COVID-19

[RealPerson]

The Netherlands and Belgium received a large quantity of masks for health care workers from a named country. The masks were shown to be ineffective as a barrier for COVID 19. The source country was China. The masks had to be removed from the supplies available. I hope they got a refund. I wonder where the contaminated test kits came from.

[7Wannabe5]

The IHME models for the U.S. and each state are now being updated daily. According to yesterday’s White House press conference, this is likely the model being most referenced for federal level decisions.

Trump seemed clearly shaken by the fact that a good friend of his is now in a Covid coma.

[Jason]

Covid coma.

If there is not an eponymously named garage band deciding on the chord to "No Respirator" I fear for the future of punk music.

[JL13]

Hmm, these estimates don’t really seem to agree. Are we still thinking 15%-20% are hospitalized 10 days on average after infection and 1% die 17 days after infection?

I think if you're going to use 1% mortality you have to use a lower number for hospitalizations. I think 20% hospitalizations and 2% mortality would get your numbers closer.

However, if the results of the Iceland study are right and 50% of people were asymptomatic rather than presymptomatic, maybe 0.8% mortality and < 10% require hospitalization.

I also wonder also if we're seeing a slowdown in new NY cases yet since the lockdown. Maybe it was doubling every 3 days wide open, and every 7 days with lockdown?

Using the above (fictional) assumptions, when do you figure we have 50% of NY infected?

[Riggerjack]

I don't think the path we're currently on is very wise. It seems like we're collectively doing our best to incur significant economic costs and ensure as few lives are saved as possible. Personally I favor a total shutdown for 3 months with contact tracing and quarantining of those who've been exposed. I acknowledge that this will neither be pretty nor pleasant, but rich countries can afford it and the track record in China, and to a certain extent SK is there. I am worried about what the increased surveillance necessary for this would imply for the future. Mostly we have to do the thing which our society is least equipped to do, make a tough decision about the future in the face of uncertainty.

Yeah, this is where my thinking was a few weeks ago. Now I am more concerned with what the second wave looks like.

We had an opportunity to get quarantine to work to shut this virus down. But that window closed about a month ago. We can't stop it, it's taking more than we are willing to pay just to slow it down.

So I see this lock down period as vital, not because it's going to stop the virus. Those days are gone. But to retool, and prepare as best we can for the virus to burn through.

Early 20th century public health practices are no match for the 21st century connectedness and a virus that is contagious before serious symptoms, coupled with such varied severity.

"You may have been exposed to a serious virus, text your public health department if you start to feel Ill, LOL" is as effective a public health plan as it was designed to be.

So now we make masks where we used to make underwear. Hand sanitizer where we used to make really good beer. The ramp up to WWII, revisited.

Yet a real estate agent is an essential service that must be maintained through this crisis, because we still, at this point, don't take this seriously.

My former neighbor is dead from this virus. He was old, and unhealthy, and diabetic with an urge to eat candy. He didn't have long for this world. But his daughter still misses him, and has to deal with not knowing there was a problem until he crashed, then not being able to see him until he was on deaths door. He died before we ran out of ventilators.

Others won't be so lucky.

And still I read about the FB bravadoes who figure their personal odds are good, and don't want to be held back by some piddly little virus.

I get it. They don't understand. They have never concentrated on their breathing for ten minutes, not as a meditation exercise, but to try to control the coughing enough to push the stars further into the edges of their vision. They are brave because they don't comprehend. They don't have a sense of responsibility when a virus is the bad actor. The downside of the bet isn't real, yet.

Odd, that these are the same people who piously condemn the European traders who sold smallpox blankets to native Americans, with no sense of irony. I guess that's what a few decades of schooling and a century and a half of germ theory bought us. The same old behavior, with a bonus of baseless self righteousness.

I find myself bouncing around a bit, emotionally. Sometimes I am frustrated that I live in a society full of people who seem to be dazed. Uncomprehending of both the dangers and the appropriate behaviors to deal with the threat. Others, I am calm, and remind myself that I am not responsible for their behavior or safety. That this really is a piddling little virus. That though millions will die, unnecessarily, billions will live. And millions die unnecessarily every year. This year, I will just know a few more.

Which brings me to my latest thought about covid 19: This is probably what the common cold was like, the first time around. It killed all that it could kill, and the survivors fought it off. Over the years, it mutated for ease of transmission. For the survivors of covid 19, it will probably be similar, after enough time passes.

I try to remind myself that as culling events go, this is still a pretty nice option.

[Papers of Indenture]

Can someone explain to me what the timeframe is supposed to be here? We're looking at 30 days of social distancing to get over the hump and flatten the curve, over which time we're going to have supposedly 100,000 -200,000 deaths in the USA. At a 1% mortality rate, that's 10m - 20m infected in the USA. How does the infection rate drop off after that? Don't we need 40%-70% of the population to be infected for herd immunity? 20m doesn't come close to that.

What is the exit scenario of these lockdowns?

An epidemiologist that I know said that the general tenor in the epidemiology community yesterday was more positive than in previous days so he was a little confused by the White House briefing yesterday. Also said the 100k-200k number would be total deaths over the next few months or all of 2020. Not just the next 30 days. He also said the model that Brix said lines up with theirs is actually about 80k deaths. He also said he thinks we come in under that 80k mark as well. He is concerned though because he feels that widespread testing for the virus and antibodies is the only way to get ahead of this after we flatten the initial curve this month and he feels that we may not be able to get it done fast enough. If we don't figure out who has immunity versus who is still highly vulnerable then we will be right back in this situation again sometime later this year. I guess that's where the higher end scenarios of 200k come into play. Also he thinks widespread testing would reveal the mortality rate to be around 0.5% maybe a tad less because a lot of folks who have it are not eligible for testing right now. So more people with immunity than meets the eye.

But coronaviruses are not his specialty and i'm an average joe so I may (probably) have misunderstood some things..

Do we have any professional infectious disease experts on this board of very smart folks?

[7Wannabe5]

The current(today-will be updated tomorrow Etc.) projection is 93,765 deaths in U.S. through August 4th. There is a good deal of uncertainty, so 95th percentile band is 42,000 to 180,000.

All this information in readily comprehensible graphical form for each state and the nation as whole, along with all relevant data and explanation of model assumptions is available at IHME Covid website which Jacob linked above. IOW, there is little need for any of us armchair epidemiologists to be constructing our own models from scratch anymore. Even if you disagree with any of the assumptions of this model, you can readily make use of it to project your own take. For instance, my anecdotal observation of behavior tends towards making me pessimistically believe that Michigan will come in within the cloud of worse than expected.

[CajunQueen]

Social Distancing Scoreboard based on cellphone data collection. Sorry if this has been posted earlier.
https://www.unacast.com/covid19

[IlliniDave]

Well, my boss's daughter is I think what people are now classifying as a "presumptive positive" pending the test results. I haven't been around her (my boss) since 2/29 so I'm not at risk from that source myself, but my boss is. It becomes a different thing when people you know personally are affected.

My employer has closed facilities and parts of facilities in other states due to a rash of positives and presumptive positives among employees and in one case a resident subcontractor. They are performing CDC Level 3 cleaning in the affected areas and have not announced when/if the facilities will be reopened in the short-term. I suspect they will. Aside from my boss's daughter, nothing else in my immediate orbit although the number of positives in the state and county, and the number of deaths sadly, moves up.

[BeyondtheWrap]

However, if the results of the Iceland study are right and 50% of people were asymptomatic rather than presymptomatic, maybe 0.8% mortality and < 10% require hospitalization.

I also wonder also if we're seeing a slowdown in new NY cases yet since the lockdown. Maybe it was doubling every 3 days wide open, and every 7 days with lockdown?

Using the above (fictional) assumptions, when do you figure we have 50% of NY infected?

OK, so the total deaths for NYC so far went like this:

3/14: 1
3/15: 5
3/16: 7
3/17: 10
3/18: 20
3/19: 26
3/20: 43
3/21: 60
3/22: 99
3/23: 131
3/24: 192
3/25: 280
3/26: 365
3/27: 450
3/28: 672
3/29: 776
3/30: 914
3/31: 1096

Assuming 0.8% mortality and 17 days from infection to death, that gives the following estimates for true number of infections:
2/26: 125
2/27: 625
2/28: 875
2/29: 1,250
3/1: 2,500
3/2: 3,250
3/3: 5,375
3/4: 7,500
3/5: 12,375
3/6: 16,375
3/7: 24,000
3/8: 35,000
3/9: 45,625
3/10: 56,250
3/11: 84,000
3/12: 97,000
3/13: 114,250
3/14: 137,000
Due to the assumed 17-day period, deaths are no longer useful here, so now I'll project with exponential growth. By this date, NYC had declared a state of emergency, gatherings over 500 were banned, and sub-500 venues had to be half capacity. The growth was already slowing, so let's assume four-day doubling period from base of 137,000.
3/15: 162,921
3/16: 193,747
3/17: 230,406
Now schools and gyms are closed and bars/restaurants are takeout/delivery only. Let's stretch the doubling to 5 days.
3/22: 460,811
Now we're at shelter-in-place, let's make the doubling every 7 days.
3/29: 921,623
4/5: 1,843,245
4/12: 3,686,490
4/13: 4,070,215 (about 50% of the city)
4/19: 7,372,980
4/20: 8,140,430
And that's about the whole city, though realistically it would have become S-curved leading up to this and I'm too lazy to calculate it that way.

Here are the reported hospitalizations for NYC:

3/3: 5
3/4: 8
3/5: 12
3/6: 17
3/7: 26
3/8: 44
3/9: 71
3/10: 107
3/11: 153
3/12: 229
3/13: 346
3/14: 500
3/15: 684
3/16: 970
3/17: 1,285
3/18: 1,639
3/19: 2,102
3/20: 2,662
3/21: 3,171
3/22: 3,700
3/23: 4,402
3/24: 5,058
3/25: 5,779
3/26: 6,524
3/27: 7,293
3/28: 7,905
3/29: 8,459
And the more recent numbers look incomplete, so I won't use them.

Assuming 10% hospitalized after 10 days, that gives us the following estimated cases:
2/22: 50
2/23: 80
2/24: 120
2/25: 170
2/26: 260
2/27: 440
2/28: 710
2/29: 1,070
3/1: 1,530
3/2: 2,290
3/3: 3,460
3/4: 5,000
3/5: 6,840
3/6: 9,700
3/7: 12,850
3/8: 16,390
3/9: 21,020
3/10: 26,620
3/11: 31,710
3/12: 37,000
3/13: 44,020
3/14: 50,580
3/15: 57,790
3/16: 65,240
3/17: 72,930
3/18: 79,050
3/19: 84,590
Looking at these numbers, the growth seems to be slower, and even linear for about the last week or so. The last doubling seems to have taken about six days, and since we were almost at shelter-in-place by then, I'll assume seven days going forward, starting from 84,590.
3/26: 169,180
4/2: 338,360
4/9: 676,720
4/16: 1,353,440
4/23: 2,706,880
4/27: 4,022,408 (about 50% of the city)
4/30: 5,413,760
5/4: 8,044,816 (almost the whole city)

Takes a little longer in this projection, but still, it looks like all of NYC will be infected around the end of April. So we might just have to wait another 2-3 weeks after that for everyone to get better and then we'll have herd immunity and could open up the city again sometime in May.

[jennypenny]

One thing that might be skewing NY's numbers is the people who fled NY when things got bad. I know of several people in my in-laws town who are infected that are only there because it's their summer home. The Jersey shore, Poconos, Florida, etc are full of people who left NY who may have carried the virus with them. That means if they get sick they are counted in another state's totals, which would also included anyone downstream of them wrt infection.

[JL13]

Takes a little longer in this projection, but still, it looks like all of NYC will be infected around the end of April. So we might just have to wait another 2-3 weeks after that for everyone to get better and then we'll have herd immunity and could open up the city again sometime in May.

Thanks, that's what I'm assuming they are saying but not saying with these models: that the majority of people are expected to contract it over the next month or two. And that's why the number of cases begin to decay in May...

[7Wannabe5]

That’s definitely not what they are saying with models. Only 3% or 4% of U.S. population is predicted to be infected by end of May. Nobody is assuming 50% hidden asymptomatic, because easily disproven with number of negatives even among symptomatic. The wave(s) will peak due to social distancing measures working. If/when social distancing measures are lifted without replacement by testing/tracing or vaccine, a second wave will begin.

[Ego]

....because easily disproven with number of negatives even among symptomatic.

How long does it take for someone who was positive with a mild infection and recovered to test negative?

[JL13]

@7wannabe5

so you're saying that these lockdowns are going to go on for YEARS, not months? I get waiting for vaccine, but is contact tracing even effective when it shows no symptoms for days/weeks and lasts for many days on surfaces? I'd imagine you'd have to lockdown entire neighborhoods/regions once someone tested positive.

[jennypenny]

They should go on until there is a reliable test for antibodies, and then only those who show immunity can go on with their lives. The rest will have to wait for a vaccine. If we lift restrictions before then, we'll have a series of lockdowns/flareups/lockdowns etc. ... I would think that would be more disruptive than waiting until we know who is in the clear and who should still take precautions. China and Japan are already locking some places down again because of flareups.

What I expect will happen is that restrictions will be lifted so that kids can return to school in the fall. We'll know by Halloween whether that prompts a deadly reemergence of the virus or not.

[Tyler9000]

Fair warning not to read much into any numbers coming out of China:

China has concealed the extent of the coronavirus outbreak in its country, under-reporting both total cases and deaths it’s suffered from the disease, the U.S. intelligence community concluded in a classified report to the White House, according to three U.S. officials.

The officials asked not to be identified because the report is secret, and they declined to detail its contents. But the thrust, they said, is that China’s public reporting on cases and deaths is intentionally incomplete. Two of the officials said the report concludes that China’s numbers are fake.

https://www.bloomberg.com/news/articles ... gence-says

[Ego]

NYC by Zip Code. Longer subway commutes correspond with more positives.

[ertyu]

I have an issue with this "china has concealed." It was known from the get-go how they report numbers. E.g. it was known they did not count asymptomatic cases. This is not new information. They wavered whether they should include them or not, decided not to. I remember reading in an article by some scientist or another, yeah this is an epidemic, how you report these numbers isn't so important as is being clear on how you're arriving at the numbers so people have a clear idea what those numbers mean.

If you think about it, the US is "concealing" asymptomatic and mildly symptomatic cases because of who gets tested. It's a limitation in testing, as was China's iffy "quadratic" curve - that was just how many tests they could process. But the thing is, none of these things are evilly obfuscated. This was all known to the international community and to the WHO. This "china has concealed" started as a PR spin after the Trump administration dropped the ball so absolutely spectacularly.

As for the deaths: this was also known. For example, it was known that if someone with heart disease dies of heart failure while having covid (covid made the heart failure worse), then this person is recorded as having died of heart failure. It was also known that, for example, people dead of "pneumonia" who couldn't be tested due to resource constraints were reported as dead of pneumonia because that's as much as was known for sure. This is what the WHO meant when they said the Chinese numbers are legit - they are internally consistent with their own reporting rules, and they have made the reporting rules transparent.

Another thing is, even Italy (Bergamo) simply lost count of the fatalities. The system was too overwhelmed, no one was counting. In the end, the recorded count is lower than the difference between the number of people who die in the city in a regular month and the number of overall fatalities now. Spain has the same issue: they can barely cope with removing the bodies of people who die at home in a timely fashion, let alone test each dead body for covid or not covid.

Tl;dr: I don't think the Chinese and Iranians were pure and innocent. I think the Iranian numbers are woefully underreported. But at the same time, the degree of deliberate maliciousness ascribed to them is exaggerated by the american PR machine. The fact is, every country will underreport, some of them due to logistical limitations, like Italy and Spain, and some of them due to political limitations, too - and btw, I include the US in this number. I don't trust Trump further than I can throw him. He himself has admitted a concern with "how the numbers look" - when he did not want to let that ship dock because he didn't want his numbers to look worse when "it wasn't his fault". If he can use political pressure at all, he will use it to prettify the numbers. Hell, at this point I wouldn't be surprised if the powers that be see such prettifying as a matter of national security/preventing civil unrest.

/rant.

Tl;dr: finger-pointing is simplistic and misinformed.

[7Wannabe5]

@Ego:

As I noted previously, I am not a math whiz, so I was having a hard time trying to explain why you can’t just sub-out higher R0 with assumption of greater percentage asymptomatic with death rate, but here goes...

Consider an easier model with R0 =2, incubation = 5 days, incubation until death = 10 days, and death rate = 50%.

If you project deaths every 5 days, starting on Day 10, it looks like .5,1,2,4,8,16...

OTOH, if you assume RO=4 (half cases asymptomatic) with death rate of only 25% then the projected deaths recorded every 5 days would be .5, 1, 4, 16, 64, 256...

So, you could go with an even lower death rate of 12.5%, but then your projected deaths would look like .25, .5, 2, 8, 32, 128...

IOW, the pattern of deaths will look different no matter how high a level of asymptomatic you assume or how low a death rate you assume in attempted coordination.

Please somebody correct my math if I am wrong!