General > General Technical Chat
Covid 19 virus
<< < (305/381) > >>
engrguy42:

--- Quote from: drussell on March 28, 2020, 01:50:24 pm ---
--- Quote from: engrguy42 on March 28, 2020, 11:07:10 am ---I'd also caution folks to try to avoid predicting what will happen. Nobody knows.
...
Or at least allow the possibility that you might be wrong.
--- End quote ---

You honestly still believe that the near-term progression in the USA cannot be predicted?  Wow.

:palm:

--- End quote ---

As usual, I provide data and you provide facepalms and unsupported, handwaving predictions.
Mortymore:
Cerebus:

--- Quote from: engrguy42 on March 28, 2020, 01:58:42 pm ---
--- Quote from: drussell on March 28, 2020, 01:50:24 pm ---
--- Quote from: engrguy42 on March 28, 2020, 11:07:10 am ---I'd also caution folks to try to avoid predicting what will happen. Nobody knows.
...
Or at least allow the possibility that you might be wrong.
--- End quote ---

You honestly still believe that the near-term progression in the USA cannot be predicted?  Wow.

:palm:

--- End quote ---

As usual, I provide data and you provide facepalms and unsupported, handwaving predictions.

--- End quote ---

No he didn't, no prediction, no handwaving. He expressed incredulity at what he sees as your beliefs. Not the same thing at all. If you're going to criticise someone, criticise them for what they actually said.
drussell:

--- Quote from: imo on March 28, 2020, 09:39:18 am ---"Using a delay-adjusted case fatality ratio to estimate under-reporting"
https://cmmid.github.io/topics/covid19/severity/global_cfr_estimates.html
--- End quote ---

I will have to read the whole paper that the calculatons in your above cited paper are using (where they estimate the 1.38%) which was based on a data from January on a few thousand cases in China:

https://www.medrxiv.org/content/10.1101/2020.03.09.20033357v1.full.pdf

At first glance of your cited paper, though, they don't really specify any backup data to show that their procedure for estimating how under-reported cases are in a given area are sound.  They're basically assuming that a) the 1.38% is a reliable reference percentage, b) that anywhere that the rate appears higher than 1.38%, it must be entirely due to under-reporting of infections.

Now, it is true that I don't expect the actual CFR in the US to be >20% when all is said and done, but there is certainly the potential for it to be higher than their "standard" of 1.38%.

As an aside, their chart essentially shows that Canada "must be" only finding 15-40% of our true cases, yet this does not seem to correlate with the 1-1.5% positive rate on all the testing we've done thus far.  Where are all these supposedly un-diagnosed symptomatic cases hiding?

It will be very interesting to see the results of some antibody tests on more of the general public once those become available to test for markers of past infections.  That will finally give us some real data on that front to make better estimates and plan who can do what, where, before a vaccine is (hopefully, potentially, eventually) available for those who have not yet been exposed.
drussell:

--- Quote from: Sredni on March 28, 2020, 03:27:51 am ---
--- Quote from: Dundarave on March 28, 2020, 01:34:13 am ---"Fairly flat" isn't how I'd define the US Covid-19 death rate, given the graphs compiled by these folks:

https://www.worldometers.info/coronavirus/country/us/
--- End quote ---

That's the total. It's the increment (new deaths day per day)  that is fairly flat and gives an almost linear progression. But at this early stage I doubt it has any relevance. Mostly these are the dead that resulted from the number of infected about two weeks earlier (as per a study on Lancet IIRC). Not enough data.
The exponential for the death is starting to kick off in these days, unfortunately.
--- End quote ---

Indeed, the flat section of the "deaths" graph corresponds to the flat section of the "infected" curve from a couple weeks ago.  It's not going to be flat again, at least not for a while:

Navigation
Message Index
Next page
Previous page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod