“Facts are stubborn, but statistics are more pliable.” – Mark Twain

Coronavirus Update: Bad Math, Dumb Reporting, Unthinking Citizens 

I had planned to publish an essay today about business strategies for surviving the Corona Crisis. I’ve decided to put that off till Wednesday to talk about the latest numbers the media has been reporting – how they don’t make sense and how the new serology tests may help us understand the two questions no one as yet has been able to answer: How many Americans will get infected? And how many will die?

If there is nothing else I will take away from following this story, it is this: Too many doctors and  scientists pretend to know more than they do, and the media’s reporting on the coronavirus and COVID-19 has been uncritical, irresponsible, and sometimes just plain dumb.

I could give you a dozen examples, but I will stick to just a few.

The Case Fatality Rate 

At the beginning of March, when the coronavirus began to be front-page news, the media was reporting that COVID-19 had a “case fatality rate” (CFR) of 10%. In other words, out of every 10 people that had tested positive for the disease at the time, one had died.

Since I didn’t understand what “case fatality rate” meant, that alarmed me. The regular old flu infects between 10 million and 50 million people each year, according to the CDC. So 10% would mean between a million and five million dead.

That was scary. And there were predictions like that at the time. But after thinking about it for about five minutes, I realized that it was impossible to project deaths based on the CFR. As I pointed out in my March 30 essay, it was an almost useless statistic.

* It was flawed because it was measuring current total deaths against current total cases. (Which makes no sense because of the lag time.)

* It was flawed because it was based first on reports from China and then from Italy. (The first were unreliable; the latter skewed by the age of the infected population.)

* It was flawed because it did not take into account the number of negative cases that were negative because the patients had already had and defeated the virus.

But the worst thing about that number was that the media never figured out or explained the difference between the CFR and the true mortality rate, which is the only statistic that ultimately matters.

I said then that I was flabbergasted that no one in the media was talking about these flaws. I became further flabbergasted when – for at least a week and maybe two – Dr. Fauci and other epidemiologists were not mentioning it themselves.

Since then, they have, but only occasionally, mentioned these important variables. But they never explained how important they were in the calculations. They never admitted that the first estimate of 10% and the subsequent estimates of 6% and 3% (the “scientific consensus”) were utterly useless and entirely misleading, based on ignoring logic and bad math.

As I write this, I’m still astonished by this – the bad science and the dumb-as-doornail efforts at reporting. Why was I able to point out the most obvious problem with the CFR – that it had to be much, much higher than the actual mortality rate?

If you want to read more about my logic and the arithmetic I used, you can do it here. Meanwhile, the media is still reporting unthinkingly on the new numbers coming from not just Dr. Fauci and Surgeon General Jerome Adams, but from Governor Cuomo and other politicians that have accepted these numbers without questioning them. Without asking one basic question.

The question is this: For every person diagnosed as positive, how many that have (or have had) the virus have not been tested?

Based on the fact that, at first, we were only testing people with all the symptoms who had been in China or in close proximity to senior centers and other “hot spots”… and that 80% of those diagnosed as positive had mild symptoms or were completely asymptomatic… how could the difference between the CFR and the real mortality rate be anything less than a multiple of 10?

That’s what I said then. And based on that (and the other flaws mentioned above), I estimated that the actual mortality rate would be between 0.85% and 1.02%.

A few days later, Dr. Fauci, the CDC, the politicians, and the media were all talking about a “fatality” rate of 1%. (They had stopped saying “case fatality,” which I didn’t notice at the time.)

In that same essay, I said that as time passes and many more tests are done, the CFR should begin to move closer to the real rate. Was I right?

Not at this point. In fact, the CFR today is actually higher than the 3% bandied about then. It is closer to 6% globally and 4% in the US!

Check it out for yourself. Right now, the number of cases globally is about 1.7 million. And the death rate is 106,000. That’s a CFR of 6%!

And in the US, the number of cases is 520,000 with 20,000 deaths. That is a CFR of about 4%!

So why are all our trusted sources saying 1%?

It could be because they realized that quoting the CFR was hugely misleading. And rather than admit it and explain the difference, they apparently decided to start quoting their estimates of the real fatality rates. That is the only explanation I can think of. Can you think of another?

So today, we are being told that the fatality rate is about 1%. But nobody in the media seems to be questioning how it went from 3% to 1%. They are assuming that the drop is due to washing hands and social distancing. But that can’t be true!

Adaptive behaviors can slow the spread of the coronavirus. But they do not account for these continued differences between a case fatality rate of 4% (or 6% globally) and the new consensus fatality rate of 1%. The difference, as I’ve explained, is in all the flaws I mentioned above.

But nobody is talking about that.

The Projected Death Toll 

The CDC’s original prediction – a worst-case scenario – was for 1.7 million deaths. On March 29, every newspaper and newscast in the nation led with an amazing update on the Corona Crisis. The new estimate was that 100,000 Americans would die from COVID-19… maybe as many as 240,000.

This was big, exciting news for everyone but little old me and anyone that had bothered to do the math I’d done. My estimate was 85,000 to 205,000.

But I didn’t pretend that my figures were anything but extrapolations based on the numbers I had to work with and the questions I had about how they were figured. I used only a few calculations. One to account for the lag time problem. Another to adjust for the difference between real and reported cases. And one that was a simple matter of multiplying my projected real mortality rate against the estimates we were getting on the number of Americans that would eventually get infected.

That third number was based on estimates that ranged from 20 million to 60 million to 200 million. For reasons I explained in my March 30 essay, I eliminated the high and the low and used the 60 million number as the factor.

But I didn’t know then and I still don’t know exactly how those estimates were arrived at. I explained that, like the government’s other numbers, they were likely derived from the number of people that became infected in China and then in Italy and then in the state of Washington.

It is, in fact, impossible to know what the infectious rate really is because it is an equation that has its own flaws. You can figure out what it could be in a regulated environment – in a lab with rats, for example (if rats responded to the virus the same as humans) or with a lesser degree of certainty in walled-off hot spots such as retirement homes and prisons.

But that number is based on unfettered movement for the virus. And since adaptive behaviors can reduce the speed at which the virus spreads, figuring out how many people will get infected will be impossible so long as society is implementing those measures.

The Arrival Date 

There is a bigger problem here, too – one that was not reported initially and is only now being touched on by marginal news outlets. (The major media are dismissing it as a conspiracy theory.)

That is the question of when the virus was first introduced. The generally accepted story is that the first case in the China was diagnosed on November 17, 2019, and the first case in the USA was diagnosed on January 20.

From the Los Angeles Times, April 11:

“The virus was freewheeling in our community and probably has been here for quite some time,” Dr. Jeff Smith, a physician who is the chief executive of Santa Clara County government, told county leaders in a recent briefing.

How long? A study out of Stanford suggests a dramatic viral surge in February.

But Smith on Friday said data collected by the federal Centers for Disease Control and Prevention, local health departments and others suggest it was “a lot longer than we first believed” – most likely since “back in December.”

“This wasn’t recognized because we were having a severe flu season,” Smith said in an interview. “Symptoms are very much like the flu. If you got a mild case of COVID, you didn’t really notice. You didn’t even go to the doctor. The doctor maybe didn’t even do it because they presumed it was the flu.”

This is one of several reports like this that are now appearing. Given how contagious coronavirus is, this means that the differential between the reported cases and the actual cases could be much higher than 10 (the number I used).

On April 8,  I reported that one epidemiologist from Harvard, Dr. Michael Mina, estimated that the differential could be 50 to 100!

Given what I’ve explained about how these projections are made, I can’t think of how he could have arrived at that number other than because he believes, as I’m beginning to suspect, that coronavirus was introduced into the US before January.

And what does that mean?

As I write this, there have been about 500,000 Americans diagnosed with COVID-19. A 10 times multiple suggests that 5 million have or have had it. A 50 times multiple would be 25 million, and a 100 times multiple would be 50 million.

Does that sound crazy?

Not if coronavirus arrived in the US a month or so earlier than reported.

So if that is true, what about the death rate? If so many millions are (or have been) infected, wouldn’t that mean that the projected death toll should go up proportionately?

But that’s not what happened. In fact, On April 8, Fauci and company revised the projected death toll down from 100,000-240,000 (the March 29 estimate) to 60,000!

As has been the protocol since day one, they didn’t explain how they arrived at that lower number. They didn’t provide the media with the analysis. And the media didn’t question it. They just announced it and, again, attributed it to the success of social distancing.

That makes no sense because social distancing only reduces the speed at which the virus communicates. It doesn’t reduce its natural infectiousness.

It could reduce the number of deaths due to lack of ICU space. And that’s why I concluded my March 30 essay by agreeing with the decision to shut down thousands of businesses and to mandate social distancing and curfews. (Not to mention criminalizing purposeful coughing.) But it turns out that the terrible predictions of patients dying in hospital hallways has not materialized. In fact, it looks like that isn’t going to be a problem.

As I said, social distancing cannot account for that 60,000 projection on April 8. Something else was going on. Could it be that Fauci and company decided to (or agreed to) reduce the high level of panic by talking only about how many would die through the summer, but not mention that by the end of the year the numbers would likely be in the 100,000-240,000 range?

To me, there is only one explanation for all this suspicious math we are being fed: The real mortality rate might be lower, even considerably lower, than 1%.

If, as suggested above, the differential between the reported cases and the actual cases is 50 or 100 instead of 10, the real lethality rate is 5 to 10 times lower than the 1% we’ve been hearing. In other words, one-tenth to two-tenths of 1%. Which is the lethality rate of the ordinary flu.

That doesn’t mean the coronavirus is the same as the flu. It is still very contagious – probably much more contagious than the flu. (The hot spot syndrome is good evidence of that.) And also, when the symptoms of COVID-9 are bad, they are sometimes much worse. That’s why it is much more dangerous than the flu for older people and people with compromised immune systems.

However, if these lower fatality rates turn out to be accurate, there is a good argument to be made that the shutdown was the wrong move. That we would have had fewer deaths and a shorter crisis if we had practiced protocols for establishing “herd immunity.” (Isolating the vulnerable only and allowing the rest of the population to interact as we do and have always done with the flu.) In theory, the “goal” of a herd immunity strategy would be to get half of the healthy population infected as soon as possible. Once that happens, epidemiologists say, the virus dies out on its own.

If we can get those new serology tests going quickly and widely, we will be better able to determine how many Americans have the virus (by comparing in a random test the percentage of those with antibodies) and have, for the first time, a good idea of what the real lethality rate is.

But again, what do I know? We’ll have to see how this plays out in the next few weeks and months and in the fall.