Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Coronavirus

This site may earn commission on affiliate links.
Also any large gatherings tends to have busy restrooms, and I have read that public restrooms are a scary place when it comes to getting exposed to this virus.

Knowing this makes it pretty easy to minimize.
a) hire restroom attendants (there are plenty of people who need jobs)
b) they insure social distancing in lines and require everyone to wash hands before and after use
c) and hand out sanitizing wipes
d) clean the restrooms every so often
e) they wear masks, gloves, etc.
f) they require people to have some kind of mask on
 
Anyway, it's a very well written paper. I'm also happy to know that they calculated the same confidence interval as me on their random population sample.

It also provides a bit more clarity on the question of whether known infections (official "test cases") are excluded.

The text in general remains obscure especially about the randomly selected group. But the flowchart (assuming the person creating the flow chart truly had more knowledge than provided in the text) clarifies that "quarantined persons" are excluded. This suggests to me that those that have been in quarantine/isolation, but are not anymore, are included, also because it is possible to test them. However that remains an open question.

This would mean that about roughly 1,000 cases country wide (the "active cases" in worldometer, as of April 1-4), as opposed to all known cases, roughly 1,400 (the "total cases" in worldometer, as of April 1-4) have been excluded.

Note that numbers in the text are for the capital area, not country wide for Iceland, which further reduces the relative number of exclusions.

So the answer seems to be somewhere between "known case are included" and "known cases are excluded".

Unfortunately, the report makes no effort to calculate the total percentage of infected+sick cases. I might try to make that calculation later on.

Also noteworthy is that only about 1/3 of all randomly selected persons have participated, so the selection is not truly random.
 
I'd expect that the projections will increase significantly this week as the lagging death numbers continue to come in higher than the current curve fit and it fits a broader sigmoid.

I think the curve is now a composite of curves for each region, although I'm not completely sure if those briefing comments applied to the IHME model specifically.

Ah. That would make sense of the humpiness. So probably it's just a composite of sigmoids, so my prior denunciations of the model for being wrong on the downward slope of the curve stand. We'll see whether my criticisms stand the test of time.

Looks like Carl Bergstrom has the same concerns as I do about the performance of this error function (a type of sigmoidal curve) on the downward slope. Here is his analysis, it is worth a read through all the tweets. Also he confirms that the curve is the composite of symmetric curves as you pointed out:

Carl T. Bergstrom on Twitter

I think it would have basically flipped the curves. You know that "flatten the curve" chart with the high one and low one. For NYC really they got the high one, though they were shooting for the low one. Letting things go per usual (with some non-disruptive measures like good hygiene) would have stretched out the curve. Peak would have come later, as is happening in Sweden. Or like in UK as opposed to Italy.

That is some pretty iconoclastic thinking there!
 
Last edited:
  • Disagree
Reactions: dqd88
No group size restriction for churches
You sure that's correct for Louisiana?

He was arrested and charged already in that state earlier (e.g. Louisiana pastor charged with defying coronavirus order against large gatherings).

Cited Central church expected 2,000-plus for Easter amid coronavirus; it drew far fewer mentions
"“We’re going to document everything and send it to the district attorney’s office,” Corcoran said. He added that the church hasn’t attracted more than a few hundred people since his office began monitoring church services."

Corcoran is the police chief.

OT, this CNBC reported survived COVID-19 (I survived coronavirus. There's no road map for what comes next) but her family who also got sick was never tested so that leaves them in a bind and unsure.
 
Was staying up to see Tuesday’s Santa Clara County’s dashboards but surprisingly not posted yet.
Coronavirus (COVID-19) Data Dashboard - Novel Coronavirus (COVID-19) - County of Santa Clara currently says
April 14, 2020 Update: Due to system issues with the California Reportable Disease Information Exchange (CalREDIE) managed by the California Department of Public Health, the County of Santa Clara's COVID-19 Cases and Laboratory Testing Dashboards will not be updated for today. We will update the data in the dashboards as soon as possible.
That message was there a few hours ago, as well.
Rural America also has fewer hospitals and beds, ICUs, and ventilators. Not forgetting to mention doctors, nurses and other healthcare workers. I think their wave is coming. Lots of good, hard working people producing our food.
Not sure what was up with the disagree. That person can read or listen to Rural Hospitals Brace For Coronavirus which I listened to weeks ago.
 
  • Like
Reactions: SMAlset
Looks like Carl Bergstrom has the same concerns as I do about the performance of this error function (a type of sigmoidal curve) on the downward slope. Here is his analysis, it is worth a read through all the tweets. Also he confirms that the curve is the composite of symmetric curves as you pointed out:

Carl T. Bergstrom on Twitter

I have also already posted being concerned about the post-peak modeling. (As did David when pointing out that the curves hit zero quite soon.)

EDIT: I'm not sure if they are exactly symmetric, but their post-initial-peak form seems too optimistic, and not just because of when they hit zero.
 
Last edited:
  • Like
Reactions: AlanSubie4Life
I have also already posted being concerned about the post-peak modeling. (As did David when pointing out that the curves hit zero too soon.)

Yeah, quite a few people have complained about the rapidity of the drop to zero. Anyway, not clear exactly where we'll end up in terms of a total. The current 68k estimate seems a bit optimistic to me. But I think after the next 5-7 days or so it should be a lot more clear. It's apparent there has been some significant undercounting in NY City as well (which may explain the relatively low mortality reported in NY state in the daily counts), though presumably there was a certain amount of that in other countries as well. I don't track each country's exact method of counting deaths.
 
Yeah, quite a few people have complained about the rapidity of the drop to zero. Anyway, not clear exactly where we'll end up in terms of a total. The current 68k estimate seems a bit optimistic to me.

The prediction of 69K from April 13 is an increase from 61K previously, down from 94K on April 1. I don't think this is just the result of changed mitigation behavior, but also of the model being highly sensitive to fluctuations (noise) in recent data.

But I think after the next 5-7 days or so it should be a lot more clear.

I agree. There may be two things: the form of the peak period, and that of the tail.

It's apparent there has been some significant undercounting in NY City as well (which may explain the relatively low mortality reported in NY state in the daily counts), though presumably there was a certain amount of that in other countries as well. I don't track each country's exact method of counting deaths.

Yes, I wonder how that recently reported missing data will be integrated into the IHME model and into the worldometer graphs.
 
1) Belgium includes each death presumably caused by Corona in our death stats. If we would use the same way of counting as in The Netherlands (i.e. only count hospital deaths with confirmed Corona test), we’d have the same death count per million as The Netherlands.
2) The majority of the deaths now comes from elderly people in nursing homes. These were the first to lock down, even earlier than the general lockdown. But the personnel of the nursing homes had no protective materials and so we got some infections in the nursing homes, and rapid spread. Apparently elderly people can very easily infect other people with Corona. I’m told this is very unexpected, for other infectueus diseases just the opposite is true, and our doctor/medical personnel didn’t expect this.
3) We have a fairly ‘lite’ lockdown (compared to Spain and Italy), people can still go to work if social distancing is obeyed, and we can still go outside to walk/bike/run.
4) Unfortunately a minority of the population violates the lockdown rules.

Some extra info about today’s record 2.5K new cases in Belgium. This record number is caused by a government effort to test all residents and personnel in nursing homes, for which the results are now coming in (results from about 10K extra tests yesterday).
The new cases from the regular tests (i.e. people that matched the normal test criteria, about 3K tests) had about 600 new cases, which is similar to the day before and at a 3 week low level.
Edit: above are approximate numbers as I remembered from the official daily press conference less than an hour ago.
 
Aside: Looks like @dqd88 may have been right - the New York Pause is likely increasing family transmission.

Sorry for commenting without having followed that particular discussion, but if you shut down all or most other transmission vectors, of course family transmission will increase percentage wise.

On the plus side, this may be one reason why the curves take longer to decay and have an unexpected "plateau" at the top. The apparent R0 stays above 1 for a bit longer while each patient infects most of their family. After that goes away after another couple weeks, we should start to see R0 decay below 1 (hopefully).

Yes, all instances that reduce R0 but leave it above 1 should have a delaying effect. An additional cause may be the distributing effect of the spread reaching hot spots and sub-hot-spots at different times (the variation in geographical or social density), which might be larger in the US than elsewhere.

How about those unknown exposures, though? Crazy.

One possible explanation might be that shopping and using services are not included in "social", but in "unknown". I'm surprised that shopping/services is not a separate category, but perhaps that is because it would be difficult to know that this was the source of infection, if it was.
 
ECDC data through Apr 14

I should follow my own advice and not pay attention to blips -- particularly when the graph is depicting mortality data
The weekend/holiday numbers are corrected now and the the countries continue on their plateaus

Screen Shot 2020-04-15 at 5.36.13 AM.jpg
 
A lot of people are looking at Sweden to see what could be the outcome of less intervention. Some people(like Trump) want to use Sweden as a warning example to argue why strong interventions are better, some look at Sweden to show that less intervention is better. Whatever the outcome, at least it’s an interesting experiment that might be useful for policy decision for other countries.

Anyway, I was surprised today when the first report after the Easter holiday came in, +169 new deaths spread out over the last few days was a lot less than I expected. Number of new confirmed cases are down, number of new deaths averaged over the last few days seems to be going down, numbers of patients currently in ICU are down while number of available ICU beds are growing. Maybe numbers are lagging, we will see over the next week, but the health organisation are optimistic and think we are past the peak. From what I have seen of this virus, I think we should be careful to think that the worst is behind us, but for now it is looking cautiously optimistic.

Some people have been saying that Sweden is going for herd immunity. The Swedish health organization has never claimed this, what they are saying is that they are trying to keep numbers low, slow down the spread and flatten the curve. Imo they have been very unclear with what is their strategy, sometimes it sounds like herd immunity, but they are denying this. Their stated goal is to protect the old and vulnerable, which imo they have failed at. They also want to have a sustainable response that we can keep for a long period of time. Also they have tried to bring into equality into the equation, as an example not everyone can work from home, thus it was wrong of Spotify to tell their workers to work from home. They got a lot of angry comments for this statement, but if one reads their mission statement, equality seems about as important as saving lives.

FWIW my own hypothesis is that Sweden has a lot going for it that will be hard to replicate in other countries. We have clean air, a high number of people with the CCR5-delta 32 mutation, we are a very individualistic country, seldom live many generations in one household, a lot of people supplement with vitamin D, and we have been practicing social distancing since before the virus. Maybe our CFR will be lower than in Wuhan, Milano, Madrid, NYC etc. Initial numbers seems to indicate that 10-20% of people in Stockholm could have had the virus already.

Tl;dr Sweden will be used as an example to argue for less intervention...
 
Last edited:
The Swedish health organization has never claimed this, what they are saying is that they are trying to keep numbers low, slow down the spread and flatten the curve
Tomaeto, Tomahto

Every place agrees that it is a bad idea to overwhelm the healthcare system. After that each jurisdiction will eventually (say, over the next ~ 2 months) choose some equilibrium between hospitalization/mortality rates and behavior modification, but they will all (obviously) open their economies.

At one extreme: Asia
At the other extreme: Trumperville

Most will be in between
 
Last edited: