Coronavirus

[AlanSubie4Life]

Another factor I didn't mention before is that because the numerator is smaller than it

No, we should not ignore this at all, this is actually the best example, since using the 14-day lag, one could see that coming already.

Really what I'm saying is that of course you get close to the right result there, since both numbers have reasonably settled. But you would have got a very pessimistic result with a 14-day lag if your case number had fallen on the exponential part of the curve. For example, March 7th and Feb 22nd. You get: 48/436 = 11%

What you're trying to do here is estimate CFR given an epidemic during the full exponential growth period. You can't do that if you're trying to test your method on a curve that isn't even close to exponential (for the cases number specifically - the death number doesn't have to fall on the exponential for the most part, though there will be some error, due to distribution of death delays).

I agree the Iceland data is too noisy to tell, and besides, their testing method is such that you'd generally expect lower CFRs.

Specifically in regards to the US data, using your 14-day method:

April 1st: 5102
March 18th: 9197

That gives you: 55% projected CFR for the US.

If you use a 6-day lag (April 1st, Mar 26th):

US: 5102 / 104125 = 4.9% (Probably still thrown off by insufficient test capacity resulting in a very low denominator...)
Germany: 931 / 43938 = 2.1%
South Korea (Mar 7th, Mar 1st) 48/3173 = 1.5%
Iceland: 2/802 = 0.25% (I think thrown off by the test method).

Just picking arbitrarily a couple countries where the outbreaks are not so severe that they overload the system leading to under-testing 6 days ago, and where they might be competent:

Israel:
6-day: 26/ 2693 = 0.9%
14-day: 26/433 = 6%
Canada:
6-day: 114/4043 = 2.8%
14-day: 114/727 = 16%
Switzerland:
6-day: 488/11811 = 4.1%
14-day: 488/3115 = 16%

In the end, there are so many sources of error with this I think it's probably futile, though. I think all you can do is take the current deaths (assuming they are a good count) and use that to back calculate approximately how many cases existed x number of days ago. It's harder to determine the number of cases today, because it depends a lot on the growth rate in the particular country in question.

 

[AlanSubie4Life]

The derivative is symmetrical around zero but I believe in real epidemics the right side will fall slower.

Probably.

Logistic Function = 150,000 / ( 1 + exp( -0.165 * t )

Can you give us some insights into the meaning of these coefficients? What do they signify, physically, regarding the epidemic dynamics?

I assume 150,000 is the expected cumulative number of cases?

 

[Norbert]

That was around the day that Fauci suggested 100k - 200k total deaths. Estimates are improving, but they still have considerable uncertainty. That should not surprise anybody, since even a single confounder like SAH orders have considerable variation in effectiveness across the nation. IHME did not know that Florida e.g. would continue to allow mass congregations for prayer.

IHME does not offer enough transparency to say *why* the estimates are going up, but so far that is the trend.

My impression/guess is that it ignores or underestimates a few things which makes it overly optimistic by nature, while also adjusting to reality as it happens, yet staying on the overly optimistic side regarding the future, as it keeps ignoring indications. Kind of like the whole thing. You won't be able to say that it is wrong, but it won't be right either. Maybe some day they will change the configuration to make it more "negative", but that doesn't necessarily increase the quality. Unless they really improve it somehow. Just so you know, I wrote this with ink that becomes invisible after two days.

I still believe that underestimating the problem is the larger danger.

 

[AlanSubie4Life]

I still believe that underestimating the problem is the larger danger.

I wholeheartedly agree. And it does appear that people who have looked closely at the assumptions behind this model are judging it to be biased towards optimism.

As we've learned so far, it really is far better to be pessimistic when dealing with exponentials. You can scale test capacity adequately, scale hospital and PPE requirements adequately, etc. It's ok to expend enormous amounts of extra money, even if it turns out to not be needed - most of the things they'll be doing can be used over the long term, anyway.

Personally feeling like we'll be at about twice what they predict, nationwide (200k deaths, ~20 million infections (infections will be heavily undercounted most likely)). Some states further along in their epidemics may end up being fairly well predicted. But ones that are heavily undersampled right now (Michigan...maybe Florida), might end up being worse than predicted.

Also I suspect the tail of the epidemic will be worse than their predictions, as people get lazy.

 

[Norbert]

Really what I'm saying is that of course you get close to the right result there, since both numbers have reasonably settled. But you would have got a very pessimistic result with a 14-day lag if your case number had fallen on the exponential part of the curve. For example, March 7th and Feb 22nd. You get: 48/436 = 11%

That's correct, but I already said and explained that especially at the early time of a spread, it requires using additional discretion and consideration, instead of blindly using the literal numbers.

 

[Norbert]

Specifically in regards to the US data, using your 14-day method:

April 1st: 5102
March 18th: 9197

That gives you: 55% projected CFR for the US.

We already talked about that, after I examined the situation of how the US numbers corresponded to the Iceland numbers, and we came to the conclusion that this relates to an an extreme under-testing (possibly under 10%) in the US. Had there been a reasonable level of testing, the numbers would have made much more sense. Even so, it helped me figure that out.

 

[AlanSubie4Life]

We already talked about that,

Honestly, can't keep track!

Here's an interesting article (opinion piece actually) on the impact of viral dose on illness severity. Somewhat speculative, it appears, but we shall see...

Opinion | These Coronavirus Exposures Might Be the Most Dangerous

 

[Buckminster]

Update from the UK. Supermarkets now have long (socially distancing) queues to get in as they are restricting the number in the store at any one time. This has only just happened fully in my town. This feels like a more significant change than it sounds.

 

[cusetownusa]

another metric for how much we've shut down the US.

US Box Office totals for the week of March 20-26 was $5,179. This time last year, it was $204,193,406 (boxofficemojo.com)

i just read an article about the hospitals in my area. They are losing millions and are furloughing some employees. Wonder if this will be a common issues for hospitals that took early initiatives to free up bed space but haven’t actually been hit with many Covid19 patients, at least not yet.

 

[SMAlset]

Title of the CNN article say it all:

Coronavirus can spread through talking or breathing, prestigious panel tells White House - CNN

 

[Unpilot]

So if I have this right the problem is X*Y=Z
we don't know X or Y and we are guessing at Z?

 

[SMAlset]

Peter Tsai, material scientist inventor of N95 electrically charged mask, suggests home made masks of the automotive shop paper toweling (typically in blue) in this Washington Post linked to article on guide to home made masks:

Your guide to DIY facial masks: What the research says and why experts are split — The Washington Post

 

[SageBrush]

I find it helpful to follow NYS stats at covid19.healthdata.org for estimates of national impact simply because the state is much further along the infection timeline than everybody else so the uncertainty is less. I expect the southern states to be ~ 2x worse due to their risk factors, and some of the flyover states to be 0.5x as bad due to actions taken earlier in their respective curves.

The middle projection is 16k deaths by the end of this wave in July in a population of 20M, so about 1.3 per thousand.

 

[mikevbf]

Here is a summary of NewYork-Presbyterian Hospital's update from yesterday:

Our states were not slow to respond to this and have been very aggressive to stem the tide here.

In NYC area Javits center is open, but patients need to be somewhat healthy to be transferred there and it is a little unclear how it will be used.

Navy ship Comfort is here, but only for non-covid patients, which means we need to do some testing before.

1,788 in patients Covid positive 482 on ventilators (27%)

Testing: all patients being admitted that we suspect of having covid, ob patients, and specific categories of patients, and patients that will be moved to Navy ship Comfort.


New: Every patient being admitted for a surgical procedure will now be tested, including those having emergency procedures.

We have been able to ramp up testing from 200 to now over 1000 cases per day. We are going on strong to increase these testing capabilities to 3000 by early next week across the enterprise.

We are validating a new test by our clinical labs to submit to the FDA that would be a shorter turnaround on test results while maintaining testing quality.

For the PCR (nose swab) test, we are now looking at a point-of-care test that is very fast. This could, if validated, be scaled up.

Serology test: uses antibodies that are made quickly against the infection and long term against the virus. Would tell you if someone has been infected.

CDC is considering more widespread use of masks for general public.

Ventilators: there are two broad categories of vents, there are invasive vents (when someone has a breathing tube down their throat), and non-invasive vents (those you don’t have the tube with). We are in need of the invasive vents. We need those intensive care type ventilators. Ventilators don’t run themselves.

Ethical considerations: the hospital will stand behind you. There may be more guidance coming out of NY state, in the meantime we are here to support you.

We continue to add capacity for ICU, also looking to expand palliative care and social workers.

 

[Sean Wagner]

Good luck. You do feel like you won some prize when you are able to find what you went for.

I know that feeling! Today it was my cherished giant Soissons beans that go into soups and chilis and what-have-yous. Previously a specific liquid soap, hand sanitizer [finally after weeks and weeks without], and of course toilet paper.

IIRC NYS has ~ 100k total "hospital beds."
Optimistically that might be 20k acute care beds available for Covid, so at the current growth rate the hospitals will be stuffed this week. Scary stuff, and we are just at the beginning

There must have been contingency plans for just such outbreaks. Congress centers etc. should be built with this in mind.

Ask and you shall receive:

A choir decided to go ahead with rehearsal. Now dozens of members have COVID-19 and two are dead

45 out of 60 attendees infected during 1 practice session. Oh and they all took distancing and hygiene precautions.

Quote from the article/
...
“One could imagine that really trying to project your voice would also project more droplets and aerosols,” he [Jamie Lloyd-Smith, a UCLA infectious disease researcher] said.

With three-quarters of the choir members testing positive for the virus or showing symptoms of infection, the outbreak would be considered a “super-spreading event,” he said.

Linsey Marr, an environmental engineer at Virginia Tech and an expert on airborne transmission of viruses, said some people happen to be especially good at exhaling fine material, producing 1,000 times more than others.
...
/Unquote.

Fascinating. Kinda. Definitely very relevant.

Up until this pandemic the only people most of us knew who died of an infectious disease were old or had some kind of serious health condition. It's alien to most of us to have someone perfectly healthy one day and dead a couple of weeks later from an infectious disease. In the arc of history this isn't new, but it is to almost everyone alive in the developed world today.

With humongous metropolitan areas growing around the world, we need to prepare for a more lethal virus too.

Quarantining a few states or cities is waste of effort; the whole country need to be locked down.

The last lockdown paces the nation. This is not the time for "feral" government.

Also I suspect the tail of the epidemic will be worse than their predictions, as people get lazy.

It will be a challenge to manage a return to broader economic activity. No doubt.


Meanwhile, Germany is testing at a rate of 350'000 - 500'000 per week by leveraging its decentralized structure and numerous laboratories. Could have happened in the US too, of course. See

Why is Germany able to test for coronavirus so much more than the UK? | BY JACK DICKENS | tweet JACKFDICKENS / 31 MARCH 2020
Why is Germany able to test for coronavirus so much more than the UK? - Reaction

 

[aubreymcfato]

A bit OT, but any take on this?
https://ftalphaville.ft.com/2020/04...omised-ventilators--These-are-BPAP-machines-/

Don't know is it's business-as-usual Elon bashing, semantics or they have a point.

 

[SageBrush]

Peter Tsai, material scientist inventor of N95 electrically charged mask, suggests home made masks of the automotive shop paper toweling (typically in blue) in this Washington Post linked to article on guide to home made masks:

Your guide to DIY facial masks: What the research says and why experts are split — The Washington Post

OUTSTANDING

 

[Sean Wagner]

Don't know is it's business-as-usual Elon bashing, semantics or they have a point.

Without speaking to the post at hand, I'm sorry to say the FT's Alphaville section is - and I mean it just so - among the rabid, positively gushing, highly biased purveyors of anti-Tesla blurbs. They will dig far and wide to "unearth" just about anything that sounds right to them, while being utterly and completely oblivious to the enterprise's successes and its mission.

Edit: I hope to have learned something just by observing these nominally sane and not-unintelligent people at "work". Not sure yet.

 

[wdolson]

I wholeheartedly agree. And it does appear that people who have looked closely at the assumptions behind this model are judging it to be biased towards optimism.

As we've learned so far, it really is far better to be pessimistic when dealing with exponentials. You can scale test capacity adequately, scale hospital and PPE requirements adequately, etc. It's ok to expend enormous amounts of extra money, even if it turns out to not be needed - most of the things they'll be doing can be used over the long term, anyway.

Personally feeling like we'll be at about twice what they predict, nationwide (200k deaths, ~20 million infections (infections will be heavily undercounted most likely)). Some states further along in their epidemics may end up being fairly well predicted. But ones that are heavily undersampled right now (Michigan...maybe Florida), might end up being worse than predicted.

Also I suspect the tail of the epidemic will be worse than their predictions, as people get lazy.

I've noted that most of the people making the 100K-200K predictions have said "if we do everything right". We're not going to do everything right. Some places are and will, but many places won't. And the federal government response is just plain awful.

 

[MXLRplus]

It's getting really worrisome in California. A celebrity was on social media telling everybody how hard it is to work a dishwasher. Her staff didn't show up. Will she survive? Who knows? Celebrities are falling like flies because the quarantine is only for the unwashed masses. The good news in this dark hour in Hollywood? We are trying to push through a large tax break for the celebrities. If you think a middle class worker has been hurt in the quarantine, some of the more important people have lost of millions of dollars. They can barely afford fuel for their yachts! I feel for them...