Coronavirus, by the numbers, as I see it, Mid July

 Coronavirus, by the numbers, as I see it, Mid July

It has been almost 3 weeks since my last update.  Since then, every state has relaxed their restrictions.  My own area was placed in the “Green” zone this morning.  I plan to not change my behavior at all. 

I am using one unit of measure. Current number of active cases.  I am thinking “Results Based” and that is the only number that matters.

Nationally, on average, the number of active cases have been going up about 1%, per day.  Two weeks ago it was .5%.  Yesterday the growth was 2%, a number we have not seen since mid-May.

The overall death rate is down to 11%, from 12% two weeks ago.  There is good news about a Remdesivir’s effectiveness to mitigate the virus that seems to reduce the morality rate by 2/3rds.  Large scale testing has been started.

Overall, I think the situation is very bad.  We avoided the dire predictions of mid March, of 8 million cases by end of April, and that was good.  We have achieved the plateau over a month ago but have not budged from it.

FYI, most of Europe also are plateaued, Some of the hardest hit countries now have new cases in the mid hundreds.  In the next few weeks we may see serious drops in their numbers.  Of western Europe, only the UK and Sweden (66 and 10 million) exceeded 1,000 new cases yesterday.  Compare that to Texas (population 29 million) adding 6,000 alone.

So how are the states doing?  Since the Federal Government has largely devolved it’s responsibilities to the states, not very well.  I was arguing with someone about my Governor doing a horrible job, but when you think about it, there are 50 Governors, and none are qualified to deal with this situation.  So the US has 50 plans.  With a wide spread of results.  The hardest hit states were mostly the densest packed.  And until a couple of weeks ago, the source of the worst numbers.

Of the 50 states, only 16 have fewer cases now then two weeks ago.  Illinois has the best reduction, 26,000 cases now down from 53,000.  Next follows North Dakota (but the total is only 140, so who cares) and Massachusetts, 8,000 from 12,000.

Other states that dropped are:

NewHampshire -25%

Pennsylvania (-5,000 cases)

Michigan  -20%

SouthDakota

Wisconsin

Nebraska  -15%

Indiana  -10%

Minnesota

Maine

NewJersey  -4% (but 5,000 cases)

Kansas

Colorado

Connecticut  -.13% (just enough to qualify as a reduction.)

 

The bottom of the list, most increases are from least to most:

WestVirginia  +20%

Georgia

Washington

NorthCarolina

Kentucky

Vermont

Tennessee

Missouri  +30%

Louisiana

California  +40%  (+40,000 cases)

Alaska

Mississippi

Utah  +50%

Alabama

Florida  +60%

Arkansas +70%

Oregon

Wyoming

Arizona  (+23,000 cases)

Nevada

Idaho  +100%

SouthCarolina  (+8,000 cases)

Texas  (+ 31,000 cases)

Hawaii  +150% (only 80 cases)

Oklahoma  +175%

Montana  +25% (only 160 cases)

 

New York still has the most cases, 295,000, and growing 3% over two weeks ago.  I have a suspicion that the this number is not accurate, to high by maybe 50%,  based on patterns I see in other states.  Time will tell.

These are the numbers, draw whatever conclusions you can from them.  My excel wookbook is attached, and yes it has macros embedded in it. 

The spread sheet DayToDay are the number of active cases, per state.  Highlights in Green mean a drop in day to day, red means a +20% increase in a single day, orange is an inexplicable drop in the numbers.

The spreadsheet RedBlue a list of all states and their case shift for the last two weeks, and annotated with their political alignment in the 2016 election. 

 

 

Coronavirus, by the numbers, as I see it. Late July

 Coronavirus, by the numbers, as I see it.  Late July

During this pandemic I studied one unit of measure. The current number of active cases.  Counter arguments have implied this is the wrong way to look at it.  So I have spent some time pondering my methodology.

The counter arguments point out that the numbers are bogus to begin with.  Often supported by some news article about this or that state ignoring important facets of the tests.  And there is no doubt this has happened and the numbers skewed.  But, I am looking at a homogenization 59 reporting bodies and they are a homogenization of at least 6,000 data points.  One off errors in reporting, even by a whole state, doesn’t change the statistics by more than 2%. 

And the argument that 59 reporting bodies just spontaneously created this as fake news to make Trump look bad is patently absurd.   

The most compelling point is to look at the mortality rate.  That comparing New York, the former #1 infected state to Florida, the current champion, needs to also compare the mortality rate. 

                Active Cases                       Deaths                  Mortality

                                                                Per million          resolved cases

NY          152,000                                 1,670                     11%

NJ           56,000                                   1,780                     12%

Florida   325,000                              242                         11 %

Texas    167,000                                 148                         2.2%

Arizona  127,000                               401                         13%

Pennsylvania 23,281                        555                         8.4%

Idaho    11,192                                   71                           2.6%

Nationwide  2,000,000                    438                         7.25%

 

Note:  I wrote this up a week ago, so the numbers are a bit dated.

 

I originally started this list just to compare NY to Florida.  But added other states of interest to me for reasons of other conversations I am having elsewhere. 

 

So anyone looking at the list at deaths per million thinks Florida, despite having more than double the current number of active cases, is doing a better job than NY with 1/7^th the deaths per million, the population of both states are close to 20 million.  But New York has resolved (survived or died) almost 300,000 cases, and those are the only cases to compare the dead to, and so NY’s mortality rate is 11%.  Florida has only resolved 45,000 cases, and has a matching mortality rate of 11%.

Nationwide, the current mortality rate was 438 per million, or 7.25% of the resolved cases. 

That 7% figure is tracking downwards due to a number of reasons.  Since the numbers include earlier higher averages (25% initially) which will drag that number upwards till more cases resolve that will pull it downwards.  Per the CDC, last week’s mortality percentage is 6.4%.  I have seen some commentary that the curve on the mortality rate looks like it is going to settle in at 4.6%.  I have had not seen any correlative evidence of this.  And since 4.6% is lower than Germanys 4.8%, I suspect it is wishful thinking.  But New Zealand’s rate is 1.6%, so a lower number is possible.

Deaths is the ultimate lagging indicator, and is really only valid as a “current state of being” for states that have had time to develop the statistics.  A number of people are claiming the low deaths per million of the sparsely populated states is due to superior medical treatments.  Based on the numbers I see, they may have a point.  Let’s hope so.

Some of the more educated, mask-denying, I will not live in fear, and/or freedom loving people will insist that I am not using the correct denominator in the calculations.   They insist I should use Total Known Cases vice Resolved Cases.  Which will generate a mortality rate closer to what they want to hear.  First, that is adding in unknowns into the equation as the fate of Total Cases will include both deaths and survived.  Second, even if I did, the most the percentage would drop is by half.  It’s not the .001% number they want to see. 

The last serious argument I have had to look at was with the Excess Deaths crowd.  How many are dead this year over last?  If the Excess Deaths are +/- 3%, then Coronavirus is a non-starter.  We are just seeing statistical variations.  But another reason for this argument is that it is a bit more invisible, and so an easy one to stand on, because it’s all conjecture.  The irony is the same group of people will not consider conjectures, also referred to as data models, on their other political talking points.

But, as it turns out, these numbers are available; the CDC keeps a running count, and has been for years!

Since end of March, 2020, and for every week till now, we are above the expected death count.  Both in raw numbers and percentages.  Worst week was April 11 with 40% extra deaths, over 23,000 people.  The numbers have improved every week since.  This chart’s accuracy lags by up to two weeks for various bureaucratic reasons.  When I added up the excess deaths from March to July 11, the number was around 151,000, higher than the official death count by Covid-19 on that day of 147,000.

See:  https://www.cdc.gov/nchs/nvss/vsrr/covid19/excess_deaths.htm#dashboard

My conclusion is to keep monitoring the number of active cases.  It is from this number, all the other numbers are eventually derived. 

 

 

 

 

 

 

 

Coronavirus, by the numbers, as I see it. Mid August.

Coronavirus, by the numbers, as I see it.  Mid August.

It has been 2 weeks since my last update.

The number of active cases continues to grow by less than .75% per day for the entire month of August.  For most of July the number was around 2%.  But I have reason to suspect that .75% is actually too high.  There are many states that lose track of “recovered” cases, and as such, their number of active cases are inflated.  The most obvious is Florida, 506,000 active cases, only 53,000 recovered.  Worst offender percentage wise is Maryland, current active count is 90,000 and recovered 6,000.  Since Florida currently represents about a quarter of all active cases, it is skewing the numbers quite a bit all by itself.  (Shades of Benjamin Harrison’s election of 1888)

This has a side effect of inflating the mortality rate.  Nationally around 5.6%

If I put corrections in for the top 5 most highest mortality rate states, the nationwide total of active cases drops by 600,000, which means nationwide the numbers have been dropping all of August.  And probably most of July.  Since I am regressing a sudden change into numbers that were not also regressed, the percentage drop is somewhat in error, but is close enough to draw a downwards vector regardless of the actual angle. 

The states I believe are most out of whack are:

State     Mortality Rate

Maryland             34.40%

Rhode Island      34.40%

Virginia                 15.00%

Florida                  14.33%

Arizona                 14.08%

 

Any state with a mortality rate over the national average of 5.6% has an active case number that is certainly out of whack; the list above are just the worst offenders.  The true national mortality rate is certainly lower as the number of recovered is undercounted.  For contrast, California, its 4.3%, which is lower than the national average, and California only has half her cases resolved, so I suspect her numbers are inflated a bit as well.  Some states have a mortality rate below 2%, Including North Carolina, Kansas, and both the Dakotas.

This observation on the uncounted recovery rate seems to have been noticed.  The CDC web site just started today reporting it as part of their standard grid. 

 

In other good news, 21 states now are showing fewer cases now, then 2 weeks ago.  When I first started tracking “Current vice two weeks ago”, there were 12 states, and that number reduced to just 6 by the end of July.  But since then new states have been added.  They are by order of reduced percentages:

NewJersey          -43.64%

NewYork             -31.45%

Louisiana             -24.39%

Utah      -24.02%

NorthCarolina    -19.32%

Connecticut        -18.67%

NewHampshire -18.43%

Ohio      -18.29%

Vermont              -16.44%

Mississippi          -13.46%

Wyoming             -11.27%

Maine   -10.05%

Iowa      -9.91%

Montana             -7.94%

Wisconsin            -6.89%

Texas    -4.53%

Pennsylvania     -4.44%

Arkansas              -4.30%

Minnesota          -3.78%

Tennessee          -1.90%

Indiana -1.66%

 

Harder numbers to find are ratios of the dead vs. crippled with lung, heart and kidney damage.  Mortality is rather Boolean, you either are, or you are not.  Damaged for the foreseeable future is harder to identify.  At the moment, my broad ass guess is that about 6% of the identified cases require hospitalization.  About a 1/3^rd of those die.  And of the recovered, most are showing damage either to lung scaring, heart damage, and kidney damage.  If anyone has better counts, I would appreciate the feedback.

Not numbers related, and slightly political.

Since I have the time, the inclination, and the numbers for the last few months I have been amusing myself by countering arguments against anti-maskers.  I made that my mission as the most valuable thing I can do to help in the situation.   

The tactics I use are to be respectful to the anti-masker, but also to never let a claim or counterclaim go unmet.  Following the axiom, a charge unchallenged, is a charge believed.  So, when I commit to battle, I must see it through. You cannot let them have the last word.  A win means the anti-masker retires from the field and anyone visiting the conversation thread sees a cogent rebuttal all the way through.  Convincing the anti-masker they are wrong is an unrealistic expectation.  Their objections are rooted in things other than science and they will not be swayed.  The battle is for the hearts and minds of everyone else. 

It’s not a satisfying victory.

Have I made a dent?  I think I have, albeit a very small one.  I have actually had 3, comments back about how respectful my arguments were and were appreciated. 

And, from my perspective, the number of vehemently opposed, anti-masking arguments has dropped considerably on the forums I peruse.

I attribute that drop to the rapid increases in the number of infections in states that had natural barriers to the initial spread of the virus due to low population densities.  The political nonsense counterpoints have dropped to near zero.  Or maybe I’m just filtering them out.  I recall one comment about the rapid rise in the number of active cases in Idaho, from 500 in June to over 10,000 mid July, “If it wasn’t for Boise ( a Democratically controlled city), we’d be all right.  The mentality that that fueled anti-masking sentiment took joy when the most infected states were Democratically run, has lost traction now that 4 of the top 5 most infected states are Republican.  The same people are that promoted anti-masking have switched to the pointing to the riots, which I am not getting into.

For full disclosure, I was a lifelong registered Republican until last year. 

Also, on 5 of my last 6 shopping expeditions, I noticed no mask violations at all.  No nostril flares, no lifting the bottom to talk.  Not one.  Some of this is behavior improvement, but I also noticed a lot of decorative masks you can buy now are designed so that you cannot wear them incorrectly.  Good going to the small businesses. 

Despite that however, yesterday, there were 60,000 new cases, 1,100 have died.  California, Texas, Florida, and Georgia lead the way with half of them. 

These are the cases we know about.  The amount of testing, the most important tool we have, week by week, has been going down for four weeks now.  And it is taking longer to get results.  My son went in for testing on Wednesday, and they told him he won’t have the results for a week or more.  It used to take 3 days.

If you want to blame politics for something, this is it.

 

 

Coronavirus, by the numbers, as I see it. Late August.

Coronavirus, by the numbers, as I see it.  Late August.

It has been 2 weeks since my last update.

The number of active cases continues to grow by less than .35% per day for the second half of the month of August.  The first half of August was .75%.

In my last missive, I mentioned that I suspected the growth in active cases to be too high because many states have lost track of “recovered” cases, and as such, their number of active cases are inflated.  Nothing I have seen in the last few weeks has dissuaded me from that opinion.  The most obvious is Florida, 529,000 active cases, up 40,000 since two weeks ago, but only 65,000 recovered, up from 59,000  two weeks ago.  Worst offender percentage wise is still Maryland, current active count is 95,000 (+ 5,000 from 2 weeks ago) and recovered 6,000 (unchanged).  However, since Florida currently represents about a quarter of all active cases, it is skewing the numbers quite a bit all by itself.

This has a side effect of inflating the mortality rate.  The official percentage nationally is around 5.4%, and falling.

This disease has a 3-6 week cycle time.  About 94% of those infected recover or die within 3 weeks.  Extreme cases up to 6 weeks.  In addition, the severely crippled can remain in the hospital even though they are cured of the infection.  I do not have a percentage of that.

Disclaimer.  The following is just an estimate based on my work.  I am not an official statistician, nor a medical expert.  This is just my observations.

Playing with the numbers, and assuming 94% of the cases from 3 weeks ago have recovered for the states that are the highest offenders, defined as having mortality rates over 9%, I get a national total of 1,400,000 current active cases.  The official number is 2,500,000.  If this is correct, then the mortality rate is reduced to around 4.2%. 

Keep in mind that a number of recovered victims will continue to suffer damage from heart, lungs and kidneys.  An estimate on that is about 100-200 percent the number of deaths.  Between 170,000 and 340,000 as of this morning

Disclaimer off.

Of states that seem to be accurately accounting for recovered patients, 17 are in an overall decline in active cases.  Two more are plateaued, dipping just above and below the zero percent mark.

The top five most declined states are:

Louisiana

Connecticut

NewHampshire

Texas

Arkansas

As mentioned above, there are a number of states that are not counting their recovered cases very well, so displaying the bottom five is kind of fictitious. 

For anyone interested, my spreadsheet is available.

Not numbers related

Regarding the future vaccine.  The USA has funneled about 5 billion dollars into 5 different pharmaceutical companies for developing a vaccine.  At least 15 other companies are also working on it in the US.

Many companies are already in trials to prove efficacy and prove no harm.  We are looking to December for approval of these.  With production lags, we are still looking to June, 2021 for enough available for the general population.

The vaccine will most likely require two shots.  And is most likely going to require yearly boosters.

This is because C-19 is like the common cold, and the antibodies produced by the body are a short time version lasting 1-6 months.

It had  been suspected that people can get C-19 twice, and for months there have been unproven examples of this, but as of this week there is the first fully documented case of someone coming down with it twice.