Man makes Covid predictions and God laughs. Yet with the stakes this high in Britain, every bit of real-world data is useful. That’s why South Africa is so important: it’s a country with a well-digitised healthcare sector that we have to thank for sequencing the Omicron variant, and has been first to experience the impact. That’s why its figures, released daily, are being watched so eagerly world over. Right now, there are two questions: is Omicron now falling? And if so, what conclusions can we draw?
The epicentre is Gauteng province: home to Johannesburg, Pretoria and about a quarter of South Africans. The below chart adjusts for population and shows that Omicron seems to have peaked in Gauteng last week — with the same trend visible in the other big provinces.
The caveat is that in South Africa, as with the UK, the latest data can be revised upwards. But it has been quite a few days now with figures below that of the peak in Gauteng. In another few days, we’ll be better able to say if this is true for Western Cape (home of Cape Town) and KwaZulu-Natal (home of Durban).
Now let’s zoom out to put Omicron in perspective: it has brought overall cases beyond its July peak. So there, as here, it sends cases through the roof.
But how does this translate into hospital care? We need to bear in mind the time lag between infection, severe illness and death. In previous waves, the average time for a Covid death has been four weeks after infection and hospitalisation is ten days. So even if Omicron cases have peaked, the below graph might go up before it goes down. But interestingly, there are signs of an early fall in hospital admissions. So we see hospitalisations at 57 per cent of the July peak, with ICU numbers barely a fifth of that peak and daily deaths (so far) just 8 per cent of what they once were.
What about lethality? The below looks at the death rate amongst those admitted for Covid, but broken down by age. This isn’t my maths: it’s a replica of a study by South Africa’s National Institute For Communicable Diseases (Figure 15 of this report). It’s for the area around Pretoria.
This 'fatality ratio’ is for people who’ve been admitted to hospital, so they’re already quite ill. The data suggests that the likelihood of patients dying from Omicron itself is less than half that of Delta — and in most cases, much less. Covid is often compared to flu which has a 0.1 per cent fatality ratio. We are pretty far from getting an equivalent figure for Omicron but flu is the obvious benchmark as it’s a risk we accept every winter, even though it kills around 25,000 annually in Britain.
Certainly if cases are falling in South Africa, Omicron’s death toll will not be anywhere near as bad as during previous waves. This would support suggestions from researchers that Omicron is more infectious — but less deadly — than the Delta variant.
So that’s the data. If you’d like to keep an eye on them, bookmark this page as the graphs will be updated daily, or visit the South Africa section of The Spectator's data hub here. There are caveats galore about the limitations of such studies and their relevance to Britain: South Africa is a younger, lower-vaxxed country going through summer right now. But if it does manage to beat Omicron without restrictions, that will be something to factor into the debate here. What real-world data we do have suggests there are limits to Omicron's growth.
Every day, The Spectator sends out an email summary of the Covid-19 news, science and data. Sign up for free here.