Coronavirus 2019-nCoV

[Aargh]

The calculated mortality rate (deaths divided by positive tests) is still rising. The only way this can happen is if the rate of infection is increasing faster than the rate of testing.

We're around 5.9% calculated mortality today, which is thought to be much higher than the actual mortality rate.

The only way that number can get bigger is if tomorrow we have a higher number of deaths compared to the number of mew confirmed cases than we had today. That means the number of people we are testing each day is smaller than the number of new infections every day. That also means that every day we have more people capable of spreading the infection than we had that we knew of the day before.

The rate of new infections is going down. The rate of potential new infections seems to be going up. Masks and mitigation have been effective at slowing the infection rate. Testing only works if the positives are quarantined and their contacts are tested. We don't seem to be close to using that method to reduce the infection rate. 32 states are testing reduced mitigation. That should give us some usable data in 4 to 6 weeks.

Texas had their highest number of C-19 deaths the day before they took an aggressive approach to reopening their state. I wouldn't be a bit surprised if their aggressive, but risky, approach worked out well. Sometimes drawing to a 16 in blackjack works out well.

[wufan]

I'd definitely be interested in knowing flu deaths, heart attack deaths, and stroke deaths to date and compared to average.

Flu deaths are way down. Not sure on heart attack or stroke. Overall deaths are up.

[Kung Wu]

Sorry for the delayed response. One side affect of COVID-19 is I am stuck at home with my kid and spent the entire freaking day helping him with a presentation on Adolf Hitler. Which was fun, but I got exactly ZERO work done yesterday. Grrr

---

Kansas overall is doing quite well.

The problem with your analysis is that all positives are not created equal, due to a change in methodology of collecting the data.

Ford, Seward, and Finney have severely over-sampled ASYMPTOMATIC cases (which skews the other county's data horribly if you don't take it into account).

Higher negatives can be drive your variability (which is good), or higher positive rates can drive your variability (which is bad)

All people fit into one of 3 classes of people:

1) No coronavirus
2) Coronavirus - Asymptomatic
3) Coronavirus - Sick

Higher negatives only tell you that you have more of 1).
But higher positives, alone, cannot distinguish between 2) and 3).

Kansas has 105 counties. In 102 of them the way tests are distributed is:

A) Person walks into a hospital with a concern. Doctor screens them and decides whether they need a test. Many get tested. If positive, the bulk of them are in category 3), because they have shown symptoms.

However in 3 counties:

B) Government randomly sampled a bunch of working age folks.

The result of B is a very high percent of people with positives that weren't sick and that would NOT have been counted in the hospital walk-in approach above.

Why does it matter?

In week 4 and week 5 the government did random sampling in Seward, Ford, and Finney and the case loads were overwhelmed dramatically with asymptomatic/mild cases (which are GOOD), that wouldn't normally be counted.

Here is the average of for the % positives

Week 1: 10.7% tested positive
Week 2: 13.3%
Week 3: 9.0%
Week 4: 16.5%

Week 3 is good trend, but week 4 is not the trend you want. You want your cases % to drop even as you ramp up testing.

The 16.5% is combining those the results of the three over-sampled counties.

If you subtract the new burst of asymptomatic cases from the case load and look at only 102 counties for an apples-to-apples comparison, the picture looks like:

Week 1: 10.7% tested positive
Week 2: 13.3%
Week 3: 9.0%
Week 4 + 5: 8.3%

PLUS

1,435 asymptomatic mild cases that would not have appeared if there had been no special study conducted by the state of Kansas and the federal government in an effort to survey the meat industry.

Ergo, KANSAS overall is doing quite well.

But how do we know we over-sampled asymptomatic mild cases severely?

This is evidenced by the hospitalization rates in those counties.

What has happened in the last week? Ford and Seward counties - 1175 positives, 1447 negatives (45% positive rate).

Proof that nearly (but not quite) ALL of those positives are ASYMPTOMATIC cases?

Yesterday in Finney County there is a 55:1 ratio between positive cases and hospitalizations.
Seward County? 51.4:1
Ford County? 100.3:1

In Sedgwick County, there is an 5.6:1 ratio between positive cases and hospitalizations.
Lyons? 4:1
Wyandotte? 4.4:1
Johnson? 3.4:1
Butler? 5.3:1
Shawnee? 5.8:1
... and so on ...

The only explanation for that discrepancy is that random sampling detects a much larger number of asymptomatic cases, while hospital driven on-demand testing skews toward symptomatic positive cases.

Early reports told us that there was a 4:1 asymptomatic:sick ratio in the community. Later reports have said 50x! If you start random testing large swaths of asymptomatic cases, it will (and in this case did) make your positives SKY ROCKET.

That said, I DO expect hospitalizations to increase some, because:

1) Only 75% (it jumped again) of the R2 is due to testing, so there is error and could be some growth.
2) When you randomly tell 1000's of people they have a virus, suddenly some of them are going to "feel" sick.
3) Natural increase in case loads in those counties.

[Kung Wu]

In regard to the R2 discussions,

If the R0 of the virus is constant, you would expect a fairly high correlation between tests and confirmed positives - correct? But as SB has stated, % positive of available tests is probably the best metric to look at.

R0 is always changing, never constant, and can be drastically different from locale to locale. It also can very widely based on the particular mutation you are dealing with.

To quote Dr. Strange, "we're in the end game now".

Looking at % positives makes no sense as methodologies are shifting. Testing is virtually worthless other than for academic interest or if you are sick or working on the front lines with at-risk people. Looking at % positives makes sense early in the exponential growth curve when you are trying to slow the disease -- we've done crossed that chasm a couple weeks ago.

[Kung Wu]

The calculated mortality rate (deaths divided by positive tests) is still rising. The only way this can happen is if the rate of infection is increasing faster than the rate of testing.

That's not a mortality rate. That's a Case Fatality Rate/Risk.

As posted on here before by someone else, deaths will lag testing/case loads by probably a couple of weeks.

[ShockerFever]

Aargh's obsession with the CFR is mind-boggling. It's pretty much a useless statistic at this point.

[ShockerFever]

Flu deaths are way down. Not sure on heart attack or stroke. Overall deaths are up.

Thanks. I definitely expected overall deaths to be up with a new disease entering the fray. But if flu deaths are way down, well you can definitely draw a lot of conclusions with that.

[wufan]

Thanks. I definitely expected overall deaths to be up with a new disease entering the fray. But if flu deaths are way down, well you can definitely draw a lot of conclusions with that.

No conclusions there. There is a negative correlation which could be caused by several different things. I did read a couple weeks ago that 10% or so of presumed flu deaths in past years were likely due to non-novel Coronavirus. It could be that social distancing decreased it. I don’t know the reason.

[SB Shock]

Here is the flu with comparison to past years.

flu2.JPGflu1.JPG f

[revenge_of_shocka_khan]

Texas had their highest number of C-19 deaths the day before they took an aggressive approach to reopening their state. I wouldn't be a bit surprised if their aggressive, but risky, approach worked out well. Sometimes drawing to a 16 in blackjack works out well.

We rank #49 out of out of 50 states in testing. Seems like they should have solved for that before opening things up.


Gov. Abbott is at odds with most, if not all big-city mayors. This is just another example of the gap between the rest of Texas and Houston, Dallas, San Antonio and Austin.

And as best I know, none of those cities had mayors who wanted to reopen now.

For instance I believe Harris county (could also be Houston proper) has had almost 5000 cases and 300 deaths. Their emergency room and ventilator supply was taxed but is now there is a little breathing room.

They dont want another round of this. Especially with the lack of reliable testing. I dont blame them.

[wufan]

Here is the flu with comparison to past years.

flu2.JPGflu1.JPG f

We were talking about Flu deaths. Please and thank you.

[wufan]

We rank #49 out of out of 50 states in testing. Seems like they should have solved for that before opening things up.


Gov. Abbott is at odds with most, if not all big-city mayors. This is just another example of the gap between the rest of Texas and Houston, Dallas, San Antonio and Austin.

And as best I know, none of those cities had mayors who wanted to reopen now.

For instance I believe Harris county (could also be Houston proper) has had almost 5000 cases and 300 deaths. Their emergency room and ventilator supply was taxed but is now there is a little breathing room.

They dont want another round of this. Especially with the lack of reliable testing. I dont blame them.

Are the cities not allowed to close if they want to?

[ShockerFever]

No conclusions there. There is a negative correlation which could be caused by several different things. I did read a couple weeks ago that 10% or so of presumed flu deaths in past years were likely due to non-novel Coronavirus. It could be that social distancing decreased it. I don’t know the reason.

Social distancing could have decreased it I guess. I'll lean more to "other factors".

[ShockerFever]

Are the cities not allowed to close if they want to?

Even further, are they forcing people to go out of their homes and eat at restaurants, go shopping, etc? People can choose to stay home if they want to. And if you're higher risk, that's probably encouraged.

[SB Shock]

We were talking about Flu deaths. Please and thank you.

I'm sure you understand that flu deaths are estimates based on how many flu cases and hospitalizations there are from year to year.

Statistically 2020 will go down (at least based on their preliminary numbers) as an average year.

flu3.JPG