COVID-19

[Mister Imperceptible]

@Dr. Fisker

It seems to me less a question of competence than one of motivation. The more byzantine the total process, the greater the separation of a Good Doctor from an executive decision made by someone whose interests and motivations are not aligned. Gauging the success of a country’s response to CV19 depends on the metric being used, and by whom.

Remember, Dr. John calculates how likely the result of 99 consecutive identical dice throws is, but Fat Tony assumes such a result means the die must be loaded.

[George the original one]

Of course, if the number of new cases per day are increasing linearly (proportional to x), then the cumulative case count is increasing proportional to x^2.

Right, we agree, but are just looking at it from different positions and I didn't grok your perspective vs. mine.

[Peanut]

https://www.nationalreview.com/2020/05/ ... -strategy/

I can't believe it's come to this, that I'm recommending the National freakin' Review, but these are strange times... apologies to those who love this rag.

I reference this because it suggests that the whole, 'no way to protect the vulnerable, so we must lockdown everyone' philosophy is wrong. I was impressed by it because like many observers in the media I had expected Florida to explode with fatalities. This article recounts DeSantis's response to Covid:

-DeSantis looked at S Korea and noted the concentration of Covid deaths in over 70s; looking at Italy confirmed that the state should focus its efforts on at-risk groups
-DeSantis researched Spanish flu pandemic and lack of precedent for national lockdown made him wary of potential negative effects from such policy
-His team lost confidence early on in infection models that looked inaccurate, and prioritized what they were seeing on the ground
-Florida's past experience in dealing with natural disasters and emergency infrastructure made it possible for them to carry out their strategy effectively (like Taiwan in this respect)
-Describes in detail how they protected nursing homes. Given that up to half of Covid deaths have happened in these settings, this is the part of the piece that really stands out. Essentially, the guiding philosophy was to prevent hospital overwhelm by getting to the most likely patients before the virus did.
-Suggests Cuomo, who has been so lauded in the media, clearly made big mistakes in his approach to nursing homes

I don't know what all the counter-points to the strategy would be, but I'm sure there are a bunch. In my view once the genie is out of the bottle all you can do is damage control. What we've had in many states (NY, NJ, IL, MI) is instead compounded damage.

[7Wannabe5]

I wonder why that strategy worked in Florida, but not in England? Population density, weather, spread of contagion before enacted?

[ZAFCorrection]

Not sure about Florida, but the nursing home situation has been a thing for awhile. I believe the NYT estimated that 30% of fatalities are linked to nursing homes when the population of nursing homes makes up like 5% of the 65+ demographic. I would call that a disproportionate impact.

Meanwhile, Cuomo was forcing the facilities to take positive patients so they could recuperate outside of hospitals. Somehow the media has been too busy giving him the Beto O'Rourke beatification to concern themselves much with that.

[tonyedgecombe]

I wonder why that strategy worked in Florida, but not in England? Population density, weather, spread of contagion before enacted?

There was no strategy to isolate care homes early on, mostly people were looking aghast at Italy and the number of relatively young people dying. I don't think it occurred to the government to look at care homes.

It probably didn't help that care homes in the UK are under resourced, especially those funded by councils.

We are also twice as densely populated as Florida.

[Peanut]

You can always count on me for another link, or two...

https://www.theguardian.com/world/2020/ ... ome-deaths
https://www.nytimes.com/2020/05/25/worl ... homes.html

"HK recorded zero deaths in care homes from Covid-19 by employing strict infection control measures that were ignored in the UK, MPs were told on Tuesday as the death toll from the virus in English and Welsh care homes reached almost 15,000." (Btw the real figure may be 22k)

-CEO of care homes testified both symptomatic and asymptomatic patients were moved from hospitals into the homes. IOW lots of people making the same mistake in place after place.
-Apparently some staff work at 3 or 4 facilities, increasing odds of transmission. They were denied tests and protective gear.

I do think the timing was a factor too. It had already spread in the UK. But also poor leadership at multiple levels.

[Riggerjack]

I wonder why that strategy worked in Florida, but not in England? Population density, weather, spread of contagion before enacted?

The virus doesn't seem to be very good at transmission outdoors. In particular, sunny outdoor places. UV, I expect.

Fl has more warm weather, outdoor activities than the UK. So not cutting off outdoor activities seems to work well.

Blue states have a different relationship with their citizens than red states. The idea that citizens could be outdoors, unsupervised, is easier to imagine in rural states. Places where people have more comfort living at their own risk than living with the expectation of lifeguards and park rangers on constant patrol.

City parks are a good example of this. Some people can't work a BBQ, let alone be trusted to walk alone in the woods. How much coddling is appropriate is a judgement call. This call will go different ways in blue states than red states.

One isn't better than the other. Just different expectations.

[7Wannabe5]

Gotcha. I was under the impression that the first British strategy based on first advice offered by epidemiologists was an effort to isolate the elderly. Then the second more grim bit of advice based on evidence that wasn’t working was to go to lockdown.

Another factor might be that Florida seniors are more likely to live at great distance from younger members of extended family, whereas intergenerational housing is much more prevalent in Italy where it spread like fire.

[Riggerjack]

Yeah. Good catch.

I don't imagine many seniors in FL are doing much commingling, regardless of lockdown or not. When they are interacting, it's likely with others who are similarly cautious. And with many of the condos set up with enclosed recreation areas, having some outside time is possible for even the condo dwellers who are being cautious.

Different from NYC and urban UK, where apts are small, crowded, and there aren't much outside spaces to enjoy without running into people desperate for the same.

As for the national review article, it seemed like a puff piece. All the effects of the policy chosen have worked well in the area they were applied... So far.

And really, that's my only objection. Treating covid like because it hasn't been a major catastrophe under lockdown conditions, it wasn't and won't be a catastrophe is... let's just say Prematurely Optimistic.

When/if someone is able to get down to contract tracing and standard public health procedures and still contain the virus, then I will relax. Until then this is still a burning/smoldering wildfire, with no firebreaks, and a mass of people desperate to call the fire out, so they can "get back to normal".

Whereas I think we have a new normal, and there is no going back. Just different paths forward, none going to where we came from.

[daylen]

The concept of "objective reality" is slowly fading back into history and it does seem that the benefit of reality based knowledge is getting reduced to having some personal satisfaction in being able to justify one's certainty about [future] reality and being more correct than wrong about it. This has to be weighed against the frustration of seeing the rest of humanity largely ignoring such knowledge and walking into predictable and preventable problems again and again. I don't have any philosophical comments about how to properly classify this kind of nihilistic post-post-modernism... but on the ground level I am observing something akin to the Kubler-Ross process in knowledge workers as the consequences of this shift are piling up.

It seems to still fit postmodernism if you consider the deconstructive/reconstructive dichotomy. Without a robust modernistic construction or institutional structure to embed themselves in, humans are susceptible to adopting a mixture of postmodern signalling and argumentative tactics that can appear(*) to deconstruct modernist narratives via a near maximum entropy word salad.

Though, you are definitely operating with more ground-level information than me. Perhaps a new classification is warranted?

(*) To others who are susceptible thereby spreading like an epidemic.

[nomadscientist]

You can always count on me for another link, or two...

https://www.theguardian.com/world/2020/ ... ome-deaths
https://www.nytimes.com/2020/05/25/worl ... homes.html

"HK recorded zero deaths in care homes from Covid-19 by employing strict infection control measures that were ignored in the UK, MPs were told on Tuesday as the death toll from the virus in English and Welsh care homes reached almost 15,000." (Btw the real figure may be 22k)

-CEO of care homes testified both symptomatic and asymptomatic patients were moved from hospitals into the homes. IOW lots of people making the same mistake in place after place.
-Apparently some staff work at 3 or 4 facilities, increasing odds of transmission. They were denied tests and protective gear.

I do think the timing was a factor too. It had already spread in the UK. But also poor leadership at multiple levels.

This verges on making the disaster a creation of government. That is about 2/3 of the total deaths. The need to shelter care homes was obvious and no one who advocates "lock downs" can argue there were insufficient resources; the UK has already spent hundreds of billions on sheltering mostly healthy young people who weren't in danger from the virus. This is not an understandable or forgivable error.

[saving-10-years]

There was no strategy to isolate care homes early on, mostly people were looking aghast at Italy and the number of relatively young people dying. I don't think it occurred to the government to look at care homes.

as @TonyE says

I'm agreeing that in the UK this was a sector that was not really seen as part of the immediate problem. Emphasis has been from the start to 'Protect the NHS' and build capacity there and only gradually looking at what is happening in the care sector. This concern seemed to be driven by news from other countries about care home tragedies and the realisation it was not only the NHS that suffered from PPE shortage, (actually it seemed that they were on the brink of running out but still had stock but care homes really had almost nothing).

I was under the impression that the first British strategy based on first advice offered by epidemiologists was an effort to isolate the elderly.

:@7w5

In the UK there was prompt action to alert and try to protect (shield) 1.5m extremely medically vulnerable people.
They receive extra help and are on lists shared with other places - e.g. supermarkets - so that they can get priority for ordering online, etc.
They qualify for food parcels, phone help, etc. The list is not age related.
So in UK 70+ year olds were asked to self-isolate (no help given centrally), but the shielded group are told to 'stay-home-don't-go-anywhere' and some help was given to do this.

I know several people in this category and they are taking it all VERY seriously.
I know several people who feel they should be on the list but are not.

People who have had a 'solid organ' transplant (kidney, liver, intestines, heart, lung and pancreas).
Pregnant women who also have significant heart disease.
People with certain cancers:
- Anyone currently having chemotherapy or radical radiotherapy for lung cancer.
- People having any treatment for blood or bone marrow cancers - leukaemia, lymphoma, myeloma, etc.
- Anyone having immunotherapy or other ongoing antibody treatments for cancer.
- People having 'targeted' cancer treatment - PARP inhibitors or protein kinase inhibitors (your team can confirm if you're on this treatment - if you're not sure, assume you are taking it until you've spoken with them).
- People who have had a bone marrow or stem cell transplant in the last six months, or who are still taking medication to suppress their immune system.
People with rare diseases that significantly affect their risk of infection (such as severe combined immunodeficiency, or SCID).
People with sickle cell disease (not sickle cell trait).
Anyone with cystic fibrosis.
People with severe asthma or COPD.
People taking medication that damps down (suppresses) their immune system significantly.
People who have had a splenectomy (spleen removal). (This group was added in April.)[/list]

https://patient.info/news-and-features/ ... o-needs-it
Methodology behind the list is published here https://digital.nhs.uk/coronavirus/shie ... ethodology

[learning]

@Jacob

Looking back in the thread I see that you recommended:

Garrett, Coming Plague.
Quammen, Spillover.
Preston, Hot Zone.

Are these still the best non-fiction w/o formulas books that you would recommend?

Among the papers that you have read so far, which are the best?

7Wannabe5 recommended:
An Introduction to Mathematical Modeling of Infectious Disease by Li
An Introduction to Infectious Disease Modeling by Vynnycky and White

Just to have some of the book recommendations in one place.

OK, so after the "herd immunity threshold" (HIT) is reached there will be some "overshoot"--up to 1/3 more of the population becoming infected--until eventually "viral death" is reached. What is really the desired end state? It could be viral death but it could also be a low prevalence endemic disease. Is there a term for this? Basically, the state where if you take these precautions then you can not worry about it.

Assuming for this discussion that there will be no vaccine available in time to make a difference, and that there will be enough immunity in the infected and recovered population to make a difference, what is your estimate for what % of the US and world population will be infected before this desired end state of low prevalence endemic/viral death is reached?

I have read numbers ranging from 50%-90% infected before reaching HIT.

HIT = 1 - (1/Rt)

Rt HIT HIT + 1/3 Overshoot
2 50% 67%
3 67% 78%
4 75% 83%
5 80% 87%
10 90% 93%

OR

Rt 2 3 4 5 10
HIT 50% 67% 75% 80% 90%
HIT + 1/3 Overshot 67% 78% 83% 87% 93%

Simply, given a fixed R0 for the virus, the more people mix and create hospitable conditions for viral spread, the higher Rt, the higher HIT, the higher HIT + Overshoot, the lower is 1 - (HIT + Overshoot) uninfected, the harder it is to avoid infection, the more disciplined one has to be in one's individual protective measures to avoid infection. For example, if ~50% will be infected and ~50% will remain uninfected, then one has a reasonable chance to avoid infection by using individual protective measures. If ~80% will be infected and ~20% will not, one must be very disciplined to avoid infection. If 90% are infected and 10% not, it becomes decreasingly likely that one can avoid infection even with extreme discipline in individual protective measures.

My non-expert unknowledgable _guess_ is that the circulation of people in the current opening in the US is creating conditions for an Rt of something like 3-4, ie, ~80% total infected. Clearly, circumstances will change with openings-closings-reopenings, adoption of individual protective measures, seasonal changes, etc., but for a big round number guess, this is what I would give now. I worry that even many who are very to extremely disciplined in their individual protective measures will be infected, like 7wb5.

What do you think?

To clarify, more than the particulars of the "model," I'm mostly interested in an estimate of the % infected and uninfected when the "desired end state" is reached and the probability of avoiding infection by using individual preventive measures that are independent of what the public and the government do. What are your thoughts specifically on this #?

[jacob]

@learning - As far as I know, the term for a low prevalence of a disease is simply what endemic means. The oscillation of breakouts happens around Rt*Susceptible = 1. The size of the overshoot would certainly depend on how large Rt is and how quickly the population of susceptible people responds to bring it down again. It doesn't have to be 1/3.

Recall that given RS=1, then a breakout decreases S which sends the reproduction under 1 per infected. Meanwhile, the act of having children increases S steadily until breakouts become possible again. Thus an endemic disease is observed in waves. Unless there's either a vaccine or the disease is completely eliminated from the planet (like smallpox), it will never go away.

The "number" depends on whether the "desired end state" involves a cultural change so that Americans wear masks when out and about or at least do not show up at work sick, etc. If so, the Rinfinity will be lower (maybe slightly above 1) and thus herd immunity is low (10-30%).

Alternatively, returning to the way things were before, the ultimate herd immunity required is higher.

Different areas will probably adopt different behavioral modifications. You'll probably see more masks in the city and indoors than you will in open spaces.

The herd immunity will therefore not be the same everywhere. This also means that spreaders (or superspreaders) can ignite pockets. E.g. consider a city which follows all the rules so herd immunity is 10% but 90% remains uninfected receiving a visitor who infects 15 people. While this outbreak will still die out on its own (because people already wear masks), such incidents will happen constantly as long as "traveling" from hotter (in the virus sense) occurs.

As we move forward, humanity will be dialing in what kind of end-state combination of infected/dead people (reduction in life expectancy) vis-a-vis inconveniences and economic damage, we're willing to accept.

I'm fairly convinced that CV19 deaths are already becoming normalized. It will just be one of those things that kill old people and random young people, like cancer, but within a couple of weeks. Note that we already accept the existence of diabetes, COPD, cancer, and cardiovascular diseases which for a large part are preventable through healthy lifestyle measures. "But who wants to live that way?" goes the counter argument. We'll simply add it to the human disease burden.

Individual preventive measures, insofar you don't want to risk ending up in the sick or dead category at all, involve avoiding close contact with other people, wearing masks, handwashing, disinfecting frequently touched areas as well as anything touched (or coughed on) by "unclean" hands, the usual stuff. You simply have to keep the dose exposure as low as possible. It works pretty well. SARS2 is not Ebola (which requires almost no dose) or airborne (which is hard to avoid). Or airborne Ebola (which actually exists but thankfully only causes mild symptoms in humans).

[learning]

@Jacob

Yes, I agree that the size of the overshoot will vary based on how people act and how they respond. You make many good points. Thank you for explaining so clearly.

You refrained from hazarding a particular guesstimate for the USA or the world. May I ask why?

To elaborate my question, in case anybody else would like to answer, assuming

1) no vaccine in time to alter the course of the pandemic, and
2) substantial and durable immunity in people who are infected and recover, for example, at least 2 years of immunity,

then: over the next 2-5 or more years, based on what you have already observed of the American government and people, what percent do you guesstimate will become infected, to the nearest 5% or 10%?

[learning]

I'm looking for a detailed analysis of the data that are emerging from covid infecting different populations, from a statistical methodological standpoint. For example, how are the populations of China, Italy, Iran, South Korea, Spain, and the United States similar and how are they different? What are the specific differences? What analysis can be performed to derive useful results for different countries, taking into consideration the differences and similarities of the populations?

Does anybody know of a thorough discussion of this by somebody who is competent in statistical methodology?

[jacob]

You refrained from hazarding a particular guesstimate for the USA or the world. May I ask why?

Because I don't know what kind of end-state people ultimately desire.

Clearly people disagree on what are acceptable losses in terms of jobs and lives. The choice of strategy has already been shown to be rather inconsistent. For example, in the US (and the UK) the strategy was initially to let the virus wash through. Then it was decided that the cost in lives was too high and a lock down was imposed. Then it was decided that the lock down was too hard on the economy and now some states are opening up despite the curve still rising in these states. We're clearly muddling our way through here and while it's relatively easy to predict what will happen for a given strategy, it's harder to predict what strategy people will choose or whether they'll change their minds.

It's possible to give you the ultimate herd immunity numbers you're asking for with an uncertainty of 10% for a given strategy, but it's impossible to predict which of these strategies people (Trump, state governors, corporations, individuals) will choose with the same level of certainty.

Countries with more self-consistent/less populist leadership are easier to predict. For example, the Swedish approach with R->1 dynamically converging on a constant number of infections will ultimately infect 60-70% of the population. The south-east Asian approach with widespread mask usage and serious contact tracing and quarantine measures for travelers will ultimately infect 5-10% of the population.

One thing to consider is that in the US, politicians currently in power have all been elected based on a pre-corona basis, whereas after November, many will have been elected based on a post-corona basis. IOW, after November, the alignment between the people and the various leaders will have been updated to reflect the more recent information. Currently, one of the issues is that leadership has broken down on several levels, e.g. governors not doing with the president wants, corporations and businesses not following the governors' orders, and individuals basically doing their own thing whether that's hunkering down or having corona parties. After November, I'd expect more alignment [between voters and leaders] although I have no idea what that alignment will be.

PS: This is all with holding currently known factors of the virus constant. It could mutate into a less lethal version (they usually do) or it could mutate into a more virulent version (like the Spanish Flu did). The first SARS (SARS-CoV1) did the former and basically faded away as did the associated vaccine research effort.

[7Wannabe5]

@learning:

No matter how many people eventually end up infected, the slower the virus transmission rate, the lower the percentage of population walking about actively infected at any given time will be, and this is the factor that matters more in terms of level of protection any individual might have to assume in order to avoid being infected with some degree of confidence.

For instance, if on average a person is only actively infectious for about 2 weeks, and it takes 2 years for the virus to spread through 80%=herd immunity level of the population then less than 1% of the people around you on average will be able to pass on the infection at any given time during that period. However, even if ultimate no behavior barred herd immunity = only 50%, it is possible to find yourself in a population in which more than 1 in 20 people in your population are currently infectious.

Even though I was taking strict precautions, I was also in a zip code that had likely already very quickly achieved 10% penetrance if the infection by mid April, so quite possible that 3% or 4% of the people living around me were actively infectious. Under those circumstances, less likely methods of infection become more likely. I still don’t know if I contracted the disease, but based on my symptoms,if I did it is likely that I ingested it and had hyper-active immune response.

[Viktor K]

https://www.npr.org/2020/05/25/86192354 ... ogist-says

7.3% of people in Stockholm had developed antibodies against COVID-19 by late April

he believed that by now "a little more than 20%" of Stockholm's population should have contracted the virus.

With 39.57 deaths per 100,000, Sweden's mortality rate is not only higher than that of the U.S. (30.02 deaths per 100,000) but also exponentially higher than those of its neighbors Norway (4.42 per 100,000) and Finland (5.58 per 100,000)