Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Covid wards 'full of children' for first time in pandemic, warn nurses

yahoo.com

Post reply on HN
Print view View original

21–23 of 23 posts

Re: Covid wards 'full of children' for first time in pandemic, warn nurses

#21

Post by DanBC »

undo123 wrote: post #3

Does anyone know where to find absolute numbers on age distributions of people currently hospitalized with C19? We’ve seen enough misunderstandings of data in the news that I prefer to see data myself. (Bonus points for “With-vs-From” data, but I haven’t been able to find that broken out in any public dataset.)

It's pretty complicated.

https://www.ics.ac.uk/ICS/News_Statements/Understanding_inte...

See also the ICNARC reports.

> “With-vs-From” data

You saying this is a pretty clear demonstration that you're not going to understand the data if it's given to you, which is one reason people haven't dug out the spreadsheets for you.

Re: Covid wards 'full of children' for first time in pandemic, warn nurses

#23

Post by dragonwriter »

undo123 wrote: post #16
67868018 wrote: post #13

Earlier quoted context omitted.

> with vs from There’s a very tired conspiracy that people can walk into an ER with e.g., a fatal gunshot wound, get tested positive for COVID, and then we are inflating the numbers and blaming their death on the virus.

Not looking to start or participate in a conspiracy, but I do think it’s valid to ask for both numbers: The number of people hospitalized (or died) who have C19, and the number of people hospitalized (or died) because of C19. Unless the coding systems simply can’t make that distinction, it seems like a no brainer to publish both numbers. The conspiracy theorists lose traction, and we get more data. Despite this, I se…

> Unless the coding systems simply can’t make that distinction

They probably can’t, assuming that they haven’t adopted a more complex (and, consequently, probably less reliable at what it purports to represent) diagnostic categorization than is generally used for other purposes in healthcare; the usual systems can distinguish admitting and primary diagnoses and diagnoses that are neither admitting nor primary (though I don’t know that the distinction is in what is shared for public health reasons), but it can’t distinguish between “diagnosis that are part of the reason that they were admitted to or not yet released from the hospital” and “diagnoses that aren’t part of that despite being present”.

At least, that’s my understanding based on a couple decades of contact with health IT systems and the associated data models.

Post reply on HN