Am American, not UK, but have worked in hospitals and surrounded by family working in hospitals during the pandemic.

This is a grey area here, because people don't always "come in because of COVID," they come in for symptom X, which might be exacerbated or caused by COVID. If someone has a set of chronic diseases, sure they have that, but the question is "why did they come into the ED today?" The answer to that is sometimes COVID even though they didn't know it. How this looks on a hospital chart is really fuzzy because it depends on all kinds of things; ICD coding can be ambiguous.

There's also been some cases of COVID contracted in the hospital (the ones I have person familiarity with), but that is much less rare, and rarer as the year has gone on.

There are also cases of people coming in for reasons unrelated to COVID, and finding out they were positive just coincidentally.

There's also cases where it's really unclear.

I mean, they may be catching covid in hospitals, there's no way to know for sure, but it seems unlikely to account for all these cases.

But yeah this is going to be picked up by covid-deniers as more "proof" that covid is exaggarated.

I am not by any means a "covid-denier," but it seems to me this is evidence that the impact of Covid may be exaggerated. It's hard to say by how much (and I would be interested to learn if this is happening in US hospitals also).

> this is evidence that the impact of Covid may be exaggerated

I don't think so. When assessing the impact of Covid, we look at the change in overall, all-cause numbers, like all-cause deaths. (Which do show a big impact.)

The reason we watch covid hospitalizations is that it is a good leading indicator of covid deaths. The "leading" part is important, because that allows health policy to react faster to what's happening, which makes it more effective at reducing the impact of Covid.

(You may also worry that covid deaths are being similarly over-counted, but the all-cause death numbers tell us we're actually undercounting the overall number.)

Anyway, I guess ultimately we can't stop people from misusing the covid hospitalizations number, but it is a very useful number.

I don't think "Covid deniers" make up any significant percentage of the population, but that doesn't change the fact that this is proof that Covid is exaggerated.

I think covid deniers do make up a significant percentage of the population, and they would take this as proof of exaggeration even though it really isn't.

US has had 1M excess deaths. That's a million people more than usual. http://www.healthdata.org/special-analysis/estimation-excess...

How is this not proof of exaggeration? Somebody in the hospital for a broken leg being counted as a "Covid hospitalization" just because they also happened to test positive for Covid means that the numbers are inflated. Period.

Covid is real. Covid numbers are inflated. You're arguing against a fictional "Covid denier" boogeyman which doesn't really exist.

Regardless of why they were admitted they still had covid.

You’re trying to answer what sent someone to hospital, not who has covid.

Your paper isn’t fact. It is an estimated published by perhaps one of the worst modelers out there. The IHME has been publishing scary models that couldn’t even accurately predict what was happening the day they were published.

Not significant percentage?

https://www.politico.com/states/florida/story/2021/07/26/sel...

"People are not catching Covid in the hospital" that is nonsense. They are, and the question is how many. The next question is if it is a significant number, is there a good way to prevent it.

I think OPs point is that this scenario is not the focus or point of the article.

I know (anecdotally) of one person who was in the hospital for an orthopedic surgery, caught Covid during their 3-day recovery stay, then got pneumonia and ended up having to stay for 10 days.

I should have said, that is not what the article is talking about. This is about miscategorization.

I haven't checked this as the full data the article was based on wasn't given in the article, but it wouldn't make much sense to count out-patients as part of the covid hospitalisation statistics, so I imagine that the word "hospitalisation" means in-patients specifically i.e. people who are sick enough to stay in a bed in a ward for a night or more, so things like pink-eye, fractures, even child birth (without complications) wouldn't count.

Very good point!

Another issue that is skewing the data:

Early in the pandemic, the United States had an undertesting problem. Now we are overtesting those who are immune and asymptomatic. A person with immunity to the coronavirus will fight off an infection. But during and after the person’s exposure to the virus, it’s common for a low number of virus particles to be detectable in the nose. In medicine, we call this virus a “colonizer” — a pathogen that does not cause illness or spread the illness. It’s an incidental finding. But in today’s world of routine coronavirus testing of vaccinated people, these positive tests are inflating the number of positive cases in a misleading way.

https://www.washingtonpost.com/outlook/2021/07/21/covid-test...

The trouble is that there's this unfortunate tendency to try and deal with this by having different standards for what gets counted as a covid case depending on whether the person's vaccinated - either by testing vaccinated people less as suggested in that article or by interpreting the test results differently. Which makes all statistics about the proportion of people who catch Covid, are hospitalized with it, etc who are vaccinated complete and utter garbage - at that point you're not finding Covid cases amongst vaccinated people simply because you're not looking. It resulted in initial numbers out of Israel that seem to have wildly overstated vaccine effectiveness against current variants, and I suspect the same might be happening in the USA right now.

> in today’s world of routine coronavirus testing of vaccinated people

It's worth noting that the CDC is explicitly telling people not to do this:

If you’ve been around someone who has COVID-19, you do not need to stay away from others or get tested unless you have symptoms. (https://www.cdc.gov/coronavirus/2019-ncov/vaccines/fully-vac...)

Agreed that people are still doing it though. That seems reasonable to me given the mixed data around to what degree the vaccine prevents spread.