Thread by @enn_nafnlaus

@TeslaStars Demonstrably, his claims aren't true, given the overloaded hospitals all over the world. But to be more specific.

1) His first *preprint* actually says that there's 95% odds that the risk of death is "1% to 81% as high as Delta" (Delta being ~2x that of Alpha)

@TeslaStars Note that that's *unadjusted* hazard ratios (single-factor), and that the adjusted hazard ratios (which they didn't even bother trying for for mortality, given the huge CIs) trend to higher hazard ratios than the unadjusted.

@TeslaStars 2) Paxlovid is awesome, but it's also unobtanium, and will continue to be for quite some time. Pfizer hopes to make 120 million treatments (for a planet of 8 billion people) this year, but mainly in the second half. Experts are dubious:
endpts.com/pfizer-says-it…

@TeslaStars Paxlovid costs over $500 per course. IF there's "only" an average of one infection per person per year, that's a good fraction of 1% of the US GDP spent just on Paxlovid. Better buy Pfizer stock.

@TeslaStars 3) You cannot just take vaccine efficacies against Delta mortality (which were similar to efficacy against hospitalization) and act like they apply to Omicron, a strain defined by its evasiveness.

We don't have mortality efficacy, but we do have hosp:

assets.publishing.service.gov.uk/government/upl…

@TeslaStars Highlighting for emphasis.

Note that combining in efficacy against infection doesn't apply when you're pursuing an "everyone gets infected" scenario.

@TeslaStars As for the IFR: IFR doubled from Alpha to Delta, and Alpha was in turn associated with a higher IFR than Wuhan-Hu-1, which was estimated at 0,7%. Our CIs on mortality from Omicron are all over the board. We can't even look to South Africa yet because deaths are still growing.

@TeslaStars (To head people off: no, you can't just compare cases to deaths, you have to do cohort studies to account for time, differences in the cohort (younger wave with Omicron, etc).

@TeslaStars You also can't factor in unobtanium Paxlovid into the equation. Tomas is also quite silent about the fact that Omicron rendered some of our best treatments, like REGN-COV2, near useless.

Lastly, the net mortality from a virus is not simply its IFR, but IFR * (R0-1).

@TeslaStars Influenza has an R0 barely over 1. Delta has an R0 of 5-9. Omicron appears to have somewhat less of an R0, but combined with a high level of immune escape. Regardless, an order of magnitude higher. Which is how it keeps overloading hospitals.

@TeslaStars Not to mention this with regards to everyone's favourite, flu comparisons:

https://twitter.com/danielgoyal/status/1464397121243828227

I'll reiterate: hospitals aren't overflowing on a lark.

@TeslaStars Tomas talks a lot about building immunity. Except the fact that past immunity offers almost no protection against Omicron. You should not expect the reverse.

imperial.ac.uk/mrc-global-inf…

@TeslaStars Bu what about that preprint citing 4.4x neutralization from Omicron against Delta?

ahri.org/wp-content/upl…

First off, that's at two weeks post-infection (peak nAbs), before the rapid post-infection waning period (followed by a slower long-term decline). But more to the point...

@TeslaStars Neutralization != Efficacy

nature.com/articles/s4159…

There's a logarithmic correlation - not a linear one - between neutralization and protection. That 2-week peak is enough to, say, boost a person from 50% to 80% protection. But then comes post-infection nAbs waning.

@TeslaStars Most critically, however, ***Evolution Isn't Planning To Stop For You***. Omicron already appears to have a more contagious sibling, BA.2:

https://twitter.com/CorneliusRoemer/status/1482833651276697612

The concerning aspect of BA.2 is that it appears to be *less* immune evasive, yet more contagious - i.e....

@TeslaStars ... a higher R0. In the past, where we've seen strains develop a higher R0, higher hospitalization and mortality rates have come with it.

But it's so early, we can say very little about BA.2; we're still learning about BA.1 and the clade in general.

@TeslaStars The simple fact is that people are not just developing antibodies to SARS-CoV-2 now. They've had them for ages. They're just having their past immunity bypassed.

Every infection is a net NEGATIVE on the global herd immunity register, a new change to evolve new strains.

@TeslaStars Take Manaus, for example. It was initially cited as an early example of a city that had achieved herd immunity.

Until Gamma came along and caused a far worse, deadlier wave.

thelancet.com/article/S0140-…

@TeslaStars Okay, but surely they got herd immunity THEN, right?

Oh wait.

I'll repeat: herd immunity through infection is a myth. Infection creates strains to BYPASS herd immunity.

@TeslaStars This thread is not intended as doom-and-gloom. We've made huge progress at lowering the mortality rate, despite our evolving foe. We will continue to do so.

But misleading about the current status helps nobody. There's still a slog ahead.