D&D and the rising pandemic

[abirdcall]

I see good leadership,

Iceland has tested more of its population for coronavirus than anywhere else. Here's what it learned

No country or scientist or doctor has all the answers about the COVID-19 pandemic that has swept the globe. Tiny Iceland may have more than most.

news.yahoo.com

I see bad leadership,
Florida Governor Spreads Dangerous Falsehood About Who Dies From Coronavirus

Yeah, this is why I don't believe the models showing the USA coming out of this any time soon.

I think a lot of people are only looking at the number of deaths too. Being on a ventilator is not a pleasant experience and could result in permanent lung damage. The young may not be dying as often but they can still be in for a rough time.

Hospitalization rates also has a huge collateral effect on people seeking treatment for other illnesses as well as putting a lot of stress on the medical system incl. staff etc. It can only take so much before breaking.

New York State is leveling off but I think other states are just starting to pick up momentum.

 

[tomBitonti]

deaths follow social distancing by a week or 2. I think we will see the deaths per day start to lower in the next few days in the U.S.

I’m seriously questioning even the current 60,000 number. I’m thinking more like 40,000.

What 60k number?

Contradicting the steady state of deaths/day, the number of active cases is still increasing. Perhaps that can be explained by increases in the number of people that are tested.

Tom Bitonti

 

[Umbran]

Yeah, this is why I don't believe the models showing the USA coming out of this any time soon.

The models are updated with data from the real world - so the pattern of what we have been doing so far is included in the results.

There will be some question around any prediction as to how well its assumptions hold. At least these assumptions are explicit. In our personal opinions, not so much.

I think a lot of people are only looking at the number of deaths too.

Duffers on the internet, sure. Folks who actually make a living at this stuff, however, are not "a lot of people".

 

[Janx]

These two things are not the same.

Just yesterday I saw the most astounding thing. People buying a new grill and carpet. Seriously?! The government sends out a stay at home request, and to only go out for bare essentials, and people still go shopping for a new grill and do home renovations?! Well I hope that new grill is worth it. Because some of those people in the same store may end up dead because you wanted to have a barbeque.

But all those fools gathering in a church during a deadly epidemic are not just going to infect themselves, but also others, like you and me.

People should be avoiding any social gatherings right now, and not just churches. For states to make an exemption for churches, is down right criminal. It is a death sentence for many people.

20 people in a room singing and flinging vocal spray at each other is not the same as a lonely ride in a truck to Home Despot, a quiet walk down the aisles to get a rug and boxed grill to a checkout counter with a plexiglass barrier and a zap gun to read the bar-code.

 

[abirdcall]

The models are updated with data from the real world - so the pattern of what we have been doing so far is included in the results.

There will be some question around any prediction as to how well its assumptions hold. At least these assumptions are explicit. In our personal opinions, not so much.



Duffers on the internet, sure. Folks who actually make a living at this stuff, however, are not "a lot of people".

It's just hard for me to see how very different models could both be correct.

Canada has projected 4400 to 22 000 deaths if we continue our lockdown.

The USA has projected 60 000 deaths?

California currently has the same population, confirmed cases, and deaths as Canada and is also under a similar lockdown. That projection that was provided is estimating 1600 deaths from California.

What is so different about Canada that they're expecting a lot more deaths comparatively than the USA which even has large parts of the country not in a lockdown?

That same site is predicting that France experiences a total of 15 700 deaths (14 000 to 20 500 range) and they are currently at 13 200 deaths with 1000 deaths per day. France deaths projected on April 10th (today) is 164 and they are currently at 987.

This is in a country which is actually in lockdown too.

 

[Umbran]

It's just hard for me to see how very different models could both be correct.

Canada has projected 4400 to 22 000 deaths if we continue our lockdown.

The USA has projected 60 000 deaths?

I can't really say much, as I don't know anything about that Canadian prediction.

One thing to note - and I have linked to articles about this a couple times in this thread already, iirc - is that there's lots of models, they don't all agree, and all the models have really wide margins of error. This is a usual bit of failure in media coverage of science - talking only about the "most likely" prediction, and not mentioning how much higher or lower it might be.

It is possible for the base predictions of two models to disagree, but fall within each other's margins for error. I don't know if that's the case here, but when some models predict "most likely a hundred thousand, but upper limit is 2 million" I could imagine that happening.

It is like - we both try to predict the number of cookies a child will eat. I predict 7 cookies, plus or minus 6. You predict 10 cookies, plus or minus 5. We fit within each other's margins of error, so who knows who is correct?

For those who just cannot believe that there may be a peak in sight, perhaps this will be informative - the absolute rate of the moment may be less indicative than the doubling time:

Are We Flattening The Curve? States Keep Watch On Coronavirus 'Doubling Times'

It is starting to take more time for cases, hospitalizations and deaths to double in several states, indicating social distancing is working. Here's how to make sense of those numbers.

www.npr.org

 

[Istbor]

The Institute for Health Metrics and Evaluation, or IHME, a research center at the University of Washington.

IHME | COVID-19 Projections

Explore forecasts of COVID-19 cases, deaths, and hospital resource use.

covid19.healthdata.org

Extensive modelling frequently updated with real data from the field.



Yes, but the rates of new cases is changing - that's what you have to watch, and may indicate that the peak is coming sooner, rather than later.

Alright thanks for the link, the article I read only eluded to this particular model without directly linking to it. Thanks media.

Sounding now like at least my states peak isn't until perhaps the first week of May. I think just trying to keep track of it in the US is going to be odd. Simply because it isn't hitting all states at the same time or at the same level of impact.

 

[tomBitonti]

I can't really say much, as I don't know anything about that Canadian prediction.

One thing to note - and I have linked to articles about this a couple times in this thread already, iirc - is that there's lots of models, they don't all agree, and all the models have really wide margins of error. This is a usual bit of failure in media coverage of science - talking only about the "most likely" prediction, and not mentioning how much higher or lower it might be.

It is possible for the base predictions of two models to disagree, but fall within each other's margins for error. I don't know if that's the case here, but when some models predict "most likely a hundred thousand, but upper limit is 2 million" I could imagine that happening.

It is like - we both try to predict the number of cookies a child will eat. I predict 7 cookies, plus or minus 6. You predict 10 cookies, plus or minus 5. We fit within each other's margins of error, so who knows who is correct?

For those who just cannot believe that there may be a peak in sight, perhaps this will be informative - the absolute rate of the moment may be less indicative than the doubling time:

Are We Flattening The Curve? States Keep Watch On Coronavirus 'Doubling Times'

It is starting to take more time for cases, hospitalizations and deaths to double in several states, indicating social distancing is working. Here's how to make sense of those numbers.

www.npr.org

Nice article! A minor data presentation nit: The “doubling lines” are better viewed as being attached to the curve endpoints, instead of at the origin. What region one of curves is in is not indicative of the current rate of growth, which is indicated by the slope at the endpoint. One might mistakenly think that because a curves lies between two of the doubling lines that the current doubling rate of that curve lies between the doubling values of those lines.

Edit: As the rate of spread is changing, there becomes a difference between the rate of doubling of the total who have become infected and the a doubling of the rate at which new persons are infected.

Thx!
Tom Bitonti

 

[tomBitonti]

Alright thanks for the link, the article I read only eluded to this particular model without directly linking to it. Thanks media.

Sounding now like at least my states peak isn't until perhaps the first week of May. I think just trying to keep track of it in the US is going to be odd. Simply because it isn't hitting all states at the same time or at the same level of impact.

These projections —IHME | COVID-19 Projections — seem dubious, with peaks being reached tomorrow. I don’t see how the projected curve matches other statistics that I’ve see. Are we anywhere close to an overall steady state (in the US), that is, where the number of new cases matches the number who are clinically post infection? A downturn requires at least that, and we don’t seem to be even there yet.

Be safe and be well,
Tom Bitonti

 

[abirdcall]

I looked into the Canada projections of 11 000 to 22 000 deaths.

They are guessing an infection amount of 2.5%-5% of the population.

Then they are taking that number and assuming roughly 1.2% die.

I am sure they are taking into account that not all cases are recorded. The actual death rate of confirmed cases will be higher than 1.2% (the current death rate of confirmed cases is 2.4%).

The USA equivalent to this forecast would be 100 000 to 200 000 deaths. The figure Dr. Fauci originally predicted before changing to 60 000.

Those projections are also showing many states hitting the bottom of their curve by May 1st with the entire country recovering by June 1st.

Perhaps Canada is far behind but they have projected 'mid-summer' for the end of the first wave only if they do a very good job of containment.