Earlier quoted context omitted.
Hm, that's [that 'bear in mind'] correct: https://coronavirus.data.gov.uk/details/about-data#daily-and... Just wanted to check that since (at at least one London hospital) they're on different wards (as in, even discounting that obviously you don't want 'broken leg with covid' on a general 'broken bones' ward with neg/query patients) so seems odd to me they'd report data like that vs. those needing treatment for it.
My understanding is, there's a fair few people at hospital with "bad Covid" who aren't in Intensive Care. For example, if you have steroids and Remdesivir given to you, you already need to be hospitalised. So there's at least three categories: people with really bad Covid, people with Covid bad enough they can't be at home, but not (at the time) awful, and people who don't suffer with Covid at all but could infect ot…
> Ultimately, if you project a complex situation onto R^1, it can't be too accurate.
Absolutely, and it's all too easy to make 'obvious' armchair policy..! I was just a bit surprised since my assumption would be it counted the relevant wards (if it were up to me, I think that would be the first three; only not the isolation-but-there-for-something-else).