COVID COVID-19 Coronavirus Containment Efforts

[collinsmark]

My iPhone is able to determine when I'm within a few feet of a device as opposed to just being in the same room, which suggests that it should be good about deciding if another phone is close enough to qualify as a possible exposure.

If you* live in an apartment building, as many do, your phone might be within a few feet from your neighbor's phone all night long, even though the phones are not even in the same room. Bluetooth signals can penetrate walls.

*I don't mean "you" in particular, but "you" in general.

 

[russ_watters]

It could help contact tracing quite a bit if many people use the app:

It still has the problems discussed a while ago, of course: Can it give a reasonable set of contacts without spamming contact tracers with everyone who visited the same mall at the same time?

No, the app is completely anonymous and all information is locally stored or encrypted, so it doesn't have the ability to do 3rd party contact tracing. There is no database of people who have come into contact with each other generated by the app/phone company/google, for the local health department to access.

The primary way it would assist in contact tracing is indirect: by telling you that you have been exposed, you then voluntarily contact your local health department. But that information is soooooo limited because you/it doesn't know who you came in contact with.

 

[russ_watters]

Apparently, many health authorities didn't have the resources and expertise to build and maintain an app, so Apple and Google have developed a simpler system. With iOS 13.7 and later, as long as your local health authority supports it, all you have to do is enable the feature on the phone. I'm not sure how Google is implementing the system on Android.

Sorry, but what app are we talking about here? You said North Dakota has two apps. As far as I know, Google/Apple only created one platform (not an app; authorities create/subcontract the app themselves). It would shock me if they also created a true tracking app, because this was a huge controversy.

The platform was pushed to phones several months ago -- I'm on Android and I got it; I assume iPhone users got it at about the same time.

The phones only exchange tokens if they're in proximity to each other for 15 minutes or more, so just walking by a person who laters reports being infected isn't going to result in a notification.

Actually, my app says it will. The tokens are saved, so if one of those contacts later reports an infection, I'll get a notification that I was exposed. I just won't know who it was.

My iPhone is able to determine when I'm within a few feet of a device as opposed to just being in the same room, which suggests that it should be good about deciding if another phone is close enough to qualify as a possible exposure.

The device doesn't need to know where you are (and for the google/Apple platform I'm aware of it explicitly does not), it just needs to know if you are near another bluetooth enabled phone. Since the range is only 30' and (I think...) they measure signal strength, that is pretty straightforward.

[edit]
Here's a timeline for what I see for N. Dakota:
4/7/2020 - "Care-19" App launched
https://www.health.nd.gov/news/north-dakota-launches-care19-app-combat-covid-19
From the description, it must use GPS (though it doesn't say "GPS") and it has nothing to do with the Google/Apple bluetooth-based platform, since it was launched long before that platform.

5/20/2020 - Apple/Google COVID tracker API released

6/24/2020 - Reference to the "second app" that was under development at the time, that uses the Apple/Google API:
https://www.statnews.com/2020/06/24/digital-contact-tracing-north-dakota/

8/13/2020 - "Care 19 Alert" App released, using the google/apple api:
https://www.kxnet.com/news/state-ne...reduce-spread-of-covid-19-as-students-return/

Looking at the google app store, and the North Dakota website, it appears to me there is only one app and that's the second one -- it appears to me that the google/apple based app replaced the gps based one:
https://www.ndresponse.gov/covid-19-resources/care19

In any case, it does surprise me that any state developed an actual location tracking app and surprises me that it didn't get more press due to the privacy concerns.

 

[mfb]

The phones only exchange tokens if they're in proximity to each other for 15 minutes or more, so just walking by a person who laters reports being infected isn't going to result in a notification.

Well, then it's completely useless for many people, me included. Public transport is the only place where I could get such a notification from - if I'm longer close to other people elsewhere then I know these people or at least can make it easy for contact tracers to find them (and vice versa). With such an app my phone would have a handful of entries from July (couldn't avoid public transport) and nothing else useful since February.

 

[OmCheeto]

Sorry, could not resist:

View attachment 269795

That's actually quite interesting about Germany. Only 1 US state had a flatter slope in that time period.
The slope being deaths per day per million.
You can also manipulate that number to yield the fraction of deaths that are due to Covid.
In Germany, 1 in 340 deaths was due to Covid.
Only Vermont had a better ratio at 1 in 380.
The next best was Maine at 1 in 68.
The other end of the scale isn't quite so good. I actually have trouble believing the numbers are that bad.
2 out of 7 deaths in Mississippi for the last 2 months have been due to Covid.

If you split the US into NY, NJ, CT, MA and everybody else, you get a very different story

If it weren't for maths, I wouldn't be able to tell the difference between the two US slopes.

 

[gmax137]

You can also manipulate that number to yield the fraction of deaths that are due to Covid.

How do you do that?

 

[jack action]

Amazon city of Manaus may have reached 'herd immunity'

 

[mfb]

0.10% of the population died in Manaus. Some parts of NYC had more deaths. It's possible that the official number underestimates the total deaths of course.

 

[russ_watters]

Well, then it's completely useless for many people, me included. Public transport is the only place where I could get such a notification from...

The intended value is that you would respond to the notification by getting yourself tested.

if I'm longer close to other people elsewhere then I know these people or at least can make it easy for contact tracers to find them (and vice versa).

Hopefully your relative would let you know, but if they register their status on the app before letting you know, the app would.

But you're right, without true contact tracing/location tracking the value is extremely limited.

 

[OmCheeto]

How do you do that?

First, you have to figure out what the average, or usual, death rate is. Globally and in the US, it's around 20 deaths per million per day.
If you add Covid deaths to that, and divide by Covid deaths, it yields the ratio.
Of course, the usual death rate varies by location, but it's close enough for my purposes.

An alternative, but dreadfully time consuming procedure:
The CDC has a list of the weekly tally of deaths for all the states from 2017 to the present.
Using Mississippi as an example, you can see that deaths for the time period from 2017 to 2019 are fairly consistent.
2020 showed a marked increase.
Dividing the total deaths by the Covid deaths for the last two weeks in this graph, you get about 3.
So although Mississippi has a death rate of 28 X/M/D, the number came out very close to my 3.5 from my previous post.

 

[Astronuc]

I posted something to this effect on my Facebook page way back on June 25th.

Study the role of hubris in nations’ COVID-19 response

I probably came to the conclusion that hubris was a big factor, as by that time, New York City was the world leader in Covid deaths, and a blurb in the Urban Dictionary said; "When I was in the Bronx I got mugged, shot, raped, and murdered in a New York Minute."
Along with everything I've ever seen about NYC characterizing the populous as "tough".

In any event, I'm guessing there's some hubris involved with the following graph.

View attachment 269794

I was trying to figure out where the data originated.

Is it https://ourworldindata.org/coronavirus-data-explorer?
https://ourworldindata.org/coronavirus

Johns Hopkins is tracking the individual states in the US.
https://coronavirus.jhu.edu/data/state-timeline/new-confirmed-cases/new-york

Currently, NY State is about 1308 deaths/million, based on 25437 fatalities reported in a population of 19.45 million (2019). California reports 15204 deaths in a population of 39.51 million, or 384.8 deaths/million. Texas report 15129 deaths in a population of 29.0 million, or 521.7 deaths/million.

I'm not sure the value of such ratios.

 

[Vanadium 50]

2 out of 7 deaths in Mississippi for the last 2 months have been due to Covid.

I get more like 20%. But New York City was a factor of six. Yet we are exhorted to be like the "success" of New York City.

 

[OmCheeto]

I was trying to figure out where the data originated.

Is it https://ourworldindata.org/coronavirus-data-explorer?
https://ourworldindata.org/coronavirus

Johns Hopkins is tracking the individual states in the US.
https://coronavirus.jhu.edu/data/state-timeline/new-confirmed-cases/new-york

Currently, NY State is about 1308 deaths/million, based on 25437 fatalities reported in a population of 19.45 million (2019). California reports 15204 deaths in a population of 39.51 million, or 384.8 deaths/million. Texas report 15129 deaths in a population of 29.0 million, or 521.7 deaths/million.

I'm not sure the value of such ratios.

Most of my data comes from another Johns Hopkins site. They track both the US states and world nations.
If you see me post a graphic without attribution, it's most likely of my own creation.

I suppose deaths/million ratio let's you know how close to achieving herd immunity you are.
The top 10 zip codes in NYC average ≈5000 deaths/million. It's just an example that things can get a lot worse, if we let our guards down.

 

[mfb]

I get more like 20%. But New York City was a factor of six. Yet we are exhorted to be like the "success" of New York City.

By whom? Who claimed NYC was a success overall? You seem to see this everywhere apparently. I do not.
The first big outbreak was a disaster.
NYC got the new infections to a very low level after its first big outbreak. That's good.

 

[OmCheeto]

I get more like 20%. But New York City was a factor of six. Yet we are exhorted to be like the "success" of New York City.

I saw your "success" post from a few days ago. The only reference I could find was a NY Post article written by Rich Lowry. I disagreed with what he stated about what those people were calling a success.

 

[mfb]

I saw your "success" post from a few days ago. The only reference I could find was a NY Post article written by Rich Lowry. I disagreed with what he stated about what those people were calling a success.

That's just one more person complaining about these mysterious people who say NYC did everything right. Where are all these mysterious people?

 

[Vanadium 50]

Where are all these mysterious people?

I answered you the last time you asked.

 

[Vanadium 50]

NYC got the new infections to a very low level after its first big outbreak. That's good.

That's like saying that once a fire is done burning that it's under control.

 

[gmax137]

First, you have to figure out what the average, or usual, death rate is. Globally and in the US, it's around 20 deaths per million per day.
If you add Covid deaths to that, and divide by Covid deaths, it yields the ratio.
Of course, the usual death rate varies by location, but it's close enough for my purposes.

An alternative, but dreadfully time consuming procedure:
The CDC has a list of the weekly tally of deaths for all the states from 2017 to the present.
Using Mississippi as an example, you can see that deaths for the time period from 2017 to 2019 are fairly consistent.
2020 showed a marked increase.
Dividing the total deaths by the Covid deaths for the last two weeks in this graph, you get about 3.
So although Mississippi has a death rate of 28 X/M/D, the number came out very close to my 3.5 from my previous post.

View attachment 269954

Thanks Om.

20 per million per day would put Mississippi (3 million population) at around 420 per week, where your plots show 560? That would be... 27 per million per day. So Miss isn't doing well in a normal year. Still that 1 in 3 deaths due to Covid is a shocker.

https://msdh.ms.gov/msdhsite/_static/14,0,420.html

shows 2870 Covid deaths in the state; if we take that as over 240 days [March thru Sept] then I get 1 in 6.7 overall. But that neglects the June-July upswing.

 

[OmCheeto]

I answered you the last time you asked.

I'd personally like to see references to where each of those people said the entire NYC fiasco was a success, rather than someone else saying they said that.

 

[OmCheeto]

Thanks Om.

20 per million per day would put Mississippi (3 million population) at around 420 per week, where your plots show 560? That would be... 27 per million per day. So Miss isn't doing well in a normal year. Still that 1 in 3 deaths due to Covid is a shocker.

https://msdh.ms.gov/msdhsite/_static/14,0,420.html

shows 2870 Covid deaths in the state; if we take that as over 240 days [March thru Sept] then I get 1 in 6.7 overall. But that neglects the June-July upswing.

I did find quite a lot of variability for death rates among the states. I guessed Utah had the youngest and Florida had the oldest populations, and was not disappointed by their associated death rates on the following graph.

Mississippi has the # 1 poverty rate in the nation, is #2 from the bottom in life expectancy, and #2 most obese, so their death rate also kind of makes sense.

 

[jim mcnamara]

CDC release - age class infection rate COVID-19 - May-Aug 2020
Population of disease morbidity burden has shifted to younger ages.

Note the large increases in cases per 100,000 in the younger populations.

Example:
 age 0-9:  May=35.0, Aug=89.0 
 age 10-19: May=74.0, Aug=246.0

See Figure 1 for a graph, or Table for data points in
https://www.cdc.gov/mmwr/volumes/69/wr/pdfs/mm6939e1-H.pdf

@OmCheeto may find this dataset useful.

 

[russ_watters]

That's like saying that once a fire is done burning that it's under control.

When I brought this up at least a month ago, I was criticized for what I guess was an over-extension of the term "herd immunity". The implication in the thread and presumably from those officials you mentioned is that New York City and surrounding areas are doing something uniquely excellent *today* to mitigate the virus spread. At the time I brought it up, several people said, 'well, it's not super unique, here are a few countries that are doing a similarly excellent job...' At the time it may not have been clear, but it should be now, that most of the world is experiencing a substantial second wave, including many of those cited as doing an excellent job: Japan, Germany, Norway, etc. At this point, if we ignore the first peak, New York's advancement would put them on the moon while everyone else is sailing square rigged frigates.

So yeah, I agree that it's a little silly to say, in effect, that they are doing a good job of hosing down the ashes after the fire went out.

[July 23]
As state hospitalizations reach their lowest levels since March, Governor Andrew Cuomo is praising the state's slow reopening process.

He says it has turned out for the best that New York took its time reopening.

States that reopened quickly are now shutting back down because of sharp spikes in COVID-19 cases.

He says New York's continued success is contingent on the rest of the country.

https://spectrumlocalnews.com/nys/c...w-york-s-slow-reopening-for-covid-19-progress

In any case, I'm sure Governor Cuomo would be delighted to receive inquiries from governments around the world on how to manage COVID and achieve his city/state's unique and spectacular success.

 

[mfb]

I answered you the last time you asked.

But not with what I asked about. You found examples of people who say NYC did a good job after the first big wave. That's very different from the mysterious people I still want to meet. Here is the post for reference

That's like saying that once a fire is done burning that it's under control.

That's what fire departments routinely do. But you are assuming general herd immunity in New York here, that's a questionable assumption. It probably contributes, okay, but it's not like NY ran out of people who can get sick.

At the time it may not have been clear, but it should be now, that most of the world is experiencing a substantial second wave, including many of those cited as doing an excellent job: Japan, Germany, Norway

You have seen Germany's "second wave" here? It's a single pixel in that graph.
Japan is at 80,000 cases. That's less than 1 in 1000.
Germany is at 280,000 cases, or ~3.5 per 1000.
Norway is at 13,000 cases, or ~3.5 per 1000.
All their new case counts are going down again and/or staying relatively low.. They are still doing a great job.
Germany's "substantial second wave" is ~1500 new cases/day. Scaled to the population of the US this would equal 6000 new cases per day.
Norway's "substantial second wave" is ~100 new cases/day, about the same per capita as Germany.
Japan's "substantial second wave" peaked at 1500/day (with 50% larger population than Germany), now they are down to ~400/day. Random comparison: Wisconsin with its population of 6 million has ~1500 new cases per day and more confirmed cases overall than Japan with its population of 125 million. Absolute deaths are about the same (1260 in Wisconsin vs. 1520 in Japan) despite a factor 20 in population.

 

[Ygggdrasil]

At the time it may not have been clear, but it should be now, that most of the world is experiencing a substantial second wave, including many of those cited as doing an excellent job: Japan, Germany, Norway, etc. At this point, if we ignore the first peak, New York's advancement would put them on the moon while everyone else is sailing square rigged frigates.

For reference, here is the 7-day rolling average of the daily new Coronavirus cases per million in Japan, Germany, and Norway compared to the US and France (which have experienced significant second waves):

(data downloaded from the ECDC on 9/24)

Cases have begun increasing slightly in Germany and Norway though not (yet?) to the extent seen in the US and France.

 

[morrobay]

For reference, here is the 7-day rolling average of the daily new Coronavirus cases per million in Japan, Germany, and Norway compared to the US and France (which have experienced significant second waves):
View attachment 269993
(data downloaded from the ECDC on 9/24)

Cases have begun increasing slightly in Germany and Norway though not (yet?) to the extent seen in the US and France.

France has had over 16000 new cases in past 24 hours. What is going on there ?

 

[Vanadium 50]

France has had over 16000 new cases in past 24 hours. What is going on there ?

These spikes are often caused by catching up with a backlog or by a redefinition of "case". (e.g. from presenting symptoms to a positive test)

 

[Vanadium 50]

So yeah, I agree that it's a little silly to say, in effect, that they are doing a good job of hosing down the ashes after the fire went out.

I think it's more than a little silly. People are making decisions based on this, decisions that affect other people's lives.

NYC has a population of 8.4 million. They have seen 24,000 deaths. If you take the 0.26% CDC "best estimate" (0.4% fatality rate if symptomatic, and 65% symptomatic) you find that 110% of the population is infected.

And yes, I know that the 0.26% has its uncertainties. Replacing 35% with 20%, as the previously posted meta-study reported, turns this to 90%. But qualitatively, it looks like yes, they are hosing down the ashes. Pretty much anyone who could catch it has caught it.

 

[Ygggdrasil]

These spikes are often caused by catching up with a backlog or by a redefinition of "case". (e.g. from presenting symptoms to a positive test)

As you can see from the graph I posted earlier, the rise in Coronavirus cases in France has been steadily rising since about August. Indeed, here's an article from Aug 17, noting the rise of Coronavirus cases in France:

Beaten Back, the Coronavirus Regains Strength in France
Faced with an increase in infections, especially among young adults, French health officials are expanding mask rules and warning against complacence.

https://www.nytimes.com/2020/08/17/world/europe/france-coronavirus.html

The rise is not some illusory result of clearing a backlog of testing either as the AP reported two weeks ago on hospitals filling up in some of the hardest hit regions of France:

While France’s daily case count climbed back up as summer vacations brought relaxed virus vigilance, the number of infected patients in hospitals and intensive care units stayed low and stable for several weeks. Until now.

Doctors in Marseille — the country’s latest virus hotspot — started sounding the alarm this week. The 70 ICU beds dedicated to virus patients in France’s second-biggest city and the surrounding Bouches-du-Rhone region were all occupied by Tuesday. The number of ICU virus patients in the region has doubled in the past 10 days and now surpasses 100.

https://apnews.com/article/marseille-virus-outbreak-archive-france-41c2741ba71ef7457caaaa3183dccd01

Testing data from the Marseille region shows a clear pattern of the outbreak starting in young 20-29 year olds then spreading to the wider community, including the elderly who are much more susceptible to bad outcomes from the disease:

(source)

Similar dynamics have also been seen in data from Florida.

 

[Vanadium 50]

As you can see from the graph I posted earlier, the rise in Coronavirus cases in France has been steadily rising since about August. Indeed

Well, I was trying to explain why sometimes you get shocking numbers.

However, I think again we are looking at something very different between now and then. August 1 to Sept. 24 France went from 165,719 cases and 29861 fatalities to 497,237 cases and 31,511 fatalities. Fatalities per case has gone from 18% to 0.5%.

Put another way, two-thirds of the cases are post July 1st. But only 5% of the fatalities are.

 

[morrobay]

Looking like natural selection , the virus evolving for more infectious and less lethal .

 

[OmCheeto]

Looking like natural selection , the virus evolving for more infectious and less lethal .

I think it's more likely that there weren't enough tests available in the beginning to count all the cases.
Comparing @Ygggdrasil 's "cases per million" chart to my "deaths per million" chart, aligning the USA July-August humps(circled in magenta) via the vertical axis shows that a significant portion of cases were missed in the first phase. The shapes of the curves should ideally be identical.

Attempting to track the case fatality ratio, it looks as though Norway did the best job in the beginning, with France doing the worst, and everyone ending up kind of in the same neighborhood.

 

[atyy]

I think it's more than a little silly. People are making decisions based on this, decisions that affect other people's lives.

NYC has a population of 8.4 million. They have seen 24,000 deaths. If you take the 0.26% CDC "best estimate" (0.4% fatality rate if symptomatic, and 65% symptomatic) you find that 110% of the population is infected.

And yes, I know that the 0.26% has its uncertainties. Replacing 35% with 20%, as the previously posted meta-study reported, turns this to 90%. But qualitatively, it looks like yes, they are hosing down the ashes. Pretty much anyone who could catch it has caught it.

The antibody positivity rate is about 25%, which is not consistent with the estimate of 90%.
https://www.nytimes.com/2020/08/19/nyregion/new-york-city-antibody-test.html
https://www1.nyc.gov/site/doh/covid/covid-19-data-testing.page

 

[Vanadium 50]

The antibody positivity rate is about 25%, which is not consistent with the estimate of 90%.

That is true. You have my numbers. You can decide which one you think is wrong: the population, the number of deaths, or the fatality rate.

If you say, it's the fatality rate, though, be aware that there are 7M positive tests in the US. If you make the fatality rate 4x higher to explain New York, you would expect no fewer than 700,000 deaths in the US, not less than 200,000.

If your answer to that is "just you wait", you can look at hospitalizations.

 

[atyy]

That is true. You have my numbers. You can decide which one you think is wrong: the population, the number of deaths, or the fatality rate.

If you say, it's the fatality rate, though, be aware that there are 7M positive tests in the US. If you make the fatality rate 4x higher to explain New York, you would expect no fewer than 700,000 deaths in the US, not less than 200,000.

If your answer to that is "just you wait", you can look at hospitalizations.

I would guess the fatality rate. In Singapore the fatality rate is 0.05% (about 27 deaths, 57000 confirmed cases), so the 0.3% of the CDC is an average across different populations and times.

 

[mfb]

That is true. You have my numbers. You can decide which one you think is wrong: the population, the number of deaths, or the fatality rate.

If you say, it's the fatality rate, though, be aware that there are 7M positive tests in the US. If you make the fatality rate 4x higher to explain New York, you would expect no fewer than 700,000 deaths in the US, not less than 200,000.

If your answer to that is "just you wait", you can look at hospitalizations.

You missed a factor 10 somewhere. A 1% infection fatality rate would mean 2.4 out of 8.4 million got it in NYC, that's a bit over 1/4 and consistent with antibody tests. It would also mean the 200,000 US deaths come from 20 million cases, i.e. the US found one in three cases overall. Currently the US has ~700 deaths per day, which would come from 70,000 cases per day, which is a factor 2 higher than the confirmed case rate. No conflict here either.

That's not taking into account that treatment has improved: The infection fatality rate in NYC was worse than the IFR in states that had their outbreaks later.

 

[Rive]

Looking like natural selection , the virus evolving for more infectious and less lethal .

This kind of thing is expected at the long run, but right now there are still several strains competing without any of them becoming dominant, while the IFR dropped almost everywhere simultaneously. Not likely that this change is about genetic change.

A more likely explanation is that it's the work of some kind of common trait of all the strains what was not recognized before.

 

[atyy]

https://www.thejakartapost.com/life...y-provide-some-immunity-against-covid-19.html
Study suggests dengue may provide some immunity against COVID-19
Pedro Fonseca

https://www.medrxiv.org/content/10.1101/2020.09.19.20197749v1
How super-spreader cities, highways, hospital bed availability, and dengue fever influenced the COVID-19 epidemic in Brazil
Miguel A. L. Nicolelis, Rafael L. G. Raimundo, Pedro S. Peixoto, Cecilia Siliansky de Andreazzi

https://www.thelancet.com/journals/laninf/article/PIIS1473-3099(20)30158-4/fulltext
Covert COVID-19 and false-positive dengue serology in Singapore
Gabriel Yan, Chun Kiat Lee, Lawrence T M Lam, Benedict Yan, Ying Xian Chua, Anita Y N Lim, Kee Fong Phang, Guan Sen Kew, Hazel Teng, Chin Hong Ngai, Li Lin, Rui Min Foo, Surinder Pada, Lee Ching Ng, Paul Anantharajah Tambyah

The paper by Yan et al was mentioned earlier in the thread. It showed that a person with COVID-19 could be mistakenly diagnosed for dengue on the basis of an antibody test for dengue.

The new preprint by Nicolelis et al suggests that having had dengue may be protective against SARS-CoV-2 infection, on the basis that in Brazil, COVID-19 rates are lower in places that had dengue outbreaks.

 

[atyy]

This kind of thing is expected at the long run, but right now there are still several strains competing without any of them becoming dominant, while the IFR dropped almost everywhere simultaneously. Not likely that this change is about genetic change.

A more likely explanation is that it's the work of some kind of common trait of all the strains what was not recognized before.

Hopefully it has dropped, because there have been improvements in how the disease is treated, as well as more resources available for treatment. However, is there data to show that the IFR has dropped?

Initial IFR estimates ranged from 0.3% to 1% averaged across age groups. Estimated IFR was lower than 0.3% for younger people.
https://www.who.int/docs/default-so...ation-reports/20200219-sitrep-30-covid-19.pdf
https://www.thelancet.com/journals/laninf/article/PIIS1473-3099(20)30243-7/fulltext

Some estimates in India are suggesting an IFR of 0.1%. It is not yet clear why, but one factor may be because the median age in India is about 28, whereas the median age in the China and the USA is about 38.
https://www.hindustantimes.com/indi...-know-today/story-mtwod5mI80yuQoWEHMPZSJ.html
https://www.bloombergquint.com/coro...-exceptionalism-may-not-explain-low-mortality
"Firstly, the age-adjusted IFR is not orders of magnitude lower in India, Juneja said: “If you were to adjust for age, the IFR for Spain or even Wuhan would be around 0.2% and what we are seeing is 0.1%,” he said. Underlying immunity, he said, could potentially explain some of this gap."

 

[morrobay]

This kind of thing is expected at the long run, but right now there are still several strains competing without any of them becoming dominant, while the IFR dropped almost everywhere simultaneously. Not likely that this change is about genetic change.

The D614G on the S-protein is the dominate strain now since May. And the glycine replacement of aspartic acid at AA 614 has increased transmission/infectivity with higher virus lodes. As well as more effective
adhesion to ACE2 receptors. (salt bridges, ionic or hydrogen bonds between respective amino acids).

 

[Vanadium 50]

You missed a factor 10 somewhere.

Where? My quations are number infected = number dead / fatality rate and fraction infected =number infected / total population. number infected = 24000/0.0026 = 9.2M. Do you doubt this?

Fraction infected = =number infected / total population = 9.2M/8.4M = 110% Do you doubt this? (The calculation, not the outcome)

I don't see a factor of ten anywhere.

 

[aheight]

Looking like natural selection , the virus evolving for more infectious and less lethal .

Looks like it to me (Washington Post link): covid evolving

In general, scientists would expect natural selection to favor mutations that help the virus spread more effectively — since that allows it to make more copies of itself — but not necessarily ones that make it more virulent. Killing or incapacitating the host would generally not help the virus spread to more people.

As would I. Really, it's all a matter of survival and reproductive success.

 

[Vanadium 50]

I would guess the fatality rate. In Singapore the fatality rate is 0.05%

The problem with that explanation is that if you make the fatality rate lower, you make the NYC incidence higher and the mismatch between antibody tests and inferred incidence rates gets even more discrepant.

So far, I think the following explanations have been suggested:

1. The disease has mutated to be less deadly. This is problematic as
   1. There hasn't been much time
   2. Coronaviruses mutate slowly.
   3. THIS Coronavirus mutates slowly
   4. We have a full RNA sequence of the virus, and would know if there were two varieties, one more dangerous than the other, and there have been no reports of such
   5. Even if there were a new strain, the old strain would still be there
2. Hospital care has improved since late spring. This is inconsistent with "there is no cure and medical care revolves around relieving symptoms.".
3. I can't do arithmetic. Fair enough. Show me where.
4. Different demographics. There is something to this, as half the fatalities have been in nursing homes, and obviously those fatalities are not in the population now. It's not this simple because the nursing-home population also has an inferred infection rate near 100% (so adding it or removing it can't change the bottom line) but maybe there's something more subtle. To take an extreme case, if my antibody sample is driven by college students, I have decoupled the antibody testing rate from the fatality rate. I haven't caused them to drift apart, but they are no longer tied together.

Let me propose two other possibilities. I don't think either is right, but I don't think we have evidence against them.

1. There are two (or more) strains, both equally contagious and equally dangerous, but only one shows up on the antibody test.
2. Antibodies only persist for 6-8 weeks post infection. People can get reinfected (and we do have some examples of that). The decrease in positivity we are seeing in NYC is not driven by testing a healthier population, but is a real change over time.

 

[mfb]

Where?

In your claim that a factor 4 higher IFR would be in conflict with overall US data. If you take an IFR of 0.0026 then New York had more sick people than people, which is obviously a nonsense conclusion. If you take an IFR of 0.01, however, everything fits.

If you say, it's the fatality rate, though, be aware that there are 7M positive tests in the US. If you make the fatality rate 4x higher to explain New York, you would expect no fewer than 700,000 deaths in the US, not less than 200,000.

That's where you got factor 10 wrong somehow. You would not expect 700,000 deaths from 7 M positive tests with an IFR of 1%. You would expect at least 70,000, and indeed the US has more than 70,000 deaths.

Hospital care has improved since late spring. This is inconsistent with "there is no cure and medical care revolves around relieving symptoms.).

Treatment is not binary. Treatment can (and did) improve without a miracle cure.

I can't do arithmetic. Fair enough. Show me where.

My second post where I do now.

 

[Vanadium 50]

That's an argument that the CDC 0.26% number is wrong. That's a position that's defensible, but should be attacked on it's merits.

 

[atyy]

The problem with that explanation is that if you make the fatality rate lower, you make the NYC incidence higher and the mismatch between antibody tests and inferred incidence rates gets even more discrepant.

I was suggesting the fatality rate was higher.

 

[atyy]

Let me propose two other possibilities. I don't think either is right, but I don't think we have evidence against them.

1. There are two (or more) strains, both equally contagious and equally dangerous, but only one shows up on the antibody test.
2. Antibodies only persist for 6-8 weeks post infection. People can get reinfected (and we do have some examples of that). The decrease in positivity we are seeing in NYC is not driven by testing a healthier population, but is a real change over time.

https://www.bmj.com/content/370/bmj.m3325
There could be false negatives (but I would guess it's not due to viral variants), since test sensitivity (lack of false negatives) is about 90% for the commonly used tests in the UK, and can be lower than 90% depending on when the person is tested, as it takes time for antibodies to build up after a person gets infected. There is evidence for the possibility that antibodies decrease after 6-8 weeks, but we can look at the data for earlier in the year.

NYC reached about 18000 deaths (13000 confirmed, 5000 probable) by May 2, and 20000 deaths (15000 confirmed, 5000 probable) by May 15.
https://www.cdc.gov/mmwr/volumes/69/wr/mm6919e5.htm
https://www.pix11.com/news/coronavirus/latest-coronavirus-updates-in-new-york-friday-may-15-2020

NYC antibody positivity rates were about 20% by May 2.
https://www.governor.ny.gov/news/am...-announces-results-completed-antibody-testing

Let's take 15000 deaths, and 25% positivity rate (to account for false negatives in the antibody testing), and an NYC population of 8400000. This gives an IFR of (15000 x 100%)/(0.25 x 84000000) = 0.7%. So it is quite reasonable that the IFR was higher than 0.3% in the early stage of the outbreak in NYC.

 

[mfb]

That's an argument that the CDC 0.26% number is wrong. That's a position that's defensible, but should be attacked on it's merits.

CDC's number being too low is more plausible than New York having more sick people than people. And yes, as discussed before, we were studying how the numbers work out if the CDC number is too low. All the US numbers fit very nicely if we assume a higher IFR.AstraZeneca, Under Fire for Vaccine Safety, Releases Trial Blueprints
Interesting article about their vaccine. Not so much about what the title says, but they have a bit of information about the two mystery patients. Two cases of transverse myelitis in the vaccine group, a relatively rare disease that can be associated with infections. One of the patients has MS which can cause it as well, for the other patient we don't know.

 

[atyy]

I can't do arithmetic. Fair enough. Show me where.

Neither @mfb nor I understand why you said that 7,000,000 confirmed cases in the US mean at least 700,000 deaths with an IFR of 1%. We think that 7,000,0000 confirmed cases mean at least 70,000 deaths if the IFR is 1%.

Overall an IFR of 1% may be a bit high, but given that the NYC health system was overwhelmed in the early stages, it seems plausible that IFR in the early stages of the NYC outbreak was higher, similar to how the confirmed case fatality rates in Hubei (where the Wuhan health system was initially overwhelmed) were 5X higher than outside of Hubei. https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(20)30746-7/fulltext

 

[russ_watters]

...but given that the NYC health system was overwhelmed in the early stages...

I don't think the evidence supports that. The media played-up busy hospitals and harried staff, but NYC added substantial emergency capacity, which went almost completely unused.
https://www.navytimes.com/news/your...s-nyc-having-treated-fewer-than-200-patients/

https://www.google.com/amp/s/abc7ny.com/amp/coronavirus-nyc-update-corona-virus-cases/6142109/

https://www.militarytimes.com/news/...orkers-are-going-straight-into-nyc-hospitals/