COVID-19 Coronavirus Containment Efforts

[DrClaude]

Here's another of the hated Sweden-Denmark comparison. I added UK and Germany for fun.

View attachment 264000

This is the 7-day running average of fatalities, normalized to the peak and (this is new) plotted against the days since the peak.

My conclusions:

1. The curves are closer to each other than I guessed before making the plot.
2. Our pariah nation, Sweden, is presently doing better than Norway, Germany and the UK. It's getting harder and harder to say "Every life matters! We need to get off the blue curve!"
3. Our pariah nation, Sweden, is not doing hugely better than Norway, Germany and the UK. It sums to 4%. or 179 people. I'm willing to believe that this is a downward fluctuation, and maybe the true value is not -4% but really +1 or +2% but not much more than that.
4. The upward part of the curve has some interesting features.

Considering the percentage of people with antibodies in Stockholm (7.3%), it really does seem that the distancing measures taken in Sweden were sufficient. The high number of death is probably a mixture of bad management of the situation with respect to the elderly and bad luck.

 

[Vanadium 50]

U.S. week/weekend of protests that have turned chaotic (sometimes with people not wearing masks in these crowds (including a reporter I saw) might lead to a COVID-19 spike.

Nonsense.

Because math.

The fraction protesting is about 10^-4. I estimate this from news reports that say "hundreds" for local protests and "tens of thousands" in total. The case fatality rate is a few 10^-4 for that age group (predominantly young), so we'll say 10^-8 in total (not all exposed become infected, and not all who are infected become cases). So if everyone protesting is infected -obviously an upper limit - that's 3 additional deaths: a 0.003% increase to the US total. Probably closer to 0.001%. (And less than the 9 killed directly in the rioting)

The protests during a medical pandemic that's killed over 100,000 seems immoral to me.

I don't wish to be lectured on morality by Mr. "Kill Off the Weak". I suspect I am not alone.

 

[Vanadium 50]

The high number of death is probably a mixture of bad management of the situation with respect to the elderly and bad luck.

Whatever the source is, policies after the peak cannot possibly affect the size of the peak.

I thought earlier someone say the US and UK were out of control, but I can't find it. The UK seems to be on the same curve as everyone else. The US is a special case, because it's a big place. New York state has 6% of the population but a quarter of the fatalities. Two-thirds of the cases are in New York City, which has 40% of the population. Sure, NYC is out of control, but the vast majority of the country is not.

 

[bob012345]

Here is an example of the strictly enforced UK COVID-19 containment efforts:

View attachment 263995

I'm not convinced this situation is so bad. People are spaced away from non relatives/friends and they don't spend a lot of time in close contact with strangers as well as being outside as opposed to a confined space. As I understand it, it takes several minutes of sustained exposed to an infected person to get infected.

 

[Ygggdrasil]

Here's the COVID-19 fatality data (downloaded from the ECDC) plotting the 7-day moving average of fatalities per 1M population for Sweden, Denmark, Germany, UK and US.

In terms of deaths per capita, the UK is clearly much worse than the other countries, while the US and Sweden have similar deaths per capita. As noted, the dynamics of the curves are fairly similar between the countries despite differences in government-imposed social distancing interventions. This observation could reflect research suggesting that (at least in the US), social distancing occurred well before governors began implementing stay at home orders, and that individuals' actions/reactions to public health advisories have been more important than strict imposition of stay at home orders by governments.

I'd agree with others that the relative sizes of the peaks likely do not reflect differences in the social distancing policies imposed by the countries, but instead may just reflect the number of unidentified cases present in each country prior to people beginning to socially distance. The number of unidentified cases likely has contributions from factors that governments can't easily control (e.g. the volume of travel from areas with transmission of the virus, population density, climate) as well as factors that reflect differences in governments' responses (e.g. testing and contract tracing in countries like Singapore, Germany and South Korea could have kept the number of unidentified cases low, whereas the lack of these capabilities in the US early in the outbreak could have contributed to higher numbers of unidentified cases).

The absolute number of cases is important as governments begin to exit from stay at home orders because we'll need to rely on contact tracing to limit the spread of new infections, and a large number of cases can overwhelm public health agencies' ability to trace contacts and isolate infectious individuals. However, the Sweden data do suggest that limited social distancing measures that keep businesses and schools open can keep the spread of the virus contained (e.g. R < 1).

 

[anorlunda]

As I understand it, it takes several minutes of sustained exposed to an infected person to get infected.

We see that quoted quite often. I've seen 15 minutes as the needed time.

But when I'm out in the grocery store, I assume that all the other people are infected. Therefore, it should be 15 minutes exposure to any collection of people taken one or more at a time, rather than 15 minutes exposure to the same person.

I also believe that talking has a lot to do with it, and the presumption is that people within 6 feet for 15 minutes are talking with each other. Passing other people in the aisles doesn't count as much if they are not talking. When I see a woman in the store yelling at her kids, I stay 30 feet away.

 

[PeroK]

I thought earlier someone say the US and UK were out of control, but I can't find it. The UK seems to be on the same curve as everyone else.

The UK was on the same curve as the other European countries, but we stepped out of lockdown earlier: in the sense that we were at least two weeks behind the other countries when we eased restrictions. The UK has more new cases than Italy, Spain, France and Germany combined. The question is whether we can continue on that downward curve, given the apparent laxity of our lockdown regime. Only time will tell, but the fact remains that we (unlike the others) did not continue with lockdown until we got the numbers right down.

The USA is, of course, about six times larger than the major European countries. Taking that into account, the numbers are still way higher. For example, in the week-ending May 26th, the USA had 22,000 new cases per day, compared to about 2,000 per day in total across Italy, France, Spain and Germany (that's 500 new cases per day per country). The UK had 2,350 new cases per day that week.

The point is clear that the major European countries have got the numbers down, whereas the UK and USA have certainly not done that yet.

 

[bob012345]

We see that quoted quite often. I've seen 15 minutes as the needed time.

But when I'm out in the grocery store, I assume that all the other people are infected. Therefore, it should be 15 minutes exposure to any collection of people taken one or more at a time, rather than 15 minutes exposure to the same person.

I also believe that talking has a lot to do with it, and the presumption is that people within 6 feet for 15 minutes are talking with each other. Passing other people in the aisles doesn't count as much if they are not talking. When I see a woman in the store yelling at her kids, I stay 30 feet away.

When I shop I'm probably getting less than one minute of total exposure within six feet of anyone. Maybe only 30 seconds worth. One caveat is if the virus is hanging around thick in the air but then I shop very early in the morning and the store has not been open too long and is basically empty. This is probably overkill but when I get home, I spray the bottoms of my shoes with a diluted bleach solution so as not to track potential contamination into the house.

 

[Vanadium 50]

The UK was on the same curve as the other European countries, but we stepped out of lockdown earlier

Compare with Sweden, they never locked down.

The USA is, of course, about six times larger than the major European countries. Taking that into account, the numbers are still way higher.

But much of this is coming from a single city. The rates are 60x lower elsewhere.

 

[PeroK]

But much of this is coming from a single city. The rates are 60x lower elsewhere.

This is simply not the case. The numbers are coming in from all over the USA. E.g. just today:

https://www.worldometers.info/coronavirus/country/us/

Florida +617
Maryland +848
Virginia +841
Indiana +407
Minnesota +300
Arizona +1127

And, in total over 6,000 new cases today already, without any figures yet from the hardest hit states like NY. By contrast:

Germany (population 80 million): +168.
Italy (population 60 million): +318.

 

[Vanadium 50]

This is simply not the case. The numbers are coming in from all over the USA. E.g. just today:

Florida +617

Florida is about the same size as New York state. FL deaths are at 2400. New York state is at 24,000.

 

[Jarvis323]

I don't understand peoples' fixation with the beach as a potential hotspot. There's not much more inhospitable a place to virus transmission than a beach, and people really do work hard to social distance on a beach.

A buddy of mine who lives in Huntington Beach, Ca complains about this constantly, but I have a suspicion it's really a cover for the parking situation.

This has frustrated me. The beaches near me that I've grown up going to are never packed with people. There are hundreds of miles of rural coastline. Often when I would go to the beach, there would be nobody, or only one or two people. Social distance is usually something like 150 feet. The cool breeze also quickly diffuses particles into the air. It is one of the least transmissible environments.

This is what they look like in terms of crowds:

Anyway, these coastlines still have been closed. Recently they opened some up, but with the caveats that you must live within 50 miles, you can only go in the morning or evening, and parking anywhere is illegal. But these beaches are basically not safely accessible without driving. Unless you have someone to drop you off, or you happen to have beach front property, you really still can't go.

It's frustrating, along with the blanket closures of outdoor recreational area's, especially when we're all taking much larger risks regularly shopping for food and doing other things. And not really that much is shut down, the skate shop is open, car dealerships are open, department stores are open, hardware stores are open, cannabis shops are open, coffee shops are open, and liquor stores are open.

Still I understand the motivation to close these beaches down and ban other safe outdoor recreation. It's really to curb travel. They don't want people coming in from other areas and bringing the virus there.

 

[russ_watters]

This has frustrated me. The beaches near me that I've grown up going to are never packed with people. There is some 500 miles of rural coastline. Often when I would go to the beach, there would be nobody, or only one or two people. Social distance is usually something like 150 feet. The cool breeze also quickly diffuses particles into the air. It is one of the least transmissible environments.

This is what they look like in terms of crowds:

The pictures we see that anger people and the media and some politicians call a problem are like this one:

https://www.theguardian.com/us-news/2020/apr/27/california-beaches-coronavirus-orange-county

There could be a thousand people (guess) in this photo ( ). But a thousand people over a (guess) length of a quarter mile and width of 100 yds is roughly 400 square feet per person or an average of 20'x20' per person! If you consider that most pare probably family units or small groups of friends in 2s, 3s, 5s, etc., the spacing is multiplied. This isn't a photo of violating guidelines, it's a photo showing excellent social distancing.

...and then there's the wind.

 

[mathwonk]

Now we are talking my kind of prevention (math themed face masks); maybe I could start up a mask service with Riemann Roch formulas on them:

https://www.etsy.com/listing/780362...her+face+mask&ref=sr_gallery-1-24&bes=1&col=1

 

[bob012345]

Now we are talking my kind of prevention (math themed face masks); maybe I could start up a mask service with Riemann Roch formulas on them:

https://www.etsy.com/listing/780362...her+face+mask&ref=sr_gallery-1-24&bes=1&col=1

What rating do you think they have?

 

[bob012345]

This has frustrated me. The beaches near me that I've grown up going to are never packed with people. There are hundreds of miles of rural coastline. Often when I would go to the beach, there would be nobody, or only one or two people. Social distance is usually something like 150 feet. The cool breeze also quickly diffuses particles into the air. It is one of the least transmissible environments.

This is what they look like in terms of crowds:

View attachment 264016

Anyway, these coastlines still have been closed. Recently they opened some up, but with the caveats that you must live within 50 miles, you can only go in the morning or evening, and parking anywhere is illegal. But these beaches are basically not safely accessible without driving. Unless you have someone to drop you off, or you happen to have beach front property, you really still can't go.

It's frustrating, along with the blanket closures of outdoor recreational area's, especially when we're all taking much larger risks regularly shopping for food and doing other things. And not really that much is shut down, the skate shop is open, car dealerships are open, department stores are open, hardware stores are open, cannabis shops are open, coffee shops are open, and liquor stores are open.

Still I understand the motivation to close these beaches down and ban other safe outdoor recreation. It's really to curb travel. They don't want people coming in from other areas and bringing the virus there.

I'm really wondering what's just beyond that dune...

 

[mathwonk]

@bob012345: "What rating do you think they have?" I'm not sure, but I was also worrying that if they are too interesting, they might attract curious people to come close and read them.

 

[mfb]

Here's another of the hated Sweden-Denmark comparisons. I added UK and Germany for fun.

View attachment 264000

This is the 7-day running average of fatalities, normalized to the peak and (this is new) plotted against the days since the peak.

My conclusions:

1. The curves are closer to each other than I guessed before making the plot.
2. Our pariah nation, Sweden, is presently doing better than Norway, Germany and the UK. It's getting harder and harder to say "Every life matters! We need to get off the blue curve!"
3. Our pariah nation, Sweden, is not doing hugely better than Norway, Germany and the UK. It sums to 4%. or 179 people. I'm willing to believe that this is a downward fluctuation, and maybe the true value is not -4% but really +1 or +2% but not much more than that.
4. The upward part of the curve has some interesting features.

If you align them by onset of the pandemic (let's say 0.2 of the peak) you get completely different conclusions. Now the path until half of the peak looks virtually identical between the four countries, but they deviate later. Suddenly Sweden dragged out the pandemic longer than Denmark and the UK (peak comes later) and it doesn't recover faster either. Funny how a different presentation can completely change the interpretation, isn't it? The day of the highest reported death count depends on details of the reporting and accounting, and a bit of randomness as well.
Normalizing to the peak could be interpreted as "here, your risk to die doesn't matter so much if many other people die around you". Don't normalize to the peak and suddenly the graph looks very different again.

The fraction protesting is about 10^-4. I estimate this from news reports that say "hundreds" for local protests and "tens of thousands" in total. The case fatality rate is a few 10^-4 for that age group (predominantly young), so we'll say 10^-8 in total (not all exposed become infected, and not all who are infected become cases). So if everyone protesting is infected -obviously an upper limit - that's 3 additional deaths: a 0.003% increase to the US total. Probably closer to 0.001%. (And less than the 9 killed directly in the rioting)

• That's assuming no one infected from demonstrations infects anyone else, which is quite an interesting assumption.
• The fatality rate looks too low as well. Reported case fatality rates for young people are ~10^-3 and higher.
• I count 13 protests described as "thousands of people", one "over 3000", one "more than 3000", one "at least 5000", and too many demonstrations with 1000-2000 people to count. And that's just the US, and only the ones I found by searching for "thousand" and "000". 10^-4 or 30,000 is too low.

Sure, NYC is out of control, but the vast majority of the country is not.

It is not out of control, but still at significantly higher rates than e.g. in continental Europe. Sure, it started later, so we can also expect it to go down later, and that's what we see.

Germany had 220 deaths in the last week in a population of 80 million, or ~3 per million per week.
France had 400 deaths in the last week in a population of 70 million, or ~6 million per week.
Italy had 560 deaths in the last week in a population of 60 million, or ~9 million per week.
Spain had 260 deaths in the last week in a population of 50 million, or ~5 per million per week.

Alabama had 70 deaths last week in a population of 5 million, or 14 per million per week.
Alaska didn't have deaths but also has a population of just 0.7 million in a giant area. 0 per million per week.
Arizona had 110 deaths last week in a population of 7 million, or 15 per million per week.
Arkansas had 16 deaths last week in a population of 3 million, or 5 per million per week.
California had 460 deaths last week in a population of 40 million, or 12 per million per week.
Colorado had 130 deaths last week in a population of 6 million, or 21 per million per week.
I didn't pick specific states, I just took the first few by alphabet. Apart from Colorado they are all below the US average: The US overall had 6500 deaths in a population of 330 million, or 20 per million per week.
The life expectancy in western countries is about 4000 weeks, so 250 deaths per million per week is the normal background rate. Which means 1 in 10 deaths in Colorado and the US overall is from COVID-19 at the moment.

 

[Vanadium 50]

If you align them by onset of the pandemic (let's say 0.2 of the peak) you get completely different conclusions.

Sure, but now you're talking about actions taken at the onset of the pandemic. That tells us something about the post-peak conditions, but not about the impact of post-peak policies. Personally, I'd think this tells us something more important: we should be looking at the leading edge and a possible second wave and not about the trailing edge.

You want to quibble that 10^-4 should be two or even three times that? Have at it.

 

[mfb]

I'm talking about actions taken over the whole period, affecting the whole period.

More infections means more people die and/or more regulations and/or a longer duration of these regulations. It's not about the few people who protest who might die (it's their own risk anyway), they might have a larger impact on the trend of infectious cases. That they are a small fraction of the population doesn't change their individual impact, which might be quite large. I don't say it has to be. I don't know. But I know it can be large. We had people who infected tens of other peoples in larger crowds.

 

[kadiot]

The pictures we see that anger people and the media and some politicians call a problem are like this one:

View attachment 264018
https://www.theguardian.com/us-news/2020/apr/27/california-beaches-coronavirus-orange-county

There could be a thousand people (guess) in this photo ( ). But a thousand people over a (guess) length of a quarter mile and width of 100 yds is roughly 400 square feet per person or an average of 20'x20' per person! If you consider that most pare probably family units or small groups of friends in 2s, 3s, 5s, etc., the spacing is multiplied. This isn't a photo of violating guidelines, it's a photo showing excellent social distancing.

...and then there's the wind.

Beach and Second Wave rhymes.

 

[kadiot]

.

Overheard: "Almost all COVID-19 patients are no longer contagious by the 11th day of illness. Korea and Singapore are changing their guidelines for clearing people from quarantine and are going for time-based clearance rather than test-based clearance. For context, RT-PCR can remain positive for up to 8 weeks, but people are no longer contagious by the 2nd week. This will allow us to record recoveries much faster (you used to have to have 2 negative RT-PCRs for clearance) and free up hospital and isolation beds.

Can you confirm this? Especially the highlighted bit.

 

[PeroK]

It is not out of control, but still at significantly higher rates than e.g. in continental Europe. Sure, it started later, so we can also expect it to go down later, and that's what we see.

I've organised the data taking population into account. Here are the average daily cases (per week) per 10 million people. I used 10 million in order to get India onto the scale.

Note: these are new cases per day averaged out over each week.

Cases per
10 million)	USA	UK	Spain	Italy	France	Germany	Russia	Brazil	India	Iran	Peru
07-Apr​	932​	633​	1,405​	704​	1,029​	611​	51​	56​	4​	306​	82​
14-Apr​	919​	813​	980​	635​	680​	418​	133​	76​	6​	209​	319​
21-Apr​	874​	740​	919​	507​	509​	277​	310​	120​	9​	169​	327​
28-Apr​	914​	675​	853​	438​	271​	195​	398​	201​	12​	133​	580​
05-May​	872​	712​	563​	272​	102​	121​	605​	281​	19​	126​	868​
12-May​	738​	662​	579​	194​	168​	105​	753​	423​	26​	184​	906​
19-May​	699​	470​	283​	129​	63​	79​	663​	634​	33​	236​	1,191​
26-May​	668​	345​	150​	91​	47​	59​	611​	810​	46​	254​	1,314​
02-Jun​	673​	268​	112​	70​	143​	48​	601​	1,105​	58​	307​	1,960​

 

[kyphysics]

0...

I mean the former. The claim was made by an Italian doctor but is still controversial.

https://www.dailymail.co.uk/news/article-8376105/Is-Covid-19-really-potent.html
https://orthospinenews.com/2020/05/...ations-may-be-weakening-this-is-virology-101/

It's a good topic to keep an eye on, but from the articles I've Googled, there is a lot of push back against these claims.

 

[kyphysics]

There is nothing to do about the riots as pertains to COVID-19.

I disagree, but will also avoid the topic, because of "social politics" involved.

I do think, as stated earlier, that allowing such mass gatherings (although, it's hard to control at this point) is dangerous during the time of coronavirus. I support the protests and their cause. I just think it's unwise and a public health hazard given what the entire world is experiencing right now with the virus. #end comments

 

[Point Conception]

I'm really wondering what's just beyond that dune...

I would guess a long remote coastline as far as can be scene until the next point. I would like to know what country? New Zealand? @Jarvis323

 

[OmCheeto]

I've organised the data taking population into account. Here are the average daily cases (per week) per 10 million people. I used 10 million in order to get India onto the scale.

Note: these are new cases per day averaged out over each week.

Cases per
10 million)	USA	UK	Spain	Italy	France	Germany	Russia	Brazil	India	Iran	Peru
07-Apr​	932​	633​	1,405​	704​	1,029​	611​	51​	56​	4​	306​	82​
14-Apr​	919​	813​	980​	635​	680​	418​	133​	76​	6​	209​	319​
21-Apr​	874​	740​	919​	507​	509​	277​	310​	120​	9​	169​	327​
28-Apr​	914​	675​	853​	438​	271​	195​	398​	201​	12​	133​	580​
05-May​	872​	712​	563​	272​	102​	121​	605​	281​	19​	126​	868​
12-May​	738​	662​	579​	194​	168​	105​	753​	423​	26​	184​	906​
19-May​	699​	470​	283​	129​	63​	79​	663​	634​	33​	236​	1,191​
26-May​	668​	345​	150​	91​	47​	59​	611​	810​	46​	254​	1,314​
02-Jun​	673​	268​	112​	70​	143​	48​	601​	1,105​	58​	307​	1,960​

Someone may want to check my maths. This can't be correct:

PeroK
new cases per day per 10 million	USA	UK	Spain	Italy	France	Germany	Russia	Brazil	India	Iran	Peru
2-Jun​	673	268	112	70	143	48	601	1,105	58	307	1,960
OmCheeto
2-Jun actual count	22,209​	1,769​	526​	420​	958​	398​	8,775​	23,205​	7,830​	2,548​	6,272​
cases not counted yet	14,800,000​	5,900,000​	4,000,000​	5,000,000​	4,300,000​	1,200,000​	360,000​	4,300,000​	700,000​	1,100,000​	550,000​
% cases not counted	89.0%​	95.5%​	94.3%​	95.5%​	95.9%​	86.3%​	46.2%​	88.6%​	77.3%​	87.3%​	76.5%​
dead	106,000​	39,400​	27,100​	33,500​	28,900​	8,560​	5,000​	31,200​	5,800​	7,900​	4,600​
years to count cases @ 2-Jun rate	1.8​	9.1​	21​	33​	12​	8.2​	0.11​	0.51​	0.24​	1.2​	0.24​

This is based on an "Infection Fatality Rate" of 0.64%, which I found in a paper yesterday.

If my numbers turn out to be correct, the point I'm trying to make is that daily case counts for these countries are somewhat meaningless, given the extraordinary number of uncounted cases.

If my numbers are incorrect, please feel free to delete.

 

[bob012345]

I would guess a long remote coastline as far as can be scene until the next point. I would like to know what country? New Zealand? @Jarvis323

I'm really wondering what's just beyond that dune... I was thinking more of this;

 

[PeroK]

If my numbers turn out to be correct, the point I'm trying to make is that daily case counts for these countries are somewhat meaningless, given the extraordinary number of uncounted cases.

If my numbers are incorrect, please feel free to delete.

The data is from:

https://www.worldometers.info/coronavirus/

This represents the number of people who have tested positive and the number who have subsequently died. The official number of cases in the UK stands at 280,000, and the number of deaths at nearly 40,000. That, however, is only those who have died after testing positive. The ONS (Office for National Statistics) is also counting the additional number of death certificates where COVID-19 is mentioned. This is, I believe, close to a further 10,000. So, we have somewhere between 40,000 and 50,000 confirmed deaths in the UK.

If you use an overall fatality rate of 0.64% (*), then this implies a huge number of unrecorded cases in most countries. I cannot comment on this, other than to say that anyone who claims such a low fatality rate has a lot of apparently contradictory data to explain. This has been raised several times in the previous posts. To repeat the one example of South Korea, who recorded 11,600 cases and 273 deaths (2.3%). Which implies that S Korean identified only one case in four - and yet effectively stamped it out.

The latest figures I have seen (for the UK) is that an estimated 7% of the population has already contracted the virus - i.e. about 5 million people. That would equate to a death rate of about 1%. In any case, this data is a significant unknown in most countries I believe.

The number of positive cases above is relevant because that represents a large proportion of those who have become seriously ill with COVID-19, which itself must be a reasonable measure of the state of the containment efforts.

(*) PS a fatality rate of 0.27% has been quoted.

 

[bob012345]

It's a good topic to keep an eye on, but from the articles I've Googled, there is a lot of push back against these claims.

There always is. It's much too easy to criticize new work. New ideas and results can get tabled if there's a rush to judgement.