Coronavirus pandemic

[kphoger]

Does anyone here know what caused the following:

Alaska – After having about three weeks of really low numbers in May, subsequently bouncing back up ever since then?

Hawaii – After having about four weeks of really low numbers in May, subsequently bouncing back up ever since then?

Montana – After having about four weeks of really low numbers in April-May, subsequently bouncing back up ever since then?

I don't know definitively, but the rebound in cases seen in these states may be attributable to one or more infected individuals traveling to the state from elsewhere and seeding a fresh outbreak.

Do you have any evidence that's what happened?  I wasn't really interested in purely hypothetical reasons.

[kphoger]

tradephoric:

Shown below is a simple illustration of how increased testing with level positivity rates doesn't mean the virus isn't spreading.  In the illustration, testing increased tenfold, but the infected population size still grew twofold.



But, as has been pointed out, more tests become available all the time and, with more tests available, a larger percentage of them are given to asymptomatic people.  Because symptomatic people are more likely to be infected than asymptomatic people, their presence in the sample size skews the results.

[NWI_Irish96]

Focus on hospitalization rates. That eliminates any bias from the differing volumes of testing over time. Regardless of testing volumes and numbers of positive tests and positivity rates, people are going to be hospitalized when they reach a certain level of illness.

[tradephoric]

tradephoric:

Shown below is a simple illustration of how increased testing with level positivity rates doesn't mean the virus isn't spreading.  In the illustration, testing increased tenfold, but the infected population size still grew twofold.

I don't understand why the total infected population just arbitrarily increased from 100k to 200k in this chart.  How did you come up with this doubling? 

[kphoger]

tradephoric:

Shown below is a simple illustration of how increased testing with level positivity rates doesn't mean the virus isn't spreading.  In the illustration, testing increased tenfold, but the infected population size still grew twofold.

I don't understand why the total infected population just arbitrarily increased from 100k to 200k in this chart.  How did you come up with this doubling? 

It was a purely hypothetical illustration, to show that actual infections can indeed be going up even as testing increases and positivity rates remain steady.

[tradephoric]

Focus on hospitalization rates. That eliminates any bias from the differing volumes of testing over time. Regardless of testing volumes and numbers of positive tests and positivity rates, people are going to be hospitalized when they reach a certain level of illness.

I agree just tell that to USA Today.  They published an article on July 15th with the headline "Spring's coronavirus surge was bad, but 32 states are doing much worse now".  The article details how 32 states are adding coronavirus cases at a rate far worse than the Spring.  But at no point in the article do they mention how much additional testing has taken place since the Spring. 

They could have just as easily wrote an article entitled "Spring's testing capacity was bad, but 50 states are seeing testing at a rate far higher than the Spring".   Of course an article focusing entirely on testing data would tell you nothing about cases, but then again an article focusing entirely on daily cases tells you nothing about testing.

Spring's coronavirus surge was bad, but 32 states are doing much worse now
https://www.usatoday.com/story/news/nation/2020/07/15/florida-texas-not-only-states-covid-spike-32-pass-spring-record/5442350002/

[NWI_Irish96]

Focus on hospitalization rates. That eliminates any bias from the differing volumes of testing over time. Regardless of testing volumes and numbers of positive tests and positivity rates, people are going to be hospitalized when they reach a certain level of illness.

I agree just tell that to USA Today.  They published an article on July 15th with the headline "Spring's coronavirus surge was bad, but 32 states are doing much worse now".  The article details how 32 states are adding coronavirus cases at a rate far worse than the Spring.  But at no point in the article do they mention how much additional testing has taken place since the Spring. 

They could have just as easily wrote an article entitled "Spring's testing capacity was bad, but 50 states are seeing testing at a rate far higher than the Spring".   Of course an article focusing entirely on testing data would tell you nothing about cases, but then again an article focusing entirely on daily cases tells you nothing about testing.

Spring's coronavirus surge was bad, but 32 states are doing much worse now
https://www.usatoday.com/story/news/nation/2020/07/15/florida-texas-not-only-states-covid-spike-32-pass-spring-record/5442350002/

That article does (briefly) mention hospitalization rates, and even though not mentioned much, hospitalization rates are skyrocketing in most of those 32 states, so the increase in cases is most certainly not just because of increased testing.

[kphoger]

Focus on hospitalization rates. That eliminates any bias from the differing volumes of testing over time. Regardless of testing volumes and numbers of positive tests and positivity rates, people are going to be hospitalized when they reach a certain level of illness.

I agree just tell that to USA Today.  They published an article on July 15th with the headline "Spring's coronavirus surge was bad, but 32 states are doing much worse now".  The article details how 32 states are adding coronavirus cases at a rate far worse than the Spring.  But at no point in the article do they mention how much additional testing has taken place since the Spring. 

They could have just as easily wrote an article entitled "Spring's testing capacity was bad, but 50 states are seeing testing at a rate far higher than the Spring".   Of course an article focusing entirely on testing data would tell you nothing about cases, but then again an article focusing entirely on daily cases tells you nothing about testing.

Spring's coronavirus surge was bad, but 32 states are doing much worse now
https://www.usatoday.com/story/news/nation/2020/07/15/florida-texas-not-only-states-covid-spike-32-pass-spring-record/5442350002/

Huh?  You agree that it's better to focus on hospitalizations, but your proposed fix to USA Today's reporting has nothing to do with hospitalizations?

[SEWIGuy]

Focus on hospitalization rates. That eliminates any bias from the differing volumes of testing over time. Regardless of testing volumes and numbers of positive tests and positivity rates, people are going to be hospitalized when they reach a certain level of illness.

I agree just tell that to USA Today.  They published an article on July 15th with the headline "Spring's coronavirus surge was bad, but 32 states are doing much worse now".  The article details how 32 states are adding coronavirus cases at a rate far worse than the Spring.  But at no point in the article do they mention how much additional testing has taken place since the Spring. 

They could have just as easily wrote an article entitled "Spring's testing capacity was bad, but 50 states are seeing testing at a rate far higher than the Spring".   Of course an article focusing entirely on testing data would tell you nothing about cases, but then again an article focusing entirely on daily cases tells you nothing about testing.

Spring's coronavirus surge was bad, but 32 states are doing much worse now
https://www.usatoday.com/story/news/nation/2020/07/15/florida-texas-not-only-states-covid-spike-32-pass-spring-record/5442350002/

Do you simply ignore arguments that point out how bad you are misreading these stats and what they imply?

[paulthemapguy]

To be fair, that statement isn't actually true.  Positive tests are only a subset of total tests, and therefore they are not an absolute measure of how widespread the virus is.

This made absolutely no sense.
In addition, everything tradephoric says makes absolutely no sense.

[kphoger]

Do you simply ignore arguments ... ?

Wait, are tradephoric and Fritzowl really the same guy, and he's been trolling us all along?

[webny99]

Not to mention that (a) there were about 2.9 million deaths in the US in 2019 (more than 20x the current COVID death toll), (b) many of those who have died were retired or not in the workforce to begin with, and (c) may have died anyways even without the pandemic.

So what? That's still 137,419 people that, this time last year, were buying food, gas, and clothing. Now they are spending nothing. Assuming they spent only $4 per meal, that's 137,419× 3 × $4 = $1,649,028 per day or $601,895,220 per year that is no longer being spent by these people because they are no longer in the market for food.

So, I don't think $601 million - or even triple or quadruple that - is very significant in a $20.5 trillion economy. Leaving aside the validity of the figures for a minute, roughly 20 times that amount of people (and thus, money) comes and goes due to the natural life and death cycle every single year.

I don't want to de-emphasize the death toll, because it is hugely problematic and we have not done well as a country in handling this virus, but the lost consumers, on their own, are a drop in the bucket from an economic perspective; obviously, it's a much different story from a social and public health perspective.

Clearly it's a better bet to go through short-term economic pain and institute sensible policies to keep people alive, so that when all is safe again, as many consumers are around to rejoin the economy as possible.

No disagreement on that point. Even though our response has been imperfect on many levels, I think the death toll would have been many, many times higher if there was no public response at all and everyone continued as normal.

[tradephoric]

tradephoric:

Shown below is a simple illustration of how increased testing with level positivity rates doesn't mean the virus isn't spreading.  In the illustration, testing increased tenfold, but the infected population size still grew twofold.

I don't understand why the total infected population just arbitrarily increased from 100k to 200k in this chart.  How did you come up with this doubling? 

It was a purely hypothetical illustration, to show that actual infections can indeed be going up even as testing increases and positivity rates remain steady.

I'm not trying to be critical but from what i see the only reason the actual infections went up from 100k to 200k in your chart is because you arbitrarily inputted them.  If a region has 1.25 million people and a positivity rate of 8%, the actual total infected population is going to equal 100k whether they are testing 10k people a day or 100k people a day.  Why did it jump to 200k in that chart you set up?

[kphoger]

tradephoric:

Shown below is a simple illustration of how increased testing with level positivity rates doesn't mean the virus isn't spreading.  In the illustration, testing increased tenfold, but the infected population size still grew twofold.

I don't understand why the total infected population just arbitrarily increased from 100k to 200k in this chart.  How did you come up with this doubling? 

It was a purely hypothetical illustration, to show that actual infections can indeed be going up even as testing increases and positivity rates remain steady.

I'm not trying to be critical but from what i see the only reason the actual infections went up from 100k to 200k in your chart is because you arbitrarily inputted them.  If a region has 1.25 million people and a positivity rate of 8%, the actual total infected population is going to equal 100k whether they are testing 10k people a day or 100k people a day.  Why did it jump to 200k in that chart you set up?

No.  If a region has 1.25 million people and a positivity rate of 8%, then you still have no idea what the total infected population is.  Positivity rate is how many people test positive out of the total tested population, not out of the total population.

You're right that the only reason total infected population increased from 100k to 200k in my illustration is that I arbitrarily did that.  But you, by the same token, are being just as arbitrary in claiming that total infections are going this way or that based on the same metrics.

[tradephoric]

Focus on hospitalization rates. That eliminates any bias from the differing volumes of testing over time. Regardless of testing volumes and numbers of positive tests and positivity rates, people are going to be hospitalized when they reach a certain level of illness.

I agree just tell that to USA Today.  They published an article on July 15th with the headline "Spring's coronavirus surge was bad, but 32 states are doing much worse now".  The article details how 32 states are adding coronavirus cases at a rate far worse than the Spring.  But at no point in the article do they mention how much additional testing has taken place since the Spring. 

They could have just as easily wrote an article entitled "Spring's testing capacity was bad, but 50 states are seeing testing at a rate far higher than the Spring".   Of course an article focusing entirely on testing data would tell you nothing about cases, but then again an article focusing entirely on daily cases tells you nothing about testing.

Spring's coronavirus surge was bad, but 32 states are doing much worse now
https://www.usatoday.com/story/news/nation/2020/07/15/florida-texas-not-only-states-covid-spike-32-pass-spring-record/5442350002/

Huh?  You agree that it's better to focus on hospitalizations, but your proposed fix to USA Today's reporting has nothing to do with hospitalizations?

My point wasn't THAT difficult to comprehend.  I agree with cabiness42 that looking at hospitalization rates is a better metric to use than daily case counts as it eliminates any bias from the differing volumes of testing over time.  But if USA Today is going to write an article focused on case counts, they really should provide that data with context in regards to testing.  It would be like writing an article arguing that automobiles were safer 100 years ago because motor vehicle fatalities were a fraction of what they are today (never mind that fatalities per 100 million VMT have dropped by 95% over the past 100 years).

[NWI_Irish96]

I agree with cabiness42 that looking at hospitalization rates is a better metric to use than daily case counts as it eliminates any bias from the differing volumes of testing over time.

and hospitalization rates are up sharply over the past couple months, therefore infection rates are up sharply over the past couple months regardless of how much testing is being done.

[hotdogPi]

Do you simply ignore arguments ... ?

Wait, are tradephoric and Fritzowl really the same guy, and he's been trolling us all along?

tradephoric tries to fight back criticism. FritzOwl ignores it entirely; I have never seen a single "you're wrong" statement from him.

I also think they live in different parts of the country.

Last, why would FritzOwl register before tradephoric if they're the same person?

I've done some research, and unless FritzOwl doesn't actually live in California, he's not an alternate account of anyone.

[tradephoric]

I agree with cabiness42 that looking at hospitalization rates is a better metric to use than daily case counts as it eliminates any bias from the differing volumes of testing over time.

and hospitalization rates are up sharply over the past couple months, therefore infection rates are up sharply over the past couple months regardless of how much testing is being done.

Nationally, the percentage positive rate began rising on June 16th.   I don't dispute that.  Still that USA Today article focuses on daily cases to write a narrative that the virus is more widespread today than it was in the Spring.  That's just not reality when looking at either the case positivity rate or the hospitalization rate.

[kphoger]

Nationally, the percentage positive rate began rising on June 16th.   I don't dispute that.  Still that USA Today article focuses on daily cases to write a narrative that the virus is more widespread today than it was in the Spring.  That's just not reality when looking at either the case positivity rate or the hospitalization rate.

If the virus is not more widespread, then how do you explain the increase in hospitalizations?

[NWI_Irish96]

Nationally, the percentage positive rate began rising on June 16th.   I don't dispute that.  Still that USA Today article focuses on daily cases to write a narrative that the virus is more widespread today than it was in the Spring.  That's just not reality when looking at either the case positivity rate or the hospitalization rate.

If the virus is not more widespread, then how do you explain the increase in hospitalizations?

Sounds like a bit of splitting hairs. Hospitalizations, at least nationally, are still below the highest rates from the spring, but have also been increasing sharply over the past month.

Some individual states do have higher hospitalizations than the spring, but nationally we aren't there . . . yet.

[tradephoric]

To be clear I've been focusing on the USA Today article that compares the daily cases today to the daily cases in the Spring.  Using that time-frame, the hospitalization rate has gone down roughly 10% since its peak in April (even after a big spike over the past month).  It's just misleading for the media to report daily cases without putting context to how much additional testing has taken place.  These selective news articles cause the public to believe things are 10x worse than reality.   

[J N Winkler]

Not to mention that (a) there were about 2.9 million deaths in the US in 2019 (more than 20x the current COVID death toll), (b) many of those who have died were retired or not in the workforce to begin with, and (c) may have died anyways even without the pandemic.

So what? That's still 137,419 people that, this time last year, were buying food, gas, and clothing. Now they are spending nothing. Assuming they spent only $4 per meal, that's 137,419× 3 × $4 = $1,649,028 per day or $601,895,220 per year that is no longer being spent by these people because they are no longer in the market for food.

So, I don't think $601 million - or even triple or quadruple that - is very significant in a $20.5 trillion economy. Leaving aside the validity of the figures for a minute, roughly 20 times that amount of people (and thus, money) comes and goes due to the natural life and death cycle every single year.

I don't want to de-emphasize the death toll, because it is hugely problematic and we have not done well as a country in handling this virus, but the lost consumers, on their own, are a drop in the bucket from an economic perspective; obviously, it's a much different story from a social and public health perspective.

What actually drives the implosion in demand is not people dying and thus no longer spending their discretionary income on restaurants etc., but rather healthy people (the vast majority of whom, based on our experience of 20th-century pandemics like the 1918 flu, will not contract the illness) staying at home and not spending money because they do not wish to court exposure to the virus without good reason.  It isn't really possible for the restaurant sector as a whole to respond to this risk aversion by going from dine-in to takeout--there is inevitably a large population of restaurants that depend on dine-in.

Economists' term for the kind of estimation USDOT does to come up with the value of a human life is called hedonic pricing.  It is tricky to do because it is based on willingness to pay, which varies by context.  It is also not, in any absolute sense, the value of a human life, or necessarily even correlated to disability-adjusted life years, which is what you are talking about when prioritizing kids higher than grandpas for rescue operations.  Its main use is as a number to rank different options for spending to save lives.

[NWI_Irish96]

To be clear I've been focusing on the USA Today article that compares the daily cases today to the daily cases in the Spring.  Using that time-frame, the hospitalization rate has gone down roughly 10% since its peak in April (even after a big spike over the past month).  It's just misleading for the media to report daily cases without putting context to how much additional testing has taken place.  These selective news articles cause the public to believe things are 10x worse than reality.   

If you look at the numbers of cases and hospitalizations, there is a strong (near 100%) correlation to when those graphs turn down and then back up. It appears that you are focused in on the fact that the number of cases has shot well past the previous peak and the number of hospitalizations has yet to do so, and concluded that testing is the primary reason why.

The fact that hospitalizations have gone back up so steeply indicate that while the magnitude of the increase in cases can be attributed to testing, the underlying hospitalization rates indicate that the raw number of cases is not nearly as horrible a proxy for the real seriousness of the current situation as you are claiming.

[jemacedo9]

To be clear I've been focusing on the USA Today article that compares the daily cases today to the daily cases in the Spring.  Using that time-frame, the hospitalization rate has gone down roughly 10% since its peak in April (even after a big spike over the past month).  It's just misleading for the media to report daily cases without putting context to how much additional testing has taken place.  These selective news articles cause the public to believe things are 10x worse than reality.   

Yes...hospitalizations are down compared to April.  Tradephoric is right.  FOR NOW.

But just wait a week or two or so.  The current trajectory is that that statement above will no longer be true.
The question is...why is the trend of hospitalizations going up?  And then, how do you stop it, if you can?

And...you can't take a long-term downward trend as meaning anything if there are erratic ups and downs and the latest shorter term trend is oppositely up.

So stop cherry-picking data to fit your narrative.

[US71]

Arkansas has just mandated masks beginning Monday (7/20)