Italy still on figurative fire (USA in trouble)

[cjrugger]

Never said doubling time was one day. I used 3 days which is on the lower end of the observed range globally (3 - 6 days in most countries excluding China). If the double time was one day then it would take 23 days to go from first case to 4 million. However, no outbreak can sustain exponential growth indefinitely as it eventually runs out of uninfected people to infect. See below for data. https://ourworldindata.org/coronavirus#the-growth-rate-of-covid-19-deaths

The number with positive tests is obviously lower than the actual number but whether it is off by 10, 100, or 1000 isn’t settled yet. Hopefully the number that had it and recovered or were asymptomatic but never tested is higher because that would depress the mortality rate and mean that we are further into the outbreak without yet overloading hospitals. That would mean fewer left to get it so we in theory are closer to the inflection point then if testing was perfect.

I don’t see how that math works. The first case in China was in late November, now over 100 days ago. With your every 3 day double billions would have it by now.

The truth is we have no idea how many people already had it, have it now, or will get it. Everyone is just plugging numbers that are almost certainly wrong into equations and spitting out answers that are almost certainly wrong

 

[Steve G]

Thanks for that link! That's definitely their paper and model!

My thoughts:

(1) We're still at a situation where a vast majority (>80%) of people tested in nearly all countries test negative. 50% having already been infected already seems high in that context, even if there are people who have had the virus and are now immune (e.g., wouldn't test positive if they tested today).

(2) The most important parameter in their model is rho (the greek letter that looks like a lower case p). rho = proportion of the population at risk to a severe disease. If you look at Figure 3, you have a significant difference between people susceptible (not already being infected) based on rho. Anywhere from ~ 96% (if rho = 0.1) to 40% (if rho = 0.001).

(3) Intuitively, a rho of 0.001 (which gives the most optimistic result) doesn't make sense. At all. 1 out of only 1000 people are at risk to severe disease? Just doesn't couch with what is happening empirically. Note that rho is just being "at risk to severe disease", not "at risk to dying", which would be a lower number (rho * theta in their model).

(4) They may actually be under-estimating R0 (the reproduction number): 2.25 is what they're generally using but I'm seeing a number of things where R0 is higher (one study from Switzerland estimated 4.26!).

(5) That said, R0 likely was lower in March vs. February. Even before all the lockdowns and such, people were getting nervous because Coronavirus was in the news, and were voluntarily doing some social distancing already. So even if R0 was, say, 2.75 in February, it likely dipped in March before any governmental guidelines were issued, then dipped a whole lot more after that.

(6) If you look at the 2 orange lines in Figure 3 - you can see this model is much more sensitive to changes in rho than R0.

(7) So, ultimately the interpretation of this model is most dependent on making a good "guess" for rho. 0.001 doesn't pass the sniff test. A rho around 0.03? I could buy that, given that (A) ~ 1 in 5 people hospitalized (e.g., at risk to severe disease) die, and (B) I think the ultimate IFR from this virus will be on the order of 0.5%-1%.

(8) Use that rho and a slightly higher R0 than 2.25, and visualizing the graph in Figure 3 we'd be at ~ 15% of the population already having been infected.

(9) That would also correlate directionally with the numbers from #1.

----------

I'm not purporting to be an expert, but those are my high-level thoughts after looking at their paper. It's always good to get all of these academic studies out there in the wild for people to look at! But I do think the Financial Times headline that came out of this study is rather optimistic.

The major missing data element for this type of modeling is an accurate serological test to capture exposed subjects. If there was a reliable antibody based test, would be possible to determine actual exposure rate. I know such tests are in rapid development and are critical.

 

[katchthis]

The Diamond Princess cruise ship had about the most "favorable" conditions possible for a spread of the virus with many infected persons, common facilities, and no ready place to escape.

Even there, the 3700 passengers produced 700-some cases of infection, and I believe the death total is 4 so far.

I'm aware of those numbers too and I'm feeling you Roy. I'm cautiously optimistic.

 

[UncleLar]

Another comparison - that might be enlightening:

The US Death rate from the 2017-2018 Flu (which was a rather bad season for the Flu) was somewhere in the neighborhood of 250 per million.... IIRC (and I think I do)

That's incorrect.

The Diamond Princess cruise ship had about the most "favorable" conditions possible for a spread of the virus with many infected persons, common facilities, and no ready place to escape.

Even there, the 3700 passengers produced 700-some cases of infection, and I believe the death total is 4 so far.

Death total is 10 (out of 712 cases).

 

[michnittlion]

The major missing data element for this type of modeling is an accurate serological test to capture exposed subjects. If there was a reliable antibody based test, would be possible to determine actual exposure rate. I know such tests are in rapid development and are critical.

You are correct - we need serological tests. I'd argue they have become more important than the PCR tests (answers the question "do you currently have the virus?").

The news reports on serology tests are hit or miss - it seems like it has been "days away" for 2 weeks now.

 

[anon_1eeb2b426hv3y]

I'm aware of those numbers too and I'm feeling you Roy. I'm cautiously optimistic.

Again, the critical piece of info is the number exposed/with antibodies.

 

[CoastGuard2018]

Italy’s percentages dropped again today in both infected and mortality. Also there is a new study out that seems to indicate Covid 19 does not mutate well. Spain and Germany down to.

 

[PSU Mike]

Please define mutate well ....

 

[Sproul]

Of Italy’s closed cases: 56% recovered, 44% died.

 

[Sproul]

Italy’s percentages dropped again today in both infected and mortality. Also there is a new study out that seems to indicate Covid 19 does not mutate well. Spain and Germany down to.

Well they went down 2 days ago and then up again yesterday and down today (I haven’t seen that data yet). 2 days a trend does not make.

 

[CoastGuard2018]

Please define mutate well ....

Study out of Italy posted today indicates it is relatively slow to mutate and if a vaccine is developed it should work well over a broad spectrum and should be viable for a significant period of time.

 

[CoastGuard2018]

Well they went down 2 days ago and then up again yesterday and down today (I haven’t seen that data yet). 2 days a trend does not make.

March .....
Here is your trend negative Nelly. Today’s numbers aren’t on the below. Special thx Michnitt
Increase in total confirmed cases: 20.2%, 22.9%, 23.5%, 24.9%, 20.2%, 26.9%, 25.4%, 24.4%, 10.7%, 22.8%, 21.3%, 16.9%, 19.8%, 17.0%, 13.9%, 12.0%, 13.2%, 14.9%, 14.6%, 13.9%, 10.4%, 8.1%, 8.2%.

Increase in total deaths: 52.9%, 51.9%, 35.4%, 38.3%, 33.1%, 18.3%, 57.1%, 26.5%, 36.3%, 31.1%, 22.9%, 24.6%, 13.8%, 25.5%, 19.3%, 16.0%, 19.0%, 14.3%, 18.4%, 19.7%, 13.5%, 11.0%, 12.2%.

 

[UncleLar]

Not according to the CDC:

https://www.cnn.com/2018/09/26/health/flu-deaths-2017--2018-cdc-bn/index.html


80,000 Deaths / 320 Million = 250 deaths per million. (Maths )


But, yeah, Santa..... I'm sure you have more reliable info.


(It's one thing to SAY something that is wrong...... it's a whole 'nother level to look at something that IS right - and spew some hot take that "That's Incorrect" . Way to go You!! )

My apologies. I didn't intend to post that. It was left over from a post that a started to make a day or so ago but when I double checked the math, I found that you were right, so I aborted the post (it's also the reason that it's so short). But when I went to specify the death total for the cruise ship, that post was automatically tacked on to the back of my old aborted post which I just didn't notice. It's a glitch in the system that I usually catch but didn't this time. Sorry. .

 

[Nittany Ziggy]

I am also reading reports from Stanford and MIT that these numbers are overblown. Numbers are all over the place.
For example in todays WSJ. Dr E. Bendavid and J. Bhattacharya, professors at Stanford school of medicine.....

Is COVID -19 as Deadly as They Say ?
(obviously just one paragraph, but it give the gist of the article )

The northeastern Italian town of Vo, near the provincial capital of Padua. On March 6, all 3,300 people of Vo were tested, and 90 were positive, a prevalence of 2.7%. Applying that prevalence to the whole province (pop.955,000), which had 198 reported cases, suggests that there were actually 26,000 infections at that time That's more than 130-fold the number of actual reported cases. Since Italy's case fatality rate of 8% is estimated using the confirmed cases, the real fatality rate could in fact be closer to 0.06%.


Sorry in advance you cannot link current WSJ articles due to their pay wall.

Great, thought-provoking article. The Comments section in the WSJ is usually fairly tame because you have to use your real name (some still slip through for a short period anonymously). The comments for this article resemble that sewer known as the Test/Politics board! LOL!!

 

[michnittlion]

Study out of Italy posted today indicates it is relatively slow to mutate and if a vaccine is developed it should work well over a broad spectrum and should be viable for a significant period of time.

Right - there have been mutations but the mutations have been slower vs. other viruses (like influenza).

This isn't a guarantee that this would happen with Coronavirus - but in many viruses the mutations tend to be less lethal over time. If a virus mutates and becomes more lethal, that's actually inefficient. It kills its host too quickly before the host can spread the virus.

 

[michnittlion]

March .....
Here is your trend negative Nelly. Today’s numbers aren’t on the below. Special thx Michnitt
Increase in total confirmed cases: 20.2%, 22.9%, 23.5%, 24.9%, 20.2%, 26.9%, 25.4%, 24.4%, 10.7%, 22.8%, 21.3%, 16.9%, 19.8%, 17.0%, 13.9%, 12.0%, 13.2%, 14.9%, 14.6%, 13.9%, 10.4%, 8.1%, 8.2%.

Increase in total deaths: 52.9%, 51.9%, 35.4%, 38.3%, 33.1%, 18.3%, 57.1%, 26.5%, 36.3%, 31.1%, 22.9%, 24.6%, 13.8%, 25.5%, 19.3%, 16.0%, 19.0%, 14.3%, 18.4%, 19.7%, 13.5%, 11.0%, 12.2%.

For today (25-March), the % increases are 7.5% (total confirmed cases) and 10.0% (total deaths).

 

[The Song Remains the Same]

The Diamond Princess cruise ship had about the most "favorable" conditions possible for a spread of the virus with many infected persons, common facilities, and no ready place to escape.

Even there, the 3700 passengers produced 700-some cases of infection, and I believe the death total is 4 so far.

Not to mention an average passenger age that was certainly in the wheelhouse of the virus...

 

[crazytoadie]

The concept of exponential increase does not mean that is precisely what will happen, but it does demonstrate why the longer you wait the less you can influence the spread.

You cannot look at the numbers as absolutely perfect because testing isn’t done until after a person develops symptoms, which may be anywhere between 5 - 14 days. So in essence the numbers you are seeing are outdated with respect to current infections. So the number of exposed today may be 2 or 4 times more than what number of positive tests. Further, maybe only 1 in 10 or 1 in 100 get symptoms severe enough to be treated or tested. The confirmed cases are severely below the actual but probably below the estimates.

 

[Nittany Ziggy]

My apologies. I didn't intend to post that. It was left over from a post that a started to make a day or so ago but when I double checked the math, I found that you were right, so I aborted the post (it's also the reason that it's so short). But when I went to specify the death total for the cruise ship, that post was automatically tacked on to the back of my old aborted post which I just didn't notice. It's a glitch in the system that I usually catch but didn't this time. Sorry. .

Congrats on being forthright, honest and contrite (a rare trifecta these days)!

 

[pointingdogsrule]

If you quarantined the positives there would have been NOWHERE NEAR the community spread.

TO THIS DAY, in the US of A, we are not testing those without symptoms. Hell in rural areas there are lots of people who are not being tested even if they have the symptoms, TODAY, if they have not been out of the country. The failure to test is what is driving this, even now.

Iowa. Are we rural enough for you?
145 positive, no deaths.
2578 negative.
What does this tell you. Keep spreading your mantra.

 

[MtNittany]

Iowa. Are we rural enough for you?
145 positive, no deaths.
2578 negative.
What does this tell you. Keep spreading your mantra.

He can't help himself.

 

[royboy]

Death total is 10 (out of 712 cases).

Thanks for the update on the Diamond Princess. Aren't we about at the point where that would be a final total or are there some still being treated?

 

[Chickenman Testa]

Thanks for the update on the Diamond Princess. Aren't we about at the point where that would be a final total or are there some still being treated?

Ironically they were probably lucky to get it early when there were enough ventilators to go around. You want to not get a serious case of this in the next month I’m thinking.

There’s going to be some pretty brutal triage decisions going down in NYC in the next few weeks.

 

[UncleLar]

Thanks for the update on the Diamond Princess. Aren't we about at the point where that would be a final total or are there some still being treated?

There were 712 positives, 10 died, 587 are listed as recovered so that leaves 115 still being treated (or probably more accurately, still awaiting a final disposition).

 

[katchthis]

There were 712 positives, 10 died, 587 are listed as recovered so that leaves 115 still be treated (or probably more accurately, still awaiting a final disposition).

Maybe like some studies , some are numbers lost to follow up exam ?

 

[CoastGuard2018]

Ironically they were probably lucky to get it early when there were enough ventilators to go around. You want to not get a serious case of this in the next month I’m thinking.

There’s going to be some pretty brutal triage decisions going down in NYC in the next few weeks.

In listening to Cuomo talk today I don’t think so. They are over halfway there according to him. They are looking at splitting, retrofitting, rolling deployment from other places and converting others that are being used in dentists offices or for elective procedures and of course more are going to be produced rapidly Things are being ramped up at existing companies and for place like Ford the contracts are being reviewed as we speak. Coronavirus has awakened a sleeping giant. Our industrial might is tremendous. They are pulling respiratory specialists out of retirement and bring them in from other states.

 

[SLUPSU]

The Diamond Princess cruise ship had about the most "favorable" conditions possible for a spread of the virus with many infected persons, common facilities, and no ready place to escape.

Even there, the 3700 passengers produced 700-some cases of infection, and I believe the death total is 4 so far.

While conditions on the ship could be ideal for transmission, do you know the specifics of how those travelers self quarantined on the ship and when? If everyone on the ship stayed in their cabins, then theoretically there's no transmission.

 

[Chickenman Testa]

In listening to Cuomo talk today I don’t think so. They are over halfway there according to him. They are looking at splitting, retrofitting, rolling deployment from other places and converting others that are being used in dentists offices or for elective procedures and of course more are going to be produced rapidly Things are being ramped up at existing companies and for place like Ford the contracts are being reviewed as we speak. Coronavirus has awakened a sleeping giant. Our industrial might is tremendous. They are pulling respiratory specialists out of retirement and bring them in from other states.

I’m hopeful that is the case. The best thing I hope we get out of this is a national consensus to bring manufacturing back here and buy American. So it may cost you an extra buck or 2. Our vulnerability has been put on full display and we need to learn from it

 

[UncleLar]

I’m hopeful that is the case. The best thing I hope we get out of this is a national consensus to bring manufacturing back here and buy American. So it may cost you an extra buck or 2. Our vulnerability has been put on full display and we need to learn from it

I would hope that what we take from this is that we are all living on one planet and we better learn how to cooperate among nations if we intend to survive as a species. Your last sentence I can fully agree with.

"Our vulnerability has been put on full display and we need to learn from it"

 

[cjrugger]

The concept of exponential increase does not mean that is precisely what will happen, but it does demonstrate why the longer you wait the less you can influence the spread.

You cannot look at the numbers as absolutely perfect because testing isn’t done until after a person develops symptoms, which may be anywhere between 5 - 14 days. So in essence the numbers you are seeing are outdated with respect to current infections. So the number of exposed today may be 2 or 4 times more than what number of positive tests. Further, maybe only 1 in 10 or 1 in 100 get symptoms severe enough to be treated or tested. The confirmed cases are severely below the actual but probably below the estimates.

I agree and my criticisms were not directed at you personally. It’s just frustrating to see so many analyses out there that are quoted and used when there data inputs are crap and thus the output is crap

 

[interrobang]

Well they went down 2 days ago and then up again yesterday and down today (I haven’t seen that data yet). 2 days a trend does not make.

4 straight days now with new cases more than 10% lower than its daily high.

 

[crazytoadie]

I agree and my criticisms were not directed at you personally. It’s just frustrating to see so many analyses out there that are quoted and used when there data inputs are crap and thus the output is crap

No worries. I appreciate the clarification as context is sometime hard to gauge in forum messages.

The number of deaths due to COVID-19 is probably pretty accurate as there probably aren’t many unexplained deaths out there. However, the absolute number of confirmed cases is underestimated but that the trend of the growth is roughly correct. Regardless, we cannot predict exactly when that trend will begin to change because the the indicator of that trend is lagging several days behind. If it is much more widespread than we think then it’s severity is much less than estimated as we are not accounting for all the exposed individuals who are pre-symptomatic, asymptomatic, mild symptomatic, or already recovered. Hopefully that is the case and we are further along as that will translate to fewer deaths that if the numbers are truly just ramping up.

The other thing is that an exponential curve might feel linear because the day over day rate will be consistent. For example, a doubling of cases every 3 days only requires an increase of 125% each day (ie cube root of two). So going from 100 to 225 to 381 to 576 in total cases seems more linear than when comparing new cases per day where it is exponential.

 

[demlion]

WV has tested 804 people. Iowa, twice the population, >3 times the # of tests. I personally KNOW people, with symptoms, who could not get a test.

 

[interrobang]

WV has tested 804 people. Iowa, twice the population, >3 times the # of tests. I personally KNOW people, with symptoms, who could not get a test.

And the huge majority of people who had symptoms and were tested ended up testing negative.

 

[pointingdogsrule]

I’m hopeful that is the case. The best thing I hope we get out of this is a national consensus to bring manufacturing back here and buy American. So it may cost you an extra buck or 2. Our vulnerability has been put on full display and we need to learn from it

Here's another scary fact. 99% of all the micronutrients used in animal feed comes from China. The USA livestock industry relies on these products. Imagine what China could do if they stopped production or deliberately contaminated these products..... African Swine Fever. Say goodbye to our food source and the entire agricultural economy.

 

[CoastGuard2018]

It takes an increase of 25% per day - ie: Day 2's figures are 25% higher than Day 1.... etc etc Or, if you like, Day 2's figures are equal to 125% of Day 1's (not a 125% increase)

[That may be what you meant, but just clarifying]


FWIW: Not to piss on the quant/analytical work that a lot of folks are doing..... but we have a scenario playing itself out with GALACTICALLY unreliable (really, meaningless GIGO) data - being used by folks all over the place. Just the nature of this current situation - not anyone's fault..... but that's why there are so many completely incongruent "solutions" being plotted.
There is - at best - ONE reasonably reliable, non-garbage, piece of relevant data - - - - - namely, the number of deaths (and this assumes that no one is deliberately fudging the numbers - which may not be the case in all countries)
THAT data (the # of deaths) is probably the only one worth doing any types of projections/trends analysis on..... IMO.

Additionally, Dr Birx said today that all these crazy projections with. 60-70 percent getting infected is only from multiple waves with no measures being taken.

 

[SLUPSU]

And the huge majority of people who had symptoms and were tested ended up testing negative.

I don't buy that, do you have a source.

 

[jwarigaku]

Roughly 90% are testing negative in Pa and 85% US wide(NY’s ~ 27% positive rate skewing the curve). California is well under 5% with ~65,000 test results logged so far. By sunrise in excess of 500,000 tests will have been processed in the United States 150,000 more than the next closest country. China has tested 320,000 total Korea about 345,000 Spain 355,000 Italy 342,000. Pennsylvania testing stats are below and updated daily. The blood serum tests for both C and G antigens are complete and have been, but awaiting clearance.

https://www.health.pa.gov/topics/disease/coronavirus/Pages/Cases.aspx

I don't buy that, do you have a source.

 

[stormingnorm]

Roughly 90% are testing negative in Pa and 85% US wide(NY’s ~ 27% positive rate skewing the curve). California is well under 5% with ~65,000 test results logged so far. By sunrise in excess of 500,000 tests will have been processed in the United States 150,000 more than the next closest country. China has tested 320,000 total Korea about 345,000 Spain 355,000 Italy 342,000. Pennsylvania testing stats are below and updated daily. The blood serum tests for both C and G antigens are complete and have been, but awaiting clearance.

https://www.health.pa.gov/topics/disease/coronavirus/Pages/Cases.aspx

But, but, but........

We Are All Gonna' Die - Because...... NO TEST KITS!!!!!!!!!


(Or maybe that was just the way to blame the Orange Man for every malady known to man?)

 

[Mister Magoo]

But, but, but........

We Are All Gonna' Die - Because...... NO TEST KITS!!!!!!!!!


(Or maybe that was just the way to blame the Orange Man for every malady known to man?)

Heard someone call him Agent Orange. I like it. Or, good boy Donny