Coronavirus Sars-CoV-2/Covid-19 Megathread

[Sheilbh]

The results are still out.  The southern states are betting that if the virus is allowed to spread uncontrollably in the under 50 population, transmission to older and more vulnerable people can still be constrained. Whether that is true remains to be seen.

Yeah - I think in the Northeast there were huge issues in failing to protect care homes. If Southern states manage to properly shield the elderly it could work - but they still need to be able to identify and control outbreaks before they turn into out-of-control transmission. I'm not clear they've done anything about that.

As I say I think this is part of the issue with the policy choice seeming to be framed as lockdown and safety v no lockdown and freedom. Those aren't the options.

Edit: And it is worth noting in that EU comparison that a number of EU states lifted lockdown at a similar time to some states. They have identified some outbreaks and contained them - like the meat-packing plant in Germany. But so far there's nowhere that's got out of control, yet.

Edit: Also, this, from the Economist is interesting - US casualties tend to be younger than in Europe. Presumably comorbidities is part of this?

[alfred russel]

AF stop with the nonsense of using Canada as a whole as a paragon comparator.  Quebec has done very poorly.  Compare Georgia with not Quebec and see how it looks.

Quebec is like a quarter or something of your population. Can I remove the worst performing quarter of Georgia's counties as well? Or is the fair comparison between all of Georgia and the parts of Canada that are doing well?

[alfred russel]

I don't think the number look damning bad for Georgia just yet, but I think you're not right to use absolute death numbers to judge policy.  I do think that the rate of change view is more relevant in this case, as long as you're careful to avoid situations such as "Mongolia has a stronger economy than the US, their GDP grows at 8% while ours grows at 3%". 

The reason why in this case first derivative view makes more sense is that once the first derivative in cases goes up, there are really only two outcomes:  loss of control over the disease, or lockdown.  The first derivative doesn't go down on its own until you get to a catastrophic situation.  Therefore, if either loss of control or a lockdown happens, we can safely call it as a failure of policy.

I'll grant that time may prove this approach wrong, but there have been a few reasons I've focused on death rather than other metrics.

-Cases are highly dependent on tests run, and who is being tested. There also seems to be variability in the quality of tests.
-Positive tests are not all equal. 100 positive tests in a nursing home is not equivalent to 100 positive tests of college students. The strategy of "lock away the old and sick, let the young run free" will look terrible if you just focus on testing.
-Hospitalizations are better, but also imperfect. At least in the US hospitals are facing an enormous financial crunch because of empty beds resulting from clearing the deck for Covid 19--they are under pressure to fill them now. Also, hospitalization criteria have been changing.
-The whole point of this enterprise is to reduce deaths. Counting bodies is most immune to issues (hard to ignore a body). It isn't perfect of course.

[alfred russel]

Oh, crap, should've checked my numbers before I posted this.  I download the public dataset every day, and do my own rate of change calculations by state.  This just in:  Georgia is now looking damnning bad.

Looking at the 7 day moving average of positive cases, it looks like Georgia is going up to the stratosphere.

Looking at the 7 day moving average of deaths, we are declining.

Looking at hospitalizations, we were at 1,500 May 1, declined to a low of 783 June 7, and have been steadily increasing from there. The rate of increase seems to be accelerating. We now have 1,124 hospitalized.

[DGuller]

I don't think the number look damning bad for Georgia just yet, but I think you're not right to use absolute death numbers to judge policy.  I do think that the rate of change view is more relevant in this case, as long as you're careful to avoid situations such as "Mongolia has a stronger economy than the US, their GDP grows at 8% while ours grows at 3%". 

The reason why in this case first derivative view makes more sense is that once the first derivative in cases goes up, there are really only two outcomes:  loss of control over the disease, or lockdown.  The first derivative doesn't go down on its own until you get to a catastrophic situation.  Therefore, if either loss of control or a lockdown happens, we can safely call it as a failure of policy.

I'll grant that time may prove this approach wrong, but there have been a few reasons I've focused on death rather than other metrics.

-Cases are highly dependent on tests run, and who is being tested. There also seems to be variability in the quality of tests.
-Positive tests are not all equal. 100 positive tests in a nursing home is not equivalent to 100 positive tests of college students. The strategy of "lock away the old and sick, let the young run free" will look terrible if you just focus on testing.
-Hospitalizations are better, but also imperfect. At least in the US hospitals are facing an enormous financial crunch because of empty beds resulting from clearing the deck for Covid 19--they are under pressure to fill them now. Also, hospitalization criteria have been changing.
-The whole point of this enterprise is to reduce deaths. Counting bodies is most immune to issues (hard to ignore a body). It isn't perfect of course.

Here is why I'm focusing on rate of growth of cases:

1)  In the long run it doesn't matter what the testing policy is.  Regardless of what share of cases you catch with diagnoses, after a one-time adjustment it will not have an effect on growth rates.
2)  Deaths follow cases, with some lag.  If your cases are growing, your deaths will be growing too in due time.
3)  Cases are more statistically credible.  Death numbers can be a little hard to work with, and in some states apparently some people come back to life due to Covid.  Metrics other than cases or death are not collected well enough on a uniform basis, so they're not even in consideration for me.
4)  There is no good scenario where case growth accelerates without the situation going unambiguously bad by all metrics, whether death or necessity of lockdown.

[Eddie Teach]

For those of you who don't want to read through the last 10-25 pages, I'll sumarise it for you:

Dorsey: The numbers don't look that bad, there should not have been strict measures
Everyone else: Yes they do look pretty bad, look here.
Dorsey: No they don't. I have found one that doesn't, see.
Everyone else: Yes but on average it's bad and its getting worse.
Dorsey: No
Everyone else: Yes, look, more numbers
Dorse: No those numbers I disapprove
Everyone else: Yes, look, more numbers


etc. etc.

More succinct-
Dorsey: The glass is half full.
Others: Bruh...

[Sheilbh]

But deaths lag cases by about, what 3-5 weeks? I agree on the other points - but I think number of cases is now valuable in the US context because there is a lot of testing. So you can probably look at that and percentage of positives.

Mexico looks worrying with it's huge percentage of positives, but their testing rate is very low so as much as anything it suggests they need more capacity to actually understand the scale. But if you already have testing at scale and you're getting a large and growing percentage of positives that suggests it's just slightly out of control.

So the 7 day moving average of deaths is a useful reflection of where you were 3-5 weeks ago, the question is whether cases are a useful indicator of where you'll be in 3-5 weeks time.

-Positive tests are not all equal. 100 positive tests in a nursing home is not equivalent to 100 positive tests of college students. The strategy of "lock away the old and sick, let the young run free" will look terrible if you just focus on testing.

-ish. Doesn't "lock away the old and sick, let the young run free" also rely on the workers who may come into contact with the old and sick being locked away? So 100 positive tests of college students looks very different if, say, 25 of them are student nurses.

[HisMajestyBOB]

Oh, crap, should've checked my numbers before I posted this.  I download the public dataset every day, and do my own rate of change calculations by state.  This just in:  Georgia is now looking damnning bad.

Looking at the 7 day moving average of positive cases, it looks like Georgia is going up to the stratosphere.

Looking at the 7 day moving average of deaths, we are declining.

Looking at hospitalizations, we were at 1,500 May 1, declined to a low of 783 June 7, and have been steadily increasing from there. The rate of increase seems to be accelerating. We now have 1,124 hospitalized.

And i cases and hospitalizations are independent of deaths, Georgia's deaths per capita will stay low. But that seems unlikely.

[HisMajestyBOB]

https://www.businessinsider.com/florida-scientist-state-stopped-tracking-icu-beds-ahead-july-4-2020-6

Jones also said multiple Department of Health sources have told her they've been instructed this week to change coronavirus numbers by "deleting deaths and cases" so it looks like Florida is improving ahead of July 4.

I bet Florida's official coronavirus numbers will continue to look surprisingly good.

Is there any other source on this than disgruntled former employee?

I haven't searched. I've only seen it attributed to that employee and to anonymous sources, but it was reported by the Post, among others.

[crazy canuck]

AF stop with the nonsense of using Canada as a whole as a paragon comparator.  Quebec has done very poorly.  Compare Georgia with not Quebec and see how it looks.

Quebec is like a quarter or something of your population. Can I remove the worst performing quarter of Georgia's counties as well? Or is the fair comparison between all of Georgia and the parts of Canada that are doing well?

Comparing a small geographic area to a country as vast as Canada without accounting for regional differences has some obvious deficiencies but if that is what you need to do to keep your narrative, carry on

[Sheilbh]

Major incident declared at Bournemouth beach which is rammed with little to no social distancing (and lockdown's not lifted yet). Also gridlock like a normal summer in Dorset with the roads filled with Grockles

Police are apparently on the beach issuing dispersal orders.

[Tamas]

Lemmings.

[crazy canuck]

Lemmings.

Rock climbers

[The Larch]

Major incident declared at Bournemouth beach which is rammed with little to no social distancing (and lockdown's not lifted yet). Also gridlock like a normal summer in Dorset with the roads filled with Grockles

Police are apparently on the beach issuing dispersal orders.

I've had to search "Grockles" to realize it means tourists in SW England from the rest of the UK.  Over here there was also quite some panic at the risk of our beach towns being flooded with tourists (mainly from Madrid) evading lockdown. We have similar epithets for them as well. 

[alfred russel]

More succinct-
Dorsey: The glass is half full.
Others: Bruh...

That isn't quite right.

I've maintained the more extreme lockdown measures were pointless and likely counterproductive. Some states implemented them late and revoked them early. The predictions on this forum were doom. They have not come to pass.

If we use April 24 as the start date (when Georgia really led the way in opening everything), for the first month plus I was hearing "cases are way up, deaths are lagging" (cases were up because testing was ramping up at that point). The doom predictions kept coming. After too long had passed for the absence of extra deaths to be attributable to lag, we get conspiracy theories about deleted or missing data or that cases really are just now spiking.

From April 24 until now, I've been hearing the excess deaths are just around the corner - and maybe they are. So far they haven't come.