March 19, 2020

"Which Country Has Flattened the Curve for the Coronavirus?"

Check out the collection of graphs at the NYT (no subscription needed).

The answer to the question is China (according to the data China reports). Also, South Korea.

Not us. Not yet.

120 comments:

Ken B said...

Since you raised the topic, which country would be first to flatten the curve if they created the virus as a weapon?

The Chinese have flattened the curve through a severe restriction on mixing. We are doing the same.

BarrySanders20 said...

Lies, damned lies, and statistics. Test more, find more. Test fewer and find fewer. The curve only represents the collected data, which may or may accurately reflect reality.

Mike near Seattle said...

"According to the data China reports." Which probably deserves a prize for fiction.

clint said...

Our "known cases" number is probably about to spike, with more widespread testing.

I expect we'll see a lot of panicked analyses of the out-of-control spread of the disease over the next week.

Ken B said...

Coming soon: the new Althouse Alibaba Portal.

Ken B said...

It is unwise to trust the Chinese government or their numbers. However if they are right it suggests social distancing works.

narciso said...

when you spread it to the rest of the world, what is the Chinese word for chutzpah,

BarrySanders20 said...

If we are talking about a nationwide epidemic, the denominator should be 330,000,000 give or take 10,000,000.

How many people last week had it? Unknown, but some did because they were diagnosed.

How many this week? Unknown, but more people were diagnosed.

What is the rate of increase or spread? Unknown. The only semi-accurate comparison is the number of people diagnosed.

Ken B said...

Your commenters seem to think they understand simple facts about testing which professional epidemiologists do not. This is bizarre.

Leland said...

You don't flatten the curve, you only delay it, unless your plan is to do the "quarantine everybody" for months. China has already had mass exposure. The are on the downside of the curve.

Here's an analysis from the Diamond Princess, which has completed the curve. Note the population numbers on a ship and then the number of cases (symptomatic and asymptomatic).

narciso said...

why would one assume that big a number, as pointed out last night, the evidence for this aggressive mitigation strategy is very scant on the ground,

Fernandinande said...

Ioannidis sez:

"If the level of the epidemic does overwhelm the health system and extreme measures have only modest effectiveness, then flattening the curve may make things worse: Instead of being overwhelmed during a short, acute phase, the health system will remain overwhelmed for a more protracted period."

Achilles said...

This is a virus. It is far more widespread than people think. There are 14 day delays in actual numbers due to the incubation period.

The only time to stop this was in December and January and China threw doctors in jail and lied to the world.

When Trump threw down the travel ban that flattened the curve as much as anything possibly could have. But even that was really too late as the Chinese New Year cycle was over before then.

After that no country "flattened the curve." Some just panicked more. Some have more test kits. Some lied about the numbers.

Once a virus is in the environment it is pervasive.

Spring is going to flatten the curve.

We are going to find out a lot of people had COVID-19 and didn't even know it.

Ken B said...

Here is a simple puzzle.

1% of people with Wisconsin virus turn blue four days after infection.
The number of infections doubles every day.
Yesterday we saw a blue person, and today two more.
Estimate the number of infections.

Answer. The evidence suggests 4 days ago there 100 newly infected, that number doubling four times since. 100*2^4 = 1600. Three blues, 1600 cases. Exponential growth with latency. I bet Most of you got it wrong.

narciso said...

haven't seen him at the good prosecutors, they relish in misery,

eric said...

Btw,

South Korea didn't keep schools, malls, theaters, restaurants, cancel events, etc.

And yet, they're fine.

I know the, "We're all gonna die!" Crowd will come up with excuses as to why (other than, this isn't as bad as we are being told it is).

But this should make many people angry. They hyped this thing and ruined a lot of lives, for nothing.

CStanley said...

Here's an analysis from the Diamond Princess, which has completed the curve. Note the population numbers on a ship and then the number of cases (symptomatic and asymptomatic).

The group on the Diamond Princess did not “complete the curve”. They got off the ship.

Balfegor said...

Re: clint

One question I have is for symptomatic carriers, what was the lag between appearance of symptoms and detection. That should give us a ballpark on what the "actual" infection spread looked like, as opposed to the "detection" spread, which is what I think we're seeing now. Are these cases all from the past few days? From a month ago?

Also, I'm seeing reports that only 64 US cases of 10,000 are serious/critical, but that seems inconsistent with the stories I am also reading about how hospitals in NYC are at the edge of their capacity. Are they burning up all their supplies and personnel just dealing with "mild" cases?

narciso said...

this provides some perspective,

Quaestor said...

Also, I'm seeing reports that only 64 US cases of 10,000 are serious/critical, but that seems inconsistent with the stories I am also reading about how hospitals in NYC are at the edge of their capacity. Are they burning up all their supplies and personnel just dealing with "mild" cases?

Normal behavior for a city that elected Bill DeBlasio as its mayor and sent AOC to Congress.

Lurker21 said...

If you trust China ...

I'm not sure I trust Russia either ...

I definitely don't trust George Conway.

"Israeli Man Walks Dog With Drone" was pretty good, though.

AZ Bob said...

China? You mean the country that covered up the spread of the disease in the first place?

Known Unknown said...

The panickers say we are going to be Italy, but I don't understand why we can't be Germany?

narciso said...

same the number of cases, are double that finding, which is possible, is that a reason to put the entire economy into hibernation,

mockturtle said...

Wow, let's hear it for China! What a testament to sound epidemic management! Maybe Xi will win the next Nobel Prize for Medicine! [The prize can't go to the Director of the Wuhan hospital because he is, unfortunately, dead].

Yancey Ward said...

"Also, I'm seeing reports that only 64 US cases of 10,000 are serious/critical, but that seems inconsistent with the stories I am also reading about how hospitals in NYC are at the edge of their capacity. Are they burning up all their supplies and personnel just dealing with "mild" cases?"

I think a lot of the deaths each day are not coming from ICUs- they are coming from people who died at home. If you look at the data of the people who have died of COVID-19 in Italy, they are extremely old- such people often choose to die at home than in a hospital, even if they might believe they could buy another year or two with intensive care. Also, a lot of them who do go there at first just demand to go home to die if that is their fate- I saw a family friend in just this kind of circumstance last month (he died of pneumonia, but had it off and on for 6 months before demanding they stop treating him). It is usually the other family members that put pressure on them to go to the hospital against their wishes.

Yes, in a panic, it is possible that doctors and patients are over-utilizing hospital beds for cases that don't actually need it- could be easily treated with in home care, and probably more safely from the perspective of every other non-Coronavirus patient. It is the precautionary principle run amok from both the patient side and the doctor side.

Francisco D said...

We are going to find out a lot of people had COVID-19 and didn't even know it.

That is a popular opinion here in Tucson. My wife thinks she had it in February (very good symptom match) and her running club buddies have similar stories

tim in vermont said...
This comment has been removed by the author.
Yancey Ward said...

It looks like Italy could be at or within a day of the peak- the 2nd derivative has changed based on the 3 day moving average, but still waiting for today's data dump, which usually arrives about 2-3 p.m. EDT. Also, Spain might be nearing it, too, but you need at least the next three days of data to make firm prediction.

stevew said...

The widespread effort to make me see this as a grave threat and to panic, as they all have, continues to fail. I'm looking at the same data as everyone else and just not seeing a justification for these action, which are increasingly imposed on me.

Balfegor said...

Re: eric:

Btw,

South Korea didn't keep schools, malls, theaters, restaurants, cancel events, etc.

And yet, they're fine.


A bit of yes and a bit of no. Schools are closed -- in fact, the closure was extended for a few more weeks just recently:

The start of the new school year has been postponed for another two weeks until April 6. Daycare centers will also stay closed until then.

It is the third postponement. Education Minister Yoo Eun-hae on Tuesday said, "We decided to delay the opening of schools further at the recommendations of the Korea Centers for Disease Control and Prevention and other experts."


Large events have also been cancelled, e.g. a major cherry blossom festival and Seoul Fashion week. Malls haven't closed across the board, but there have been repeated closures as employees have been found to be infected and the facilities have gone through rounds of disinfection (anecdotally, I had heard that malls were totally abandoned in February, although people are probably venturing back now). Churches were cancelling Sunday services.

That said, a key element of Korea's (largely successful) struggle to regain control of the outbreak has been their dystopian public tracking of infected persons. I don't think that would be possible here -- Americans aren't going to stand for the government publicising their detailed GPS location history that way (especially not the way Korean netizens have been mocking the people whose private lives have been exposed through this system). Or at least, I would think Americans would rebel against that. But in a lot of ways Americans evince less real concern about privacy than foreigners, so maybe I am misreading the situation.

tim in vermont said...

"That is a popular opinion here in Tucson. My wife thinks she had it in February (very good symptom match)”

From the recent presser 90% of people with severe symptoms tested tested negative for ChiCom Flu.

tim in vermont said...

"I'm looking at the same data as everyone else and just not seeing a justification for these action,”

Sigh...

tim in vermont said...

If we all agree to be tracked by the FBI and NSA as if we were running against Hillary, then we could have the flat curve they have in Korea.

Achilles said...

Ken B said...
Here is a simple puzzle.

1% of people with Wisconsin virus turn blue four days after infection.
The number of infections doubles every day.
Yesterday we saw a blue person, and today two more.
Estimate the number of infections.

Answer. The evidence suggests 4 days ago there 100 newly infected, that number doubling four times since. 100*2^4 = 1600. Three blues, 1600 cases. Exponential growth with latency. I bet Most of you got it wrong.



That is the thing. People keep throwing out the word exponential.

All of this is being done to cause harm.

People should watch this video.

It does a good job.

The moral of the story is many have probably been exposed already.

tim in vermont said...

"It looks like Italy could be at or within a day of the peak- the 2nd derivative has changed based on the 3 day moving average, “

After a national lockdown.

Yancey Ward said...

Germany and the US are not going to see confirmed cases peak before the end of next week- both are still in the testing ramp-up phase- most tests, more confirmed cases.

tim in vermont said...

If the cholorquinine works, we can let the kids live their lives and restart the economy. That would be great.

Yancey Ward said...

Aunty Trump, you had better hope it has nothing to do with a national lockdown, because this will just happen again when it is lifted.

tim in vermont said...

"because this will just happen again when it is lifted.”

That’s what I expect, and what the modeling shows. This is like war except there is nobody to negotiate with and there will be no Christmas Truce.

narciso said...

germany has 1/5th our population, so it would stand to reason their numbers would not be nearly the same, there are some cases in the middle east, outside of iran, but few in sub-Saharan Africa, maybe it will get there, but considering their population densities,

Jersey Fled said...

I'm simply amazed at the number of posters here who seem to be totally invested in the worse case "we're all going to die" scenario.

Carona virus is still a virus. Past corona viruses included MERS and SARS. 858 people died from MERS. 774 people died from SARS. In fact, corona viruses are pretty common. According to WebMD:

"Almost everyone gets a coronavirus infection at least once in their life, most likely as a young child."

This virus will run its course, just like every other flu outbreak. So far the U.S. experience looks far more like South Korea than China or Italy.

Wash your hands, don't sneeze on people, avoid crowded places if you can, and get on with your life. If you get the virus, you will likely have a mild case, and very few of you will die.

Be smart, and don't panic. And stop acting like Chicken Little.

Curious George said...

" Francisco D said...
We are going to find out a lot of people had COVID-19 and didn't even know it.

That is a popular opinion here in Tucson. My wife thinks she had it in February (very good symptom match) and her running club buddies have similar stories"

My boss in Scottsdale also.

Ignorance is Bliss said...

Epidemiologists predicted that deaths would rise exponentially unless we took significant measures to slow the spread.

Deaths in the US are currently rising exponentially.

I'm happy to listen to anyone who thinks this is being over-hyped, if they also predicted that deaths would rise exponentially.

If you didn't predict that, in writing, what would lead me, or anyone, to believe you know better than the epidemiologists?

stevew said...

Seems it is agreed that stopping the spread of the virus and letting it die out are impossible goals. Then the stated goal is to flatten the curve of the spread of the virus and thereby prevent the healthcare system from being overwhelmed, presumably saving lives. I would suggest that inciting a panic is the wrong way to accomplish that goal.

Jersey Fled said...

My wife is also convinced that she had it about two weeks ago. So far, I've had nothing.

I'm Full of Soup said...

So far 133 confirmed cases in PA which has 12.6 Million people. That is about 1 case per 100,000. Are we scared yet?

Jersey Fled said...

Deaths from influenza do not rise exponentially forever, only in their early stages. Then they flatten out and eventually go away.

Otherwise we would all be dead from last year's virus.

Achilles said...

Aunty Trump said...
"I'm looking at the same data as everyone else and just not seeing a justification for these action,”

Sigh...

The numbers just aren't there.

Orders of magnitude more people are dying right now of the flu.

Orders of magnitude more people died of swine flu.

The numbers being reported just do not match how these viruses spread. If this virus actually has a higher r0 than the flu we have all been exposed.

I believe it does have a higher r0. I believe we have had wide exposure. I believe isolating certain populations makes a lot of sense. But COVID-19 will not even sniff swine-flu numbers or flu-season flu numbers.

I believe a lot more people are going to die from the coming food and product shortages than from COVID-19.

This will happen on the margins where you all wont have to acknowledge the costs of these actions.

narciso said...
This comment has been removed by the author.
I'm Full of Soup said...

Don't we give the CIA about 50 billion a year to uncover stuff like this and give us a heads up ? They missed 911, ISIS and now this. W hat good are they?

narciso said...

1/1000 of one percent,

tim in vermont said...

"Don't we give the CIA about 50 billion a year to uncover stuff like this and give us a heads up ? “

They were buys with a coup against Trump at the time, so cut them a break.

mockturtle said...

I haven't seen the panic. Haven't seen any hysteria about the virus. Bhe media keep telling people that the stores are being cleaned out so people are scrounging up what's left in fear of being without.

tim in vermont said...

I don’t read your posts anymore Achilles. I noticed that one because my handle was right next to yours. Just so you know.

Achilles said...

Jersey Fled said...
My wife is also convinced that she had it about two weeks ago. So far, I've had nothing.

You were asymptomatic. A carrier. There were a lot of us.

But at that time democrats were busy with impeachment and calling Trump's travel ban on China Xenophobic and Racist.

Jersey Fled said...

I calculated a few days ago that if Italy's death rate continued to increase exponentially (R=0.5) everyone in the country would be dead in a few hours short of 34 days.

I'm betting that isn't going to happen.

Any takers?

Achilles said...

Aunty Trump said...
I don’t read your posts anymore Achilles. I noticed that one because my handle was right next to yours. Just so you know.

That's good.

Sheep don't like numbers.

mockturtle said...

Don't we give the CIA about 50 billion a year to uncover stuff like this and give us a heads up ? They missed 911, ISIS and now this. W hat good are they?

None whatsoever, Soup. Get rid of them. Think of the $$$ we'll save [actually, their real budget is classified information and could be much more]. Drain the swamp!

narciso said...

is the panic in the store shelves, proportional to the epidemic? is flatlining the entire economy the right thing to do?

mockturtle said...

Achilles asserts: Sheep don't like numbers.

If that's so, Achilles, why do we count them when we're trying to sleep? ;-)

tim in vermont said...

"My wife is also convinced that she had it about two weeks ago. So far, I've had nothing.”

Like I said, 90% of the people who have symptoms have been testing negative, but I am sure your psychic powers are much stronger than those other self deceiving fools.

This is the most interesting study in denial, if it is nothing else.

David53 said...

As reported by many news outlets today,

"China reported no new local infections for the first time since the coronavirus crisis began three months ago"

Do you believe China?

tim in vermont said...

I have a feeling that what we are seeing is a divide between those who have taken college calculus and those who have not.

Francisco D said...

This is the most interesting study in denial, if it is nothing else.

What is your problem, Aunty Trump?

You seem to be convinced that your guess/opinion is the only one that counts.

I'm Full of Soup said...

Yesterday in Philly, they set up tents to administer tests. I wish someone would give us a simple daily recap with # of tests given and # of confirmed cases. How hard would that be?

chuckR said...

Aunty -
Not just calculus but also a course in (how to mislead with) statistics. A plot of new cases per day per capita identifies the countries with the biggest problems. These include Iran, Spain and Italy.
You can change the impact of the plots by using linear-linear or log-linear or log-log scales. The last flattens everything, the first is extra alarming. Powerline blog has a good series of posts starting with an FT statistician's log-linear presentation.

GingerBeer said...

Professor, On Jan. 14th WHO announced "...the Chinese authorities have found no clear evidence of human-to-human transmission of the novel #coronavirus (2019-nCoV) identified in #Wuhan, #ChinaFlag of China." What is in the bottled water in Madison, WI that would cause you to accept at face value anything the Communist Chinese government says?

https://twitter.com/WHO/status/1217043229427761152

Achilles said...

Aunty Trump said...
I have a feeling that what we are seeing is a divide between those who have taken college calculus and those who have not.

I have a feeling we are seeing a divide between people who know that there are trade offs in any situation and sometimes there are no good answers.

I also have a feeling we are seeing a divide between people who can deal with criticism and people who can't.

People are going to die every day.

More people died of the flu today than from Corona Virus.

More people will die of the flu tomorrow.

More people will die of the flu this year.

If COVID-19 was moving through Australia and Africa and India the way it is following specific temperature and humidity bands around the globe there would be more room for concern.

But it just isn't.

Achilles said...

narciso said...
is the panic in the store shelves, proportional to the epidemic? is flatlining the entire economy the right thing to do?

Some people will not accept responsibility for their actions.

That is a bigger number than those who will die of COVID-19.

Known Unknown said...

I'm sure psychosomatic cases will reach approximately 99% of the population within days.

tim in vermont said...

"You seem to be convinced that your guess/opinion is the only one that counts.”

Sorry, I am happy to hear criticism based in logic and/or mathematics, even discussions of assumptions, especially discussions of assumptions. Criticisms based on feelings or “senses”? No. Arguments that are based on insult and invective? Not so much. Anecdotes about Aunt Sally’s bad flu? Don’t even start.

If my opinions are so weak, it seems like there must be dozens of people with better training than my own in mathematics and statistics lined up to knock me down with valid arguments. They don’t because no sensible person is going to advocate risking hundreds of thousands of extra lives lost when the facts aren’t known based on a “feeling” that this has been around forever.

Of course there are assumptions involved that may turn out to be wrong, but we need to be on guard against wishful thinking now more than ever.

Known Unknown said...

"I calculated a few days ago that if Italy's death rate continued to increase exponentially (R=0.5) everyone in the country would be dead in a few hours short of 34 days."

There are more than 34k cases in Italy. Probably over 100k easy, so their death rate is much lower than what is represented. It's probably not 7% but a much lower 1-2% range. Germany's rate is 0.29%.

elkh1 said...

Korea, I wholeheartedly believe.

China? Want to buy a bridge in Brooklyn? They are re-closing their towns, and re-opening their temporary hospitals.

Known Unknown said...

"Yesterday in Philly, they set up tents to administer tests. I wish someone would give us a simple daily recap with # of tests given and # of confirmed cases. How hard would that be?"

This is government work we are talking about. Prepare to be underwhelmed.

tim in vermont said...

I blame the global warming alarmists who refused to acknowledge the obvious weaknesses in their arguments, and their fan boi press coverage, that thought that covering the field really meant “covering for” the field, in a lot of ways.

They squandered their credibility. So has the press. It would be nice to have back, but they don’t deserve to be believed.

Browndog said...

What is your problem, Aunty Trump?

You seem to be convinced that your guess/opinion is the only one that counts.


No shit.

Remember a coupled days ago when everyone said "Hey, this is a brand new virus, Nobody knows how it's going to act, and we won't know a damn thing until there's widespread testing"

Since then apparently we've all agreed that no matter what the first thing we should do is destroy our economy and put everyone out of work.

tim in vermont said...

Speaking of bullshit media, Bloomberg just said that Trump contradicted his FDA chief when no such thing happened. Trump said that we were going to be testing medications that have already been approved on the virus and they said that he said that the medications were approved to treat the virus.

What bullshit. Still going on right now as “breaking news”

"Lie lie lie lie lie lie” - Bloomberg

Ken B said...

I see comments have disappeared. A claim was made that chloroquine was not approved by the FDA for covid19. That is false. The drug is already approved and can legally be prescribed for covid19.

tim in vermont said...

OH my f*cking god! Either this guy is a fucking moron, Josh Wingrove, or he is a liar.

tim in vermont said...

"Thao Nguyen
@helloitsthao
·
42s
trump’s approval on handling of the chinese coronavirus is surging in yougov polling

disapproval has dropped 10 points in the two weeks”

Now the real panic begins. I think this is why Bloomberg has gone back to anti-trump propaganda, they are shitting a brick.

https://twitter.com/helloitsthao/status/1240706433395507202

pacwest said...

is the panic in the store shelves, proportional to the epidemic? is flatlining the entire economy the right thing to do?

A day late and a dollar short narciso. Too late to deviate from the path we are on now. It's just a matter of how long. Best case May. Government estimates August. Very best case is everyone on this board apologizes to Achilles in a couple of months.

Ken B said...

I agree with Achilles that this is an excellent video on exponential growth and infections. I repeat his link

https://www.youtube.com/watch?v=Kas0tIxDvrg

I am just amazed that Achilles thinks it vindicates his complacency. The nice thing about the video is how well it shows reducing Contacts, exactly what Achilles opposes.

mandrewa said...

Ken B. said,

"Here is a simple puzzle.

1% of people with Wisconsin virus turn blue four days after infection.
The number of infections doubles every day.
Yesterday we saw a blue person, and today two more.
Estimate the number of infections."

Ok.

The underlying sequence is 1, 2, 4, 8, 16, …

Where each number in the list is that day's new cases.

Alternatively we could write that as 2^^n , where n is the nth day of the epidemic, and 2^^n is the number of new cases on that day.

2^^0 = 1, which is the first case on the 0th day, or the start of the epidemic.

Ken, your simple problem assumes that we have continuously had good data on the epidemic since it began. This "simple problem" therefore doesn't reflect the reality we face.

But to continue, since in the simple problem we know exactly when a blue person appears, and for each blue person, that means a hundred were infected, and for each new blue person, a hundred people were actually infected 4 days previously, we can, with but two data points and the knowledge of the exponent (which is simply given to us in the problem), figure out where we are on the sequence given to us.

The two data points are for day n, when a 100 people were newly infected, and for day n+1, when 200 people were newly infected.

So where does that put us on the sequence 1, 2, 4, 8, 16, 32, 64, 128, 256, …. ?

Well we can just read it from the sequence. n is pretty close to 8 as on the 8th day there are 128 new cases and on the 9th day there are 256 new cases. (To match it more precisely we have to find the hour of the 8th day that matches a 100 new cases.)

Note that this simple problem implicitly assumes that the epidemic started 13 days earlier.

Note that the total number infected (which would include the dead, the sick, and the recovered) would be 1+2+4+8+16+32+64+128+256+512+1024+2048+4096, or 8191. Plus or minus a couple of thousand, because I haven't figured out the hour of the day.

Oddly enough, this does not match your estimate of 1600 infections.

Probably I made a mistake. Would you explain?

Now I studied this sort of thing a long time ago. I think this was what we used to call pre-calculus. And I did not remember all of this just now. I had to look up some of the details. The benefit of having studied it previously, though, is that it all comes back.

Now as I have said before, I don't there think there is any doubt that there is an exponential series underneath all of this. But, I do think there is some doubt about what that exponent is, even though I think it should be possible to estimate it. I also have some doubt that it is really being accurately estimated.

I mean someone somewhere is surely doing so. But it isn't necessarily the people being cited by the media.

Ken B said...

Aunty Trump
Trump is right. Once a drug is approved for use then except in special cases it is approved for use on other diseases. Chloroquine was approved decades ago. It can be legally prescribed for flu, colds, piles, ... and covid19.

Ken B said...

Mandewa
Don’t be an ass. It’s a puzzle to make a point about exponential growth not a model of covid19. Nothing happened on the previous days. People are yelling, only X deaths so it’s not a big deal. Wrong.

tim in vermont said...

""Here is a simple puzzle.”

What you have to remember is that at the beginning, virtually none of the people you encounter have been exposed to the virus so it skyrockets, but as the infected, and subsequently recovered population grows, each new case has a harder time finding unexposed people to infect, then you let the old people out of the attic.

That’s the British strategy. It might have been the best strategy scientifically, if you were dealing with livestock and not people in terms of fewest deaths at the end of the day. I like the US strategy of building capacity and delaying the peak.

Francisco D said...

Aunty Trump wrote: If my opinions are so weak, it seems like there must be dozens of people with better training than my own in mathematics and statistics lined up to knock me down with valid arguments. They don’t because no sensible person is going to advocate risking hundreds of thousands of extra lives lost when the facts....

I have a great deal of training in social science statistics and taught grad-level stats 30+ years ago. An experienced statistician evaluates the validity and reliability of the data before trying to analyze it. The numbers are meaningless when there are systematic biases in data collection. Hence all subsequent analyses are meaningless (i.e., garbage in, garbage out).

I don't get the sense that you understand that. In the absence of reliable and much more complete data (that we will have when this is over), we are all just spitballing. Science often works that way to develop hypotheses that are later tested.

mandrewa said...

Ken B.,

If day 0 of the epidemic for the United States was four months ago, or in other words 120 days ago, back in December, and on the 110th day, we found 20 new cases, but those 20 new cases were not what actually what happened on the 110th day (we can add in a delay or not, it is still the same point), but rather a sum of new cases that actually started on the 90th, 91st, 93rd, and 96th days, and a similar thing for the 111th day, where we are actually mostly catching up with cases that we were previously not aware of, and you try to estimate what the underlying exponential series is by assuming that all these new cases really go with the 110th and 111th days, then we are going to get a much larger exponent than the underlying series actually has.

Matt Sablan said...

I think the most frustrating thing is that this is a no-win situation.

If the situation doesn't turn out nearly as bad as we were told, we'll never know how necessary the actions were, and both sides will say, "See? We didn't need to do anything/Our actions saved countless lives!"

It doesn't help that back in February I was being told by March, we'd be making choices on who lives and dies, with people dying in the streets. There was way too much early panic, which was incongruous with people's reactions to things like the travel ban, that I think there'll never be a good way to know what actually was the right response.

Operaman said...

The skepticism about the numbers of reported cases is warranted but not for the reasons that most comments suggest. We are, unfortunately, unable to test most of those who are manifesting symptoms. The majority of the sick seeking tests are turned away, even if, by clinical examination, a person has advanced symptoms unless they are elderly or meet other triage criteria, which vary widely.

Consequently, the actual number of infected persons in the U.S.is almost certainly far higher than is currently reported. Thankfully, the government is finally acting on this reality. At this point, however, given the spread of the virus, even with extreme measures such as those put in place in the Bay Area, we may be dealing with the virus for another year or more until a vaccine or other treatments can be implemented.

Taking a cavalier approach - I.e. “it’ll pass like viruses do and a few will die but not me or you” - will kill a lot of people. Yes, avoiding the maximum number of deaths will devastate our economy, which in turn will affect even more people than the virus. This dilemma leaves us with no good options. Personally, I favor saving lives. We can band together to mitigate the economic hardships but, until we have a vaccine, we have no means of saving ourselves from the worst - a few million deaths - except for stringent measures to delay the spread.

Francisco D said...

Known Unknown wrote: I'm sure psychosomatic cases will reach approximately 99% of the population within days.

It is allergy season here in Tucson, and sometimes I wonder if sniffles means that I have COVID-19.

My wife and I test drove a 4Runner yesterday (with all appropriate precautions). The no-pressure Toyota salesman was a nice mature man from the ME. When he sneezed (outside the window) it was like someone farting loudly and crapping their pants.

We used up a lot of Lysol disinfectant wipes that day

Meade said...

"It is allergy season here in Tucson, and sometimes I wonder if sniffles means that I have COVID-19. "

My understanding is that it is just the opposite, no sniffles is what should worry you. Dry cough, shallow breathing, feeling easily winded, spiking fever over 100F — these are the most common symptoms of COVID-19.

D.D. Driver said...

"That’s the British strategy. It might have been the best strategy scientifically, if you were dealing with livestock and not people in terms of fewest deaths at the end of the day. I like the US strategy of building capacity and delaying the peak."

Time will tell. I will say if we spiral into a global depression (which for the first time in my life seems possible) because we overreact the death, suffering, and misery that the world endures might make the time when we were panicking about corona virus seem quaint. Right now we have ridiculous calls for martial law and a total shutdown of the US for a month. I think Trump will resist this silliness but I wish I had 100% confidence.

We won't know for years, but my personal suspicion is that the UK did the right thing. As an aside, as each country is rebuilding the its economy, all the folks accusing the Brexiters of being a bunch of racists will be (at least secretly) relieved that the UK is not burdened down by the EU. And if we hit a global depression I wonder how many Bernie Bros will be happy to finally make progress on reducing wealth inequality, but now I'm just ranting.

WhoKnew said...

The whole point of the government crashing the economy by restricting any gatherings and closing tens of thousands of businesses is to prevent the hospitals (especially the ICU units) from being overwhelmed. Therefore, it seems to me the most important numbers that the media should be reporting are those that tell us how many hospital and ICU beds are being taken up by Wuhan flu patients. But good luck finding that number in the news.

Meade said...

"Dry cough"

I should add: dry unproductive continuous cough.

(Note: I am a retired landscape professional, not a doctor. For medical diagnosis and advice, consult a licensed physician.)

mockturtle said...

When he sneezed (outside the window) it was like someone farting loudly and crapping their pants.

Maybe it was!

Achilles said...

Ken B said...
I agree with Achilles that this is an excellent video on exponential growth and infections. I repeat his link

https://www.youtube.com/watch?v=Kas0tIxDvrg

I am just amazed that Achilles thinks it vindicates his complacency. The nice thing about the video is how well it shows reducing Contacts, exactly what Achilles opposes.


You are not applying the math being proposed.

This virus started in November.

Doctors were being jailed in December.

China made some moves to limit the Chinese New Year celebrations but they were not substantial.

All those people flew home in February. They were in the US for 2 14 day incubation periods before Trump dropped the travel ban the democrats called xenophobic.

It was during this period that everyone returned to Italy from the Chinese New Year celebrations. This pattern repeated all over the world.

Looking back at the timeline if I had to guess whether I got COVID-19 or not looking at where I work, who I work with, and everything else I would bet that I was at least exposed. Millions of people in the US have been exposed. I am not going to be at my Dad's 70th birthday next weekend.

But closing down all "non-essential" economic functions has costs too. There will be famines and hard times on the margins.

Nothing people here need to worry about.

But bad things happen everyday and sometimes you choose the less bad things.

Like we did when 100,000-500,000 people died of swine flu.

Or 300,000-700,000 people die of the flu every year.

The real damage is when COVID-20 is actually bad and credibility of these institutions is all gone.

mandrewa said...

Aunty Trump said, If the cholorquinine works, we can let the kids live their lives and restart the economy. That would be great.

Chloroquine, or more likely hydroxychloroquine, is not going to be the magic bullet that stops this in it's tracks.

We know this because South Korea has been treating most of their patients with hydroxychloroquine almost from the beginning.

But hydroxychloroquine may be a part of the reason that South Korea has a fatality rate of 1% (or maybe really 0.5% if we could accurately count all the infections) rather than the seemingly much higher fatality rates of Iran and Italy.

But hydroxychloroquine by itself isn't going to stop everyone who has Wuhan Coronavirus from dying.

And by the way, maybe we should let the kids restart the economy anyway. They are not seriously threatened by this either way.

Francisco D said...

Dry cough, shallow breathing, feeling easily winded, spiking fever over 100F — these are the most common symptoms of COVID-19.

My wife had those exact symptoms in February. She was really sick for a week and not steadily back on her feet for another two weeks. Now she is back running. Several members of the club reported the same symptoms.

She has a past history of childhood asthma, but without that COVID-19 symptom cluster. Her asthma medication helped her breath just before I was going to take her to the ER in mid-February.

My symptoms are just like every Spring but with obsessive worry and hand washing.

Spiros said...

I think the real danger is a second, deadlier wave of Covid 19 (think 10 to 20% fatality among under 30s, with strong immune responses causing "cytokine storm syndrome"). The Spanish influenza came at us in waves. Thankfully, a small study suggests that macaque monkeys can't become reinfected with Covid 19. And there are, at least a few, scientists who believe that an increase in temperatures and relative humidity (i.e., spring time) will substantially lower the virus’s transmission. Other scientists are speculating that only a few thousand deaths will occur in the United States. I just don't think the virus will pass like a destructive wave around the world. I think we're nearing the end. That's probably bad news for the media (who want to inflict pain on the country for the audacity of electing a slob like Trump).

Francisco D said...

"When he sneezed (outside the window) it was like someone farting loudly and crapping their pants."

Blogger mockturtle responded... Maybe it was!

Well, we didn't sniff the air for obvious reasons.

In sympathy, he was of an age when farting and sneezing have unintended consequences.

rcocean said...

I wouldn't trust anything the ChiComs say. Korea I would trust. I believe Singapore has been VERY successful in containing the virus.

Achilles said...

rcocean said...
I wouldn't trust anything the ChiComs say. Korea I would trust. I believe Singapore has been VERY successful in containing the virus.

Africa was just as successful as Singapore and India were and for the same reasons.

tim in vermont said...

"I don't get the sense that you understand that. In the absence of reliable and much more complete data “

Did you miss this part of the post you quoted?

" when the facts aren’t known”?

We have the data we have. We have to make the most prudent judgements we can from that data. Or we can just cast the fate of hundreds of thousands of Americans to the wind and hope for the best.

Your argument, I guess, is to do nothing until it’s all over and we can figure out what happened.

tim in vermont said...

"My wife had those exact symptoms in February.”

Very scientific! You know that the figure given from the news conference today was that 90% of the people with symptoms that were tested for it didn’t have it? But on this site, everybody, it would seem, has already had it and these moves are all stuff and nonsense.

“Isn’t it pretty to think so.” - Hemingway.

Lucien said...

Would it not be possible to estimate prevalence of the disease by testing random population samples? How long do people continue shedding virus after infection if: a) they become symptomatic and then recover, or b) they never become symptomatic?

tim in vermont said...

""I don't get the sense that...”

Anybody who read my earlier posts knows why I am laughing out loud right now.

Francisco D said...

@Aunty Trump

Being a word Nazi is not the same as making a useful counterargument.

My "sense" was about your understanding or lack thereof. I cannot verify that you do not understand my perspective, but based on your comments it is my strong inference.

Do you like that word better?

Please feel free to debate my point about systemically unreliable and invalid data and whether it is useful to analyze it. That would be more to the point.

mandrewa said...

Lucien said, "Would it not be possible to estimate prevalence of the disease by testing random population samples?"

Exactly, Lucien, exactly.

I think the problem is that they don't have enough tests. The tests are rare enough still that they feel like they have to spend them on people who are sick and who might have the coronavirus, rather than expend a hundred thousand or so tests on a random sample to estimate the percentage of the population that is infected.

tim in vermont said...

"Please feel free to debate my point about systemically unreliable and invalid data and whether it is useful to analyze it. “

Please feel free to explain how we are supposed to manage this crisis while waiting for perfect data. This isn’t an academic exercise.

Italy sure seems to think it’s a crisis. I know, Italy is very inconvenient, let’s talk about anything else.

Paul said...

China???? And Ann, you BELIEVE the Chinese??? Hahahahahaha.... Might as well also believe North Korea has zero Coronavirus. And I have a bridge to sell you!

tim in vermont said...

"Exactly, Lucien, exactly.”

I am not saying you are wrong. I am just saying that you should read this. It comes into play even if the test is 99% accurate. I can’t find the accuracy stats for the various tests going on right now in the various countries for ChiCom flu, so it’s just something to think about.

https://en.wikipedia.org/wiki/Base_rate_fallacy#False_positive_paradox

tim in vermont said...

"Might as well also believe North Korea has zero Coronavirus.”

And there are no gay men in Iran.

mandrewa said...

"I am not saying you are wrong. I am just saying that you should read this. It comes into play even if the test is 99% accurate. I can’t find the accuracy stats for the various tests going on right now in the various countries for ChiCom flu, so it’s just something to think about."

That was the issue I wondered about when Ann Althouse wrote a few weeks ago about the researcher studying influenza who had all these throat swab samples and then tested them for the coronavirus. It was inevitable that she would have false positives, and unless she had a good way to deal with that issue, that may have been all she found.

We can get around part of the problem of false positives by retesting anyone identified as having the virus a second time. Two positive identifications would greatly raise the odds that they actually had it. And as long as only a small percentage of the population has it, this does not take that many more tests.

It seems to me that false negatives are a bigger problem and harder to get around.

MikeR said...

We can't know yet if we've bent the curve, not until we at least have testing surveys of the country. It takes a week or two for changes to show up.

Fernandinande said...

Most of you got it wrong.

LOL you sure got it wrong. Or better put "not even wrong".

Achilles said...

Aunty Trump said...

Please feel free to explain how we are supposed to manage this crisis while waiting for perfect data. This isn’t an academic exercise.

We should look at infection patterns, infection rates, death rates and known information and weigh costs and and benefits rationally.

We know what the numbers are from the Diamond Princess. 100% of the people were tested. ~3700 people. Just under 700 contracted the disease, 7 died all over 70 with pre-existing conditions. Enclosed environment and universal exposure.

We know what ages were most affected. We have an idea of incubation times. We know that Coronviruses persist strongly in air-conditioned environments.

We also know it spreads particularly fast in a certain temperature and humidity band and it has been in the US in high numbers since the Chinese New year travelers returned.

Given a 14 day incubation period there should be (N * r0)^4 of total cases before the March 5 travel ban to China as all those travelers returned from China in January/February. I am not sure how many thousands of people traveled to China for the new year. If r0 is over 5 that is a stupid large number now that we are at ^8-10.

A lot more people in the US have had COVID-19 than is reported. Orders of magnitude more people.

Death Rates in certain populations are higher than normal flu. Those people need to be isolated. I am telling my parents to stay away from their grand daughter who is going to school every day and traveling to see her Mom on weekends.

The rest of us need to wash our hands and go to work because this happens every year.

If the rates being quoted to us are accurate I would not be surprised if infection totals for COVID-19 in Seattle were already approaching flu totals this year.

If they are not accurate then they are lying.

Either way the only thing we actually know is how many people are dying.

We are not a belt and road country. Italy is. Their per capita N was much higher than ours.

Francisco D said...

Aunty Trump said: Please feel free to explain how we are supposed to manage this crisis while waiting for perfect data. This isn’t an academic exercise.

I have not commented on how the crisis should be managed, but to say that I support diverse methods in different countries because we will discover whose methods were most and least effective.

Treating data appropriately is not an academic exercise. It is a very practical one. It sounds academic to you because you seem to have little appreciation for the importance of reliability and validity.

Remember: Garbage In = Garbage Out.

Marcus Bressler said...

The Hostess (and others) BELIEVE China. I don't believe a SINGLE thing China says or purports to do or reports as the "truth".

THEOLDMAN

The Chinese Wuhan Media Virus also known as The Chinks Jinx