Coronavirus

[jerry33]

I doubt a year in jail for leaving your house fits into our current culture.

Are you kidding? The for-profit prisons want as many prisoners as they can get--each prisoner increases their income.

 

[AlanSubie4Life]

Headline on this one is that 88% of people put on a ventilator die.

A little clarification on this one. The numbers are bad, but maybe not quite as bad as that number makes it sound:

John W Scott, MD MPH on Twitter

'A better conclusion? - "Among the 1,151 pts who were mechanically ventilated, at a median follow up of 4.5 days, 3% had gone home, 25% had died, & 72% were still in the hospital". - "The eventual mortaily rate will thus range between 25% (282/1151) and 97% (1113/1151)."'

 

[MXLRplus]

Why do you keep bringing politics into this ?

May I also note we kill kids who are playing with plastic guns ? Police can do no wrong. We can go to another country and kill a million people with impunity ? A homeless mother gets a jail term of 6 years for lying about the address and a rapist gets probation ?

It's sort of obvious isn't it? SK and Taiwan have a different style of lawmaking than the US does. Politicians are lawmakers. Perhaps I should have used the L term since the P word offends you? Americans often resist authoritarian actions. Look at the huge fight we had to go through to even reduce airline traffic from Wuhan. Now you want to spy on 330,000,000 people?

 

[Papafox]

Interesting presentation by William Bryan of DHS at today's presser.

Looks like outdoors in the summer sun, the virus degrades significantly faster than indoors. High humidity also adversely affects the virus.

 

[Daniel in SD]

ps : NONE of the methods in which these serology tests are sampling would be accepted in a simple election poll.

Someone just needs to figure out how to do an antibody test over the phone.

Specificity of 93% to 100%.

3.6% antibody rate in Upstate New York suggests a specificity of greater than 96.4% and the fact there are a lot of cases Upstate means it's probably higher than that.

A little clarification on this one. The numbers are bad, but maybe not quite as bad as that number makes it sound:

John W Scott, MD MPH on Twitter

'A better conclusion? - "Among the 1,151 pts who were mechanically ventilated, at a median follow up of 4.5 days, 3% had gone home, 25% had died, & 72% were still in the hospital". - "The eventual mortaily rate will thus range between 25% (282/1151) and 97% (1113/1151)."'

Isn't that really bad?
25/(3+25)=89% They're still dying at the same rate.

 

[EVNow]

Specificity of 93% to 100%.

I reserve my judgement, until we can see the actual study.

http://dmna.ny.gov/covid19/docs/all/DOH_COVID19_AntibodyTestingFactSheet_041420.pdf

This Professor at the Department of Microbiology Icahn School of Medicine at Mount Sinai is not convinced.

Florian Krammer on Twitter

Sorry, Governor, but that is BS. Which test is this based on?

I think this is too high. It is possible. But a 20% plus infection rate seems too high for NYC due to a number of reasons. I would think 6-8%, maybe 10% are closer to the truth. It would be nice to know more about the test, its sensitivity and specificity and the test population.​

NEW: The first phase of results from a statewide antibody study are in. We collected approximately 3,000 antibody samples from 40 locations in 19 counties. Preliminary estimates show a 13.9% infection rate.
​

ps : NONE of the methods in which these serology tests are sampling would be accepted in a simple election poll.

Trevor Bedford thinks the numbers look ok. That gives an IFR of 1%.

Trevor Bedford on Twitter

Also, as I said, the sampling needs improvement.

Bianca on Twitter

at my local store a lot of people were going specifically to take the test and lining up instead of it being randomized, so prob plenty of people who thought they had it before or had been around someone but had no symptoms so couldn’t get a normal test​

 

[Daniel in SD]

Interesting presentation by William Bryan of DHS at today's presser.

View attachment 535128

Looks like outdoors in the summer sun, the virus degrades significantly faster than indoors. High humidity also adversely affects the virus.

Restaurants and gyms are allowed to open in Georgia but they must not have A/C and must install one 500W grow light per square meter of floor area.

 

[Doggydogworld]

I am not so sure about that. I expect, except in states that have positivity rates below 4% or so, that the actual cases will be about 10x the diagnosed cases. I think there's just sort of a floor on it - people who know they've been exposed to a known case self-isolate, ride out minor symptoms, never get tested. And that multiplier will be fairly constant on average.

I'd divide testing into three basic levels:

1. Aggressive testing, even of symptom-free people found via contact tracing
   
2. Test most/all with symptoms, plus some screening tests for healthcare workers and suc
   
3. Only test those with severe symptoms

South Korea and other APAC countries are level 1. Low single digit positivity and near 1% nCFRs. A few US states might qualify, e.g. UT and probably HI.

NYC/NJ/CT are level 3. They have 40-50% positivity and only catch 10% of cases.

Most of the rest of the US is close to level 2. The largest states (CA, TX, FL) have ~10% positivity vs. 40-50% and nCFRs of 3.8%, 2.6% and 3.3% respectively vs. NJ at 5.6% and NY/CT above 7.5% (and MI above 8%, ugh). To me this makes a strong case they catch at least twice as many of their cases.

I'd be surprised if we see any instance of the multiplier being less than 5x (except perhaps in Montana which has (had?) a very low positivity rate).

I'd expect 5x to be pretty normal outside the known problem states NY/NJ/CT and a couple others like MI. That said, the problem states really dominate the national numbers. So even if the median state is at 5x we're probably still close to 10x nationally, e.g. 7-8m infections vs. 875k confirmed cases.

 

[sparcs]

Two different articles related to Bill Gates about the virus. One about the innovation we will need and another that gives a hint at what life might be like in the months and years ahead.

The first modern pandemic | Bill Gates

I break the innovation into five categories: treatments, vaccines, testing, contact tracing, and policies for opening up. Without some advances in each of these areas, we cannot return to the business as usual or stop the virus.

Bill Gates: As things get back to ‘semi-normal,’ it’s impossible to overstate the pain that lies in the years ahead

The good news, he said, is we can look forward to a “semi-normal” world over the next two months.

“People can go out, but not as often, and not to crowded places,” Gates wrote. “Picture restaurants that only seat people at every other table and airplanes where every middle seat is empty.”

He said he believes schools will reopen, but stadiums won’t.

“The basic principal should be to allow activities that have a large benefit to the economy or human welfare, but pose a small risk of infection,” Gates said.

“The bulk of the story will be what happens next,” he wrote. “Even if governments lift shelter-in-place orders and businesses reopen their doors, humans have a natural aversion to exposing themselves to disease. Airports won’t have large crowds. Sports will be played in basically empty stadiums. And the world economy will be depressed because demand will stay low.”

He said life will only return to normal when most of the population is vaccinated, and that could take awhile, though he hopes that one will be in mass production by the second half of 2021.

 

[dfwatt]

This has turned into the hypochondriac thread...

Almost all of the antibody tests now seem to suggest a much lower IFR than previously thought, but you seem to ignore them all... It almost seems like you want the disease to be more deadly than it is.

Hmm. Not sure how you get the idea that anybody is cheering for the disease to be any deadlier than it already is. Or that anything anyone is saying makes them "hypochondriacal." Frankly, your kind of late to the party on this issue and perhaps you should read whole lot of the posts before you jump to conclusions. You seem either a bit slow on the uptake or , , , , just missing whole lot of context. I'll assume it's the latter and try to provide you with some.

The only credible antibody tests (not the sketchy Santa Clara study), just completed in New York show essentially what a lot of people here predicted, namely that the rate of actual case to detected case is something on the order of 10 to 1. This does not turn the IFR into some kind of cakewalk number which you seem to think it does. It's still in a bad range. somewhere between 0.6 and perhaps 1.3 or so, after you make the adjustments that you have to to the various metrics.

Hope that's helpful.

 

[AlanSubie4Life]

Isn't that really bad?
25/(3+25)=89% They're still dying at the same rate.

Yeah, that was what was headlined. The thing is, the vast majority of the patients were still on ventilators at the end of the study. It's possible that people who died quickly on ventilators were more likely to die on a ventilator than the remaining patients.

It's not like there is a lot of good news here, though. That's true.

I'd expect 5x to be pretty normal outside the known problem states NY/NJ/CT and a couple others like MI. That said, the problem states really dominate the national numbers. So even if the median state is at 5x we're probably still close to 10x nationally, e.g. 7-8m infections vs. 875k confirmed cases.

Unfortunately we'll probably never have a good enough study or specificity in low prevalence areas to ever know what the actual multiplier is in most places. It'll have very wide confidence intervals I would think. Really only useful in the places where actual prevalence starts to exceed about 2-3%. But that probably would account for most of the cases nationwide.

 

[EVNow]

How would you guys like to give your location & health data to Peter Thiel ?

Trump Admin Gives Coronavirus Tracking Contract to Peter Thiel's Palantir: Report

In other unrelated news, Trump suggests "The disinfectant, where it knocks it out in a minute, and is there a way we can do something like that by injection inside, or almost a cleaning. It gets in the lungs".

Not a fan of either of these.

 

[sparcs]

How would you guys like to give your location & health data to Peter Thiel ?

Trump Admin Gives Coronavirus Tracking Contract to Peter Thiel's Palantir: Report

I think we are steadily heading towards a world where it is a cross between Westworld: Season 3 and The Purge: Pandemic Edition

 

[jhm]

Yes, you are right. In fact, the linear projection from the last 14 days points to 117,444 deaths.

The chart below shows how these ultimate projections along with 14-day intercept and slope estimates have evolved over the last month. Prior to this the slope sometimes goes positive, so the ultimate projection are too unstable even to graph meaningfully.

View attachment 535009


From 4/10 to 4/18, projected ultimate deaths are relatively low and stable. This is a time when the NY outbreak is dominating national numbers and coming into a little more control. The uptick in recent 4 days is worrisome. It could be that a broadening of the epidemic in most states is starting to dominate the growth of deaths.

Thanks for the suggestion. This could be an interesting diagnostic to watch.

I've been doing this in a spreadsheet. I looked back at my formulas and found that the above was actually based on 12-day rolling windows, not 14-day. The difference matters because of day of week periodic behavior. So below is an updated chart, including today at 49,736. Note that the last estimate of ultimate deaths is 97,821 and for yesterday 91,020.

Fortunately for my dumb-ass model, these true 14-day projections yield more comparably ultimates than the 12-day projections. The difference between the dumb-ass and 14-day methods can be seen in the second chart below.

 

[SageBrush]

My epidemio-arithromancy came to 1% IFR for NYC.
Using that as an upper bound country wide, I decided to try and estimate the acute care hospital bill for a trumper scenario of letting the epidemic play out until herd immunity.

Starting points:
300E6 infections
IFR = 1E-2
Hospital admission = 5x mortality
Hospitalization cost: 1E5 dollars

3E8*1E-2*5*1E5 = 15E11 = $1.5 Trillion

 

[jhm]

My epidemio-arithromancy came to 1% IFR for NYC.
Using that as an upper bound country wide, I decided to try and estimate the acute care hospital bill for a trumper scenario of letting the epidemic play out until herd immunity.

Starting points:
300E6 infections
IFR = 1E-2
Hospital admission = 5x mortality
Hospitalization cost: 1E5 dollars

3E8*1E-2*5*1E5 = 15E11 = $1.5 Trillion

$1.5T is nothing to sneeze at in an economy of only $21.4T in 2019.

 

[AlanSubie4Life]

If I understood him correctly, Trump is hoping that sunlight and disinfectants can be used on the body to help eliminate the infection. Says Dr. Birx maybe should look into it. It's not clear whether he was speaking of an injection, ingestion, or a more topical treatment.

My go to is Tide Pods.

 

[EVNow]

My go to is Tide Pods.

Cue : There is a run on Tide Pods.

 

[EVNow]

Interesting presentation by William Bryan of DHS at today's presser.

View attachment 535128

Looks like outdoors in the summer sun, the virus degrades significantly faster than indoors. High humidity also adversely affects the virus.

The 2 biggest hotspots in India
Delhi : very Hot but Dry
Mumbai : very Hot and Humid

Any questions ?

 

[sparcs]

The 2 biggest hotspots in India
Delhi : very Hot but Dry
Mumbai : very Hot and Humid

Any questions ?

India looks scary. How many tests per day will they have to run to try to slow this down? Five to Ten million per day?

I don't see how a sustained lockdown in a country with such a large population, population density, and so much poverty is going to be possible.