Post #1380: Omicron, I sure got it wrong.

Posted on January 2, 2022


I expected the U.S. Omicron wave to be short, sharp, and with very low average case severity.  That’s what occurred in South Africa, and that’s what I expected to see here.

If we’d repeated the South African experience, we’d have peaked by now. 

Instead, the Omicron wave in America is continuing longer, and moving  higher, than it did in South Africa.  And average case severity for new cases, relative to Delta, is higher here than it was in South Africa.  As a result, what I thought was going to be a fairly benign wave of COVID in the U.S. is starting to show some potential for turning into a true disaster.

In short, the South African experience was not a good model for what’s happening in the U.S., and the U.K., and some other European countries.

What are some possible reasons for that?  Turning that around, what are some major differences between South Africa and the U.S. that might have caused the Omicron wave to have differed?

Delta never left us.

Let me start with the most obvious contrast between South Africa and the U.S./U.K./Europe.

South Africa had a “pure” Omicron wave, in the sense that there were almost no pre-existing Delta cases circulating in the country.  All you ever saw there, for their short, sharp wave, was Omicron.  As you can see, the daily new case count was practically zero prior to the start of their Omicron wave.

In the U.S., by contrast, we never really finished our Delta wave.  You can see a significant case count of Delta already occurring by the start of the U.S. Omicron wave.  The U.S. — and most of Europe — was already in the middle of a mild (U.S.) to quite severe winter (U.K.) wave of Delta, when Omicron came along.


We know that Omicron is displacing Delta as a fraction of all cases Here’s the most recent CDC estimate:

Source:  CDC COVID data tracker accessed 1/20/2022.

(This is something that I’ve puzzled over before.   Somehow, each new strain manages to kill off the prior strains.  Alpha displaced the native (Wuhan) strain, Delta displaced Alpha, and in each case, more or less 100% of new cases end up being the new strain.  The older strain disappears.  But the mechanism behind that has never been clear to me.)

So, Omicron has been displacing Delta.  Or has it?  Delta has been falling as percent of cases.  But what has it been doing in terms of the actual number of cases?

Let’s convert the CDC’s Omicron-as-a-percent-of-new-cases numbers above to actual counts of Omicron and Delta cases.  That’s easy enough to do — just multiply the total new case count by those variant percentages. (There are a few fine points of method here — I filled in the daily percentages using the week-to-week growth rates above, I slid the whole assembly of daily percentages back by three days prior to the listed end-of-week dates, and and so on.  Those were all obvious things to do if you’d thought about it for five minutes.)

And despite how obvious it was to do this, I got a pretty big surprise.  Through Christmas 2021, Omicron did not displace Delta, it mostly added on top of Delta.

Source:  Calculated from case counts as variant estimates from the CDC COVID data tracker.

Above, the orange line is my estimate for the count of new Delta cases.  As you can see, that hadn’t really budged, as of Christmas 2021.  Omicron accounted for more than half of new U.S. cases by Christmas.  But it did that by growing on top of a fairly stable population of daily new Delta cases.

As simple and obvious as this now appears, I haven’t seen this point made elsewhere yet.  So I wonder if this is a new phenomenon.  Maybe this is some unique consequence of Omicron being so vastly more contagious than Delta.  As some point, if I can dig up the data, I may want to go back and see what happened at the Wuhan/Alpha and Alpha/Delta transitions.

The upshot is that one major difference between South Africa and the U.S. is the Delta wave.  South Africa didn’t have one.  All they had to deal with was Omicron.  But as of Christmas 2021, in the U.S., that winter Delta wave has not gone away.  It’s still here, hiding beneath the Omicron wave.  If anything, it appears to be fading slowly.

Difference sources of COVID-19 immunity

Another glaring difference between South Africa and the U.S./European experience is the mix of sources of COVID-19 immunity within the population.  Roughly speaking, most of our immunity comes from vaccination, most of theirs comes from prior infection.  Approximately like this:

Sources of information:

U.S. percent with prior infection is the count of positive cases as of 12/15/2021, times 1.9 (based on the most recent seroprevalence data from CDC, to account for cases not formally diagnosed), divided by total U.S. population.  All the COVID data are from the CDC COVID data tracker.

U.S. percent fully immunized and boostered is from the CDC COVID data tracker as of 2/1/2022.

South African percent fully immunized is from Our World in Data, rounded down to 25%. 

South African percent with prior infection is from this analysis from this  undated PowerPoint from the South African National Institute for Communicable Diseases.

To be clear, Omicron appears to be able to infect both the fully-vaccinated and prior-infection populations easily.  That is, based on the South African experience, there were a lot of breakthrough infections (among fully-vaccinated) and reinfections (among those recovered from prior infection).

If there were some difference in protective ability, then that might influence the course of the Omicron pandemic in these countries.  But the bottom line is that there is no hard evidence one way or the other, on this point.  Near as I can tell, all the evidence says that neither prior infection nor full (two-shot) vaccination provides significant protection against Omicron.  Which one is the lesser of two zeroes cannot be determined from existing data.

In theory, this ought to be roughly knowable from population-based administrative data available in the U.S.  In U.S. states, on a person-by-person basis, state health departments know who has tested positive for COVID-19 in the past, and who has been vaccinated (partially, fully, or boostered).  From that, they can (and some do) flag cases that are breakthrough (in a fully-vaccinated person) and reinfection (in a person with prior positive test for COVID-19).  If there is some new, significant different in immunity across those populations, then the ratio of identified reinfections to identified breakthrough infections should change as the Omicron fraction of all infections rises.

There is no doubt that a state health department could do this analysis.  That said, as an outside, I have yet to find a state health department that has published the relevant, current data, let alone done the analysis.

Instead, I have to rely on various scholarly studies.  Near as I can tell, these boil down to:

  • Neither prior infection nor standard (two-dose) vaccination provides much protection against Omicron, if any.
  • Studies disagree whether or not prior infection or vaccination provides better immunity.
  • There’s enough uncertainty around the estimates that it’s probably not possible to answer this question from existing research data.

The first study comparing natural immunity and vaccine immunity against Omicron is a study of infection rates of persons in Great Britain, using data from the end of November into early December.  Based on their sample sizes, they are unable to rule out zero protection from Omicron, from either source of immunity.  Omicron.  As reported at this link, based on the original research as reported at this link:

The reinfection risk estimated in the current study suggests this protection has fallen to 19% (95%CI: 0-27%) against an Omicron infection.
vaccine effectiveness estimates against symptomatic Omicron infection of between 0% and 20% after two doses, and between 55% and 80% after a booster dose.

A different study, using in vitro assays of immune response, again found little response from either prior infection or full vaccination, but suggested that the prior infection response was, if anything less than that of antibodies from fully-immunized individuals.  That’s reported at this link, with the original research at this link.

Antibodies from people double-vaccinated with any of the four most widely used vaccines—Moderna, Pfizer, AstraZeneca, Johnson & Johnson—were significantly less effective at neutralizing the omicron variant compared to the ancestral virus. Antibodies from previously infected individuals were even less likely to neutralize omicron.

I’m going to give up on finding anything more definitive for now.  Near as I can tell, neither prior infection or full vaccination provides much protection against Omicron, if any.  It seems unlikely that any small difference in effectiveness between these two sources of immunity would be able create such a large difference between the U.S. and South African experience.

COVID hygiene.

The research above suggests that neither full vaccination nor prior infection provides any significant protection against Omicron.  The logical implication is that for the roughly 70 percent of the U.S. population that has not received a booster dose, the only thing that slows spread of Omicron is COVID-19 hygiene:  Wearing a high-quality mask, staying out of high-transmission-risk situations, and the like.

Restated:  If most of the population has virtually no immunity to Omicron, then differences in COVID hygiene can play a significant role in determining variations in Omicron spread across nations.

For the U.S., as I have shown repeatedly here, mask use has hardly changed over the past month, based on data from Carnegie-Mellon University.

Source:  Carnegie-Mellon COVIDcast.

It’s tough to find hard data on the extent to which South Africans increased use of masks.

News reporting makes it clear that they take mask mandates seriously.  (Thousands arrested for not wearing masks, dated Feb 2021).  I don’t believe I’ve ever heard of an American arrested or even fined for failing to wear a mask.  And, while South Africa is now lifting other restrictions, failure to wear a mask in public remains a criminal offense.

Historically, rates of mask use in South Africa appear to have been comparable to those in the U.S., based on self-reported survey data from the end of 2020.  That’s reported at this link.

I can’t seem to find objective data.  But the presence of a national law that makes it a criminal act to be outside the home without a mask — actually enforced with arrests and penalties — suggests a far more serious attitude in South Africa than in the U.S.

So this remains a guess, but I’m guessing that South Africans take mask wearing more seriously.  Certainly compared to a country with no national mask mandate, where state mandates seem to be completely unenforced by law officers, and where many states have chosen to prevent localities from passing their own mask mandates.


There’s no simple answer as to why South Africa’s experience with Omicron has been so vastly better than our own.

In part, we were already burdened with Delta, and we continue to be burdened with that.  But that, by itself, wouldn’t explain prolonged growth of Omicron here, compared to South Africa.  More total cases, perhaps, but not more growth.

In part, prior pandemic waves seem to have been somewhat shorter, consistent with the U.S. being a geographically sprawling country compared to South Africa.

In part, for both countries, out existing immunity defenses have largely failed.  Near as I can tell, neither prior infection (the main South African source of immunity to COVID) nor full vaccination (the main U.S. source of immunity to COVID) does much of anything against Omicron.  There, only the boostered population has significant resistance to symptomatic infection.  And that would argue for a smaller wave in the U.S. than in South Africa, as about 30% of our population has been boostered (against what I believe to be a negligible fraction of the South African population.)

Finally, in that case, if our main defenses against COVID-19 fail against Omicron, we’re down to our backups.  That’s mask wearing, distancing, and avoiding high-risk situations.  More-or-less, it’s as if we’re back to the start of the pandemic, and the only real barrier to transmission is COVID hygiene.

South Africa never took down its backup systems.  It had kept restrictions on some forms of social activity in place since the start of the pandemic.  It made mask-wearing outside the home a legal requirement, and enforced that.  And it kept that legal requirement in place, even after the end of the Delta wave.

In the U.S. by contrast, we dismantled the backups.  And that seems to be permanent.  The issue of mask-wearing has been so poisoned by the Republican party that few governors have the stomach for putting any sort of mask mandates back in place. (Or, alternatively, the U.S. population is so full of snowflakes on this issue that few governors dare to do that.)

And so, as a country, we’ve gone into the Omicron wave with no effective vaccine (without a booster shot), no effective protection based on prior infections, and a Republican-driven culture that prevents mere re-imposition of a mask mandate, let alone enforcement of it.

In effect, we have made ourselves purposefully defenseless.  No immunity, no mask.  Nothing except the booster shots, and the common sense of the people.

Once I boil it down that way, I guess it all starts to make sense.  We’ve already had far more cases that I would have remotely believed possible.  And it’s not clear that there’s a light at the end of the tunnel yet.

I keep asking “are we done being stupid yet”?  And the answer keeps coming back, “No”.  I got my prediction wrong.  But as a country, I think we’re fundamentally getting Omicron wave wrong.  If we have no effective immunity, and hospitals are starting to fill, we’re right back where we were in the winter of 2020.  And out COVID-19 hygiene ought to reflect that. But it doesn’t.

To sum up the U.S. situation:  We have

  • no immunity from prior infection,
  • no immunity from two-shot vaccination,
  • no mask use or other COVID hygiene policies,
  • no way to reimpose any such hygiene, and
  • a new variant that is off-the-charts contagious. 

In that context, an overwhelming U.S. Omicron wave starts to make sense.

Maybe things will turn out OK anyway.  Maybe we won’t generate enough cases to overwhelm the hospital system.  Maybe there will be a huge number of hospitalizations, but not so many ICU cases that hospitals run out of ICU beds.

Maybe we’re due for a break any day now.  But in the middle of what I’ve termed the “don’t give a damn” wave, we sure aren’t doing one whole heck of a lot to make that happen.