[OzyM]

How likely are these numbers to be accurate? This Swiss study of 431 adults gives a 26% incidence of "Long Covid" at 6-8 months.

But another study that I'm more familiar with that tracked 4,182 people (mostly from the U.K., U.S., and Sweden) says only 2.3% of participants reported symptoms past the three month point. [0]

I feel like every time I look into this topic I find a new study with a wildly different estimate.

[0] https://www.nature.com/articles/s41591-021-01292-y

[amirhirsch]

>>Symptoms were present in 385 (89%) participants at diagnosis and 81 (19%) were initially hospitalized.

The US hospitalization rate is (2,297,764 hospitalized / 33,604,986 tested positive) 6.8% so the study is a cohort of extra-sick folks.

Also consider the selection bias of those who want to tell the surveyor about it. I've been thinking about how to correct for this in cohort studies, like perhaps ask people if they are vegan.

[OzyM]

You lost me at the vegan part? I agree that this is a major source of error in a lot of studies like these, though.

[amirhirsch]

The joke is that they go out of their way to tell people about it. Like, I tried the vegan diet, but gave it up after telling my cousin.

[peteretep]

I am partial to the Zoe studies, as their cohort seems to be large, stable, and include people who haven’t been infected.

Write-up: https://covid.joinzoe.com/us-post/long-covid Paper: https://www.medrxiv.org/content/10.1101/2020.10.19.20214494v...

[ping_pong]

The numbers were self-reported, so it's not going to be very accurate.