What do I need to know about coronavirus?

[dadaddadaBATMAN]

---

Members do not see this ad.

Actually, the people who live IN the city need it more.
What would be an efficient way, beyond what they have done with essentially a shelter in place order, to dissuade people from using the subway? Issue IDs to grocery clerks?

they likely already have an id or similar, or a letter.

People could easily synthesize similar badges but it would be more inconvenient.

Doesn’t seem much crazier than what’s been done

 

[lp430]

If they ground domestic flights, these hospitals that rely on locums coverage are going to be SOL. I feel very fortunate that I moved out of NYC right before all this started. Much respect to the docs out in the trenches there now.

 

[bky3c]

I wonder how our country is going to address the fact that NYC residents will soon start leaving the city in droves...assuming that has not already started to happen. All those lovely personalities from Manhattan, Queens, and the Bronx seeding infection across the country as they scatter across America. The same goes for Detroit and New Orleans which are right behind NYC in terms of new cases.

I think the Escape from New York is on Netflix. Pretty soon it may be time to embrace our inner Snake Plissken.

Oh, it's started. They recently arrived in my parents' neighborhood several states away -- adult children and their little kids from NYC decided to go stay with Grandma and Grandpa in the country until this blows over. None of them had symptoms on arrival, but who knows about asymptomatic spread. And of course they drove there, which means they had to stop for gas, restrooms, maybe food, etc.

 

[bravotwozero]

Governor Orders Limited Gatherings, Declares Most Businesses 'Essential,' Supersedes Local Safety Efforts

Mississipi needs to get a clue

 

[Birdstrike]

Oh, it's started. They recently arrived in my parents' neighborhood several states away -- adult children and their little kids from NYC decided to go stay with Grandma and Grandpa in the country until this blows over. None of them had symptoms on arrival, but who knows about asymptomatic spread. And of course they drove there, which means they had to stop for gas, restrooms, maybe food, etc.

I live in a tourist town in the Southeast and people are coming here from NY in droves. In fact, our governor has already issued an order that they must quarantine themselves for 14 days after arrival. I'm pretty sure that will be near impossible to enforce.

 

[Birdstrike]

I think the Escape from New York is on Netflix. Pretty soon it may be time to embrace our inner Snake Plissken.

Great movie

 

[thegenius]

I live in a tourist town in the Southeast and people are coming here from NY in droves. In fact, our governor has already issued an order that they must quarantine themselves for 14 days after arrival. I'm pretty sure that will be near impossible to enforce.

Corona is coming to you! You can't escape it.

Corona
Corona
CORONA!!!

 

[VA Hopeful Dr]

I live in a tourist town in the Southeast and people are coming here from NY in droves. In fact, our governor has already issued an order that they must quarantine themselves for 14 days after arrival. I'm pretty sure that will be near impossible to enforce.

My governor just did the same, I wonder if we live in the same state. Though outside of Charleston nothing we have is a tourist town and we have no large PP multispecialty groups to my knowledge.

 

[miacomet]

they likely already have an id or similar, or a letter.

People could easily synthesize similar badges but it would be more inconvenient.

Doesn’t seem much crazier than what’s been done

You are going to make every bodega, corner store, and grocery worker, as well as every delivery person, make their own unofficial ID? Or you think bodega workers already have a special bodega ID? Who writes the bodega letter?

What's to stop everyone from making a grocery clerk ID?

You do understand that you are requiring tiny independent grocers and clerks to come up with some sort of official ID?

Who checks this? Who pays them? How do they know if the ID is fake or real?

 

[thegenius]

You are going to make every bodega, corner store, and grocery worker, as well as every delivery person, make their own unofficial ID? Or you think bodega workers already have a special bodega ID? Who writes the bodega letter?

What's to stop everyone from making a grocery clerk ID?

You do understand that you are requiring tiny independent grocers and clerks to come up with some sort of official ID?

Who checks this? Who pays them? How do they know if the ID is fake or real?

You asked 8 questions in one post. That is too many.
Please reduce them down to 3 next time. 🙂

 

[Birdstrike]

Corona is coming to you! You can't escape it.

Corona
Corona
CORONA!!!

Me yesterday: “We can handle Coronavirus guys. We got dis!”

Me right now after seeing all the NY/NJ plates at Costco: “Where’s the plaquenil? Where’s the !*#!*#ing
PLA-QUE-NIIIIIIIILLLL. !!!!???!!!!!”

“

 

[miacomet]

You asked 8 questions in one post. That is too many.
Please reduce them down to 3 next time. 🙂

It would help if people thought out their oh-so-easy solutions.

 

[turkeyjerky]

I live in a tourist town in the Southeast and people are coming here from NY in droves. In fact, our governor has already issued an order that they must quarantine themselves for 14 days after arrival. I'm pretty sure that will be near impossible to enforce.

China absolutely shut down a city the size of Chicago and imported a number of healthcare workers equivalent to 50% of the number of ultimate cases. They built special hospitals to quarantine suspected cases. They only considered someone ruled out if they had 2 negative PCRs and a negative CT scan. They voluntarily tanked their economy.

And they did all this when they only had about 1000 cases.

I just don't think we're up the task here. We're cutting hospital staffing, running out of PPE, and having hospitals try to out-bid each other for critical equipment.

Highly recommend the following article on admin responses across the country:

COVID Pandemic Exposes the Ugly Secrets Hidden in America’s Healthcare System:

 

[Birdstrike]

China absolutely shut down a city the size of Chicago and imported a number of healthcare workers equivalent to 50% of the number of ultimate cases. They built special hospitals to quarantine suspected cases. They only considered someone ruled out if they had 2 negative PCRs and a negative CT scan. They voluntarily tanked their economy.

And they did all this when they only had about 1000 cases.

I just don't think we're up the task here. We're cutting hospital staffing, running out of PPE, and having hospitals try to out-bid each other for critical equipment.

Highly recommend the following article on admin responses across the country:

COVID Pandemic Exposes the Ugly Secrets Hidden in America’s Healthcare System:

Yeah, man. What do you want me to say. It’s a real ****-show out there, man. A real **** show.

 

[evilbooyaa]

NYC should be treated as Italy or China was a few weeks ago. If you come from NYC to X, you are required to stay on a 14-day quarantine. Of course people won't ****ing listen and will spread it, but NYC, if it isn't already, should be on lockdown like Italy and Wuhan were. The national guard should be clearing out the streets of NYC, setting up road blocks and only allowing those doing essential things - going to work, going home, going to the grocery store, going to pick up medication from a pharmacy. One of those 4 things or turn your ass around.

But we're not, b/c 'Murica, and therefore it's going to keep spreading. Completely agree that still being able to fly out of NYC is a public health debacle.

De Blasio has just given up, Cuomo is fighting to make sure people don't die, and Trump is going to get people killed by the thousands if he sticks with his Easter 'deadline'.

That being said, the paper about perhaps we've developed herd immunity is interesting, and hope we can get some data over the next few days if that's actually real. I agree that if it is real, maybe time to pull back.

One thing I want to say about the difference between exponential and logistic growth for the people who keep getting into a pissing match over it:

This will be logistic growth. However, the first 2/3rd of an exponential graph and a logistic graph can look quite similar. Even if it is a logistic growth curve, it can still grow exponentially right now, until it hits the carrying capacity (the horizontal asymptote) and burns itself out. TBD what that is. We don't know if this rate of growth will continue for 2 weeks or 2 months or longer because we don't know what the point it's going to level off at will be.

 

[miacomet]

Do you think we should do the same thing with New Orleans? Apparently that's the epicenter now....

 

[Birdstrike]

This will be logistic growth. However, the first 2/3rd of an exponential graph and a logistic graph can look quite similar. Even if it is a logistic growth curve, it can still grow exponentially right now, until it hits the carrying capacity (the horizontal asymptote) and burns itself out. TBD what that is. We don't know if this rate of growth will continue for 2 weeks or 2 months or longer because we don't know what the point it's going to level off at will be.

Go to Worldometer and pay special attention to the logarithmic graphs. They are much better at bringing into focus when this is going to happen. They give your different information than the linear graphs.

 

[evilbooyaa]

Do you think we should do the same thing with New Orleans? Apparently that's the epicenter now....

Have heard anecdotally about NO, haven't seen hard numbers from it. Worldometer is great but doesn't have those cool graphs for each individual state so hard to see acceleration of case rate.

 

[dadaddadaBATMAN]

You are going to make every bodega, corner store, and grocery worker, as well as every delivery person, make their own unofficial ID? Or you think bodega workers already have a special bodega ID? Who writes the bodega letter?

What's to stop everyone from making a grocery clerk ID?

You do understand that you are requiring tiny independent grocers and clerks to come up with some sort of official ID?

Who checks this? Who pays them? How do they know if the ID is fake or real?

For some reason my notifications aren’t showing me when people are replying, I might need to check my settings.

anyway, you seem very emotionally invested in this, and you are needlessly abrasive.

In answer to your question(s), the method employed in Italy for preventing unauthorized use was to have people carry a form with them where they stated their need to leave their home or go on public transport.

it’s far from perfect, and it may not prevent use of the subway system or leaving the house. But at least it’s some attempt at controlling public transit and putting massive numbers of people in direct contact with one another and countless fomites where the virus lives for days at a time.

your alternative seems to be to whine, hand-wring about the state of the country, and jump down the throats if anyone who has a slightly different viewpoint than you rather than considering any different view or solution.

if your behavior in life resembles your behavior on this board, I have little trouble figuring out why you are experiencing so much dissatisfaction and burn out. Your attitude is horrible and you seem miserable. Good luck to you.

 

[evilbooyaa]

Go to Worldometer and pay special attention to the logarithmic graphs. They are much better at bringing into focus when this is going to happen. They give your different information than the linear graphs.

I've looked at them, do you mind explaining further what I should be looking at? They show a linear pattern on a log scale, which can be indicative of either exponential or logistic growth, IIRC. Been a long time since whatever math course taught us all this stuff.

 

[Matthew9Thirtyfive]

Even if it is a logistic growth curve, it can still grow exponentially right now, until it hits the carrying capacity (the horizontal asymptote) and burns itself out. TBD what that is. We don't know if this rate of growth will continue for 2 weeks or 2 months or longer because we don't know what the point it's going to level off at will be.

If you track the rate of change ratio, once it hits 1 and stays below 1 (it was actually .9 the other day but jumped up to 1.2 the next day), that is the inflection point, and if the growth follows logistic growth, the horizontal asymptote will be about twice the number of cases at that inflection point. This is sometimes easier to see on the log scale, because you can tell that it is growing exponentially as long as it is a straight line.

 

[evilbooyaa]

If you track the rate of change ratio, once it hits 1 and stays below 1 (it was actually .9 the other day but jumped up to 1.2 the next day), that is the inflection point, and if the growth follows logistic growth, the horizontal asymptote will be about twice the number of cases at that inflection point. This is sometimes easier to see on the log scale, because you can tell that it is growing exponentially as long as it is a straight line.

Thanks. Good point.

So right now, since the logarithmic scale is basically a straight line, we have not yet reached the inflection point, and we have no means of knowing when we will reach said inflection point?

 

[Birdstrike]

I've looked at them, do you mind explaining further what I should be looking at? They show a linear pattern on a log scale, which can be indicative of either exponential or logistic growth, IIRC. Been a long time since whatever math course taught us all this stuff.

Okay. Take Italy for example. Go over to the Italy Worldometer page. You'll see that currently there are 80,589 cases and 8,215 deaths.

Then if you look at the linear graph of that, you can see they are still rising:

But looking at those case numbers and the linear graph, gives you no idea of where that increase is likely to level off. Where are the cases and deaths likely to level off at? There are no indications at all, in those two forms of looking at the info. Will the cases level off at 90,000, at 100,000 or 1,000,000? Are the deaths going to rise to 10,000, or 100,000 or 3,000,000? Higher? Neither of those graphs give you any hint. Now, there's no way for any of us to predict for sure. But look at the logarithmic graphs of the same data:

Now looking at the data this way, it's a totally different story, isn't it? Again, none of us know for sure where these numbers will end up, but if you were to forecast Italy leveling off somewhere around 100,000 cases and 10,000 deaths, you'd have a heck of a lot more confidence in those educated guesses by looking at the data in this form.

If you go back and look at the data in the same way, at China for example, which is much further along with this, and you'll see that those logarithmic graphs honed right in around just under 100,000 for cases and around 4,000 for deaths. And that's just about exactly where they're at and seem to be stuck.

Bottom line, those logarithmic graphs give you a much better idea of where that logistic growth is likely to level out at. It's not perfect, but its much better than looking at the "exponential" part of the linear graphs. It give the impression if infinite rise, with no end point, and that's not how these things work.

Will continue in a new post, because there's max number of pictures SDN will let you post in 1 post...

 

[turkeyjerky]

Dammit Jim, I'm a doctor not a mathematician!

It looks like the logarithmic curves model caseload w/ extreme control measures. Unfortunately, we don't know if that'll be possible here. And we don't know what happens once we loosen social distancing. And we know that the growth in both Italy and China was enough to overwhelm the local health systems. I don't know about you, but I'd prefer to avoid extreme rationing of ventilators.

 

[evilbooyaa]

Sure, fair enough, and I get those points for Italy and China. I get what you're saying, Birdstrike, but we're not at a point yet where we will know what that number will be in the US.

 

[Matthew9Thirtyfive]

Bottom line, those logarithmic graphs give you a much better idea of where that logistic growth is likely to level out at. It's not perfect, but its much better than looking at the "exponential" part of the linear graphs. It give the impression if infinite rise, with no end point, and that's not how these things work.

The problem is that with logistic functions, which are generally of the form y(t) = C/(1+a*e^(-kt)), you can have very different growth curves just by altering the value of k and C, and they can look very similar.

 

[Birdstrike]

Of course, the obvious next question is, how does USA look through that lens?

Answer: It's too early to say. Let's do the same thing for us, as of today:

So, looking at our data in that form, looks pretty bad right? If you looked at this and said, "We're effed, with no end in sight!" I'd have no argument with you and you might be right (and you might be).

But let's take a look at this "logistic growth" through a logarithmic lens:

It looks better on this view, doesn't it? But not leveling off yet, as much as we want. That's going to take a little more time. Some of the other countries have gotten there faster than us. But if I had to guess where these would level off, looking at the data this way, in this form, I'd guess that the cases would level out somewhere between 250,000 and 750,000 and the US deaths somewhere between 3,000 and 9,000. Now, I could be totally wrong. It's just an educated guess. It's too early to say for sure. And the people saying, "We're totally f---ked," could be totally right, because as we all know, the future is difficult to predict. But when I made those exact predictions on this thread many, many posts back, this is one of the data sets I used to choose those numbers.

If you follow the patterns of the other countries ahead of us through this data though this lens, keeping in mind logistic growth (not "exponential" growth) is the pattern these outbreaks follow, that seems to be a much better way, from what I've seen than simply plugging and chugging "Death rate x population x estimated total population affected." Those calculations are how you get the numbers like "2.2 million Americans dead" and "1.6 million Americans dead"

0.01 [1%] death rate x 331,000,000 Americans x 0.50 [50% infection rate] = 1.6 million Americans dead. Does that number sound familiar?
Yes, it's the number Fmr CDC Dir Tom Frieden and the media took hold of and spread virally. But have you stopped to wonder why they haven't gone back and applied the numbers to China, where the infection has already run it's course? Because if you do the number you get is 7,000,000 deaths. 0.01 x 1,400,000,000 x 0.5 = 7,000,000.

Has China had 7 million deaths?
No.

Is China going to hit 7 million deaths?
No

I just showed you and pretty much everyone agrees now, that at least at the moment, China has leveled off just under 3,500 deaths. I know some people have questioned China's data, but do you really think they're hiding 7 million deaths? Do they have 200,000% more deaths than they're reporting (that's the actual number, not a typo, 200,000% more; i.e. 7,000,000 / 3,500 x 100 = % increase)? Do they have 2,000 times more deaths than we think? I doubt it. I could see them being off a few percent, hell, even 50% or even double at 7,000 deaths. But nowhere near the millions, which would be 7,000,000 for their population, which is the equivalent of 1.5 million for ours.

 

[Birdstrike]

Sure, fair enough, and I get those points for Italy and China. I get what you're saying, Birdstrike, but we're not at a point yet where we will know what that number will be in the US.

Was typing part 2, which has the answer to this, as you were typing, but it took a while to finish and hit enter. My answer, one post above. I think it'll make more sense when you read part 2.

 

[Birdstrike]

The problem is that with logistic functions, which are generally of the form y(t) = C/(1+a*e^(-kt)), you can have very different growth curves just by altering the value of k and C, and they can look very similar.

A formula is great, but show us some examples like I did above.

 

[Birdstrike]

Dammit Jim, I'm a doctor not a mathematician!

It looks like the logarithmic curves model caseload w/ extreme control measures. Unfortunately, we don't know if that'll be possible here. And we don't know what happens once we loosen social distancing. And we know that the growth in both Italy and China was enough to overwhelm the local health systems. I don't know about you, but I'd prefer to avoid extreme rationing of ventilators.

Honestly, I'm with you all at this point. I've tried to spell out in detail above where I "think" we may be headed. But I fully admit I could be wrong at this point. I'm beyond the point of saying this isn't a big deal or trying paint a rosy picture of this. Early on, I thought (and hoped) SARS-COV-1 (COVID-19) would fizzle out quickly like SARS-COV in 2003-2004. That hasn't happened. I was wrong. 100% of what I've written on this thread could end up being 100% wrong. It could get W A Y worse before it gets better. We could 100% be f--cked.

Or maybe not. Maybe we end up nowhere near the worst case scenarios and some of the stuff I said on this thread will turn out to be right. But I'll be honest with you, I have no f--cking idea at this point, other than my best, amateur guess 3, 4, 5 posts above, that come with no guarantees whatsoever. So, if you say, "We're f--ked," then we're f--ked. You'll get no further arguments from me. You guys win. You broke me. From now on, I'm assuming you're all right and, "We're all screwed." From now on I'm going into, I-just-hope-we-get-lucky mode. Don't expect any more reassurance from we, because I don't have it. But I'm not sure what else we can do at this point. This virus is here to stay. I'm probably going to get it. Hopefully my family won't. Hopefully I'll fight it off. Make sure you've got some Plaquenil for yourself and your family. Every man for himself.

 

[Matthew9Thirtyfive]

A formula is great, but show us some examples like I did above.

So I just plugged in all the data from the JHU site into Origin and did some curve fitting. Using a logistic fit for growth, we get this:

Which then, plotted out looks like this:

This model puts the carrying capacity at 138,603. Now, that means our inflection point will happen at 69,302, which is just north of the upper limit of the data. Since the latest data on JHU has us at 83,836, that means theoretically we should be past the inflection point and into the slowing down phase of the growth.

If I add that extra 15k or so cases to the data, we get:

The carrying capacity now is 222,009, which puts our inflection point at 111,005, so we haven't hit it yet. If we solve for that, we can see that we should theoretically hit the inflection point on just about day 67, which is 2 days from now.

Those curves look very similar, but just adding a little more data, we get an increase in the carrying capacity by 83,406 cases. As more data gets added, we can get a clearer picture of how the curve will look, but it's really impossible to say with any certainty. There are tons of more complex models out there, and just adjusting some things a little bit, you can make it predict 10s of millions of infected people with 4-6 million dead in the US. Or you can change some things and drop those numbers by a lot.

 

[RustedFox]

So I just plugged in all the data from the JHU site into Origin and did some curve fitting. Using a logistic fit for growth, we get this:

View attachment 299901

Which then, plotted out looks like this:

View attachment 299902

This model puts the carrying capacity at 138,603. Now, that means our inflection point will happen at 69,302, which is just north of the upper limit of the data. Since the latest data on JHU has us at 83,836, that means theoretically we should be past the inflection point and into the slowing down phase of the growth.

If I add that extra 15k or so cases to the data, we get:

View attachment 299905
View attachment 299906

The carrying capacity now is 222,009, which puts our inflection point at 111,005, so we haven't hit it yet. If we solve for that, we can see that we should theoretically hit the inflection point on just about day 67, which is 2 days from now.

Those curves look very similar, but just adding a little more data, we get an increase in the carrying capacity by 83,406 cases. As more data gets added, we can get a clearer picture of how the curve will look, but it's really impossible to say with any certainty. There are tons of more complex models out there, and just adjusting some things a little bit, you can make it predict 10s of millions of infected people with 4-6 million dead in the US. Or you can change some things and drop those numbers by a lot.

Wow.
Great post.

If SDN had "Golds" to award like Reddit does, this gets a Gold.

 

[Matthew9Thirtyfive]

Wow.
Great post.

If SDN had "Golds" to award like Reddit does, this gets a Gold.

If you’re being serious then I think I just hit the highlight of my SDN career. Never thought I’d get an RF compliment lol.

 

[Birdstrike]

So I just plugged in all the data from the JHU site into Origin and did some curve fitting. Using a logistic fit for growth, we get this:

View attachment 299901

Which then, plotted out looks like this:

View attachment 299902

This model puts the carrying capacity at 138,603. Now, that means our inflection point will happen at 69,302, which is just north of the upper limit of the data. Since the latest data on JHU has us at 83,836, that means theoretically we should be past the inflection point and into the slowing down phase of the growth.

If I add that extra 15k or so cases to the data, we get:

View attachment 299905
View attachment 299906

The carrying capacity now is 222,009, which puts our inflection point at 111,005, so we haven't hit it yet. If we solve for that, we can see that we should theoretically hit the inflection point on just about day 67, which is 2 days from now.

Those curves look very similar, but just adding a little more data, we get an increase in the carrying capacity by 83,406 cases. As more data gets added, we can get a clearer picture of how the curve will look, but it's really impossible to say with any certainty. There are tons of more complex models out there, and just adjusting some things a little bit, you can make it predict 10s of millions of infected people with 4-6 million dead in the US. Or you can change some things and drop those numbers by a lot.

I appreciate it. Thank you, because I asked for you to write this post. I have no disagreement with any of it. Read my update, which I was writing when you were writing this (ended further up the thread). The ball is in your court: What do I need to know about coronavirus?

 

[RustedFox]

If you’re being serious then I think I just hit the highlight of my SDN career. Never thought I’d get an RF compliment lol.

Lol.
I'm not famous.
I'm just a loudmouth with some creative writing skills.

Truly, a great post.

 

[Matthew9Thirtyfive]

I appreciate it. Thank you, because I asked for you to write this post. I have no disagreement with any of it. Read my update, which I was writing when you were writing this: What do I need to know about coronavirus?

Yeah, tbh I’m not in the end of the world camp. I’m really just in the we don’t know camp. I expect there to be a few hundred thousand cases. That model up there could be right, and we could see this thing slow down tomorrow or the next day. Or it could continue to increase by an order of magnitude every 8 days until half the country is infected. One thing I know from getting my math degree is that it is really great at modeling things once we have a lot of data. Unfortunately, we just don’t have a lot of data right now.

 

[Birdstrike]

So I just plugged in all the data from the JHU site into Origin and did some curve fitting. Using a logistic fit for growth, we get this:

View attachment 299901

Which then, plotted out looks like this:

View attachment 299902

This model puts the carrying capacity at 138,603. Now, that means our inflection point will happen at 69,302, which is just north of the upper limit of the data. Since the latest data on JHU has us at 83,836, that means theoretically we should be past the inflection point and into the slowing down phase of the growth.

If I add that extra 15k or so cases to the data, we get:

View attachment 299905
View attachment 299906

The carrying capacity now is 222,009, which puts our inflection point at 111,005, so we haven't hit it yet. If we solve for that, we can see that we should theoretically hit the inflection point on just about day 67, which is 2 days from now.

Those curves look very similar, but just adding a little more data, we get an increase in the carrying capacity by 83,406 cases. As more data gets added, we can get a clearer picture of how the curve will look, but it's really impossible to say with any certainty. There are tons of more complex models out there, and just adjusting some things a little bit, you can make it predict 10s of millions of infected people with 4-6 million dead in the US. Or you can change some things and drop those numbers by a lot.

Okay. I see what you did here. I see the large amount of variation possible when you vary the inputs with data from one snapshot in time like the Johns Hopkins site. But we have more than snapshot. We have data over time and data that's falling into a pattern already. See if you can take the data from the Worldometer site and see if you can use that as a starting point. Because it seems clear that the trends are falling into some pattern, without such a wide ranging margin of error as the wild extremes you're talking about. The further we get into this, the margin of error should shrink, sort of like the cone of uncertainty gets more narrow as a hurricane gets closer. You have the math background and software that I don't. so see if you can refine it with some of the data and graphs from worldometer. If you can. Or if it's too hard, don't bother. But I wonder if that would help refine it. I don't know.

 

[Matthew9Thirtyfive]

Okay. I see what you did here. I see the large amount of variation possible when you vary the inputs with data from one snapshot in time like the Johns Hopkins site. But we have more than snapshot. We have data over time and data that's falling into a pattern already. See if you can take the data from the Worldometer site and see if you can use that as a starting point. Because it seems clear that the trends are falling into some pattern, without such a wide ranging margin of error as the wild extremes you're talking about. The further we get into this, the margin of error should shrink, sort of like the cone of uncertainty gets more narrow as a hurricane gets closer. You have the math background and software that I don't. so see if you can refine it with some of the data and graphs from worldometer. If you can. Or if it's too hard, don't bother. But I wonder if that would help refine it. I don't know.

I used the data from JHU from the first case in the US until what we have now (64 days of data from the first confirmed case they have recorded). So the fit is based on all the data we have, not just a snapshot. The difference between the two models are just that I added one more day of data. I could do the same with the worldometer data, but it would be about the same. The numbers might be slightly different, but we’d see the same issue—as we get more data, we refine the model. That’s really my point.

We will know more as we get further into this pandemic. But I also think we are also trending toward something, though that’s just gestalt.

 

[alpinism]

NYC has definitely gotten busier over the past few days but its still a long way from becoming a war zone.

We're probably seeing on average roughly 100 cases a day at this point but the vast majority have mild symptoms and can be safely discharged. I'd estimate that probably only 10% require supplemental oxygen and another 5% require intubation. The main problem is that many hospitals are still operating business as usual and triaging these patients into the main ER where they sit for hours instead of quickly sending them home and preventing it from spreading to other patients. In addition many docs are doing full sepsis workups on young healthy patients then will perform walking sats measurements and if they ever drop below 95% will admit them to the hospital.

 

[zurned]

NYC has definitely gotten busier over the past few days but its still a long way from becoming a war zone.

We're probably seeing on average roughly 100 cases a day at this point but the vast majority have mild symptoms and can be safely discharged. I'd estimate that probably only 10% require supplemental oxygen and another 5% require intubation. The main problem is that many hospitals are still operating business as usual and triaging these patients into the main ER where they sit for hours instead of quickly sending them home and preventing it from spreading to other patients. In addition many docs are doing full sepsis workups on young healthy patients then will perform walking sats measurements and if they ever drop below 95% will admit them to the hospital.

What is your ED policy on nebs and Bipap on these people?

 

[miacomet]

For some reason my notifications aren’t showing me when people are replying, I might need to check my settings.

anyway, you seem very emotionally invested in this, and you are needlessly abrasive.

In answer to your question(s), the method employed in Italy for preventing unauthorized use was to have people carry a form with them where they stated their need to leave their home or go on public transport.

it’s far from perfect, and it may not prevent use of the subway system or leaving the house. But at least it’s some attempt at controlling public transit and putting massive numbers of people in direct contact with one another and countless fomites where the virus lives for days at a time.

your alternative seems to be to whine, hand-wring about the state of the country, and jump down the throats if anyone who has a slightly different viewpoint than you rather than considering any different view or solution.

if your behavior in life resembles your behavior on this board, I have little trouble figuring out why you are experiencing so much dissatisfaction and burn out. Your attitude is horrible and you seem miserable. Good luck to you.

I just think you people should have done more research. Subway and bus traffic is down 75-90% in NYC, so it seems the problem has mostly been resolved.

Who issues these forms? How do people apply? It seems very difficult, for example, to get the undocumented guy at the bodega to register for or carry a form.

 

[miacomet]

No PAPRs at my dysfunctional FSEDs. I tried wearing an N95 yesterday for 4 hours and it was torture - just ripping in to the back of your scalp. Will probably start wearing my own respirator with P100 filters.

Hospital I work at uses 90% locums/firefighters from out-of-state. Admin has said that any physicians recently in CA, WA, or NY need to quarantine in-town for 14 days before working a clinical shift. That's basically our entire roster. Quarantine in a hotel, where they don't live, for 2 weeks, at their own expense. When we told them that this would be impossible/a total non-starter, we were told "figure it out". This is what I'm up against. Idiots.

My buddy is an intensivist at a large academic MICU in the Midwest. Has one patient in early 40s, no PMHx, on VV ECMO. Multiple people intubated under 40 years old. I barely got a flight home from a locums assignment yesterday. It's getting hairy out there.

Are you sticking with this job? It seems insane. I'm so sorry.
Agreed the quarantining at one's own expense seems to be a thing; I don't get it. Have they since rectified this?
I don't know how places with IC and locums are going to survive. I'm getting tons of locums calls, but there is no provision as to what to do if you get sick, or if the planes stop flying, or if you are quarantined.

Awful about young folks. I don't know why people still think this only affects the elderly.

 

[TooMuchResearch]

Add to the anecdote pile:
We have a previously healthy person under age 30 in the ICU, intubated with high COVID-19 suspicion.

Are you sticking with this job? It seems insane. I'm so sorry.
Agreed the quarantining at one's own expense seems to be a thing; I don't get it. Have they since rectified this?
I don't know how places with IC and locums are going to survive. I'm getting tons of locums calls, but there is no provision as to what to do if you get sick, or if the planes stop flying, or if you are quarantined.

Awful about young folks. I don't know why people still think this only affects the elderly.

 

[thegenius]

If you go back and look at the data in the same way, at China for example, which is much further along with this, and you'll see that those logarithmic graphs honed right in around just under 100,000 for cases and around 4,000 for deaths. And that's just about exactly where they're at and seem to be stuck.

It's stuck because they are a bunch of LIARS. No way they are only getting 50ish new cases a day, a country of > 1,000,000,000 people, as they loosen the borders.

 

[thegenius]

So I just plugged in all the data from the JHU site into Origin and did some curve fitting. Using a logistic fit for growth, we get this:

View attachment 299901

Which then, plotted out looks like this:

View attachment 299902

This model puts the carrying capacity at 138,603. Now, that means our inflection point will happen at 69,302, which is just north of the upper limit of the data. Since the latest data on JHU has us at 83,836, that means theoretically we should be past the inflection point and into the slowing down phase of the growth.

If I add that extra 15k or so cases to the data, we get:

View attachment 299905
View attachment 299906

The carrying capacity now is 222,009, which puts our inflection point at 111,005, so we haven't hit it yet. If we solve for that, we can see that we should theoretically hit the inflection point on just about day 67, which is 2 days from now.

Those curves look very similar, but just adding a little more data, we get an increase in the carrying capacity by 83,406 cases. As more data gets added, we can get a clearer picture of how the curve will look, but it's really impossible to say with any certainty. There are tons of more complex models out there, and just adjusting some things a little bit, you can make it predict 10s of millions of infected people with 4-6 million dead in the US. Or you can change some things and drop those numbers by a lot.

Ooohhhh
This is a statistics fight.
I love it.

it looks a little weird though, but adjusting numbers is considered within the rules of statistical fighting.

 

[dadaddadaBATMAN]

I just think you people should have done more research. Subway and bus traffic is down 75-90% in NYC, so it seems the problem has mostly been resolved.

Who issues these forms? How do people apply? It seems very difficult, for example, to get the undocumented guy at the bodega to register for or carry a form.

I did do research on it. though honestly I’m not proposing policy at the state house, I’m shooting the **** on an online forum.

you’re viewing it from the perspective that everyone has to be issued something for it to work. That’s not the deal. The deal is that if you don’t have it, you don’t get on the train. You make it available online or on a smart phone, which most people have (including >90% of homeless from my ed encounters, though may be a biased/savvy sample) . You have to mark down a date, time and reason for leaving the house. Any cop anywhere can stop you and tell you to go home if you don’t have the paper/form.

admittedly, this did not work great in Italy, but honestly I don’t know what else they can do short of marshal law, which few people are probably willing to do. It beats doing nothing though.

I was surprised that traffic only seemed to be down 15% by comparison to last year. Admittedly my source was a random article and I don’t have access to reliable information on ridership.

Redirect Notice

This link was my source on ridership. Do you have one for it being down 90%? That would be great news

 

[Lawpy]

So I just plugged in all the data from the JHU site into Origin and did some curve fitting. Using a logistic fit for growth, we get this:

View attachment 299901

Which then, plotted out looks like this:

View attachment 299902

This model puts the carrying capacity at 138,603. Now, that means our inflection point will happen at 69,302, which is just north of the upper limit of the data. Since the latest data on JHU has us at 83,836, that means theoretically we should be past the inflection point and into the slowing down phase of the growth.

If I add that extra 15k or so cases to the data, we get:

View attachment 299905
View attachment 299906

The carrying capacity now is 222,009, which puts our inflection point at 111,005, so we haven't hit it yet. If we solve for that, we can see that we should theoretically hit the inflection point on just about day 67, which is 2 days from now.

Those curves look very similar, but just adding a little more data, we get an increase in the carrying capacity by 83,406 cases. As more data gets added, we can get a clearer picture of how the curve will look, but it's really impossible to say with any certainty. There are tons of more complex models out there, and just adjusting some things a little bit, you can make it predict 10s of millions of infected people with 4-6 million dead in the US. Or you can change some things and drop those numbers by a lot.

ngl i'll be surprised if carrying capacity is less than 200k

 

[bravotwozero]

What are you guys thoughts on the whole 'asymptomatic carrier' business? What is the evidence for this? If the infection is so widespread, that HAS to mean that a huge chunk of us who had no symptoms got it, recovered, and are now immune. But is that really the case? Do we even know how long one can remain a carrier?

 

[thegenius]

What are you guys thoughts on the whole 'asymptomatic carrier' business? What is the evidence for this? If the infection is so widespread, that HAS to mean that a huge chunk of us who had no symptoms got it, recovered, and are now immune. But is that really the case? Do we even know how long one can remain a carrier?

I suspect they have mild symptoms and are not reporting it.

For instance, I've blown my nose 3 times today. Maybe I have corona virus.
Or a tickle in the throat.

I have a hard time believing the virus is multiplying madly inside your respiratory tract, and one therefore spews forth huge viral load droplets without having ANY symptoms. Sure there is a small majority where that is the case but I'm sure most people probably have some symptom.

 

[thegenius]

I'll tell you this. There should not be a single region in the world that should ever experience what Spain and Italy are going through right now with COVID-19. From this date on. Unfortunately Spain and Italy had to go through this, and maybe NYC.

If regions aren't locking down now they are idiots.