D&D and the rising pandemic

[FrogReaver]

You had the hypothesis: "If 5 times the people have it than are confirmed as having it (presumably most of those cases would be milder) then you have a much lower mortality rate."

In other words, you are suggesting that far more than 80% of the cases are mild, and that in fact the outbreaks are 500% large than have been documented. This would suggest that perhaps 96% of the cases are mild, and we are catching only the 4% most severe and only like a 5th of the mild cases. And it's possible to believe that in say the USA where testing has been limited so far,

I'm with this so far.

but in cases like South Korea that pattern of the disease is in fact impossible. Because if the disease had that pattern, South Korea with its wide nets of testing would be detecting all of those mild cases.

So just read an article to find out how many they have tested compared to how many have been confirmed.

South Korea's rapid coronavirus testing, far ahead of the US, could be saving lives

It's an amazing response overall in South Korea!

But more on topic the article says they've tested 240,000 and found 8000 with it. They have had 800+ recoveries and less than 80 deaths. That's less than a 1% mortality rate (unless I'm calculating that incorrectly). If their hospitals get overloaded that number could start to drastically rise.

With an R0 that high and a disease spreading that widely, they'd get lower and lower numbers of negative tests. Most tests would be turning up positive. But that isn't being observed. Instead, as they casts their nets wider, they get fewer and fewer positives.

That's fair but from what I read it's suspected that even with their testing that they have about twice as many people with it than they have confirmed.

And again, in fact R0 was high enough that the disease was spreading everywhere sufficiently to see the real numbers of cases be 5 times that observed, then we'd see very different things than we see. The R0 is in fact fairly well constrained at this point. It might be 2.5, it might be 4, but it very unlikely to be wildly off that.

Or South Korea had it for a week or 2 before the first few cases were confirmed - which throws that timeline off a bit and thus the rate of infection values off as well - right?

Similarly, so is the mortality rate assuming sufficient care is available. It might be 2%. It might be 3.6%. But it's very unlikely to be wildly off that.

In South Korea it's looking like less than 1% at least for now?

Now, left untreated, we have very good reason so suspect that it's much higher than 3.4%. Those numbers assume that those requiring oxygen can receive it. If they can't, experience shows mortality rate shoots up over 5%. And of course, different demographics in a population will effect the observed rate as well.

Based on the South Korea model I'm not seeing how we get to 3.4%.

Now if hospitals get flooded and can't keep up I have no idea how high it could go. 5% might not be unreasonable in that scenario.

Well in the sense that we would expect early on in an epidemic if you have spotty documentation for cases to be about 240 times more than deaths if in fact the R0 is around 3, and the mortality rate is around 3.4, and the disease takes 5-7 days to incubate, and then another 12-14 days to become lethal, then yes.

However, the R0 and mortality rate that we have are being increasingly tightly constrained.

Seems to depend a lot on early social distancing especially of the most at risk populations.

 

[FrogReaver]

I don't think it is inaccurate, but I'd also like to not be as harsh on frogreaver as all that.

The questions he's having and the models he's offering up are indeed models that some real professional epidemiologists did offer up as possibilities earlier in the pandemics spread. And many of those claims did filter into mainstream journalism through editorials, interviews with experts, social media and so forth. Unlike a lot of theories out there, these aren't views of reality based on wild conspiracy theories.

Thank you.

I wanted to add though - I can't swear I wasn't influenced by some other expert on this. If I was I don't recall. These are questions and concerns I came up with on my own after reading the articles in question. It's nice to know they aligned with some initial theories though. I very much appreciate you taking the time to answer these questions and concerns instead of just screaming at me that some authority said X and therefore since I'm not an authority I can't question it and have no hope of independently understanding it - as too many on this thread are doing.

The high R0 low mortality rate model, or what you might call the optimistic model, was in fact very plausible and very much promoted by some experts back in January and early February. So it's entirely reasonable for a layman to latch on to it. In fact, in my experience with people, regardless of the level of their education, they tend to latch on to the first thing that is explained to them in a forceful and authoritative manner. They they go about defending that initial impression to the death (as it were). Indeed, even professionals who form these hypothesis act in very much the same manner, defending their model of reality even as new evidence appears.

And there is a huge mental cost to letting go of this particular mental model, even more so than the usual.

Yea that is a phenomenon. I don't believe that's what happened to me in this case. I'm still struggling to see how South Korea can have around 80 deaths to 800 confirmed cases and the projected mortality rate sit at 3.4%. I can see how this implies it's not a high R0 and low mortality rate situation but what explains their low mortality rate numbers in comparison to the higher 3.4% estimate?

Or is what is being claimed is that it's at 3.4% if we do nothing (pre hospital overload)?

 

[Dannyalcatraz]

It's not panic buying, it's greed and the desire to instigate and profit from the fears of the pandemic.

He isn’t alone, but this behavior seems to be the exception, not the rule. Most of the buying is people who are irrationally buying something to feel like they’re taking some kind of real prep for the pandemic.

The ones really profiteering off the panic are guys like Jim Bakker and Alex Jones. Both are being investigated for peddling fake Covid-19 cures right now- Bakker is selling colloidal silver as a cure at 25x its regular price; Jones‘ product is anti-coronaviruses toothpaste.

This, for example, is not how profiteers behave:

Amid coronavirus panic-buying, 2 injured in fight in Sam's Club wine aisle

The dispute began when a man in a motorized shopping cart bumped into a couple's cart, which was carrying their child. It ended with two men wielding broken wine bottles.

www.yahoo.com

 

[Celebrim]

I'm still struggling to see how South Korea can have around 80 deaths to 800 [sic] confirmed cases and the projected mortality rate sit at 3.4%. I can see how this implies it's not a high R0 and low mortality rate situation but what explains their low mortality rate numbers in comparison to the higher 3.4% estimate?

So, let's look at today's numbers for South Korea.

They have 8162 cases, of which only 76 have died. That makes for a transient case fatality ratio of 0.9%. (Notice that it has climbed by 0.3% since the article you quoted. This is not unexpected, and I think you'll soon see why.)

So how do we know that it will converge (or in this case climb) to 3.4%?

Well, so far of the 8162 cases, 834 people have recovered and are free from the illness. That means that there are 834 + 76 closed cases, or 910 cases that are complete. Of those 910 cases that are complete, 8.3% have died. That 8.3% number is the current mortality rate in South Korea. Recall that I told you that at the time the CFR was 0.6% the mortality rate was actually a scary sounding 24%. But it's decline down to at present 8.3% is also not unexpected.

What is going on here?

Well, there are 7253 cases of people that have neither gotten well nor died. Some of those people will get better, and some of them will die. When all the cases are closed, we'll know the final mortality rate - predicted to be around 3.4%. The two numbers are converging toward some number between them, and the reason they have the particular pattern they do is that this disease takes longer to recover from on average than it does to kill you. It usually takes 20-40 days to recover. But it usually only takes 12-17 days to die. So deaths tend to be front loaded resulting in a high mortality rate in the early going (governed by a ratio of deaths to recoveries). But transient case fatality ratios tend to be very low if you are finding the cases quickly, because the deaths lag symptoms by such a long length of time.

Or is what is being claimed is that it's at 3.4% if we do nothing (pre hospital overload)?

Oh no, it is MUCH worse if you do nothing. The highest observed rates have been around 5.9% in Wuhan itself where the medical system all but collapsed and they were forced to provide only triage to the most ill patients. In theory, if you could do nothing at all, the mortality rate could be much higher than that. About 10% of patients get very very ill and have to be on oxygen for 2-4 weeks. That's why the talk about ventilators is so important and everyone is talking about "flattening the curve".

(Now, all that said there are some weird data sets out there that seem to show you can get significantly below 3.4% mortality rate, though I don't think South Korea is necessarily one of them. However, the two data sets that I know of that suggest that have some huge problems.)

 

[FrogReaver]

So, let's look at today's numbers for South Korea.

They have 8162 cases, of which only 76 have died. That makes for a transient case fatality ratio of 0.9%. (Notice that it has climbed by 0.3% since the article you quoted. This is not unexpected, and I think you'll soon see why.)

So how do we know that it will converge (or in this case climb) to 3.4%?

Well, so far of the 8162 cases, 834 people have recovered and are free from the illness. That means that there are 834 + 76 closed cases, or 910 cases that are complete. Of those 910 cases that are complete, 8.3% have died. That 8.3% number is the current mortality rate in South Korea. Recall that I told you that at the time the CFR was 0.6% the mortality rate was actually a scary sounding 24%. But it's decline down to at present 8.3% is also not unexpected.

What is going on here?

Well, there are 7253 cases of people that have neither gotten well nor died. Some of those people will get better, and some of them will die. When all the cases are closed, we'll now the final mortality rate - predicted to be around 3.4%. The two numbers are converging toward some number between them, and the reason they have the particular pattern they do is that this disease takes longer to recover from on average than it does to kill you. It usually takes 20-40 days to recover. But it usually only takes 12-17 days to die. So deaths tend to be front loaded resulting in a high mortality rate (governed by a ratio of deaths to recoveries). But transient case fatality ratios tend to be very low if you are finding the cases quickly, because the deaths lag symptoms by such a long length of time.



Oh no, it is MUCH worse if you do nothing. The highest observed rates have been around 5.9% in Wuhan itself where the medical system all but collapsed and they were forced to triage only the most ill patients. In theory, if you could do nothing at all, the mortality rate could be much higher than that. About 10% of patients get very very ill and have to be on oxygen for 2-4 weeks. That's why the talk about ventilators is so important and everyone is talking about "flattening the curve".

So reading the first part of this response and I realized my mistake. 80 compared to 800 is 10% not 1%. DOH!

 

[Dannyalcatraz]

This gets a whole lot of “Nope!“ from me:

Beijing is promoting traditional medicine as a 'Chinese solution' to coronavirus. Not everyone is on board | CNN

Xiong Qingzhen, a drone engineer in the central Chinese metropolis of Wuhan, spent more than two weeks in a makeshift hospital in February receiving treatment for Covid-19, the respiratory disease causing a global health crisis.

www.cnn.com

 

[Zardnaar]

South Korea (and Taiwan) have good health systems. And a decent government response.

China lolbad, USA participation trophy.

 

[FrogReaver]

South Korea (and Taiwan) have good health systems. And a decent government response.

China lolbad, USA participation trophy.

U.S. did pretty good with travel bans. U.S. doing a crap job of testing.

 

[FrogReaver]

So, let's look at today's numbers for South Korea.

They have 8162 cases, of which only 76 have died. That makes for a transient case fatality ratio of 0.9%. (Notice that it has climbed by 0.3% since the article you quoted. This is not unexpected, and I think you'll soon see why.)

So how do we know that it will converge (or in this case climb) to 3.4%?

Well, so far of the 8162 cases, 834 people have recovered and are free from the illness. That means that there are 834 + 76 closed cases, or 910 cases that are complete. Of those 910 cases that are complete, 8.3% have died. That 8.3% number is the current mortality rate in South Korea. Recall that I told you that at the time the CFR was 0.6% the mortality rate was actually a scary sounding 24%. But it's decline down to at present 8.3% is also not unexpected.

What is going on here?

Well, there are 7253 cases of people that have neither gotten well nor died. Some of those people will get better, and some of them will die. When all the cases are closed, we'll now the final mortality rate - predicted to be around 3.4%. The two numbers are converging toward some number between them, and the reason they have the particular pattern they do is that this disease takes longer to recover from on average than it does to kill you. It usually takes 20-40 days to recover. But it usually only takes 12-17 days to die. So deaths tend to be front loaded resulting in a high mortality rate (governed by a ratio of deaths to recoveries). But transient case fatality ratios tend to be very low if you are finding the cases quickly, because the deaths lag symptoms by such a long length of time.



Oh no, it is MUCH worse if you do nothing. The highest observed rates have been around 5.9% in Wuhan itself where the medical system all but collapsed and they were forced to provide only triage to the most ill patients. In theory, if you could do nothing at all, the mortality rate could be much higher than that. About 10% of patients get very very ill and have to be on oxygen for 2-4 weeks. That's why the talk about ventilators is so important and everyone is talking about "flattening the curve".

(Now, all that said there are some weird data sets out there that seem to show you can get significantly below 3.4% mortality rate, though I don't think South Korea is necessarily one of them. However, the two data sets that I know of that suggest that have some huge problems.)

Very good explanation. Thanks.

 

[Celebrim]

South Korea (and Taiwan) have good health systems. And a decent government response.

Right now almost everyone doing a good job of this is in Asia, in part because I think they've been through this before with SARS and actually absorbed the lessons of that.

Nations/territories doing a decent job:
Taiwan, Hong Kong, South Korea, Japan, Israel, Singapore, Bahrain, Qatar, New Zealand

Surprisingly, if you get past all the initial denial, corruption, and cover up (and granted, it's really hard to get over that), China has done a really good job once they started to take it seriously. It's been authoritarian and brutal, but its been effective and many nations have done much worse.

Hitherto I might give C's to Germany, Denmark, Poland, and Norway. Germany has numbers that look a lot like South Korea. Denmark and Norway had the bad luck to be in the Schengen zone, but seem to be recovering from that. They have proportionately some of the worst outbreaks in the world, so things could get really bad for them for a while but they seem to be containing now which is impressive because the list of countries that appear to be is very small.

Pretty much the rest of Europe gets an F. That March 4th emergency meeting will go down in the history books as one of the worst civic reactions in the history of government. I had to pick myself off the floor when I read the outcome, and the more I thought about it the more I went from dumbstruck to angry. Well, enough of that. Not my country.

If you were holding a civil trial, China, Italy, and Iran would need to pay damages. Between the three of them, the rest of the world is screwed.

I was giving a D to USA up until about 5 days ago when we finally got serious about this. Final grade to be evaluated when the SHTF some time in the next week or two.