News Article: Good luck dealing with a pandemic, depending on this administration...

[gtowntide]

Steve Mnuchin Thinks $1200 Will Last Families Up To 10 Weeks

I wonder if he could live on 1200 for ten weeks? How out of touch can these people be?

 

[DzynKingRTR]

Steve Mnuchin Thinks $1200 Will Last Families Up To 10 Weeks

Not exactly what he said, but ok.

“I think the entire package provides economic relief overall for about 10 weeks. Hopefully we will kill this virus quicker and we won’t need it, but we have liquidity to put into the American economy to support American workers and American business.”

I can see the small business loans helping for companies for 10 weeks. As a matter of fact that is exactly what my boss said (boss actually said 2 months)

 

[MattinBama]

I wonder if he could live on 1200 for ten weeks? How out of touch can these people be?

 

[PacadermaTideUs]

The trend on new daily deaths - 2,407 today. Not seeing a plateau yet

View attachment 7438

A few pages back, I posted three weeks of 7-day running average confirmed case growth factors.

News Article: - Good luck dealing with a pandemic, depending on this administration...

They eat some seriously weird stuff there. One Chinese guy said "If it's live and not human, we'll eat it"...

www.tidefans.com

In summary, it showed that spread of covid-19 is no longer accelerating since approximately April 10th, and has been essentially slowing since April 11th. Though the virus continues to spread, each day’s spread is less than the previous day’s. In other words, we’ve passed the peak of the bell curve.

Since most people who eventually succumb to the disease were diagnosed with it some time beforehand, it stands to reason that death numbers will lag some length of time behind case numbers.

I originally assumed that the lag would be 10 to 15 days since I believe that previous studies have shown an approximate 10 to 15 day average time between diagnosis and death among those who’ve died. But looking at the 7-day running average death growth factor, it looks more like a 7-day lag.

The last 3.5 weeks for deaths:

March 22: 1.46
March 23: 1.42
March 24: 1.47
March 25: 1.37
March 26: 1.33
March 27: 1.42
March 28: 1.47
March 29: 1.22
March 30: 1.26
March 31: 1.26
April 1: 1.27
April 2: 1.25
April 3: 1.19
April 4: 1.18
April 5: 1.21
April 6: 1.15
April 7: 1.14
April 8: 1.11
April 9: 1.12
April 10: 1.12
April 11: 1.07
April 12: 1.06
April 13: 1.05
April 14: 1.05

Since confirmed case count plateaued around April 10th or 11th, and deaths appear to lag behind diagnosis by approximately 1 week, I would expect daily death counts to plateau around April 17th or 18th, with growth factors around 1.0.

That deaths vs recovered stat is ominous. I think recovered means those that got out of the hospital alive.

That’s a comparison that I’ve been watching since the outbreak began.

There is a mistaken tendency for people to try to key in on deaths divided by total cases to calculate CFR. This presently yields a CFR of 4.4%.

That’s a valid method after a pandemic has run its course. But in an ongoing pandemic, it may be wildly inaccurate (likely low) since the vast majority of cases are unresolved. We don’t yet know how many of these unresolved cases are going to result in deaths.

A different methodology in an ongoing pandemic would be deaths divided by all known resolved cases (resolved = deaths + recoveries), which presently yields a CFR for the USA of 37%.

That, as well, is likely to be wildly inaccurate (high this time) due to the fact that we don’t truly know how many recoveries there have been since we don’t truly know how many cases there have been.

A German study has estimated that only 1.6% of American cases have been discovered. If that’s true, 98.4% of covid-19 cases (presently around 35 million unresolved cases) are not known about, likely because symptoms are mild or nonexistent. It is likely that a large majority of those cases will result in recoveries.

Bottom line: recovery vs death is important to watch because it may shed light on CFR among those cases which cause serious enough symptoms to earn test-confirmation and therefore diagnosis. But it’s going to badly miss the mark on true CFR. We’re not really going to have an idea of that until well after the fact, maybe never. There are still widely varying CFR estimates for the 1918 Spanish Flu outbreak.

 

[B1GTide]

A few pages back, I posted three weeks of 7-day running average confirmed case growth factors. Not going to repost the whole list here, but in summary, it showed that spread of covid-19 is no longer accelerating since approximately April 10th, and has been essentially slowing since April 11th. Though the virus continues to spread, each day’s spread is less than the previous day’s. In other words, we’ve passed the peak of the bell curve.

Since most people who eventually succumb to the disease were diagnosed with it some time beforehand, it stands to reason that death numbers will lag some length of time behind case numbers.

I originally assumed that the lag would be 10 to 15 days since I believe that previous studies have shown an approximate 10 to 15 day average time between diagnosis and death among those who’ve died. But looking at the 7-day running average death growth factor, it looks more like a 7-day lag.

The last 3.5 weeks for deaths:

March 22: 1.46
March 23: 1.42
March 24: 1.47
March 25: 1.37
March 26: 1.33
March 27: 1.42
March 28: 1.47
March 29: 1.22
March 30: 1.26
March 31: 1.26
April 1: 1.27
April 2: 1.25
April 3: 1.19
April 4: 1.18
April 5: 1.21
April 6: 1.15
April 7: 1.14
April 8: 1.11
April 9: 1.12
April 10: 1.12
April 11: 1.07
April 12: 1.06
April 13: 1.05
April 14: 1.05

Since confirmed case count plateaued around April 10th or 11th, and deaths appear to lag behind diagnosis by approximately 1 week, I would expect daily death counts to plateau around April 17th or 18th, with growth factors around 1.0.



That’s a comparison that I’ve been watching since the outbreak began.

There is a mistaken tendency for people to try to key in on deaths divided by total cases to calculate CFR. This presently yields a CFR of 4.4%.

That’s a valid method after a pandemic has run its course. But in an ongoing pandemic, it may be wildly inaccurate (likely low) since the vast majority of cases are unresolved. We don’t yet know how many of these unresolved cases are going to result in deaths.

A different methodology in an ongoing pandemic would be deaths divided by all known resolved cases (resolved = deaths + recoveries), which presently yields a CFR for the USA of 37%.

That, as well, is likely to be wildly inaccurate (high this time) due to the fact that we don’t truly know how many recoveries there have been since we don’t truly know how many cases there have been.

A German study has estimated that only 1.6% of American cases have been discovered. If that’s true, 98.4% of covid-19 cases (presently around 35 million unresolved cases) are not known about, likely because symptoms are mild or nonexistent. It is likely that a large majority of those cases will result in recoveries.

Bottom line: recovery vs death is important to watch because it may shed light on CFR among those cases which cause serious enough symptoms to earn test-confirmation and therefore diagnosis. But it’s going to badly miss the mark on true CFR. We’re not really going to have an idea of that until well after the fact, maybe never. There are still widely varying CFR estimates for the 1918 Spanish Flu outbreak.

Your number for yesterday has to be incorrect. It was, far and away, the highest death total day and today may be even higher.

 

[TexasBama]

A few pages back, I posted three weeks of 7-day running average confirmed case growth factors. Not going to repost the whole list here, but in summary, it showed that spread of covid-19 is no longer accelerating since approximately April 10th, and has been essentially slowing since April 11th. Though the virus continues to spread, each day’s spread is less than the previous day’s. In other words, we’ve passed the peak of the bell curve.

Since most people who eventually succumb to the disease were diagnosed with it some time beforehand, it stands to reason that death numbers will lag some length of time behind case numbers.

I originally assumed that the lag would be 10 to 15 days since I believe that previous studies have shown an approximate 10 to 15 day average time between diagnosis and death among those who’ve died. But looking at the 7-day running average death growth factor, it looks more like a 7-day lag.

The last 3.5 weeks for deaths:

March 22: 1.46
March 23: 1.42
March 24: 1.47
March 25: 1.37
March 26: 1.33
March 27: 1.42
March 28: 1.47
March 29: 1.22
March 30: 1.26
March 31: 1.26
April 1: 1.27
April 2: 1.25
April 3: 1.19
April 4: 1.18
April 5: 1.21
April 6: 1.15
April 7: 1.14
April 8: 1.11
April 9: 1.12
April 10: 1.12
April 11: 1.07
April 12: 1.06
April 13: 1.05
April 14: 1.05

Since confirmed case count plateaued around April 10th or 11th, and deaths appear to lag behind diagnosis by approximately 1 week, I would expect daily death counts to plateau around April 17th or 18th, with growth factors around 1.0.



That’s a comparison that I’ve been watching since the outbreak began.

There is a mistaken tendency for people to try to key in on deaths divided by total cases to calculate CFR. This presently yields a CFR of 4.4%.

That’s a valid method after a pandemic has run its course. But in an ongoing pandemic, it may be wildly inaccurate (likely low) since the vast majority of cases are unresolved. We don’t yet know how many of these unresolved cases are going to result in deaths.

A different methodology in an ongoing pandemic would be deaths divided by all known resolved cases (resolved = deaths + recoveries), which presently yields a CFR for the USA of 37%.

That, as well, is likely to be wildly inaccurate (high this time) due to the fact that we don’t truly know how many recoveries there have been since we don’t truly know how many cases there have been.

A German study has estimated that only 1.6% of American cases have been discovered. If that’s true, 98.4% of covid-19 cases (presently around 35 million unresolved cases) are not known about, likely because symptoms are mild or nonexistent. It is likely that a large majority of those cases will result in recoveries.

Bottom line: recovery vs death is important to watch because it may shed light on CFR among those cases which cause serious enough symptoms to earn test-confirmation and therefore diagnosis. But it’s going to badly miss the mark on true CFR. We’re not really going to have an idea of that until well after the fact, maybe never. There are still widely varying CFR estimates for the 1918 Spanish Flu outbreak.

I had stopped looking at cases mainly because of the errors/assumptions in the number. Change in deaths is about as error free as is readily available. I think this curve will look like a left-skewed curve rather than a standard bell, jmo. Given that, it may be near mid-May before we get back to "only" 500 deaths per day, assuming no major decrease in the day-day cases.

 

[DzynKingRTR]

Your number for yesterday has to be incorrect. It was, far and away, the highest death total day and today may be even higher.

Yesterday was the highest total death day yet. I would imagine the deaths are going to improve too much. The real number we should be paying attention to is the "new cases". Provided those numbers are even accurate.

 

[IM4UA]

Yesterday was the highest total death day yet. I would imagine the deaths are going to improve too much. The real number we should be paying attention to is the "new cases". Provided those numbers are even accurate.

My assumption is the level of testing is still far too limited and inconsistent to provide any meaningful data. Unfortunately, tracking the number of fatalities is the only relatively accurate gauge at this point.

 

[B1GTide]

Yesterday was the highest total death day yet. I would imagine the deaths are going to improve too much. The real number we should be paying attention to is the "new cases". Provided those numbers are even accurate.

Yeah, I only look at deaths because we are not testing so the number of new cases reported is garbage. Heck, even the number of deaths is very low.

 

[PacadermaTideUs]

Your number for yesterday has to be incorrect. It was, far and away, the highest death total day and today may be even higher.

The numbers I posted are 7-day running averages of growth factor (GF). GF is calculated by dividing the day’s total by the previous day’s total. For example, yesterday’s total deaths (2407) divided by Monday’s total deaths (1535) = 1.57. If GF > 1.0, it indicates that daily numbers are growing. If GF < 1.0, it indicates that daily numbers are shrinking. It’s an important stat to watch because it shows you where you are on the bell curve.

Due to wide day-to-day variability, and noise caused by reporting tendencies (weekends, etc), it can be helpful in seeing trends to smooth out the data. Hence the 7-day running average.

The number listed for yesterday is the 7-day average GF between April 8th (0.98), April 9th (0.98), April 10th (1.07), April 11th (0.90), April 12th (0.83), April 13th (1.00), and April 14th (1.57). Average = 1.05.

Hope that clears it up.

 

[MobtownK]

I had stopped looking at cases mainly because of the errors/assumptions in the number. Change in deaths is about as error free as is readily available. I think this curve will look like a left-skewed curve rather than a standard bell, jmo. Given that, it may be near mid-May before we get back to "only" 500 deaths per day, assuming no major decrease in the day-day cases.

I have been looking for a model that shows what happens when we open back up in May. I can't find one anywhere.

Reason being, if it is estimated that only a % of the population has been exposed, we re-open, and more people are exposed... Seems like the curve would then go up.
It's been a looong time since I took statistics, but to me this is the logical outcome. I am not factoring in hospital capacity.
This is just a rough sketch of what it could look like.
What do y'all think?

 

[B1GTide]

The numbers I posted are 7-day running averages of growth factor (GF). GF is calculated by dividing the day’s total by the previous day’s total. For example, yesterday’s total deaths (2407) divided by Monday’s total deaths (1535) = 1.57. If GF > 1.0, it indicates that daily numbers are growing. If GF < 1.0, it indicates that daily numbers are shrinking. It’s an important stat to watch because it shows you where you are on the bell curve.

Due to wide day-to-day variability, and noise caused by reporting tendencies (weekends, etc), it can be helpful in seeing trends to smooth out the data. Hence the 7-day running average.

The number listed for yesterday is the 7-day average GF between April 8th (0.98), April 9th (0.98), April 10th (1.07), April 11th (0.90), April 12th (0.83), April 13th (1.00), and April 14th (1.57). Average = 1.05.

Hope that clears it up.

It does, but it also demonstrates how flawed it is. I disagree with your methodology.

 

[B1GTide]

 

[PacadermaTideUs]

It does, but it also demonstrates how flawed it is. I disagree with your methodology.

Ok. Doesn’t really affect me one way or the other. There are certainly flaws with just about every methodology. Care to expound though so that I might learn something from your post?

 

[B1GTide]

Ok. Doesn’t really affect me one way or the other. There are certainly flaws with just about every methodology. Care to expound though so that I might learn something from your post?

The number of deaths is rising. That is a known. That is the only known. Nothing is flattening until that number flattens.

 

[PacadermaTideUs]

View attachment 7467

So looking at this graph, do you have any idea where we are on the bell curve?

 

[MobtownK]

So looking at this graph, do you have any idea where we are on the bell curve?

I would assume it would depend on where in the US you live. AL only shut down recently, so the asymptomatic period could just now be coming to haunt us. Washington may be on the downslope.
Our population is so varied, it might be better to look at regionally.

 

[B1GTide]

So looking at this graph, do you have any idea where we are on the bell curve?

Look, we have no idea what kind of bell curve applies here. We are guessing. And the numbers in NJ are still going up, up, up. We are hopeful that they will level out and start to drop soon. But it is only hope.

 

[DzynKingRTR]

The number of deaths is rising. That is a known. That is the only known. Nothing is flattening until that number flattens.

Or maybe some of our fellow Tidefans that actually work in hospitals could give us better numbers. I would be more inclined to believe them than any numbers put out.

 

[PacadermaTideUs]

Look, we have no idea what kind of bell curve applies here. We are guessing. And the numbers in NJ are still going up, up, up. We are hopeful that they will level out and start to drop soon. But it is only hope.

No methodology is an unerring crystal ball. But there are tools that can help us see trends, which are indicators. Nothing more. Growth Factor is a staple in the epidemiology tool belt. And in this particular case, using this particular tool, a clear trend is evident.

Nothing can tell us if the trend will continue. There are too many changing factors that can greatly affect it (loosening of distancing, travel restrictions, etc). But the GF trend, which I believe is a positive and direct result of distancing and state/local shutdowns, does continue, barring any new explosive hot spot a la New York, I believe it is very possible that we as a nation will reach the peak of the bell curve in the very near future. That’s not based on hope. That’s based on statistical tools.

I’ve not looked at any particular regional trends (New Jersey, Alabama, etc), nor made any predictions regarding them.

But to each, his or her own. I’m not selling anything.