Coronavirus

[bkp_duke]

There is a unique symptom of COVID-19 that some people have reported. They lose their sense of smell and often taste just before coming down with it. They are not stuffed up or anything that would normally cause a loss of sense of smell. That happened to me mid-February for no reason. I normally have an unusually good sense of smell and I could just barely smell many things I knew smell strongly. The weird asthma-like symptoms started right about that time too. My sense of smell came back, but I don't remember when, probably only a few days.

Just for the record, loss of smell and taste is not a unique symptom of SARS-CoV-2 infection. It doesn't come with influenza, but it is pretty well-known to happen with other viruses. The MSM has latched on to this and overblown it as a symptom.

Until you provide an antibody status that proves otherwise, as a physician I strongly question your assertion that you had a SARS-CoV-2 infection. I hear people near me stating the same thing, and every one of them keeps forgetting, or ignoring, that there is NOT a unique specific symptom or constellation of symptoms for infection of this virus.

I'm not saying ignore the symptom of loss of smell and taste, but it needs to be taken with a grain of salt and put into perspective. This is NOT a binary indicator.

 

[bkp_duke]

The current PCR test uncovers hidden infections because it tests for the virus. Antibody tests test for antibodies which you only produce after you've been infected for some period of time (don't know how long).

IgM antibodies are produced during infection and used to help fight off the virus. IgG antibodies are produced at the end of the infectious cycle.

That is just the antibody branch of the specific immune response, which researchers are telling us is not that good with most coronaviruses (and why immunity wanes over months to years). We don't know exactly yet how good or bad it is with this one.

With coronaviruses, the T-cell response is historically more robust, specifically the T-helper 2 (Th2) cells. There is no ready-made test to check for Th2 population (you can do FACS, but that's research lab stuff). This population of cells is also known to tail off in a few months after exposure.

 

[nervus]

Actually, the user that posted that is well-known for advocating for very socialistic policies. That's fine, that's his right, but they are "pie in the sky" ideas with no real rational thought put forward on how to pay for them.

The title he posted was essentially click bait, and it IS a very important consideration for how something will be paid for and changes the discussion considerably.

There are VERY few countries that if I, as a US citizen, go to and get sick that they will pay for my healthcare. That's just a fact of how our world works.

Politically, we are having enough problems figuring out how to pay for US citizens healthcare. And while I sympathize with the plight of a visitor getting sick and needing care, I do believe that if they are going to visit, they should have the forethought to purchase temporary insurance to cover the potentiality of getting sick here.

In my country we would, as well as in most of the northern European countries. Not because we are liking you, but just because it is the decent thing to do (Luke 10 : 33).

 

[Daniel in SD]

was the spread attributed to just 1 person bringing it onboard?

Yes.

 

[jhm]

I hate to say it but as far as I can tell with the current death rate and trajectory in Italy a fatality rate in the range of 200-500 DPM in the relatively short term is a near certainty. It could even be higher with the hospitals completely overwhelmed in some areas. The only thing I can see that will stop that is some miraculous treatment breakthrough in the very short term, which seems highly unlikely. One crude model I posted recently estimated that deaths in Italy would be in the range of 25000-32000, which would put the DPM at the high end of your range, and it is not clear this model adequately factors in the added fatalities from an overwhelmed medical system. https://www.medrxiv.org/content/10.1101/2020.03.25.20041475v1.full.pdf

I would appreciate your thoughts about this as I have great respect for your modeling, and I'm basically spitballing here.

The reason I am interested in IFR and/or serological testing is to have some sense of where we will be in different parts of the world after social distancing and other policies have taken effect and the virus is under control at least temporarily. Will 50% of the population already be exposed and herd immunity in effect, or will we barely be scratching the surface with only 1-5% exposed? That's a critical question and we don't seem to have a clue.

One of the problems with deciding how to tackle this virus or plan for it is that almost all of the attention is focused on the short term (understandable in a crisis but still a problem). We lack the most basic information necessary to predict how things will play out in the medium to long term. If we tamp it down in the next couple months in the Northern hemisphere, is it going to come roaring back with equal or greater force after we lift restrictions, or next winter (assuming it is temperature sensitive -- IMO probably yes but another issue where the jury still seems to be out).

Thanks. I wish the authors of that note had included a mathematical specification of their model. From the verbiage, I can't really tell what they are doing, i.e., I could not replicate their results without having access to their actual code.

Here's what I would do. First I would work with log(N_t) rather than N_t. This transformation roughly linearizes the counting process.

Next, the focus is on differences in log(N),

dlogN = log(N(t+1)) - log(N(t))

If one can form a reasonable forecast for dlogN, you can accumulate that back to logN and transform to N. So the strategy is to form a reasonable local (recent case) model for dlogN.

The simplest model is

dlogN = a + b t + error

This can be estimated on the most recent 5 or more cases. The b coefficient is related to the second derivative or curvature of N. It tells us how much the growth rate is declining each day.

Note also that t = -a/b is the time when reaches its peak. This is interesting as rough estimate of when deaths halt. But it is only good as a local model. In actually dlogN could remain strictly positive for the foreseeable future, but become small enough that the epidemic no longer disrupts society. So other kinds of models can be formulated to deal with longer term dynamics.

But for our purposes here, a quick estimate for the next 7 days or so, the linear model above can work reasonably well. Note that if b >= 0, there is no stabilization in the near term; the outbreak is accelerating, not stabilizing!

So that gives you an idea of how I would approach this. I have not actually started playing with the data. Rather I am content to look at charts below to suggest where we are headed.

Note that dlogN is just log(GrowthFactor) in this chart. For the most recent 2 weeks taking the log of growth factors can help make this a little more linear. At any rate, Netherlands needs to decrease case growth rates even faster to cool this off in the near term.

 

[bkp_duke]

In my country we would, as well as in most of the northern European countries. Not because we are liking you, but just because it is the decent thing to do (Luke 10 : 33).

And that's great that you do that, but it's not the norm throughout the world. I've traveled enough, and gotten sick abroad, to know that it is always in my benefit to have a travel insurance policy.

I'm NOT ARGUING against the MORAL argument of taking care of your brother. I agree with that philosophy.

I'm stating that in reality, unfortunately, that's just not how the majority of our world works. For better for for worse. Those that are "disagreeing" with me I hope are disagreeing based upon their wishes, because the facts are . . . simply the facts.

 

[Daniel in SD]

Let's play Armchair Epidemiologist.

On what day will the US cross the 10 deaths per million inhabitants? (To be precise, 3310 deaths as published here https://www.washingtonpost.com/graphics/2020/world/mapping-spread-new-coronavirus/)

Post your guesses.

My guess is March 31.

March 30th (on worldometer, add whatever lag to the WP)

 

[EinSV]

Thanks. I wish the authors of that note had included a mathematical specification of their model. From the verbiage, I can't really tell what they are doing, i.e., I could not replicate their results without having access to their actual code.

Here's what I would do. First I would work with log(N_t) rather than N_t. This transformation roughly linearizes the counting process.

Next, the focus is on differences in log(N),

dlogN = log(N(t+1)) - log(N(t))

If one can form a reasonable forecast for dlogN, you can accumulate that back to logN and transform to N. So the strategy is to form a reasonable local (recent case) model for dlogN.

The simplest model is

dlogN = a + b t + error

This can be estimated on the most recent 5 or more cases. The b coefficient is related to the second derivative or curvature of N. It tells us how much the growth rate is declining each day.

Note also that t = -a/b is the time when reaches its peak. This is interesting as rough estimate of when deaths halt. But it is only good as a local model. In actually dlogN could remain strictly positive for the foreseeable future, but become small enough that the epidemic no longer disrupts society. So other kinds of models can be formulated to deal with longer term dynamics.

But for our purposes here, a quick estimate for the next 7 days or so, the linear model above can work reasonably well. Note that if b >= 0, there is no stabilization in the near term; the outbreak is accelerating, not stabilizing!

So that gives you an idea of how I would approach this. I have not actually started playing with the data. Rather I am content to look at charts below to suggest where we are headed.

View attachment 526372
Note that dlogN is just log(GrowthFactor) in this chart. For the most recent 2 weeks taking the log of growth factors can help make this a little more linear. At any rate, Netherlands needs to decrease case growth rates even faster to cool this off in the near term.

Italy's growth factor looks a little more promising, but ICU trends are disturbing. Something to keep an eye on as a hint of where we might be heading in other hot spots.

Coronavirus Italia

 

[jbcarioca]

Actually, the user that posted that is well-known for advocating for very socialistic policies.
...
There are VERY few countries that if I, as a US citizen, go to and get sick that they will pay for my healthcare. That's just a fact of how our world works.

Politically, we are having enough problems figuring out how to pay for US citizens healthcare. And while I sympathize with the plight of a visitor getting sick and needing care, I do believe that if they are going to visit, they should have the forethought to purchase temporary insurance to cover the potentiality of getting sick here.

Apart from policy characterization as a politcal attitude, he is my actual outlook.

1. I do advocate treating everyone, regardless, for any communicable disease with Ro higher than 1.
2. It is almost always far cheaper to prevent a communicable disease than it is to treat one.
3. It is nearly always cheaper to treat one rather than allow unchecked spread.
4. The societal consequences of unmitigated spread are almost always reflected in economic costs, often not quantified.

Those four points can be debated, but can also be quantified. I do not suggest I have done that in this case. In numerous cases from typical influenza, yellow fever, 'childhood' diseases much of the world routinely vaccinates nearly everyone thought to be susceptible. Even in poor countries (Brazil is an example) most vaccinations are free and available to everyone. In the present case everyone symptomatic is tested if they show up.

It is simple economics for me, driven primary by Ro.

On most health topics I am distinctly more nearly agnostic. My personal experience under both 'socialistic' and 'capitalistic' systems has been a bit mixed, but I have never been in a situation in which economic rationing would have had any effect on my life.

Undoubtedly most MD's prefer a system that allows them to set their own prices. Globally private services are typically more pleasant than are public, from both patient and physician perspective. This debate is active among my physician friends, but all of them benefit greatly by individual price setting. Clearly there are limits when dealing with health plans, rarely as onerous as those of a nationalized system. None of that has anything at all to do with managing epidemics the most effectively and efficiently.

It is interesting how much these issues surface when an epidemic hits. BTW, in what systems are countries best equipped to deal with epidemics and pandemics? Think economically and from a public health perspective. Think a bit about supplies of critical equipment stored 'just in case'.

 

[Daniel in SD]

No!! It does NOT '...atke a week to be contagious". The gestation period varies from <one day to ~three weeks. The fourteen day is arbitrary; after all there must be a convention for such rules, and outliers will happen in any rule at all.

The Imperial College study used a time from infection to infectiousness of 4.6 days. The only point I was trying to make is that if you've got 100 people standing next to each other for 30 minutes it's not going to hop from person to person and infect all 100 people. I was arguing against the idea that spreading from 1 person to 676 in as little as two weeks on a cruise ship means it's not all that contagious.

https://www.imperial.ac.uk/media/im...-College-COVID19-NPI-modelling-16-03-2020.pdf

 

[bkp_duke]

Apart from policy characterization as a politcal attitude, he is my actual outlook.

1. I do advocate treating everyone, regardless, for any communicable disease with Ro higher than 1.
2. It is almost always far cheaper to prevent a communicable disease than it is to treat one.
3. It is nearly always cheaper to treat one rather than allow unchecked spread.
4. The societal consequences of unmitigated spread are almost always reflected in economic costs, often not quantified.

Those four points can be debated, but can also be quantified. I do not suggest I have done that in this case. In numerous cases from typical influenza, yellow fever, 'childhood' diseases much of the world routinely vaccinates nearly everyone thought to be susceptible. Even in poor countries (Brazil is an example) most vaccinations are free and available to everyone. In the present case everyone symptomatic is tested if they show up.

It is simple economics for me, driven primary by Ro.

On most health topics I am distinctly more nearly agnostic. My personal experience under both 'socialistic' and 'capitalistic' systems has been a bit mixed, but I have never been in a situation in which economic rationing would have had any effect on my life.

Undoubtedly most MD's prefer a system that allows them to set their own prices. Globally private services are typically more pleasant than are public, from both patient and physician perspective. This debate is active among my physician friends, but all of them benefit greatly by individual price setting. Clearly there are limits when dealing with health plans, rarely as onerous as those of a nationalized system. None of that has anything at all to do with managing epidemics the most effectively and efficiently.

It is interesting how much these issues surface when an epidemic hits. BTW, in what systems are countries best equipped to deal with epidemics and pandemics? Think economically and from a public health perspective. Think a bit about supplies of critical equipment stored 'just in case'.

I'm not arguing that in this acute setting, that this approach is invalid. And when I was practicing we always were "treat first, figure out how to pay later". Many times we could simply convince our hospital to eat the cost and not go after the family (it was, however a Children's hospital and perhaps they are a bit more kind).

But when we look as a society, there is a (political) discussion that has to be had do we cover healthcare for visitors. There will be vehement arguments on both sides that I don't need to rehash here.

 

[ggr]

There are VERY few countries that if I, as a US citizen, go to and get sick that they will pay for my healthcare. That's just a fact of how our world works.

How, exactly, do you claim to know that? It isn't true. I've been treated for free in two countries when visiting for conferences (Australia and Germany, so I think most of EU probably have the same policy).

 

[Dr. J]

There was no real push to go down the path of testing and isolation

This part of your post is true.

 

[linux-works]

DIY oriented article on uvc, from today:

https://hackaday.com/2020/03/27/measuring-uv-c-for-about-5/

(hack a day is actually a good 'maker' site with lots of ideas shared; code, circuits, enclosures, etc)

this may be useful - I was asking just recently how we are going to calibrate our uvc lightboxes and now there's a sensor that at least gives us a starting value to measure.

sensor is in the $10 range and code is available.

mine will be here early next week

 

[bkp_duke]

How, exactly, do you claim to know that? It isn't true. I've been treated for free in two countries when visiting for conferences (Australia and Germany, so I think most of EU probably have the same policy).

Travel to south america, central america, and less developed parts of asia.

EDIT - GB didn't when I visited a few years ago.
https://www.quora.com/Which-countries-offer-free-healthcare-to-visitors

EDIT2 - 26 European countries require you to buy healthcare coverage (travel insurance) before entering:
Countries Which Won't Let You In Without Health Insurance

 

[Daniel in SD]

A ray of hope I'm seeing is the liberal community has now taken the denier role in American society. Trump's moronic denial if urgency has created a massive liberal reaction very similar to the act of voting for Trump.

We've shut down the entire nation (which I'm 100% FINE with) with no real scientific indication it's anywhere near a net positive health policy. There was no real push to go down the path of testing and isolation that was not only recommended strongly by the WHO, but was extremely successful in other countries. We're basing policy on a doomsday scenario that was never real

If you think of a Trump voter as a person who's thrown up their hands and sat down in the middle of the path, you could draw a pretty good parallel to left leaning America right now.

I take that as a good indicator we're moving toward a time of greater understanding and cooperation. Though I'm an irrational optimist.

We have not shut down the entire nation. Some states have done shut downs, others have not. How can you do testing and isolation when you don't have the testing capacity? Now that we have more testing capacity the virus is out of control and I haven't seen anyone credible say that it would be possible to do in our current situation. What are people denying? The doomsday scenario is only not real because many people are not in denial and the path towards that scenario would bring even more people out of denial.
Where are those SARS numbers? Don't you want to know if you were wrong?

 

[Lessmog]

No support in reality? We have no idea how many people have caught the virus and recovered. Everyone agrees that the official number of cases in the US is understated

So, you actually agree then that "your totally unqualified assertion that the fatality rate will be "0.2% or less", "has no support in reality"? In that case, why argue?

 

[bkp_duke]

How, exactly, do you claim to know that? It isn't true. I've been treated for free in two countries when visiting for conferences (Australia and Germany, so I think most of EU probably have the same policy).

Interestingly enough, I would like to know how Germany treated you for free. According to this, you are supposed to come with insurance that covers medical problems if you enter Germany:

Countries Which Won't Let You In Without Health Insurance

"The Schengen Area is a zone in Europe where 26 countries have acknowledged the abolishment of their internal borders. Anyone who needs to apply for the Schengen visa to enter Europe must have international health insurance.

The 26 countries in the zone are Austria, Belgium, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Italy, Latvia, Liechtenstein, Lithuania, Luxembourg, Malta, Netherlands, Norway, Poland, Portugal, Slovakia, Slovenia, Spain, Sweden, and Switzerland."

 

[1208]

Biggest daily increase.

Coronavirus: UK deaths rise by 185 to 769

 

[SageBrush]

They add new data (yellow/orange) retroactively to previous days.

Thanks -- now the graph makes sense.

For anybody still unclear, Norbert is saying that it takes from 1 - 3 days for a sample to be tested and the positive cases show up in the data on the day the sample was collected once the test is performed