By using this site, you agree to our Privacy Policy and our Terms of Use. Close

Forums - General Discussion - Coronavirus (COVID-19) Discussion Thread

JRPGfan on 25 March 2020
Teeqoz said:
JRPGfan said:

The point is that we have tested more people.... That is why our data is more accurate.... which is why our data is useful for extrapolation, while Italy's - and Denmarks - isn't.

A 15% mortality rate isn't a worst case scenario. It's just wrong. Any projection - even a worst case scenario projection - has to look at the data. If not you might as well say that the worst case is a 100% mortality rate. Both 15% and 100% are nonsense and equally worthless as worst case secnarios.

If you don't care about the quality of the dataset, you can draw whatever conclusion you want. I could use only the cases that have died and say "oh, in my dataset, everyone with the disease died, so the worst case is a 100% mortality rate". Obviously, that is misleading at best, and malicious at worst. That is what you are doing, though taken to the extreme.

You seem to understand that the Norwegian dataset is of higher quality, so I can't understand why you'd disregard it and use lower quality datasets.

I guess I can understand your point of view.
The truth is probably somewhere it the middle though, norways data could be a outlier too (too many younger/fit people tested)?

Its hard to say.


*edit:
1,5% of norway is around 80,000+ tested.
in denmark that number is only ~14,000 (im not actually 100% sure)

However 1,5% (of norway) is still a small sample size, depending on how you choose who to test.
Like women are alot less likely to get hospitalised or die from this than men.
Same with younger people, or ones that are in great health.
Smokers, diabetes, blood pressure, circulation, weaken immunesystem,... theres plenty of factors that effect things.

So if you randomly test 80,000 school kids..... your numbers are going to be very differnt than other places in the world.


Also even though we all call it Covid-19, theres many differnt mutations of it.
On island there were over 40 differnt version of it alone.

Maybe the virus in denmark, is more dangerous than the one in norway? all it takes it a mutation from the ones that carried it back and spread it here.


*edit2:
If you want to use big data sets, use China's.
No country has been as strict as them with testing.

They meassure your temperature everytime you exit your apartment buildings, everytime you enter a shop, everytime you return ect.
Ontop of testing for the virus, everytime anyone had symptoms or fever.

This ofc means they "found" alot of sick people really early, and could treat them early.
They have a Mortality rate of over 4% currently.

Last edited by JRPGfan - on 25 March 2020

Around the Network
SvennoJ on 25 March 2020
Teeqoz said:
JRPGfan said:

That is not a worst case.

Worse case (70% infection rate) the health care system collapses.
By then its not 2,5% mortality rate, but something likely well over 10%.

You likely wont even have people to give beds/care, or air to the sick.
Not to mention enough ventilators, in such a situation.

In italy its around ~9,8% currently (confirmed vs mortalities):
Believe it or not, there health care system still funktions pretty well.

At a 70% infection rate, any system would totally collapse..... at that point hospitals would be like? whats the point.
Just stay home and die there, we cant help.


Upto ~20% require medical attention aided air (not ventilators) but a sick bed to lay down in, and 100% oxygen to help them get air through their lunges.

Thats why I dont think Mortality rates will go over 20%.
However you *could* see something like 10-15% mortality rates, if things get really nasty and health care collapses completely.



Your worse case should be:

70% infected + ~15% mortality rate =  ~34,4 million people dieing in the US.

Thats the absolutely worst, this virus can do.
Hopefully that doesnt happend but.... it gives you perspective.

My country, Norway, has tested roughly 1.5% of the population, which is probably one of the highest test coverages of any country (more than twice as many per capita as South Korea for instance). We have 3k confirmed cases, which should be a decent enough sample size to draw some conclusions. 

So far, we have 14 dead, which is a less than 0.5% CFR. We have 237 hospital admissions, which is a less than 8% hospitalisation rate. 2.5% of the cases have required intensive care (some of which have sadly perished).

While these numbers might increase a bit (notwithstanding a huge amount of undiagnosed cases, though that would bring the numbers further down), as some cases will progress and get worse, and certainly if our health system gets over capacity, it's easy to see that your estimates are absolutely ludicrous and not based on good data. Data sets from other countries that have done extensive testing tell the same story. Italy's CFR and hospitalization rate are so high because they haven't tested enough people.

This doesn't mean this isn't a serious illness. But you still shouldn't make up numbers, or make wild guesses without having done proper research.

When you are on the ball with testing, your estimates will be too low.

The mean incubation time is 5.2 days.
Mean time from illness onset to death was 20 days (In China)

So you need to compare the 14 dead to the sample size you had 15 to 20 days ago.
Hospitalizations don't happen straight away either and ICU admissions come after that.
Recovery can last a long time, 25% was still not recovered 6 weeks after the Diamond Princess went into quarantine

Plus when the health care system collapses, the ICU admission rate will the lower estimate for death rate, while a percentage of those that required hospitalization without ICU will also perish.

A worst case scenario where 8% of those that get infected don't make it through is very possible.


Even your 8% hospitalization rate would put 300K Norwegians in the hospital when 70% of the population gets infected, while Norway only has a bit over 20K hospital beds (which are likely mostly in use already) So what's going to happen with 290K Norwegians that need help?


The mortality rate entirely depends on how much medical help is available.



Teeqoz on 25 March 2020
JRPGfan said:
Teeqoz said:

The point is that we have tested more people.... That is why our data is more accurate.... which is why our data is useful for extrapolation, while Italy's - and Denmarks - isn't.

A 15% mortality rate isn't a worst case scenario. It's just wrong. Any projection - even a worst case scenario projection - has to look at the data. If not you might as well say that the worst case is a 100% mortality rate. Both 15% and 100% are nonsense and equally worthless as worst case secnarios.

If you don't care about the quality of the dataset, you can draw whatever conclusion you want. I could use only the cases that have died and say "oh, in my dataset, everyone with the disease died, so the worst case is a 100% mortality rate". Obviously, that is misleading at best, and malicious at worst. That is what you are doing, though taken to the extreme.

You seem to understand that the Norwegian dataset is of higher quality, so I can't understand why you'd disregard it and use lower quality datasets.

I guess I can understand your point of view.
The truth is probably somewhere it the middle though, norways data could be a outlier too (too many younger/fit people tested)?

Its hard to say.

Well, luckily for you, Norway releases the age distribution of our positive cases, so we can compare it to our overall population age distribution to see if it is a representative sample.

90-99 : ~0.9% of cases    ~ 0.9% of overall population (representative)

80-89 : ~3.3% of cases    ~3.3% of overall population (representative)

70-79 : ~5.9% of cases    ~7.9% of overall population (underrepresented)

60-69 : ~ 11.7% of cases ~ 10.8% of overall population (slightly overrepresented)

50-59 : ~ 21.4% of cases ~ 13% of overall population (very overrepresented)

40-49 : ~ 17.8% of cases ~ 13.6% of overall population (overrepresented)

30-39 : ~ 13.9% of cases ~ 13.6% of overall population (representative)

20-29 : ~ 10.5% of cases ~ 13.5% of overall population (underrepresented)

10-19 : ~ 3% of cases      ~ 12% of overall population (very underrepresented)

0-9     : ~ 0.5% of cases   ~ 11.5% of overall population (extremely underrepresented)

in conclusion, except for the age group 70-79, which is somewhat underrepresented, every age group over 30 is either overrepresented or accurately represented. Ages under 30 are in general very underrepresented compared to the population.

So on the contrary to what you're suggesting, younger and more healthy people  actually seem to be quite underrepresented in our data.



JRPGfan on 25 March 2020
Teeqoz said:
JRPGfan said:

I guess I can understand your point of view.
The truth is probably somewhere it the middle though, norways data could be a outlier too (too many younger/fit people tested)?

Its hard to say.

So on the contrary to what you're suggesting, younger and more healthy people  actually seem to be quite underrepresented in our data.

I HOPE your right.
If the rest of the world is as lucky as Norway is currently (14 deaths out of 3,066 people infect = 0,4%) then that doesnt sound so bad.

I suspect your wrong, and this is more dangerous than norways data seems to indicate so far.

Time will tell.

So far, china has had this virus the longest, had the most cases, and tested the most.
They have over 4% of people that get it die.

In italy its over 10% of the confirmed cases, that got the infected that died.

Norways current ~0,4% death rate is very good though.



jason1637 on 25 March 2020

Younger folks are underrepresented because even if they might have the virus some dont show symptoms and aren't able to get tested.



Around the Network
haxxiy on 25 March 2020
SvennoJ said:


Even your 8% hospitalization rate would put 300K Norwegians in the hospital when 70% of the population gets infected, while Norway only has a bit over 20K hospital beds (which are likely mostly in use already) So what's going to happen with 290K Norwegians that need help?

Well, even if it were to spread uncontrolled over the course of about two months, the estimated peak would be less than 10% of total number of infections ocurring at the same time. So, that's more like 30,000 hospital beds needed, rather than 300,000. And that's assuming the sample is statistically relevant and requirements for hospitalization remain more or less the same as they are now.

@JRPGfan - Italy only tests high risk persons since February 27. They estimate there are some ten other cases for every one they are diagnosing. Same thing happened in Wuhan: 86% of estimated undiagnosed cases. When we'll be sure? Only when we're able to test a statistically significant sample of the population for Covid-19 antigens.

That's why estimating from a 2.5% death rate, or any other death rate one might choose, for 70% of the US population seems disingenuous. Even if 70% of the population were to get it, the number of diagnosed cases would be off by an order of magnitude or more. Specially if the elderly remain in isolation and it disproportionally targets younger, mostly assymptomatic people, as it would be the case with vertical isolation.



 

 

 

 

 

JRPGfan on 25 March 2020
haxxiy said:
SvennoJ said:


Even your 8% hospitalization rate would put 300K Norwegians in the hospital when 70% of the population gets infected, while Norway only has a bit over 20K hospital beds (which are likely mostly in use already) So what's going to happen with 290K Norwegians that need help?

Well, even if it were to spread uncontrolled over the course of about two months, the estimated peak would be less than 10% of total number of infections ocurring at the same time. So, that's more like 30,000 hospital beds needed, rather than 300,000. And that's assuming the sample is statistically relevant and requirements for hospitalization remain more or less the same as they are now.

@JRPGfan - Italy only tests high risk persons since February 27. They estimate there are some ten other cases for every one they are diagnosing. Same thing happened in Wuhan: 86% of estimated undiagnosed cases. When we'll be sure? Only when we're able to test a statistically significant sample of the population for Covid-19 antigens.

That's why estimating from a 2.5% death rate, or any other death rate one might choose, for 70% of the US population seems disingenuous. Even if 70% of the population were to get it, the number of diagnosed cases would be off by an order of magnitude or more. Specially if the elderly remain in isolation and it disproportionally targets younger, mostly assymptomatic people, as it would be the case with vertical isolation.

This is all just guess work, some say x5 others x10.

What we do know is that of the people we suspect have it, and test possitive for it, in denmark, ~20% end up going to the hospital and ~5,1% go to the ICU.

How many more are running around out there, with the virus without symptoms? Probably as they say 5 for every 1 showing symptoms.

And Yes your right, that number should be factored into a situation where you assume 70% of the population has it (which I clearly forgot).



Those would still be massive numbers though.
No nation wants to see 70% of their population get this, the herd immunity route is crazy risky.

^ this was what started the best case/ worst case debate.

It was a matter of economy vs risk of letting it spread, debate (a few pages back in this thread, which lead us here).


edit:
"Mostly" is the keyword.
Yesterday a 28year old marathon runner, in great shape without any underlying risk factors passed away to this.
Today in the papers, I read a 5year old got so sick (in the UK) he ended up on a ventilator.

"Specially if the elderly remain in isolation"

Dr John Campbell, went over some statistic yesterday breaking down the mortality rates.
Apparently obesity is a big factor (~71% of people that it kills are overweight), people immuno-compromised, renal (kidneys), cancer, ect.

Are you fat? do you smoke? are you a male? (apparently it kills alot more men than women).

source: https://www.youtube.com/watch?v=TJ3QqNeeWUA

50-59 = 1% mortality
60-69 = 3,5% mortality
70-79 = 12,5% mortality
80-89 = 19,7% mortality

Campbells numbers for Italy's mortality statisics.

Over 50 years old? stay home.
Smoke / astma? ect stay home.
Fat? stay home
Cancer? stay home.
kidney issues? stay home
High blood pressure? stay home.


I'm sure what you said, could be done, in a inteligent way, and you could make the best of the situation by only haveing the population that statistically should survive it, get it.

I still think its a very risky way to go, no "isolation" is perfect, and the more of a population total has this virus, the higher the risks those groups likely to die from it, at put at.

Last edited by JRPGfan - on 25 March 2020

SvennoJ on 25 March 2020
haxxiy said:
SvennoJ said:


Even your 8% hospitalization rate would put 300K Norwegians in the hospital when 70% of the population gets infected, while Norway only has a bit over 20K hospital beds (which are likely mostly in use already) So what's going to happen with 290K Norwegians that need help?

Well, even if it were to spread uncontrolled over the course of about two months, the estimated peak would be less than 10% of total number of infections ocurring at the same time. So, that's more like 30,000 hospital beds needed, rather than 300,000. And that's assuming the sample is statistically relevant and requirements for hospitalization remain more or less the same as they are now.

@JRPGfan - Italy only tests high risk persons since February 27. They estimate there are some ten other cases for every one they are diagnosing. Same thing happened in Wuhan: 86% of estimated undiagnosed cases. When we'll be sure? Only when we're able to test a statistically significant sample of the population for Covid-19 antigens.

That's why estimating from a 2.5% death rate, or any other death rate one might choose, for 70% of the US population seems disingenuous. Even if 70% of the population were to get it, the number of diagnosed cases would be off by an order of magnitude or more. Specially if the elderly remain in isolation and it disproportionally targets younger, mostly assymptomatic people, as it would be the case with vertical isolation.

The people that need hospital care can be in there for weeks:

Of 138 hospitalized patients from January 1 to January 28, 2020; final date of follow-up was February 3, 2020. The median time from first symptom to dyspnea was 5.0 days, to hospital admission was 7.0 days, and to ARDS was 8.0 days. Patients treated in the ICU (n = 36), 4 (11.1%) received high-flow oxygen therapy, 15 (41.7%) received noninvasive ventilation, and 17 (47.2%) received invasive ventilation (4 were switched to extracorporeal membrane oxygenation). As of February 3, 47 patients (34.1%) were discharged and 6 died (overall mortality, 4.3%), but the remaining patients are still hospitalized. Among those discharged alive (n = 47), the median hospital stay was 10 days (IQR, 7.0-14.0).
https://jamanetwork.com/journals/jama/fullarticle/2761044

A shame there is no follow up of what happened to the remaining 85 patients. Your active cases keep piling up in the health care system. The recovery time works against you.

86% of undiagnosed cases, where do you get that from?

The Diamond Princess suggests 46% of people infected that show no symptoms (331 of 712)
https://www.statista.com/statistics/1099517/japan-coronavirus-patients-diamond-princess/
But the median age aboard a cruise ship is likely higher.
15 are still hospitalized btw and another 115 still quarantined.

The fact remains that cases that require hospitalization will climb much faster and higher than our healthcare system can cope with. Italy is enough proof of that, no matter how many undiagnosed cases there are.



crissindahouse on 25 March 2020
Teeqoz said:
JRPGfan said:

That is not a worst case.

Worse case (70% infection rate) the health care system collapses.
By then its not 2,5% mortality rate, but something likely well over 10%.

You likely wont even have people to give beds/care, or air to the sick.
Not to mention enough ventilators, in such a situation.

In italy its around ~9,8% currently (confirmed vs mortalities):
Believe it or not, there health care system still funktions pretty well.

At a 70% infection rate, any system would totally collapse..... at that point hospitals would be like? whats the point.
Just stay home and die there, we cant help.


Upto ~20% require medical attention aided air (not ventilators) but a sick bed to lay down in, and 100% oxygen to help them get air through their lunges.

Thats why I dont think Mortality rates will go over 20%.
However you *could* see something like 10-15% mortality rates, if things get really nasty and health care collapses completely.



Your worse case should be:

70% infected + ~15% mortality rate =  ~34,4 million people dieing in the US.

Thats the absolutely worst, this virus can do.
Hopefully that doesnt happend but.... it gives you perspective.

My country, Norway, has tested roughly 1.5% of the population, which is probably one of the highest test coverages of any country (more than twice as many per capita as South Korea for instance). We have 3k confirmed cases, which should be a decent enough sample size to draw some conclusions. 

So far, we have 14 dead, which is a less than 0.5% CFR. We have 237 hospital admissions, which is a less than 8% hospitalisation rate. 2.5% of the cases have required intensive care (some of which have sadly perished).

While these numbers might increase a bit (notwithstanding a huge amount of undiagnosed cases, though that would bring the numbers further down), as some cases will progress and get worse, and certainly if our health system gets over capacity, it's easy to see that your estimates are absolutely ludicrous and not based on good data. Data sets from other countries that have done extensive testing tell the same story. Italy's CFR and hospitalization rate are so high because they haven't tested enough people.

This doesn't mean this isn't a serious illness. But you still shouldn't make up numbers, or make wild guesses without having done proper research.

You could also just use numbers from single regions of countries like Italy because you can't just say Norway is a better representation as Italy because of a higher percentage of testing when different regions test differently.

Italy tested over 80k people in the Lombardy which has 10m people. Of those 80k tested, you have around 4,5k deaths and like 33k positive cases.

Lombardy has around the same amount of tests and you have like a 30x higher death rate as Norway has.

Italy might be a very negative scenario but Norway is at least right now a very positive one. But the death rate will increase there as well like it also did everywhere else. That's sadly the truth like we also see in countries like Germany now (which does still look good compared to many others)



Teeqoz on 25 March 2020
SvennoJ said:

When you are on the ball with testing, your estimates will be too low.

The mean incubation time is 5.2 days.
Mean time from illness onset to death was 20 days (In China)

So you need to compare the 14 dead to the sample size you had 15 to 20 days ago.
Hospitalizations don't happen straight away either and ICU admissions come after that.
Recovery can last a long time, 25% was still not recovered 6 weeks after the Diamond Princess went into quarantine

Plus when the health care system collapses, the ICU admission rate will the lower estimate for death rate, while a percentage of those that required hospitalization without ICU will also perish.

A worst case scenario where 8% of those that get infected don't make it through is very possible.


Even your 8% hospitalization rate would put 300K Norwegians in the hospital when 70% of the population gets infected, while Norway only has a bit over 20K hospital beds (which are likely mostly in use already) So what's going to happen with 290K Norwegians that need help?


The mortality rate entirely depends on how much medical help is available.

I did say these numbers will rise, and that includes the CFR and the hospitalization rate. However the current numbers are already enough to say that a 15% mortality rate is completely ridiculous, even as a worst case scenario, which was what I was responding to. However the progression of cases turning more serious might just as well be countered by undiagnosed cases, as we certainly have plenty of them in Norway in too.

As for the rest you are saying, everyone won't be infected at the same time, and everyone won't require hospitalization at the same time. Even if they would, it's not like every single person that would be hospitalized will die if they aren't. People are hospitalized to keep them under observation in case their condition worsens, and they might be treated with antiobiotics to prevent opportunistic bacterial infections, but everyone hospitalized case won't automatically turn critical without hospital care.

There are so many ludicrous assumptions to get to an 8% mortality rate that it's a scenario that's not worth thinking about - not in relation to the Covid19 pandemic anyway.

JRPGfan said:
Teeqoz said:

So on the contrary to what you're suggesting, younger and more healthy people  actually seem to be quite underrepresented in our data.

I HOPE your right.
If the rest of the world is as lucky as Norway is currently (14 deaths out of 3,066 people infect = 0,4%) then that doesnt sound so bad.

I suspect your wrong, and this is more dangerous than norways data seems to indicate so far.

Time will tell.

So far, china has had this virus the longest, had the most cases, and tested the most.
They have over 4% of people that get it die.

In italy its over 10% of the confirmed cases, that got the infected that died.

Norways current ~0,4% death rate is very good though.

Suspicions, hope, guesses.

I prefer data. The data shows that 15% is ridiculous.

China has done far fewer tests per capita than Norway. South Korea is another country that has a good dataset, and which seem to have gotten some control of the epidemic. They currently have a ~1.3% Case Fatality Rate, and of cases that had an outcome, 3.31% died, though that number is dropping (just 4 days ago it was 4%. Recoveries take more time than fatalities).