COVID-19 Coronavirus Containment Efforts

[Vanadium 50]

Same source different chart: top 13 states for death rate are republican and 13 out of 15 of the states with the lowest death rate are democratic.

You've seen the full data. It doesn't show what you want, so you can cut it until it does. "We're going to make it political if it kills us! Literally."

Top and bottom states? NJ and VT. Both Deomcratic. That doesn't work. Top two? Add ME and MA. Both Democratic. That doesn't work either.

OK, let's go to 3. HI and CT. No, that doesn't work. Both Democratic. Number 4? AK and ND. Both Republican. Fifteen states? Still even, although you can make the difference a state depending on what you do with NH (Cook index is zero) and whether you include DC or not.

The full dataset shows no effect. It's possible to get any effect I want by a posteriori deciding my cuts and metric. I can pick my starting date. If June doesn't show what I want, I can try May. Or July. Ten states don't? Try fifteen. Maybe eight. Whatever it takes.

We wouldn't accept that in "real science" and we shouldn't accept it now. (I know, "...but politics is important!")

 

[fresh_42]

We have a mutation in GB!

 

[OmCheeto]

We have a mutation in GB!

This one?: New Coronavirus strain spreading in UK has key mutations, scientists say

Odd. It's dated two days ago. Seems like we would have heard about this immediately.

 

[anorlunda]

What's with those Vermonters, not wanting to join the party?

I'm a Vermonter. I'm in Vermont June-October and in Florida November-May.

All summer in Vermont I saw nearly 100% compliance with social distancing and masks. Instantly when I returned to Florida on November 1 I was shocked to see maybe 2/3 of the people in public with no masks. I take that as a clue to how much people changed their behavior.

By the end of November, Vermont (population 500K) had only 2 people hospitalized with COVID.

p.s. The NY Times has a dashboard you can customize to the state/county level. It's useful for comparisons. I'm not sure if it has a paywall.
https://www.nytimes.com/interactive/2020/us/covid-cases-deaths-tracker.html

 

[fresh_42]

I'm not sure if it has a paywall.

It worked without paywall or other restrictions, even for me.

 

[Borg]

I think they allow a certain number of visits before you're blocked.

 

[mfb]

This one?: New Coronavirus strain spreading in UK has key mutations, scientists say

Odd. It's dated two days ago. Seems like we would have heard about this immediately.

Why? There is no indication that this particular mutation is different from the thousands of others people have seen.

Looks like new daily cases have stopped increasing in the US in December, stable at ~200,000. That means new deaths could stabilize at ~2500/day soon, too (as weekly average).

 

[OmCheeto]

Why? There is no indication that this particular mutation is different from the thousands of others people have seen.
...

Exactly!

 

[Astronuc]

Why? There is no indication that this particular mutation is different from the thousands of others people have seen.

I posted about that somewhere, either in this thread, or in one of the threads in Biology and Medical forum. The significance is that "The new variant, which UK scientists have named “VUI – 202012/01” includes a mutation in the viral genome region encoding the spike protein, which - in theory - could result in COVID-19 spreading more easily between people."

 

[mfb]

"In theory" every mutation can do that, but without any evidence that this particular one does anything it's just one of many.

 

[Ivan Seeking]

You've seen the full data. It doesn't show what you want, so you can cut it until it does. "We're going to make it political if it kills us! Literally."

Top and bottom states? NJ and VT. Both Deomcratic. That doesn't work. Top two? Add ME and MA. Both Democratic. That doesn't work either.

OK, let's go to 3. HI and CT. No, that doesn't work. Both Democratic. Number 4? AK and ND. Both Republican. Fifteen states? Still even, although you can make the difference a state depending on what you do with NH (Cook index is zero) and whether you include DC or not.

The full dataset shows no effect. It's possible to get any effect I want by a posteriori deciding my cuts and metric. I can pick my starting date. If June doesn't show what I want, I can try May. Or July. Ten states don't? Try fifteen. Maybe eight. Whatever it takes.

We wouldn't accept that in "real science" and we shouldn't accept it now. (I know, "...but politics is important!")

Are you arguing that masks don't work or that party has no bearing on behavior? Are you denying that events like political rallies and Sturgis were likely superspreader events? I don't understand what you are trying to deny here.

Clearly you can't look at the first wave as you suggested given that no best practices were in place yet. How is that logically consistent? And infections are still spreading wildly, showing by far the highest density in areas that focused on rallies and public events.

 

[Vanadium 50]

Are you arguing that masks don't work or that party has no bearing on behavior?

Did I say any of that? Tell ya' what. You make your points and I'll make mine. That way you don't have to

Clearly you can't look at the first wave as you suggested given that no best practices were in place yet.

A posteriori argument. Feel free to make it, but don't pretend it's anything else.

 

[Ivan Seeking]

Did I say any of that? Tell ya' what. You make your points and I'll make mine. That way you don't have to
A posteriori argument. Feel free to make it, but don't pretend it's anything else.

I posted a map showing the CURRENT infection rates. You are using data going back to the beginning before best practices. Use the data I posted and your argument is moot.

 

[collinsmark]

I can't believe people are still arguing "My political tribe is doing better! You just have to look at the data the right way!"

OK, so my numbers proved unconvincing. Here's the scatter plot. Cook Political index is on the x-axis and deaths per million is on the y-axis.

View attachment 274435Do you see a trend there? I sure don't. People who are arguing that their political tribe is doing best are starting with the conclusion and fiddling with the data to support it. They sure aren't looking at the above data and drawing conclusions from it.

There is a striking correlation (at the moment) in the US between political trends and Covid-19 outbreak severity. But you won't find it that way for a couple of reasons:

1. Trying to find correlation between US states is too broad of strokes. A given state can have large metropolitan areas with a given political leaning while having large swaths of counties in other areas with the opposite leanings; all within the same state. It's better to get a bit more granular, like comparing counties.
2. It's not useful to include data early on in the pandemic. Very early on, the pandemic was spread by travel hubs, people were unaware of the proper precautions, and none of it was politicized.

Do you remember what it was like for those couple of weeks back in March? The whole country banded together. We were going to beat this virus, or at least pound it into submission for awhile to gain some time. We were all going to do it together. As a team. (Well, everybody except Trump who was spewing out some sort of unsubstantiated nonsense about how we should figuratively stick our heads in the sand because the virus is going to magically go away by Easter, "like a miracle.") Oh, I haven't forgotten those times.

I took the liberty of gathering some information summarizing the virus' spread and mortality rate by date and county, and comparing that to the 2020 presidential election results, also by county. I focused on a narrower timeframe, starting at Oct. 1st, up to the present. That let's us focus on the latest surge. The results mostly speak for themselves.

Figure 1. Scatter plot showing correlation between recent Covid-19 cases and voting trends, per county. Each dot represents a single US county/parish/municipality. (Note: Data for individual counties in Alaska were not available, thus all of Alaska is represented by a single dot.)

The correlation becomes pretty obvious when you look at Fig. 1. Each dot represents a single county. Counties on the right are those that showed larger support for Trump in the 2020 general election. Those on the left showed more support for Biden. A dot at 80 means that in that county, Trump had 80% of the vote. A dot at 20 means Trump only had 20% of the vote for that county.

Dots that are higher on the graph are counties that are getting hit hard by the pandemic right now.

The green, dashed line is a linear trendline confirming the correlation. If the pandemic was independent of voting trends, that line would be level with the horizontal. It's not even close to level.

You might be wondering why most of the dots are to the right of the 50% mark since Biden won the election. That's merely because there are more numerous, sparsely populated counties that leaned politically to Trump. Those to the left of the 50% mark, supporting Biden, while fewer in number, are more heavily populated on average.

Fig. 2 shows a similar plot but for Covid-19 deaths.

Figure 2. Scatter plot showing correlation between recent Covid-19 deaths and voting trends, per county (Note: Data for individual counties in Alaska were not available, thus all of Alaska is represented by a single dot.)

I've also attached a spreadsheet where one can analyze the data, if one wishes.

So what about causation? Well, that's complicated. I mean we all might be tempted to say that the unity the country embraced at the beginning of the pandemic fell apart when the red and tinfoil hat, conspiracy theorists started shouting bloody murder and mouth breathing about how a small piece of cloth on their nose is the most intense hardship they've ever experienced since they lost Civil War I. But I might caution against that since the relationship between correlation and causation can be a tricky one. (But, yeah.)

But since this thread was questioning the correlation aspect of it, I can and will say this: Yes, the correlation is there. It is real and undeniable according to my data.

----------------------------------------------------------------------
Sources:

Covid-19 data taken from the New York Times Covid-19-Data repository on Github. It's updated every day, usually around midnight, Greenwich Mean Time.
https://raw.githubusercontent.com/nytimes/covid-19-data/master/us-counties.csv

Population data was taken from US Census estimate for 2019 (it would appear the 2020 Census data isn't available just yet).
https://www.census.gov/data/tables/time-series/demo/popest/2010s-counties-total.html

Finding 2020 election data by county was a bit tougher. I mean it's widely available, but finding it all in one place was the thing. I ended up using this data from Tom McGovern on Github
https://github.com/tonmcg/US_County...2020_US_County_Level_Presidential_Results.csv
I compared a sample of counties against websites such as this one (https://abcnews.go.com/Elections/2020-us-presidential-election-results-live-map/), and it seems to check out from my limited sample. If you have any questions about the data, again, I've included the Excel file so anyone can critique it.

 

[russ_watters]

It's not useful to include data early on in the pandemic. Very early on... and none of it was politicized.

Hehe.

Even if that were true, which is isn't, there's an underlying assumption that the first wave has no impact on subsequent waves. But each wave necessarily lessens the severity of subsequent waves by reducing the pool of potential victims. The more severe the prior waves, the less severe subsequent waves will be.

 

[mfb]

I suggest to end the Democrat/Republican discussion here. It's questionable if it is within the forum rules, and it's getting really repetitive and it seems to take over the thread. Posting largely the same arguments yet another time isn't helping anyone.

 

[fresh_42]

If this ...

The significance is that "The new variant, which UK scientists have named “VUI – 202012/01” includes a mutation in the viral genome region encoding the spike protein, which - in theory - could result in COVID-19 spreading more easily between people."

... is true, then this ...

"In theory" every mutation can do that, but without any evidence that this particular one does anything it's just one of many.

... is unfortunately wrong, simply because the currently hyped vaccines other than other vaccines primarily attack named spike protein, i.e. it is not clear whether currently planned vaccinations protect against that new strain. However, I saw this on a news channel and the professor who was interviewed said, that it is unlikely that a mutation changes the spike. Maybe he had been caught on the wrong foot and he was not prepared. But he expressed his concerns that it won't remain in UK over the holidays. "It's just one of many" is an a priori assessment. Let's hope it will become an a posteriori assessment, too.

 

[atyy]

... is unfortunately wrong, simply because the currently hyped vaccines other than other vaccines primarily attack named spike protein, i.e. it is not clear whether currently planned vaccinations protect against that new strain. However, I saw this on a news channel and the professor who was interviewed said, that it is unlikely that a mutation changes the spike. Maybe he had been caught on the wrong foot and he was not prepared. But he expressed his concerns that it won't remain in UK over the holidays. "It's just one of many" is an a priori assessment. Let's hope it will become an a posteriori assessment, too.

There are already mutations in the spike protein. Currently, they are thought not to affect the spike protein enough to affect vaccine effectiveness. These mutations do affect the behaviour of the virus in cell culture, but at the clinical level these are thought to be of minimal significance.

https://www.nature.com/articles/s41541-020-00246-8
Experimental and in silico evidence suggests vaccines are unlikely to be affected by D614G mutation in SARS-CoV-2 spike protein

In principle, spike protein mutations can cause a problem for vaccines, however, back of the envelope estimates by Trevor Bedford suggest that this is likely to take a few years:
"If I had to guess, I would predict that #SARSCoV2 will behave similarly to existing seasonal coronaviruses in its ability to mutate to avoid vaccines and immunity.
...
So, my prediction is that we should see occasional mutations to the spike protein of #SARSCoV2 that allow the virus to partially escape from vaccines or existing "herd" immunity, but that this process will most likely take years rather than months."

 

[Astronuc]

United (I believe other airlines require the same) requires all its passengers to complete a "Ready-to-Fly" checklist before checking into a flight, which prompts them to acknowledge they have not tested positive for COVID-19 nor have they had any related symptoms in the past two weeks.

https://abcnews.go.com/Politics/man-dead-medical-emergency-united-flight/story?id=74802890

Yet, a man with symptoms of Covid-19 died on a flight.

United Flight 591 (Orlando (MCO) to Los Angeles (LAX)) was diverted to New Orleans on Monday after a male passenger became ill. Paramedics transported the individual to a local hospital where he was pronounced deceased, the carrier said.

If he had symptoms, he shouldn't have been flying.

 

[fresh_42]

If he had symptoms, he shouldn't have been flying.

This virus is an excellent measure on how far a society is altruistic or selfish. All individual measures are mainly to protect others rather than oneself. E.g. people who refuse to wear a mask and protest against it justify that with their personal freedom. A freedom to the expense of others. One should not complain about it because it is in accordance to the general commitment on individualism versus social responsibility.

 

[nsaspook]

Boris Johnson is holding a press conference amid rising Coronavirus infections. Some info about the mutation and its effect on current and future containment methods.

 

[Ygggdrasil]

Regarding the new N501Y mutation in the spike protein of SARS-CoV-2 identified in the UK, here's a press release from the scientific consortium that identified the mutation: https://www.cogconsortium.uk/news_i...ant-and-how-cog-uk-tracks-emerging-mutations/

The variant described today in the House of Commons contains a novel set of mutations associated with a lineage spreading rapidly in the South East of England (and more widely) that is the subject of ongoing investigations by the UK Public Health Agencies, coordinated by Public Health England and supported by COG-UK. This variant carries a set of mutations including an N501Y mutation in the receptor binding motif of the Spike protein that the virus uses to bind to the human ACE2 receptor.

Efforts are under way to confirm whether or not any of these mutations are contributing to increased transmission. There is currently no evidence that this variant (or any other studied to date) has any impact on disease severity, or that it will render vaccines less effective, although both questions require further studies performed at pace. We will provide further updates as our investigations proceed.

 

[Astronuc]

A severe variant of Coronavirus has been detected in South Africa which could explain the rapid spread of a second wave affecting younger people, the health minister said today.

Known as the 501.V2 Variant, it was identified by South African researchers and details have been sent to the World Health Organization, Zwelini Mkhize said in a statement.

https://www.rte.ie/news/world/2020/1218/1185398-south-africa-covid/

As of December 18, 2020 , South Africa had recorded 892,813 known cases of Covid-19 with 24,000 known deaths.
https://www.theeastafrican.co.ke/te...rus-variant-confirmed-in-south-africa-3233502

Looking at the picture from Soweto, it seems like an large proportion of folks are not wearing masks.

 

[Astronuc]

so far, more than 400,000 patient courses of both the monoclonal therapies from Eli Lilly and Regeneron have been allocated to the states, and more than 250,000 have been delivered -- but very little of what has been distributed is actually being used. ABC News confirmed the findings of a new U.S. Department of Health and Human Services report which shows that only 5-20% of that available supply has been used -- strikingly low uptake of what could be a “lifesaving intervention” for some COVID-19 infections as case counts surge across the country.

https://www.msn.com/en-us/health/medical/why-are-authorized-monoclonal-antibodies-rarely-being-used-to-treat-covid-19/ar-BB1c2QUn

Data for December 18, 2020
Currently hospitalized 113,955
Currently in ICU 21,745
Currently on ventilator 7,786

Deaths 305,127
Date, Deaths (cumulative), New Deaths

Dec 18, 2020	305,127	2,866
Dec 17, 2020	302,261	3,438
Dec 16, 2020	298,823	3,448
Dec 15, 2020	295,375	2,971
Dec 14, 2020	292,404	1,358
Dec 13, 2020	291,046	1,494
Dec 12, 2020	289,552	2,494
Dec 11, 2020	287,058	2,749

 

[atyy]

Further discussion by Trevor Bedford on antigenic drift.
- Some variants like mutations at position 501 (one of the features in the new UK variant) may be emerging independently in different part of the world, perhaps consistent with weak selection pressure
- In spite of the antigenic drift, he still estimates that if the vaccine will need to be updated, it will still be on the time scale of a few years (~ 3 years, is mentioned for the antigenic drift of other coronaviruses)

 

[atyy]

Here is the paper from Rachel Eguia and colleagues from group of Jesse Bloom, the basis of the ~3 years of other coronaviruses to escape immunity by antigenic drift. They've also tweeted is a very readable summary.

A human Coronavirus evolves antigenically to escape antibody immunity
https://www.biorxiv.org/content/10.1101/2020.12.17.423313v1

 

[OmCheeto]

It looks as though the UK is being heavily isolated this morning.



I noticed an upswing in deaths here in my state of Oregon last month, that is continuing. Even Vermont has an upswing. Not sure if these are due to the cold weather or this the new English mutant. From the following graph I made, it looks like it may have originated in France.

Deaths/million/day

 

[PeroK]

It looks as though the UK is being heavily isolated this morning.

It's a trifle harsh to close the English Channel to haulage (between the UK and mainland Europe), given the relatively low risk of letting haulage through and the huge impact this is having. But, if you elect a government that openly boasts contempt for our soon-to-be-ex European partners, then perhaps we can't complain if they put the boot in given the chance.

We are an Island Nation once more!

 

[fresh_42]

We are an Island Nation once more!

Well, you can't change geography. The resulting isolation, however, is self-imposed. Eire is an island, too, but ...

As far as the new strain is concerned, the current actions are pure populism and can only slow down spread. It is merely the assurance against a later "they knew and haven't done anything"-complaint. AFAIK it is already in Italy, Belgium, the Netherlands, and some I have forgotten. It is the kind of political marketing which brought you the Brexit. I guess it is a natural consequence of evolution: the virus adjusts its genome to his main host.

 

[mfb]

I noticed an upswing in deaths here in my state of Oregon last month, that is continuing. Even Vermont has an upswing. Not sure if these are due to the cold weather or this the new English mutant. From the following graph I made, it looks like it may have originated in France.

In a world where the virus is almost everywhere you don't need a specific source for a new larger outbreak. Just poorer control in the existing place is sufficient. The precise timing will depend on many things - local holidays, weather, end/begin of government intervention, ...

 

[OmCheeto]

In a world where the virus is almost everywhere you don't need a specific source for a new larger outbreak. Just poorer control in the existing place is sufficient. The precise timing will depend on many things - local holidays, weather, end/begin of government intervention, ...

Well, I'll be interested in reading the paper on why Germany, Oregon, and Vermont tracked so closely, for this current period.
They were all previously doing exceptionally well, and then, all hell seemed to break loose.

 

[russ_watters]

Well, I'll be interested in reading the paper on why Germany, Oregon, and Vermont tracked so closely, for this current period.
They were all previously doing exceptionally well, and then, all hell seemed to break loose.

Luck? Selection bias? The fact that they were doing well increased the likelihood of doing poorly?

 

[fresh_42]

Well, I'll be interested in reading the paper on why Germany, Oregon, and Vermont tracked so closely, for this current period.
They were all previously doing exceptionally well, and then, all hell seemed to break loose.

I doubt that you can formulate a hypothesis which survives any statistical test, means I assume this is independent from a random coincidence.

Informally and personally I assume that people who got away with a controlled environment became sloppy in their efforts due to the illusion of a false safety. Moreover there have been holidays between March and October, so people took the chance to get infected somewhere else! A second wave was likely from the beginning. And we should not forget that this is a highly dynamic system. The virus learns to adapt to his main host.

 

[OmCheeto]

...The virus learns to adapt to his main host.

Idiots?

 

[PeroK]

Well, I'll be interested in reading the paper on why Germany, Oregon, and Vermont tracked so closely, for this current period.
They were all previously doing exceptionally well, and then, all hell seemed to break loose.

Exponential growth and critical mass. Like a Uranium bomb?