by Pieter Streicher
Wednesday, 29
December 2021

Omicron in South Africa: even milder than expected

Every number has come in under the projection
by Pieter Streicher

Two weeks ago, when panic about the Omicron variant of Covid was at its peak and it felt imminent that the UK would be going back into lockdown, I gave an interview to UnHerd’s Freddie Sayers. I am a PhD in engineering and a research associate at the IFK University of Johannesburg — I am not an epidemiologist — but I had been closely following the data of the Omicron wave that started in Gauteng province in my home country of South Africa.

On 4th of December I put together some basic projections for all the main Omicron variables — cases, hospital beds, ICU beds, ventilated beds, deaths — using simple growth rate extrapolations. Each variable goes through an exponential phase during which daily growth peaks, followed by a period with a consistent decline in growth rate. This consistent decline makes it possible to predict the date of the peak with reasonable accuracy. Once a likely date range for a variable has been determined, it becomes easier to gauge the likely peak levels of all variables. It was clear already by then that the Omicron wave was going to be extremely mild compared to the Delta wave, which my forecasts confirmed.

All the figures have come in under the projected values

What has been remarkable in the intervening weeks — despite the very clear data trends and statements of practitioners like Dr Angelique Coetzee, head of the South African Medical Association, that the variant was mild in all groups — is how slow other countries have been to accept this as fact. Sir Patrick Vallance and Chris Whitty, the UK’s chief scientific and medical officers, seem to be the last to admit what has been evident to most observers for almost a month: namely, that Omicron is much milder than Delta. With numbers in UK hospitals rising, a degree of caution is of course understandable, but the refusal to acknowledge the evidence from South Africa has been odd.

It is true that South Africa is a very different country to the UK, with a younger population and a different profile of prior immunity. But the population also has a major obesity problem, poor metabolic health, widespread HIV and had been hit relatively badly by previous waves of the virus. So the idea that ‘just because Omicron has not done much damage in South Africa it could still be devastating in the UK’ was always hard to believe.

In fact, the Omicron wave in Gauteng turned out to be even milder than my forecasts suggested; since the interview, every single one of the variables came in below projected levels. My estimate that deaths would be 25 times lower than the Delta wave still holds, and I have revised my estimate for the infection fatality rate down from 0.053% (11th December) to 0.036%.

As all variables seem to have peaked, we can also make estimates of the peak numbers per 100,000 — which should be useful for planning purposes in other regions and nations.

Join the discussion

  • The issue with the IC model as well as all the other models is not how well or badly the code is written, or whether the code is spaghetti code, but rather an understanding of what modeling is about under circumstances where there are many unknowns (both known and unknown). Ferguson et al. have fooled the politicians and public health authorities by making the simplest model of infectious disease spread as complicated as possible, thereby giving the appearance of increasing sophistication and accuracy (just as the climate modelers have done). Unfortunately nothing could be further than the truth.
    You don’t need a model to know that if you isolate everybody from one another, the number of cases will go down, but as long as immunity within the community is not established, then once the mitigation procedures are removed things will start up again just as they did from the get go.
    Take masks, for example. It was argued that these are a very effective method of source control based on idealized lab experiments. Let’s assume this is correct and let us assume that masks are close to 100% effective in this regard. Then clearly after 6 weeks or so where the infected at the time of mask introduction will have either survived or died, and one might therefore argue that the respiratory tract infection, in this case COVID, would all but disappear. But this argument has a major and fatal flaw because the uninfected population would still be naive to the virus, so that the moment strict mask mandates were lightened up, all it takes is just one infected person to start the whole thing back up again. And that’s precisely why masks, social distancing, and other mitigation procedures don’t work. All they can do is flatten the curve while leaving the area under the curve unchanged. And the only escape is either through natural immunity, sterilizing (or close to sterilizing) vaccination (which we clearly don’t have as the current crop of vaccines have clearly failed), or effective early at-home treatments.
    In other words, fighting VOCID is a bit like fighting a war. If one is going to get into a war, one has to have an end-game (and as recent history has shown the US has failed to have an end game in their middle east adventurers, and hence each one has ended up in disaster and left the place worse off than when they came). The end-game with COVID that Fauci et al envisaged was vaccination. Clearly if one can make use of mitigation measures to hold things down for a sufficient length of time to enable an effective vaccine to be developed, the mitigation measures would have been worthwhile. But this requires an effective vaccine to be developed, not one that lasts at best for 6 months after 2 shots, and only provides limited protection for 10 weeks after a 3rd shot. Further, it is not evident that the mandated mitigation procedures introduced, for example, in the US and UK, were any more effective than the completely voluntary changes in behavior that took place in Sweden.
    Fortunately for us, it would seem that the virus will have done the work of the public health authorities for them by mutating to a less virulent, albeit more transmissible (supposedly) variant. That’s what happened with the Spanish Flu, and it’s clear that Omicron represents just such a development with COVID. And if that is indeed the case, as it certainly looks as if it is from not just the South African data but also the Danish experience, then we should all be rejoicing the advent of Omicron as a great Christmas and New Year’s present to all.

  • “….but the refusal to acknowledge the evidence from South Africa has been odd.”
    The takeaway from this piece is that the sky is not falling, the virus is mutating–that’s what viruses do– but the variations are becoming less virulent, and humans are learning to live with it. That there is a “refusal to acknowledge the evidence” is not odd, but de rigeur. Whatever happened to follow the science?
    Because of the measured tone backed up by (seemingly) reliable data, this piece should be banned as the wacko propaganda of the extreme right. It might people less afraid, and the boffins cannot have that. FaceBook–please take this down!
    Taken to its logical conclusion, it would seem that somewhat normal life can be resumed, which is not what the boffins–Neal Ferguson, Jacinda Ardern, Fauci, Sleepy Joe Biden and their ilk–want to hear. If there is no emergency, then it follows that there can be no emergency measures. Freedom, not an inherent right but something granted by the leave of our betters, might be returned to the people. This is something that the nanny state will fight to their last breathe….
    This piece shows what an absolute clown show the “scientific” advisors have run, and the BBC is one of the worst offenders. Their “scientific” experts have touted “vaccine equity” at every turn, virtually demanding that we in the West not have second shots, not have booster shots until the corrupt and feckless Third World attains “equity.” THE “EXPERTS” still push this argument to an extent, but since it has failed to gain much traction on the ground, speak wistfully that this has not happened. Simply disgusting!

  • Agreed.
    It’s not my speciality, but the company I work for develop material and stress analysis software and it is highly useful in a range of engineering applications. As you say, the ability to feed results back into the system from testing is how they become more accurate.
    The IC model was apparently designed from the Spanish flu (which is questionable in itself due to the lack of genuinely good data from that period) and just given a few different variables for the viruses it’s been used to model in recent years. My suspicion is that viruses have a lot more variability between strains than say different types of metals.

  • To get involved in the discussion and stay up to date, become a registered user.

    It's simple, quick and free.

    Sign me up