2021-12-25/26/27/28 General


There has been difficulty figuring out how badly Omicron is going to hit the hospital system first because it is a lagging indicator, second because Omicron affects vaccinated and unvaccinated people so differently, and third because of poor-quality data around the holidays. Still, there are hints that, as this tweet from New York says, “Calling the outbreak milder is correct (relative to cases); Calling it mild is WRONG.”

This Twitter thread by a New York City Emergency Room doctor describes what he’s been seeing. Briefly, people with boosters, two mRNAs, or one shot of J&J have genuinely mild cases (in ascending order of unpleasantness), none have needed hospitalizations, none have trouble breathing. All the unvaccinated people he sees are in bad shape, with trouble breathing. This is anecdotal, but it matches the other anecdotes I’ve been seeing on Twitter.

This article says that pediatric hospitalizations in the USA for COVID-19 are up 30% in one week.

This pre-preprint says that Omicron infections provide some cross-protection against Delta. That’s really good news.


Today, India gave an Emergency Use Authorization to Corbevax. This is a huge fucking deal.

  • Corbevax was developed by Baylor University, who is granting a free license to developing countries.
  • It uses old, well-understood technology. While it’s not quite easy enough for me to make in my kitchen, just about any drug company ought to be able to make it. And that also means it will be CHEAP!
  • Their Phase 3 studies looked at the levels of neutralizing antibodies (nAbs), i.e. in test tubes and not in humans. This is defensible, faster/cheaper, but not quite as trustworthy as randomized clinical trials (because antibodies are only one piece of the immune system). I actually think that’s a good thing for the world: I think rich countries are going to turn their noses up at it which means they won’t poach it from poor countries.
  • From the nAb studies, they estimate that Corbevax’ effectiveness against symptomatic Delta infection is over 80%. That’s good!
  • The nAbs only wane in protection against infection ~30% over six months, which is actually quite a bit better than the mRNAs.

I haven’t seen any data on how well Corbevax works against Omicron, alas.

Not all is awesome: the Sinovac vaccine — which ~2.3B people have gotten — isn’t effective against Omicron infection, even with three doses. Also, the Vietnamese government gave up on trying to find enough volunteers for Covivac — the Vietnamese HexaPro vaccine. I had high hopes for Covivac.


This preprint says that an earlier study which said that Long COVID only affected 1.8% of kids was total horseshit had some serious limitations. They point out that a different study said it was 7x higher, with the strong implication that they thought the other study was not horseshit more believable.


This study found that cancer patients were at a way high risk of hospitalization/ICU/death from COVID-19:

Partially vaccinated48%14%<6%
Fully vaccinated65%19%13%

I have looked at the table multiple times, and yes, the odds get WORSE the more vaccinated the patients were. I’m hoping they mislabeled the table, otherwise it’s reeeeeally strange.

Note that this was survey information, so it is possible that there is a selection bias of some sort. Also, the sample size was small. Still.

This preprint says that there is a protein (TMPRSS13) that is on some white blood cells, and that protein can cleave the Spike protein, which helps SARS-CoV-2 enter cells. A piece of the innate immune system called “enzyme interleukin four-induced gene 1” (IL4I1) binds to TMPRSS13, so maybe flooding the zone with ILRI1 would help stave off COVID-19.


This pre-preprint says that the viral load in infected people’s breath wasn’t higher in Omicron than previous strains. This suggests that it’s greater infectivity per virion and/or greater immune evasion that’s responsible for the Omicron spike, not more virus floating around.


I (and others) have been noticing a fair number of anecdotes on Twitter from people who tested negative with a rapid antigen test when they followed the directions in the packet, but tested positive when they swabbed their tonsils. SWAB YOUR TONSILS!!!

This preprint says that rapid antigen tests are generally less sensitive to Omicron than other variants. In this picture, the roman numerals correspond to different brands of rapid antigen test.

Numeral I is Panbio/BinaxNow, theAbbott labs test widely used in the USA and I think Canada where rapid tests are found. VII is Flowflex, which I don’t know much about but which looks awesome.


I don’t understand this preprint completely, but I think what it says that they have made a pseudovirus of SARS-CoV-2 which glows when it is viable. This is safer to study than the actual virus (which has to be kept in Biosafety Level 3 labs) and easier than evaluating non-glowing pseudoviruses. (You can just plug a light sensor in to see how good your treatment is!)