Omicron update with some caveats.

in covid •  2 months ago 

image.png

Your Local Epidemiologist published an Omicron update today, and IMO she missed a couple of important factors about the data. Maybe I'm wading into my Dunning Kruger zone, but here's her link followed by my comments.

https://yourlocalepidemiologist.substack.com/p/omicron-update-dec-4

In section 1, that graphic is excellent, she doesn't point out that it's raw numbers against rapidly changing population sizes. In the first wave, everybody was getting their initial infection; there were no reinfections possible. (She mentions that.) But in the second and third wave, the population of never-infected people was getting smaller, and the population of previous-infected (and therefore eligible to become a reinfection stat) was growing rapidly.

This means if all the waves were absolutely identical, we'd see the primary infection curves getting smaller and smaller, and the re-infections dominating more and more. That would give the false illusion that each wave was getting better at reinfecting, when it was actually just a side effect of the shifting population pools.

So while it does look like Omicron is better at reinfecting than Delta was, it's probably not as much as the raw-numbers graph would suggest.

In section 2, she presents the data for hospital admissions one section of one state, showing a much larger share of young adults and children. This is true, but they're small numbers from the start of an outbreak that happened overwhelmingly in young people in a few clusters. We really need to compare hospitalizations to case rates by the same age bands before drawing conclusions. The date range is also ending November 27, which means it's almost entirely too early for Omicron to be driving the hospital admissions. (This would have been for infections acquired mostly in the first two weeks of November, too early for Omicron to have much impact.)

She also comments in section two that it's good news that we're not seeing vaccinated people making up a large portion of the people in the hospital yet. While that's true, and she does add a caveat to be careful...this is entirely meaningless for Omicron. (Hospital admissions lag cases by about a week, total number in the hospital lags at least a week farther back. The patients in the hospital over the past week are overwhelmingly from before the Omicron wave.

She knows way more about epidemiology than I ever will, but it looks like she overlooked a couple of data timing issues in the rush to get all the info out.

Authors get paid when people like you upvote their post.
If you enjoyed what you read here, create your account today and start earning FREE STEEM!