COVID-19 Coronavirus Containment Efforts

[Dr.AbeNikIanEdL]

There is a noticeable periodicity in this record of daily Covid-19 deaths in Sweden. I don't think other nations show that kind of thing. What would be causing this?

Germany looks similar https://www.worldometers.info/coronavirus/country/germany/ ...

I think the usual explanation is that there is e.g. less testing/reporting on weekends etc (or more complicated technical reasons why the reports are clustered without the cases/deaths necessarily being clustered), and those numbers are then reported later in the week. Seems to fit that one cycle is roughly 7 days.

 

[OmCheeto]

I like the "no singing in churches" recommendation.
https://www.theguardian.com/world/2...-for-religious-services-including-singing-ban

One of my theories on disparities in infection rates between locales is "how loudly the locals speak".
Having never been to New York City, I can't really say if they talk louder than people in the rest of the world, or if that's just Hollywood.
In any event, if you've missed the profanity laced video titled "Ticked off Vic", then you'll have missed probably the most efficient human nebulizer/atomizer on the planet. He (Vic DiBitetto) could probably put out small fires with the amount of spittle he generates.
I seriously think singing, shouting, talking at Jersey levels, and laughing without face masks should be considered crimes until this is over with.

Of course, there are lots of other variables, also.
I discovered that one possible reason why Sweden and Denmark have different rates is that Copenhagen, the capital and most populous city of Denmark, appears to be the bicycling capital of the world.

 

[gmax137]

I can't really say if they talk louder than people in the rest of the world, or if that's just Hollywood.

That's just Hollywood.

 

[DennisN]

Update on the numbers from Sweden

Two interesting recent news articles in Swedish news:

(1)

FOHM (The Swedish Health Agency) has calculated that the R-number (reproduction number) is now below 1 in Sweden:

Public Health Agency: Sweden's R-number is now below 1.0

Sweden's R-number has been below 1.0 for a week, according to a calculation made by the Public Health Authority. If the trend continues it means the pandemic will gradually ebb.

[...]

(Google translation to English, with some corrections by me)

Source: Folkhälsomyndigheten: Sveriges R-tal nu under 1,0 (DN, May 2 2020, Swedish only)

(2)

A short interview with Anders Wallensten from FOHM (The Swedish Health Agency) regarding the high number of deaths in nursing homes in Sweden:

Folkhälsomyndigheten (FOHM) investigates high death rates

Sweden's high death toll due to Covid-19 is significant in comparison with our Nordic neighboring countries. An important explanation is that the infection came into the country's nursing homes early, says Anders Wallensten from FOHM.

According to the latest public health statistics, a total of 2,679 people have died due to Covid-19 in Sweden. This is more than three times more compared to our Nordic neighbors - in total - and the figure will certainly be adjusted upwards when the weekend's backlog in reporting has been entered.

According to Sweden's Deputy State Epidemiologist Anders Wallensten, the main reason for this big difference is that the infection entered our nursing homes. Dagens Nyheter has produced data from the country's regions that show that at least 541 nursing homes have been affected.

- It is highly unfortunate that there has been such a large spread of infection there. We are investigating what has failed, what can be done better and in what way more support is needed, in order to improve this, says Wallensten.

TT: What is the spread of infection at the country's nursing homes at the moment?

- I don't have the current number. But it has been very large and it has not changed overnight. Unfortunately, it is true that once you have got the infection, it is difficult to manage in a nursing home. Great efforts are needed to ensure that no more people become infected, says Anders Wallensten.

TT: Whose responsibility is it that the infection has entered the nursing homes?

- It is surely a shared responsibility between everyone involved with the elderly. After all, there are many principals who work with elderly care and it is important that routines are working, but I cannot comment on whose responsibility it is specifically, says Anders Wallensten.

TT: Does the Public Health Agency have any responsibility for not having provided sufficiently clear guidelines?

- The Public Health Agency does not manage elderly care. Basically, it is about issues that always should be in place, even when it is not a pandemic, such as basic hygiene routines etc.

TT: Who should have been aware of shortcomings in basic hygiene practices?

- Well, it is the business owner who should have that competence. But as I said, it is too early to say exactly what has not worked out. We are going through it now and will report more during the week, says Anders Wallensten.

(Google translation to English, with some corrections by me)

Source: http://www.sydsvenskan.se/2020-05-03/folkhalsomyndigheten-granskar-hoga-dodstal (SDS, May 3 2020, Swedish only)

 

[Vanadium 50]

Yeah, nursing homes are important. I looked at the Massachusetts data, and the average age of a Covid-19 fatality is 82. 98.4% had identified underlying conditions, and 60% were in nursing homes.

I played around with a simple model, just at the Excel level. The idea is you have a large population A with a small probability of death, and a small population B, with a large probability of deaths. Instead of R's, I worked with probabilities: p_AA is the probability someone in group A is infected by someone else in group A, p_AB is the probability probability someone in group B is infected by someone in group A and so on. In this model, overall R varies depending on the relative sizes of group A and group B even for the same probabilities, and of course it depends on the p's.

The most important is p_BB, and the next most important is the product p_AA p_BA. p_AA by itself has less of an impact.

 

[mfb]

It's not so much the precision that I find surprising, it's the speed. A week after the changes are in place one can see this level of change? Contrast that with Sweden where we were told we had to wait more than a month to make any comparison.

It was an estimate, not a precise measurement. Clearly they vary quite a bit depending on the methods and so on.
https://www.cnbc.com/2020/04/28/germanys-coronavirus-infection-rate-has-edged-up.html

 

[nsaspook]

Let's hope the trend in Sweden continues as a possible way to a responsible exit strategy . The daily death rate and COVID-19 case increases are dropping. Unfortunately, "flatten the curve" IMO has morphed into "stop the virus", which is unattainable and has no realistic exit strategy other than wait to next year at the absolute earliest for a vaccine.

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

 

[Swamp Thing]

Germany looks similar https://www.worldometers.info/coronavirus/country/germany/ ...

I think the usual explanation is that there is e.g. less testing/reporting on weekends etc (or more complicated technical reasons why the reports are clustered without the cases/deaths necessarily being clustered), and those numbers are then reported later in the week. Seems to fit that one cycle is roughly 7 days.

I was initially convinced by this, but now I'm not too sure.

It would certainly explain periodicity in the number of new cases each day, if each test outcome is date-stamped with the day on which the test was processed. But in the case of deaths, each data point would be date-stamped with the day on which it actually took place (presumably). So when you plot the daily deaths, it would reflect the actual number of patients who died on a particular day, irrespective of variations in processing throughput.

Perhaps it is the effect of some real phenomenon like less staff being available in nursing homes over weekends (Someone has pointed out on this thread that a large percentage of deaths involves patients who were already receiving care in nursing homes).

 

[russ_watters]

I was initially convinced by this, but now I'm not too sure.

It would certainly explain periodicity in the number of new cases each day, if each test outcome is date-stamped with the day on which the test was processed. But in the case of deaths, each data point would be date-stamped with the day on which it actually took place (presumably). So when you plot the daily deaths, it would reflect the actual number of patients who died on a particular day, irrespective of variations in processing throughput.

I don't think that's likely as it would require retroactive edits to the data instead of just reporting a new number each day, and that would take a lot of work. What the data (on positive tests and deaths) tells us is as of that day, X many are known to have happened.

 

[Swamp Thing]

Perhaps it is the effect of some real phenomenon

Or maybe it's a complex phenomenon, as in, the probability of succumbing is the square of a complex number.

Please excuse the quantum graveyard humor.

 

[Swamp Thing]

it would require retroactive edits to the data

All it would take is that properly date-stamped data should propagate from hospital to area to town to region to country. Once that happens, the graph would just be based on a query off a database and would be free from processing artefacts.

 

[Dr.AbeNikIanEdL]

It would certainly explain periodicity in the number of new cases each day, if each test outcome is date-stamped with the day on which the test was processed. But in the case of deaths, each data point would be date-stamped with the day on which it actually took place (presumably).

That would certainly make sense, however:

(1) I don't see any indication that is what they do. As far as I can tell the number for each day is frozen at 0:00 GMT. Exceptions are usually explicitly mentioned in the Updates section. At least for Germany, the source is just a newspaper quoting the total number of deaths, presumably the new deaths every day is just the difference to the last day.

(2) It is not clear to me (again in particular for Germany) that such numbers would be officially reported anyway. I see (understandably) a great deal of trying to estimate when people actually got sick. For deaths however only the total number, and differences to the previous day.

(3) Look e.g. at China, they corrected the number of deaths on April 17 by a significant amount. Presumably no one of those was actually declared dead on April 17, yet the graph shows over 1200 new deaths on that day.

 

[russ_watters]

All it would take is that properly date-stamped data should propagate from hospital to area to town to region to country. Once that happens, the graph would just be based on a query off a database and would be free from processing artefacts.

I understand it could be done, but I'm pretty sure it isn't being done, which is what your question was about. Heck, even if the data was collected that way, it still wouldn't change the way it is primarily reported. That just isn't what the reported data is for.

 

[Vanadium 50]

These people: https://arxiv.org/abs/2004.07208 say the 7 day effect is real.

I'm not agreeing with them, just saying that's their claim. It's certainly not impossible for this behavior to occur in other systems: e.g. pogo oscillations.

 

[DennisN]

Coronavirus: UK hospital trials new treatment drug (BBC, 4 May 2020)

A new drug developed by UK scientists to treat Covid-19 patients is being trialled at University Hospital Southampton.

Developed by UK bio-tech company Synairgen, it uses a protein called interferon beta, which our bodies produce when we get a viral infection.

Initial results from the trial are expected by the end of June.

[...]

Interferon beta is part of the body's first line of defence against viruses, warning it to expect a viral attack, explains Richard Marsden, chief executive of Southampton-based Synairgen.

He says the Coronavirus seems to suppress its production as part of its strategy to evade our immune systems.

The drug is a special formulation of interferon beta delivered directly to the airways when the virus is there, with the hope that a direct dose of the protein will trigger a stronger anti-viral response even in patients whose immune systems are already weak.

[...]

Synairgen's drug trial is the template for a new fast-track clinical scheme that has just been set up by the government.

The Accord programme, as it is known, is designed to accelerate the development of new drugs for patients with Covid-19.

The first phase of the programme involves six other drugs.

[...]

 

[Ygggdrasil]

There is a noticeable periodicity in this record of daily Covid-19 deaths in Sweden. I don't think other nations show that kind of thing. What would be causing this?

https://www.worldometers.info/coronavirus/country/sweden/
View attachment 261996

In the data from Sweden, the 7 day periodicity is almost certainly due to reporting. In a previous post, I noted differences in the daily deaths data from different sources:

Data from the European Centre for Disease Prevention and Control (ECDC), show the periodicity in deaths while data from the Public Health Agency of Sweden (FOHM) do not. It looks like the FOHM data attribute the death counts to the dates the individuals died (with a lag time of ~1-2 weeks for reporting) reflecting the actual number of deaths per day while the ECDC just scrapes the daily death count totals and reflects the number of new deaths reported per day.

It seems like the worldometers site uses the ECDC (or similar) source for their data while sites like Wikipedia use the FOHM data.

 

[Vanadium 50]

regarding the high number of deaths in nursing homes

I mentioned the Massachusetts numbers. I looked at them again and, wow. Massachusetts has about 38000 nursing home residents. They have about 2400 deaths in nursing homes, and looks like they are about 2/3 of the way through the pandemic. Plug in a 10% CFR for people that age, and you get about 36000 cases: pretty much everyone who could get infected did.

 

[kyphysics]

CDC says:

If you develop any of these emergency warning signs for COVID-19, get emergency medical attention immediately:
Trouble breathing
Persistent pain or pressure in the chest
New confusion or inability to arouse
Bluish lips or face

This is a dumb question, but what do they mean by "arouse" there? Is that a way of saying standing up or rising? Or, is it something else?

 

[kyphysics]

As Germany opens more and more things the estimated reproduction rate went from 0.7 to 1. Still enough to keep the disease at a low level, and with the delay between infections and confirmed cases this means the confirmed cases still go down. If we take 2 weeks between confirmed infection and death we can expect daily deaths to shrink by another factor 2, to ~50, maybe even a bit better. Or 18,000 in a year if it is kept constant. That's still 5 times as many as traffic accidents (with normal traffic). If the reproduction rate is a bit lower that number can go down a lot, if it is higher we'll probably see some restrictions coming back.

Overall I like the German strategy. The result is not as good as in NZ/Australia/Iceland, but Germany isn't an island.

R0 of 1 means their growth rate becomes exponential. You have to have an R0 of below one (and sustain that!) to stop growing cases over time.

I am actually disturbed by Germany's spike in new cases, as they are back to exponential growth!

 

[DrGreg]

CDC says:This is a dumb question, but what do they mean by "arouse" there? Is that a way of saying standing up or rising? Or, is it something else?

I would assume "arouse" means "be woken up", i.e. gain conciousness. That's something for someone else in your house to observe rather than yourself.

 

[bob012345]

Let's hope the trend in Sweden continues as a possible way to a responsible exit strategy . The daily death rate and COVID-19 case increases are dropping. Unfortunately, "flatten the curve" IMO has morphed into "stop the virus", which is unattainable and has no realistic exit strategy other than wait to next year at the absolute earliest for a vaccine.

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

I noticed that too, the strategy of "flatten the curve" has morphed. They told us at the beginning that we were just slowing the rate we were going to get sick so as to not overwhelm the healthcare system. We did so now the strategy needs to be opening up at a rate such that the system can handle the rise in new cases. But now some see the predicted rise in new cases as a 'failure' when actually it's a success. Also, comparing Sweden to its immediate Nordic neighbors is misleading. Overall, compared to all other nations as a whole, Sweden has done quite well especially since their economy has remained open.

 

[russ_watters]

I noticed that too, the strategy of "flatten the curve" has morphed. They told us at the beginning that we were just slowing the rate we were going to get sick so as to not overwhelm the healthcare system. We did so now the strategy needs to be opening up at a rate such that the system can handle the rise in new cases. But now some see the predicted rise in new cases as a 'failure' when actually it's a success.

I never liked the "flattening the curve" strategy -- I always thought it indicated failure and wasn't much better than just letting the virus run wild. "Flattening the curve" (without a significant reduction in the total number who become infected) basically means being ok with many hundreds of thousands of deaths in the US. Are we really ok with that? I'm not. I think we should have chosen to try to do better.

 

[atyy]

There seem to be several different usages of "flatten the curve". In the strict sense, it means the same number of deaths over a longer time, where the main aim is prevent the healthcare system from being overwhelmed.

However, some people use the term loosely to mean slow the spread. If the spread is slowed enough (say over 5 years instead of 1 year), and a vaccine comes in before that, then "flatten the curve" would mean fewer deaths.

Also there are several meanings of the terms "suppression", "containment", "mitigation". Sometimes, "containment" has meant the same thing as "suppression", with the effective reproduction number < 1, while at other times "containment" has referred to contact tracing and quarantine of infected people and close contacts, without an increase in social distancing.

Also suppression (R < 1) and mitigation (R > 1, but low, with a healthcare system that can cope) strategies can be a continuum. One could attempt a suppression strategy, with the understanding that it might not work, and the failed suppression strategy would be a mitigation strategy.

 

[Vanadium 50]

However, some people use the term loosely to mean slow the spread. If the spread is slowed enough (say over 5 years instead of 1 year), and a vaccine comes in before that, then "flatten the curve" would mean fewer deaths.

I don't think a state of emergency can be sustained for five years.

 

[kyphysics]

There seem to be several different usages of "flatten the curve". In the strict sense, it means the same number of deaths over a longer time, where the main aim is prevent the healthcare system from being overwhelmed.

However, some people use the term loosely to mean slow the spread. If the spread is slowed enough (say over 5 years instead of 1 year), and a vaccine comes in before that, then "flatten the curve" would mean fewer deaths.

Also there are several meanings of the terms "suppression", "containment", "mitigation". Sometimes, "containment" has meant the same thing as "suppression", with the effective reproduction number < 1, while at other times "containment" has referred to contact tracing and quarantine of infected people and close contacts, without an increase in social distancing.

Also suppression (R < 1) and mitigation (R > 1, but low, with a healthcare system that can cope) strategies can be a continuum. One could attempt a suppression strategy, with the understanding that it might not work, and the failed suppression strategy would be a mitigation strategy.

https://www.bbc.com/news/health-52473523

I feel like flatten the curve was a rallying cry to:

i.) at minimum, try to prevent a overwhelming of the healthcare system (via a slower transmission rate)
ii.) on the more optimistic side, try to give the virus nowhere to spread

Per the article's chart here, an R0 (r "naught") value of lower than 1 means the virus dwindles down over time.

Why is a number above one dangerous?

If the reproduction number is higher than one, then the number of cases increases exponentially - it snowballs like debt on an unpaid credit card.

But if the number is lower, the disease will eventually peter out as not enough new people are being infected to sustain the outbreak.

Governments everywhere want to force the reproduction number down from about three to below one.

This is the reason you've not seen family, have had to work from home and the children have been off school. Stopping people coming into contact with each other to cut the virus's ability to spread.

If we had an aggressive nation-wide lockdown for 45 - 60 days (except for absolutely essential workers) and strong enforcement of various safety protocols (social distancing, mask wearing, non-large group gatherings, etc.), then we could have conceivably driven the R0 value to a very low level to give it few places to spread. Maybe it'd still be around, but in much fewer infectious people. And opening up from that point, while maintaining safety protocols and awareness, could help keep the spread minimal.

I think we botched a lot of things in that regard and I am scared of an immediate second wave after opening things back up again. We never really shut down. Parts of Texas and other areas of the country never really closed. Some states seem to have increasing case counts just as they're reopening.

 

[Rive]

I always thought it indicated failure and wasn't much better than just letting the virus run wild.

But it is. Just apart from preventing healthcare to be overrun by patients it also means that at the top you have less active cases => with R0 falling below 1 it will vanish faster (with less additional cases) from a lower base compared to the 'wild' top with much higher numbers.

Of course the ideal solution would be an extinction, but that's unrealistic since it would require strict cooperation of every country affected (with wide type/strength of economies).
Another 'ideal' solution is to develop a vaccine - but not every country can pay the price of a lockdown that long.

 

[atyy]

I don't think a state of emergency can be sustained for five years.

Let's see how China, South Korea, Hong Kong, Taiwan, Australia and New Zealand do. The have single-digit or near single-digit new cases per day, and many businesses can function at some level.

https://www.scmp.com/economy/global...s-most-us-firms-taiwan-feeling-pinch-covid-19

https://edition.cnn.com/travel/article/new-zealand-australia-travel-bubble-intl-hnk/index.html

 

[russ_watters]

But it is. Just apart from preventing healthcare to be overrun by patients it also means that at the top you have less active cases...

And how many would that save? I don't think I've seen an estimate of the difference in mortality between a healthcare system that is overrun and one that isn't. On the Diamond Princess, 1.7% of the infected died (12 people). Early estimates from Wuhan indicated 2.3% iirc. In the US we have 5.8% in a non-overwhelmed healthcare system, which means we're missing at least half the infections. These are scary-high numbers.

So what's a reasonable expectation for the death rate under an effective social distancing scenario? 1%? 2%?

=> with R0 falling below 1 it will vanish faster (with less additional cases) from a lower base compared to the 'wild' top with much higher numbers.

But that's not how "flattening the curve" was described. I'm seeing estimates including one from a Harvard epidemiologist saying 40-70% of the world population may eventually become infected in a year, and 1-2% die.
https://www.cbsnews.com/news/corona...ldwide-virus-expert-warning-today-2020-03-02/
This was pre-social distancing and this is one of the types of predictions that led to it. That's 1.3 to 4.6 million people.

The crude graphs and statements I've seen on the impact of social distance literally just show the curve flattening, without noticeably decreasing the area under it. I've seen no estimates of how many fewer people a "flattened curve" would infect. So again: it appears to me that as-sold to the public, a successful outcome would kill more than a million people while not specifying how many could be saved. Maybe it's the bottom-end vs the top-end of that range. I don't know.
https://www.uchicagomedicine.org/forefront/coronavirus-disease-covid-19/social-distancing-psa

I posted about this a few weeks ago and someone replied something to the effect of "do you know what an overwhelmed healthcare system looks like?" Sure -- I watch the news. It's a half hour of human-interest stories where altogether a dozen people are interviewed to discuss how hard they are working, and some photos/video of hospital beds in corridors. I didn't respond to it at the time because that means very little. This is a numbers game, not a human-interest story -- that's largely the flaw in how its being discussed. What I need to see are the numbers: how much extra harm would be caused by an overwhelmed healthcare system?

The only number I know of that answers that question is 80%. That's the percentage of people on ventilators who die. That tells me that an overwhelmed healthcare system matters very little because the deaths delta is only 25% of the excess of critical patients. But, you might ask, what about all those other patients not being served who might die -- a car accident victim, a heart attack victim? It doesn't affect the upper-bound. The upper-bound is this: everyone infected with COVID-19 stays home; Nobody gets a ventilator. Under that scenario, the deaths from other causes are unaffected and the COVID-19 deaths increases by some portion of 25%. I say less than 25% because that assumes everyone who could be saved by a ventilator is identified in time to save them. I'm sure many people die without ever being put on ventilators, after rapidly crashing. They probably couldn't be saved, but are included in the total anyway.

Of course the ideal solution would be an extinction, but that's unrealistic since it would require strict cooperation of every country affected (with wide type/strength of economies).

Now, 25% is a lot of people. But the problem is; 25% of what? It's not 25% of 80,000 (the hospital system isn't overwhelmed at that level) or 25% of 1,000,000?

What's more important is the order of magnitude between 80,000 and 1,000,000. That's the difference we should be talking about and the focus of our efforts.

Country-by-country extinction is possible because countries have borders and can isolate themselves.

Every country makes their own choices and the outcome is going to be based on those choices. I think extinction in the United states should have been the goal, and other countries have shown it's achievable. But we're not even trying -- we're barely even allowed to discuss it.

Another 'ideal' solution is to develop a vaccine - but not every country can pay the price of a lockdown that long.

Nobody can pay the price of an 18 month lockdown. That's not an option that anyone has seriously considered, as far as I'm aware.

 

[russ_watters]

Let's see how China, South Korea, Hong Kong, Taiwan, Australia and New Zealand do. The have single-digit or near single-digit new cases per day, and many businesses can function at some level.

Should we really wait and see or should we try to duplicate their success?

 

[Rive]

In the US we have 5.8% in a non-overwhelmed healthcare system, which means we're missing at least half the infections.

I'm not sure about that. Death makes its way into statistics faster than recovery, so in the expanding phase the mortality looks far worse than what it really is, even without much missed infections.

But that's not how "flattening the curve" was described.

Memes works best with a single, contagious piece of information. So yes, it's kind of the way it was sold.

But, you might ask, what about all those other patients not being served who might die -- a car accident victim, a heart attack victim?

I think they should be considered as victims of the overran healthcare: victims of the pandemic.

 

[mfb]

R0 of 1 means their growth rate becomes exponential. You have to have an R0 of below one (and sustain that!) to stop growing cases over time.

I am actually disturbed by Germany's spike in new cases, as they are back to exponential growth!

No, R=1 means a constant rate of new infections. Every infected person infects (on average) one other before they recover.

Please explain where exactly you see a spike in new cases, because I'm really curious. I see the lowest new case counts since mid March:

then we could have conceivably driven the R0 value to a very low level to give it few places to spread. Maybe it'd still be around, but in much fewer infectious people. And opening up from that point, while maintaining safety protocols and awareness, could help keep the spread minimal.

In other words: Lowering the cases, then opening up just enough to keep R not going above 1. Germany does exactly what you suggest in this post, despite you being disturbed by it earlier.

I never liked the "flattening the curve" strategy -- I always thought it indicated failure and wasn't much better than just letting the virus run wild. "Flattening the curve" (without a significant reduction in the total number who become infected) basically means being ok with many hundreds of thousands of deaths in the US. Are we really ok with that? I'm not. I think we should have chosen to try to do better.

What made you change your mind? A month ago you seemed skeptical of all the proposed and actual measures to "do better", citing their not well-known impact on the economy. Attempts to keep everything running as normal was exactly what prevented people from doing better.

The only number I know of that answers that question is 80%. That's the percentage of people on ventilators who die. That tells me that an overwhelmed healthcare system matters very little because the deaths delta is only 25% of the excess of critical patients.

You assume that everyone who doesn't get a ventilator in a hospital would survive fine at home. I would like to see something backing that assumption.
I would also like to see where this 80% number comes from. In China it was ~50%, and 1/3 of people admitted to ICU.
In the UK only 1/5 of people requiring "basic respiratory support" died. In the subset requiring mechanical ventilation it was 2/3. If everyone requiring basic respiratory support dies without a hospital your deaths increase by 400%, not 25%. I don't say that is true, but for an upper estimate that's certainly something to consider. And that's assuming a hospital does nothing but providing respiratory support, which is clearly not correct.

These people: https://arxiv.org/abs/2004.07208 say the 7 day effect is real.

I'm not agreeing with them, just saying that's their claim. It's certainly not impossible for this behavior to occur in other systems: e.g. pogo oscillations.

And they'll keep claiming that forever because apparently it's their personal pet hypothesis that they are unable to give up.
We know that reporting depends on the day of the week. We even have German states reporting zero on some days (i.e. not reporting the numbers in time). It's also something that only appears in some countries but not elsewhere.

 

[russ_watters]

I'm not sure about that. Death comes faster in statistics than recovery, so in the expanding phase the mortality looks far worse than what it really is, even without much missed infections.

That number is deaths over infections: it is an underestimate based on the available data, for the other side of the coin from what you describe.

Memes works best with a single, contagious piece of information. So yes, it's kind of the way it was sold.

That's fine, but I'd like to see evidence that there was a real plan/goal behind it, and I'm having a lot of trouble finding one - I'm mostly just guessing. Because independent of how it is being sold, if there's not plan/goal, why are we doing what we are doing?

I think they should be considered as victims of the overran healthcare: victims of the pandemic.

I agree, but you misunderstood what I was doing there. If you keep everyone who has Covid-19 out of hospitals, the number of additional non-covid-19 deaths should be zero.

 

[Rive]

If you keep everyone who has Covid-19 out of hospitals, the number of additional non-covid-19 deaths should be zero.

On the other hand, the Covid-19 deaths would go up. I don't think there is any 'good' solution for this.

Because independent of how it is being sold, if there's not plan/goal, why are we doing what we are doing?

No idea. I don't know if there is a plan, or the biggest goal is just (political) survival.
Kinda' feels like way back watching the BSG series. Was that 'they have a plan' stuff, can you recall? Later on it turned out that there wasn't, but it sold itself well anyway.

But I can tell you one thing: I can't get through to people even the basic price of 'herd immunity'. When I just multiply the 60% with the mortality and apply it to the population they call me alarmist and worse, but none dares to take it seriously.

Ps.: it is the same with hospitalization rate.

 

[atyy]

Should we really wait and see or should we try to duplicate their success?

Well, speaking selfishly, I think it would benefit us (Singapore) if the US tried to replicate their success.

But the US system is very complicated. Maybe even more complicated than the German system (probably an understatement) - I mention Germany, because I think its health system is also one in which each state is responsible for its own contact tracing etc. Germany seems to have done decently so far, even if it's not as well as South Korea.

At the start of this epidemic, I had expected the US CDC (at that time it had tremendous reputation throughout the world) to coordinate US efforts. The CDC made some big mistakes, notably on the development of testing in the US. However, now I wonder whether the CDC had any power to coordinate the efforts of different states in the first place. I know we are way past it now, but what would have an optimal US response looked like?

 

[gmax137]

But that's not how "flattening the curve" was described.

Right, as I understood it, the idea was that, while the number of infections would be the same (area under the curve), the number of deaths would be lower since at any time, fewer people would be hospitalized. If too many enter the hospital at once, they don't get the care they need to survive.

This makes sense, since without a vaccine, eventually everyone will be infected.

But it stops making sense if it is true that 80% of the hospitalized die anyway. I don't know if that number is right, I have heard (TV news reports) even higher numbers.