COVID infodump, March 2023

Posted by Patrick Lam on Wednesday, March 8, 2023

Table Of Contents

I was writing on Facebook that the rate of new information about COVID that’s been coming out has really slowed down since March 2020. But there are a bunch of new factoids of various levels of reliability. Let’s talk about them here. I’m still hoping for a nasal vaccine, but there’s no evidence that is going to be available anytime soon.

March 1099, 2020

jwz points out that we’re now three years into the pandemic; in his words, “If you choose to stand around inside a crowded room without wearing a mask – I think you’re a fucking idiot.” (He runs a nightclub, which is the definition of a crowded room.)

Oh yeah, there was that Cochrane review. What is it with Cochrane reviews? I don’t super believe their bicycle helmet review either. My belief is that masks have a small impact at a population level (10%) but much better at an individual level. It’s hard to force people to wear masks. But, if I choose to diligently wear a mask, I’m much more likely to dodge COVID. Indeed, I’m going to continue usually masking, bringing around my CO2 meter, and definitely masking in sketchy places. And planes. No reason to not mask on a plane.

Here’s a Twitter thread by Prof. Trisha Greenhalgh that is a quite detailed discussion of masks in light of the Cochrane review. Maybe the threadreaderapp link will continue working; who knows how long Twitter will be around.

She reiterates the argument that Randomized Controlled Trials are not the only evidence we should believe and anyway they need to be consistent with mechanistic explanations. Also in the thread there are links to non-RCTs which show a bigger-than-10% effect of mask mandates. She does say that masks should be limited to times where there is a lot of COVID going around.

And, this blog post is pretty consistent with what I believe to be true in general. Maybe I disagree in that I think that repeat infections are not more likely to disable, in the general population. The VA study that shows more badness in a more vulnerable population is perhaps not applicable to the general population. (Not less likely, but not more likely).

New Zealand excess deaths

Avoiding COVID was easy mode for me through November 2021, since I was in New Zealand and there was effectively no COVID where I was in New Zealand. For January 2020 through January 2023, New Zealand still has negative excess deaths (fewer flu deaths than usual). These days, NZ is experiencing more COVID deaths in a week than it did during all of 2020 and 2021.

Other bad results of COVID

There are a lot of studies now that show that people have at least 1.5× more risk of heart attacks and strokes for at least a year after COVID.

I’ve also read about elevated diabetes risk. And all this putting aside long COVID. Of which…

Endemic? and long COVID

Eric Topol is fairly balanced and I don’t think he’s previously said that COVID looks endemic. Now he has. Endemic isn’t great. Tuberculosis was endemic and regularly killed, say, 3 white people/1000/year in New York/New Orleans and 5-12 black people/1000/year. I guess at least it doesn’t have huge waves anymore? There haven’t been significant new variants for a while. New Zealand seems to be having a bit of a wave. Waterloo has been stable at a relatively high level since December, with the wastewater numbers between 100 and 300 average copies/mL.

However, experts say that, with 10-20% probability (averaged over multiple experts), they expect another Omicron event within the next two years. No expert said 0% probability.

The other exciting news is that metformin, which is pretty safe, appears to reduce long COVID risk by half. Well, half overall. The hazard ratio is 0.85 in the vaccinated population; I think that means that 85 vaccinated people treated with metformin get long COVID compared to 100 people not treated. The half is the top-line result combining vaccinated and unvaccinated.