COVID-19 Megathread 3: Third time's a charm

[Meclazine for Israel]

A projection meter for COVID-19 in the US.

https://covid19.healthdata.org/projections.

Predicted peak ventilator requirements: 20,000.

Start of June looks promising for an end to this thing. I am still thinking mid June to see cases drop to 10% of their peak in the USA. South Korea is currently dropping past 50% of their peak.

Do they have active cases anywhere, or just deaths?

[GoTfan]

A projection meter for COVID-19 in the US.

https://covid19.healthdata.org/projections.

Predicted peak ventilator requirements: 20,000.

Start of June looks promising for an end to this thing. I am still thinking mid June to see cases drop to 10% of their peak in the USA. South Korea is currently dropping past 50% of their peak.

Do they have active cases anywhere, or just deaths?

Just deaths from what i can see.

The people in Washington state have been using this model and it seems to have been fairly accurate so far.

[brucejoel99]

The “pro-life” Party.

https://twitter.com/propublica/status/1243727045680680964?s=21

One of the biggest problems this disaster has uncovered is the lack of preparation for people in hospitals with special needs, and not just COVID-19 patients. As soon as a ward takes its first coronavirus patient, it has to suspend all visitation, & in many cases, people whose family or a caretaker have been sitting by them & representing them for sometimes weeks or months in a hospital simply disappear. Our most vulnerable have unfortunately been placed in a very ugly position right now.

[emailking]

People that think many parts of the USA won't turn into a situation lile Italy soon have no clue. COVID-19 is spreading faster in America than it did in Italy, and Italy had good leadership and a national lockdown, neither of which America has.

This is really sad for me to say, but in 2 or 3 weeks, Americans will be praying that the situation improves to a level like Italy had. It's going to be worse. Many are in denial but they'll see. I'm so sorry for the American people and their families for what's coming. God bless your souls.

I'm staying inside except fir the grocery store so hopefully I'll be ok. Today when I went, I avoided my face and washed my hands 3 times when I got back and threw my clothes in the laundry right away.

[jimrtex]

https://twitter.com/ScottGottliebMD/status/1243696001958981632

How can it take longer to get a negative result than a positive one?  If after 5 minutes, you don’t have a positive result, what else could the result be?

Abbott Launches Molecular Point-of-Care Test to Detect Novel Coronavirus in as Little as Five Minutes

Abbott's press release says "delivering positive results in as little as five minutes and negative results in 13 minutes".

The equipment uses polymerase chain reaction to make millions to billions of copies of a DNA sample. I'm assuming that depending on the initial concentration, the duplication might be sufficient to detect in 5 minutes, but in lower concentrations might take 13 minutes.

I'm not a scientist, but I did read a Wikipedia article while staying at Holiday Inn Express last night.

[True Federalist (진정한 연방 주의자)]

The “pro-life” Party.

https://twitter.com/propublica/status/1243727045680680964?s=21

To be fair, people with cognitive disabilities often have difficulty accepting intrusive measures, even when they are well-intentioned. In a case where there is a shortage of the devices, it does make a degree of unfortunate sense to prioritize supplying them to people who won't actively resist being ventilated or try to take them out.

[emailking]

Trump's tweet was deleted. Here's the new one.

https://twitter.com/realDonaldTrump/status/1243726993537073152

[Frodo]

Trump's tweet was deleted. Here's the new one.

https://twitter.com/realDonaldTrump/status/1243726993537073152

The projection is so strong it's giving me a headache... 

[HagridOfTheDeep]

Forgive me for circling back, but I'm a bit confused about the projection model on this site posted earlier: https://covid19.healthdata.org/projections.

If my math is right, with total American deaths coming in at around 81,000, a death rate of 1% means we will only see ~8 million Americans get infected with this virus. How is that possible in a country as populous as the United States? Either the death rate is wrong and there is a ridiculous proportion of asymptomatic cases, or the virus will come back with just as much force as before... right?

[GoTfan]

Forgive me for circling back, but I'm a bit confused about the projection model on this site posted earlier: https://covid19.healthdata.org/projections.

If my math is right, with total American deaths coming in at around 81,000, a death rate of 1% means we will only see ~8 million Americans get infected with this virus. How is that possible in a country as populous as the United States? Either the death rate is wrong and there is a ridiculous proportion of asymptomatic cases, or the virus will come back with just as much force as before... right?

One thing the US has working its favour is population density.

The reason why NYC in particular is getting hit hard is because everyone is packed tightly in. The US outside of the coast and industrial states is reasonably spread out.

[Meclazine for Israel]

Forgive me for circling back, but I'm a bit confused about the projection model on this site posted earlier: https://covid19.healthdata.org/projections.

If my math is right, with total American deaths coming in at around 81,000, a death rate of 1% means we will only see ~8 million Americans get infected with this virus. How is that possible in a country as populous as the United States? Either the death rate is wrong and there is a ridiculous proportion of asymptomatic cases, or the virus will come back with just as much force as before... right?

81,000 deaths is a model prediction. I find it ghastly to predict any death to be frank.

But I am predicting a decrease of active cases of between 90-95% of the number of "peak Active cases" (top of the curve) by early to mid-June based purely by modelling the curves from China and South Korea.

https://talkelections.org/FORUM/index.php?topic=355847.msg7257352#msg7257352

Can further back that up when I find the quarter point of the curve. That is the point of maximum rate of change on the way up the slope of Active Cases.



That will not predict the top, but it will predict the time this thing lasts. As yet, the USA numbers have failed to stop accelerating. I will present some data tomorrow that shows that, in the USA, the foot on the accelerator is still firmly pressed to the floor.

Under those circumstances with no drop in growth, it is impossible to predict anything except that the end point is going to be further out in date than first expected. Maybe July-August 2020 for example.

USA has good and bad curve attributes. The good being an accelerated sharp entry may mean a sharp decay and an early finish. The bad being the effect on the healthcare system. The main episode of the Spanish Flu in the USA in 1918 was all over within 45-60 days in October-November of that year.

Be very skeptical of the data being presented by the media as fact.

[It’s so Joever]

A projection meter for COVID-19 in the US.

https://covid19.healthdata.org/projections.

I have my doubts about that projection.
Even if we implemented a full nationwide Wuhan style lockdown across the entire country, case growth itself wouldn’t decline up to April 10, and active case numbers would probably peak 7-10 days after.
At best, we will have a peak by mid-April, but I think early May is more likely. Also I do think more people will die and be infected without further intervention.

[Grassroots]

A projection meter for COVID-19 in the US.

https://covid19.healthdata.org/projections.

End by early June?

That sounds pretty good.

[GoTfan]

A projection meter for COVID-19 in the US.

https://covid19.healthdata.org/projections.

I have my doubts about that projection.
Even if we implemented a full nationwide Wuhan style lockdown across the entire country, case growth itself wouldn’t decline up to April 10, and active case numbers would probably peak 7-10 days after.
At best, we will have a peak by mid-April, but I think early May is more likely. Also I do think more people will die and be infected without further intervention.

it also does have projections for each state.

[It’s so Joever]

A projection meter for COVID-19 in the US.

https://covid19.healthdata.org/projections.

I have my doubts about that projection.
Even if we implemented a full nationwide Wuhan style lockdown across the entire country, case growth itself wouldn’t decline up to April 10, and active case numbers would probably peak 7-10 days after.
At best, we will have a peak by mid-April, but I think early May is more likely. Also I do think more people will die and be infected without further intervention.

it also does have projections for each state.

The model isn’t updated.
Several states that have recently implemented a stay-at-home order are shown as not having done such.

[Meclazine for Israel]

Forgive me for circling back, but I'm a bit confused about the projection model on this site posted earlier: https://covid19.healthdata.org/projections.

If my math is right, with total American deaths coming in at around 81,000, a death rate of 1% means we will only see ~8 million Americans get infected with this virus. How is that possible in a country as populous as the United States? Either the death rate is wrong and there is a ridiculous proportion of asymptomatic cases, or the virus will come back with just as much force as before... right?

81,000 deaths is a model prediction. I find it ghastly to predict any death to be frank.

But I am predicting a decrease of active cases of between 90-95% of the number of "peak Active cases" (top of the curve) by early to mid-June based purely by modelling the curves from China and South Korea.

https://talkelections.org/FORUM/index.php?topic=355847.msg7257352#msg7257352

Can further back that up when I find the quarter point of the curve. That is the point of maximum rate of change on the way up the slope of Active Cases.



That will not predict the top, but it will predict the time this thing lasts. As yet, the USA numbers have failed to stop accelerating. I will present some data tomorrow that shows that, in the USA, the foot on the accelerator is still firmly pressed to the floor.

Under those circumstances with no drop in growth, it is impossible to predict anything except that the end point is going to be further out in date than first expected. Maybe July-August 2020 for example.

USA has good and bad curve attributes. The good being an accelerated sharp entry may mean a sharp decay and an early finish. The bad being the effect on the healthcare system. The main episode of the Spanish Flu in the USA in 1918 was all over within 45-60 days in October-November of that year.

Be very skeptical of the data being presented by the media as fact.

The reason they go with deaths and not 'raw cases' is that deaths are more accurate, particularly in Iran where they obey strict burial procedures.

The Iran death curve for example:

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

looks more consistent than the 'raw cases' data.

That is, they may not trust the data on 'Active Cases' to make predictions. Reverse-engineering everything from the resulting death rate is possible if you know what you are doing.

[It’s so Joever]

https://twitter.com/JesseLehrich/status/1243619964252282886?s=20

The Michigan Genocide will be forever remembered.

[HagridOfTheDeep]

Lower population density is all well and good, as is basing this model on trends we have seen elsewhere, but I guess I am having a hard time reconciling this model with logic. Unless every last stitch of this virus is cut off from the potential to spread and then dies out, it is going to linger. 8 million infections by the end of that curve in the model is nowhere near enough to achieve herd immunity. So isn't it, like... extremely likely it will return?

And when it does, and we start to open things back up, are we really going to be able to do enough tracking, contact tracing, and quarantining of resurgent cases/exposures to put a lid on things before  we get another curve? I'm very pessimistic, sorry.

[Co-Chair Bagel23]

Trump is such a disgusting obese f#ck that i wouldnt be surprised if he needed one ventilator for each lung

[Fmr. Gov. NickG]

I do wonder if some of the need for ventilators is a function of American healthcare culture in general, specifically our propensity to using prolonged extraordinary measures to extend a very poor quality of life to patients with no hope of recovery, largely for the sake of the family’s piece of mind.

I don’t know how much to trust the Chinese data, but what I have seen of that suggests that less than 5% of covid patients who are placed on ventilators ever recover.

There is also anecdotal evidence of many patients in Italy being placed on ventilators for weeks.  And almost no one on a ventilator for weeks ever recovers.   It doesn’t just use up ventilators, but it takes up ICU bed space for long periods.

This seems like a very poor use of scarce resources that doesn’t actually save any lives, perhaps with only some psychological benefit.  Am I missing something?

[Florida Man for Crime]

I do wonder if some of the need for ventilators is a function of American healthcare culture in general, specifically our propensity to using prolonged extraordinary measures to extend a very poor quality of life to patients with no hope of recovery, largely for the sake of the family’s piece of mind.

I don’t know how much to trust the Chinese data, but what I have seen of that suggests that less than 5% of covid patients who are placed on ventilators ever recover.

There is also anecdotal evidence of many patients in Italy being placed on ventilators for weeks.  And almost no one on a ventilator for weeks ever recovers.   It doesn’t just use up ventilators, but it takes up ICU bed space for long periods.

This seems like a very poor use of scarce resources that doesn’t actually save any lives, perhaps with only some psychological benefit.  Am I missing something?

IIRC ~50% or so of patients on ventilators recover. In that case, it would be enough to make a significant difference (as opposed to only saving 5%). But I don't remember which study I read that in or when, and it is possible I might misremember.

[Southern Senator North Carolina Yankee]

It is true that the longer someone is on a ventilator, their chances of recovery dwindle over time. That being said, the idea that should just terminate people who need a ventilator is a mistake. A lot of people have been saved from pneumonia in the current age because we have ventilators whereas previously, it was almost a death sentence.

[Southern Senator North Carolina Yankee]

We are going into a recession and car sales are plummeting. Getting DPAed is probably the best thing that could have happened for GM, because now they can at least make something and have a sure buyer for it. Even if they end up not being needed, then we will have them on hand.

[GoTfan]

Lower population density is all well and good, as is basing this model on trends we have seen elsewhere, but I guess I am having a hard time reconciling this model with logic. Unless every last stitch of this virus is cut off from the potential to spread and then dies out, it is going to linger. 8 million infections by the end of that curve in the model is nowhere near enough to achieve herd immunity. So isn't it, like... extremely likely it will return?

And when it does, and we start to open things back up, are we really going to be able to do enough tracking, contact tracing, and quarantining of resurgent cases/exposures to put a lid on things before  we get another curve? I'm very pessimistic, sorry.

Currently, it seems to be barely outpacing their pessimistic projections.

[T'Chenka]

Forgive me for circling back, but I'm a bit confused about the projection model on this site posted earlier: https://covid19.healthdata.org/projections.

If my math is right, with total American deaths coming in at around 81,000, a death rate of 1% means we will only see ~8 million Americans get infected with this virus. How is that possible in a country as populous as the United States? Either the death rate is wrong and there is a ridiculous proportion of asymptomatic cases, or the virus will come back with just as much force as before... right?

You're corrrect.

Any model showing less than 250,000 deaths need to explain why the following isn't true:

- AT LEAST 0.5% of infected Americans will die (probably more)
- AT LEAST 25% of all Americans will eventually get COVID-19 (probably more)
- If herd immunity doesn't materialize, simply containing the virus is a stopgap

All three are true... no?

Punch in the US population on a calculator, multiplied by 0.25 infected, multiplied by 0.005 death rate, and keep in mind I'm using conservative numbers here. I'd be glad to be wrong about this, but nobody has shown me why I'm wrong as of yet. I'm open-minded to changing my mind here.