V2]. We have very limited SARS-CoV-2 seroprevalence data from South Africa to help understand the true extent of the epidemic. In studies using residual blood samples from routine public sector antenatal and HIV care, seroprevalence in parts of the City of Cape Town was estimated at approximately 40% in July-August, towards the end of the first epidemic wave in that area. We have shown that EC, and Nelson Mandela Bay in particular, were worse affected than City of Cape Town in the first wave, and therefore we believe that population immunity could have been sufficiently high in this region to contribute to population-level selection. Whilst there have been no confirmed re-infections (supported by whole genome sequencing) in South Africa, the true extent of re-infections is unknown and this is now the focus of urgent investigation. [emphasis added]

In plain English, the hypothesis is that SARS-CoV-2 spread so fast in South Africa that herd protection became a predominant factor, incentivising adaptations to dodge recovered immune systems, producing a second-generation viral mutant.

Since then, a preprint by Professor Constatinos Kurt Wibmer et al (including Professor Penny Moore) found no reinfections in test tubes for the Wuhan variant (more technically full neutralization of the virus) but 48% reinfections for the South African variant (no neutralization).

The evidence is starting to mount. The conclusion, not yet certain, is nevertheless dire.

South Africa’s lockdown, the world’s longest, arguably its most irrational, possibly its most murderous and economically screwed up, not only coincided with reaching the first tier of herd protection first, at 25% of the population infected in the spring or 2020, but for that very reason also drove the virus to evolve in ways that decreases herd protection both from recovery and vaccination.

If South Africa has been the petri dish for breeding the most toxic form of Covid-19 to spread globally then Pretoria will find good company with Beijing, but where else?

Moreover, if it took eight months for the virus to evolve to a ‘second generation’ with minor, but significant, ability to dodge herd protection it could plausibly evolve an even more dodgy ‘third generation’ mutant form by late winter.

Maybe this is less of a million-dollar question and more of a trillion-dollar question, because if we keep breeding mutant SARS-CoV-2, that is what we could cost the world. The impact of the HRSC-led stage-one 19 000-person seroprevalence study could produce the most politically damaging evidence of maladministration since the Guptaleaks.

Or it could to some extent vindicate the National Coronavirus Command Council, and embarrass Professors Madhi, Hsiao, Discovery Health and others by finding that only 20% of the population has been infected to date, which means that the springtime estimates were totally off.

The political stakes are high. Happily the HRSC and Prof Puren have serious reputations for seeking truth from facts. Real science will be badly needed on this; if the news is good we need to know, and if it is bad the whole world will need to know that, too.

Read also: