Mega-Thread: Coronavirus Pandemic

[Katchii]

You do realize that we only have limited data and estimations about the C19 mortality opposed to decades of hard data on flu, yes?

Hell, RKI for instance calculates death rates only via #tested/#dead, which is absolute nonsense, since it completely ignores the # of people that had this disease w/o it ever being known.

Not that I disagree that it's off, but how do you accurately account for the number of deaths from COVID if you don't have accurate number on the number of people with COVID? You can only perform calculations with numbers you know.

[segara82]

Not that I disagree that it's off, but how do you accurately account for the number of deaths from COVID if you don't have accurate number on the number of people with COVID? You can only perform calculations with numbers you know.

Similar to what the Economist did a little while ago: compare the number of deaths over the last years and calculate the difference in growth. The difference in the usual growth of death and the current numbers is referred to as Excess mortality. While the numbers are not 100% foolproof (some countries are terrible at keeping their books) it gives you a better idea how many people really died because of Covid19.
https://www.economist.com/graphic-de...ross-countries
Based on that we have guess-timates that we have 1.5 million dead or more.

[Katchii]

Similar to what the Economist did a little while ago: compare the number of deaths over the last years and calculate the difference in growth. The difference in the usual growth of death and the current numbers is referred to as Excess mortality. While the numbers are not 100% foolproof (some countries are terrible at keeping their books) it gives you a better idea how many people really died because of Covid19.
https://www.economist.com/graphic-de...ross-countries
Based on that we have guess-timates that we have 1.5 million dead or more.

While true, even then it's just an estimate based on a number of factors and the number can't conclusively be attributed to COVID. It's likely, but it's still just "best guess."

Granted, it's better than nothing for sure, but it's hardly conclusive evidence of the number of deaths explicitly from COVID. Especially since there's a lot more issues aside from health going on because of COVID, mental health being one of them.

[Granyala]

You can only perform calculations with numbers you know.

Sure but these values are very misleading, especially if you deal with a sickness that has a high probability of going undetected due to a large %age of asymptomatic (or very mildly symptomatic) cases that never get recognized.

You could see these numbers as a "worst case" but even for that they are simply put: way too inaccurate.
They are very useful to scare people into obedience though. *sadistic chuckle*

[Katchii]

You could see these numbers as a "worst case" but even for that they are simply put: way too inaccurate.
They are very useful to scare people into obedience though. *sadistic chuckle*

I'm a stickler for accuracy too, as anyone should be honestly, but I'll admit in this situation where an undetermined and incalculable (because you literally don't have the numbers to do the calculations) number of lives are at risk if people don't follow the guidelines, I'm OK with using numbers like this to scare people into line, especially because they're not false, they just don't tell the whole story because they can't.

One of the main problems right now is that you have outright science deniers latching on to this narrative that the numbers are inaccurate, without fully understanding WHY they're inaccurate, to downplay the whole thing as some kind of hoax or minor thing. And because their argument sounds somewhat coherent and intelligent, people who don't know any better, or for whatever reason refuse to actually use their brain and think for themselves instead of follow their cult, also downplay it.

[Forogil]

While true, even then it's just an estimate based on a number of factors and the number can't conclusively be attributed to COVID. It's likely, but it's still just "best guess."

Granted, it's better than nothing for sure, but it's hardly conclusive evidence of the number of deaths explicitly from COVID. Especially since there's a lot more issues aside from health going on because of COVID, mental health being one of them.

Obviously, but there are a number of factors that make it likely that it is closer to the truth than just the reported deaths in many countries.
In Europe one could see that the overall deaths rose together with the pandemic deaths earlier this; and it doesn't fit mental health issues (lockdown weren't stepped up and down so quickly; and one would expect a longer delay for them).

However, in some countries it seems even overall deaths (for all causes) are not that well known, so even that method will fail.

But you are right that it's more complicated, and lockdowns are also more directly responsible for some deaths, I believe I saw numbers close to 1,000 in India (due to police brutality, people run over by trains, etc) - but there 106,000 reported deaths in India so it is still a minor effect. There are also the economic impacts - with 150,000,000 persons set to fall into extreme poverty due to the covid-situation primarily due to the global situation.

For the number of cases you can similarly sample the population - either for current infection with PCR, or for anti-bodies for a past infection; but it can be complicated - and the accuracy of the tests are unclear for current or past infections (not to be confused with accuracy in terms of detecting virus and anti-bodies).

That's why reporting that we have 36,255,199 reported cases and 1,063,780 reported deaths doesn't say much; when the actual cases and deaths are perhaps 700 millions and 2 millions (as some report).

- - - Updated - - -

Multiple points, me thinks:
a) # of tests, you need to find data on that. I seriously doubt that you increase testing throughput at the above rate but it may be a contributing factor.
b) a younger part of the population gets infected. C19 isn't very deadly to young people compared to the 80yr + segment.
c) this is the ugly one: dying can take time. Death numbers can lag behind the case numbers for up to 4 weeks.

d) We are better at treating the disease, using steroids when appropriate, and also other minor improvements, and in some cases because hospitals aren't overfull which drastically reduce the quality of care.

[Katchii]

For the number of cases you can similarly sample the population - either for current infection with PCR, or for anti-bodies for a past infection; but it can be complicated - and the accuracy of the tests are unclear for current or past infections (not to be confused with accuracy in terms of detecting virus and anti-bodies).

That's why reporting that we have 36,255,199 reported cases and 1,063,780 reported deaths doesn't say much; when the actual cases and deaths are perhaps 700 millions and 2 millions (as some report).

The logistics of sampling the entire population, or even just a representative sample size large enough to actually BE representative of the population, is a nightmare, and honestly not a realistic expectation. Having the number of required tests available (who pays for it?), actually performing the tests (on which people? Only those who voluntarily come in? how is that representative because those kinds of people are likely already pretty vigilant about their care and therefore aren't representative of those who aren't).

Again, I'm not disagreeing that the numbers are far from accurate, but given the circumstances and logistics involved with actually getting a truly accurate count, I think this is just about as good as we're going to get.

[Granyala]

d) We are better at treating the disease, using steroids when appropriate, and also other minor improvements, and in some cases because hospitals aren't overfull which drastically reduce the quality of care.

Absolutely.
I read somewhere that some hospitals ventilated too early and too aggressively, causing more deaths as well.

- - - Updated - - -

Again, I'm not disagreeing that the numbers are far from accurate, but given the circumstances and logistics involved with actually getting a truly accurate count, I think this is just about as good as we're going to get.

Honestly, I think we will only get somewhat more accurate data in a post mortem analysis (no pun intended), after the whole shabang has been dealt with.

This situation really demonstrates how much we depend on the availability of accurate data.

[Edge-]

Absolutely.
I read somewhere that some hospitals ventilated too early and too aggressively, causing more deaths as well.

https://www.statnews.com/2020/04/21/...n-ventilators/

Seems to be accurate, but this was discovered pretty early on in April. It seems like it was a reasonable response given all the unknowns, low blood oxygen can get really dangerous, but that as they learned more about the virus the urgency to get folks with low blood oxygen lowered due to the added risks of the ventilator.

[Forogil]

The logistics of sampling the entire population, or even just a representative sample size large enough to actually BE representative of the population, is a nightmare, and honestly not a realistic expectation.

It's not that hard. Pollsters regularly do it, and you shouldn't need massive numbers.

Having the number of required tests available (who pays for it?), actually performing the tests (on which people? Only those who voluntarily come in? how is that representative because those kinds of people are likely already pretty vigilant about their care and therefore aren't representative of those who aren't).

Governments regularly do exactly that:
https://www.cdc.gov/coronavirus/2019...e-surveys.html
https://www.thehindubusinessline.com...e32724694.ece#!
https://www.ukbiobank.ac.uk/2020/07/...atest-updates/

Others combine that with a simple estimate by using blood donors (as the blood is already collected):
https://www.eurosurveillance.org/con....25.28.2001285
(The downside is that blood donors differ a bit from the general population.)

I'm not saying that this gives a correct count down to several decimal places, merely that it gives a better ballpark figure.

[Katchii]

Honestly, I think we will only get somewhat more accurate data in a post mortem analysis (no pun intended), after the whole shabang has been dealt with.

This situation really demonstrates how much we depend on the availability of accurate data.

10000% agreed

It's not that hard. Pollsters regularly do it, and you shouldn't need massive numbers.

I wouldn't exactly call that data representative though, as it only represents the portion of the population that actually answered those questions

Governments regularly do exactly that:
https://www.cdc.gov/coronavirus/2019...e-surveys.html
https://www.thehindubusinessline.com...e32724694.ece#!
https://www.ukbiobank.ac.uk/2020/07/...atest-updates/

Just because they do it doesn't mean it's highly accurate or representative, it just means they're doing their best with what they can get their hands on.

Others combine that with a simple estimate by using blood donors (as the blood is already collected):
https://www.eurosurveillance.org/con....25.28.2001285
(The downside is that blood donors differ a bit from the general population.)

You already stated how this could be a bit off, because the difference in population of those who do and don't donate blood.

I'm not saying that this gives a correct count down to several decimal places, merely that it gives a better ballpark figure.

And I agree with this, but it is still a "ballpark" figure. There's a LOT of caveats to the data set and therefore isn't truly representative, it's just as close to representative as we're realistically going to get. That's my point.

[Keosen]

Fellow Americans, please tell me, how did you vote this clown? Isn't he, seriously, making you feel embarrassed as a citizen?
He can't even form a coherent sentence, he is using the vocabulary of a 10yo, it's literally unbelievable that a guy like him became president.

https://twitter.com/realDonaldTrump/...59702104023047

[Zan15]

Fellow Americans, please tell me, how did you vote this clown? Isn't he, seriously, making you feel embarrassed as a citizen?
He can't even form a coherent sentence, he is using the vocabulary of a 10yo, it's literally unbelievable that a guy like him became president.

https://twitter.com/realDonaldTrump/...59702104023047

i am embarrassed it happen and apologize to the world.

[tromage2]

i am embarrassed it happen and apologize to the world.

Dont, Biden will be a bigger joke to the world.

[Machismo]

Dont, Biden will be a bigger joke to the world.

Considering polling about the standing of the United States regarding this administration and the last (when Biden was VP), I'd say you are dead wrong.

[PhaelixWW]

Dont, Biden will be a bigger joke to the world.

The rest of the world disagrees with you.

[Forogil]

Just because they do it doesn't mean it's highly accurate or representative, it just means they're doing their best with what they can get their hands on.

And they are willing to pay to get better numbers.

And I agree with this, but it is still a "ballpark" figure. There's a LOT of caveats to the data set and therefore isn't truly representative, it's just as close to representative as we're realistically going to get. That's my point.

I'm not denying that it's a bit inaccurate.

However, that's why we estimate that there have been perhaps 700 or 800 million cases of covid in the world.
Obviously that's a large margin of error, but it's a lot more than the 36 million reported cases.

[Granyala]

Fellow Americans, please tell me, how did you vote this clown? Isn't he, seriously, making you feel embarrassed as a citizen?
He can't even form a coherent sentence, he is using the vocabulary of a 10yo, it's literally unbelievable that a guy like him became president.

https://twitter.com/realDonaldTrump/...59702104023047

The FDA approves of "things" within weeks that used to take years.... uhuh, what could possibly go wrong?

[Zan15]

Dont, Biden will be a bigger joke to the world.

I apologize for you and the impact you and your ilk have falling upon this world.

[Katchii]

And they are willing to pay to get better numbers.

Again, better doesn't necessarily mean representative or highly accurate....just MORE accurate.

I'm glad they're working on getting as much, and as accurate amount of data as possible, though.

I'm not denying that it's a bit inaccurate.

However, that's why we estimate that there have been perhaps 700 or 800 million cases of covid in the world.
Obviously that's a large margin of error, but it's a lot more than the 36 million reported cases.

I'm skeptical of this, not because I think they're wrong or they don't know what they're talking about, but because it's based on models and estimations that are already using inaccurate, incomplete and/or non-representative data. I don't doubt the numbers are far larger than 36 million, though, and I'll rely on the WHO to do their best and still listen to them.