Would you have had this reaction if the same thing had been tweeted by anyone else? When you’re looking at constant news stories about increased childhood hospitalizations with COVID, it seems important (to me) to understand the extent to which hospitalizations are directly due to COVID. For example:
From that story:
Also complicating the picture: Alarming hospitalization figures can be misleading because they sometimes include all children who have tested positive for the coronavirus upon admission.
Some hospitals around the country have reported positivity rates as high as 20 percent among children. But the vast majority were asymptomatic and arrived at the hospital with other health problems, officials say.
So what is it about this tweet that is bad, other than E***y O***r wrote it?
There’s probably a few other people, but when it’s someone who writes for the Atlantic with fans here after being laughably wrong the entire pandemic, it seems reasonable to dunk on her again
I didn’t even see the dunk. The posted tweet seemed like a legit question. I thought there might be more if I actually looked at the feed (maybe she is drawing some ridiculous conclusions from that data, which would be on brand for her) but I couldn’t be bothered.
And it would be very interesting and important if that changed with Omicron being the dominant strain. And it seems possible that might happen, so it’s worth keeping an eye on.
I thought it was a legit question even then. However, you wouldn’t have to dig too deep into the data to see that nearly all the excess deaths were due to COVID. So, the problem is not really the question itself, it’s where they (and perhaps Oster) run with it after that.
Not really familiar with her objectionable child covid takes but the one you posted today seems quite reasonable. I mean it’s not even a take, she’s just asking about how to collect with/for hospitalization data.
The COVID deniers were constantly talking about deaths with the “with COVID” vs. “due to COVID” attitude, with claims that most COVID deaths were from things like motorcycle accidents or whatever being attributed to COVID. And that’s obviously nonsense.
But in a world where some/many hospitals are testing every patient for COVID regardless of why they’re there (so that increased COVID spread among the population would naturally lead to higher hospital-associated rates even if all of those cases were asymptomatic), it seems completely appropriate to ask how much of COVID-tagged hospitalizations are actually due to COVID.
I haven’t seen any hint of COVID denialism tied to this Oster discussion, and I’m not sure why it’s assumed in this case.
Osters work is well known. Feel free to review it yourself. This sudden interest “with Covid” is an attempt to salvage an intellectually dishonest “win” after her work has fallen apart even more
Sure if you want to link to the work of hers that you think has most disastrously fallen apart I’ll be happy to review it. I’m not really up to date on Oster’s previous work.
But it’d be foolish to think ‘oh hey there’s a big spike in pediatric hospitalizations right now and a bunch of them are testing positive for covid, maybe covid is unrelated’.
Covid in kids is tricky. It rarely causes the respiratory issues classic in adults. What it does cause is bronchiolitis and asthma exacerbations, diarrhea, refusals to eat maybe cause taste is gone.
How you code these types of encounters is going to make a big difference.
The last covid+ kid I admitted was admitted for dehydration. Kiddo had a bunch of diarrhea and wasn’t eating/drinking. Came to me lethargic and severely dehydrated. Covid+. The reason for admission was dehydration. It wasn’t covid. But covid caused that dehydration.
This preprint illustrates some of the differences between kids and adults:
Length of hospital stay was short (mean 3.2 days), and in 44% COVID-19 was the primary diagnosis. Most children received standard ward care (92%), with 31 (25%) receiving oxygen therapy. Seven children (6%) were ventilated; four children died, all related to complex underlying co-pathologies. All children and majority of parents for whom data were available were unvaccinated.
25% is insanely low for o2 therapy. For adults its 90%+ I’m sure. But the kid with T1DM who gets admitted with DKA, his primary process isn’t the covid, but covid almost certainly contributed to the DKA. There’s a lot of room to fudge the numbers here and I can’t wait for some preprint to do exactly that like we’ve seen in other awful preprints.
The authors of this preprint go into this, 18.8% are cases where covid is ‘contributory’. The remaining 37% are incidentals like kids with appendicitis or whatever.
SA has a nice advantage right now, it’s summer so I don’t think they’re getting the massive amounts of RSV we’re getting right now so it should be easier.
Regardless, it’s incredibly foolish to expect that a spike in hospitalizations that coincide with covid positivity as a coincidence.
