Omicron and South Africa

 Michael - Thursday

And the question is...

A couple of days ago I asked a friend why the West (at least) didn’t take South Africa’s experience with omicron as relevant in the world context and downplayed it by commenting, for example, that “the South African experience is very different”. He thought for a few moments and suggested that it was because (most) of the people in South Africa are black. I don’t think that’s right (at least I certainly hope it isn’t), so I thought about it a bit more seriously.

So here are the facts of the South African context.

Omicron was actually discovered not in SA but in Botswana, and even then it was found in a group of conference delegates from Europe. I think almost everyone would accept that it’s irrelevant where it originated, and, in fact, it is most likely to have evolved in an area with low vaccination rates and broad covid spread. Although South Africa is well-off in the African context (28% fully vaccinated and moving to a booster vaccine for people who had their second shot six months ago), it is quite likely that it developed here. Although understandable, the knee-jerk response of some countries of closing their borders to southern Africa has been shown to be exactly what the scientists said at the time it would be. Pointless.

NICD SA data for 2020 and 2020

Omicron spread very quickly, much more quickly than the earlier forms of the disease. In just two weeks it climbed from a couple of thousand new cases per day to nearly 40,000 cases. It’s been computed that this may be an underestimate by a factor of 5. (More on that later.) So let’s say we went from 10,000 new cases to 200,000 new cases per day in fourteen days. From there, the numbers have declined over thirty days back to 10,000.

Deaths peaked at 140 per day. Of course, one doesn’t want anyone at all to die, and using the times 5 rule, that would be 700, but in the scheme of things that’s not a large number. It’s about 25% of the peak in the initial wave which was attacked with as strict a lockdown as anywhere in the world. (Stricter if you note the ban on the sale of tobacco products and alcohol.) Hospitals have not been strained by this wave - there are around 7,000 patients admitted with all variants of covid countrywide and very few of them are in ICU. The government has made no interventions beyond the requirement to wear masks in public, which has been in force all along. In fact, it relaxed the gathering rules for the Christmas period, and has now dropped mandatory self-isolation. And recall that less than 30% of the population here is vaccinated.

It seems obvious then that omicron is both much more contagious and much less dangerous than the other variants. Yet while the former was accepted with alacrity, little credence was given to the latter experience. So why?

Argument #1: It’s highly unlikely that a variant would develop like that.

Wrong. It’s an evolutionary direction that many common diseases have followed. They are common because they don’t kill their hosts. The hosts recover and so are available to catch yet another variant after a while. Recall that the “common cold” is a type of coronavirus.

Argument #2: South Africa has a much more youthful population than, say, the UK.

This is true, but anyone who knows a bit of statistics and certainly any scientist who studies population dynamics knows how to correct for that. If you do so, it doesn’t change the conclusions much.

Argument #3: South Africa’s numbers are totally unreliable so you can’t deduce much from them anyway.

The first part of this is correct because many rural people find it hard to be tested and so basically get sick and recover on their own (or die). Tests are not easy to get even in urban centers. However, deaths are recorded properly throughout the country. So at the end of 2020, it was possible to estimate the increase in the total deaths that would not have been predicted by population developments. A factor of around 5 was required to explain that i.e. in order to explain the unexpected increase in deaths in the data, one had to suppose five times as many deaths from covid as formally recorded. That suggested five times as many cases. Thus the numbers are not exact, but the trends and orders of the numbers can be clearly seen.

Argument #4: South Africans have a higher natural immunity due to the prevalence of AIDS and TB and the widespread vaccinations against the latter.

There may be something in this, but TB is a caused by a bacterium not a virus and early experiments with TB vaccines had no statistical effect. This seems similar to be the Invermectin argument.

So if we apply Occam’s Razor, it seems that the most reasonable explanation is indeed that omicron is less virulent. But why were spokespersons so keen to brush that under the carpet? For once I’m going to toy with a conspiracy theory although I don’t suggest it was a consciously planned one. I think what public health and government officials desperately wanted (and still want) is for everyone to take reasonable precautions (such as distancing and wearing masks), and above all to get vaccinated. If we assume those are the public health priorities, then it’s absolutely clear that the last thing you want people to believe is that the disease is becoming something like a flu – not worth worrying about. So the news from South Africa was actually quite unwelcome. And thus downplayed.

Please explain to me why I’m wrong!