August 2, 2020: How Deadly is COVID-19? Let's Count the Deaths....

The Wall Street Journal keeps its content behind a paywall, so I’m putting this link here knowing that you likely won’t be able to get to it. I got to it somehow once, despite not subscribing, but I can’t get back.  The article was called “How Deadlyis COVID-19; Researchers Are Getting Closer to an Answer.”

The authors argued that COVID-19 is less deadly than the flu, but that it looks more deadly because it is so contagious.

I tried hard to grasp what they meant.  As best I could discern, the authors were saying that more people have caught COVID-19 this year than have caught the flu , and so the higher death rate from COVID is because it is more widespread, not more deadly.

But hang on.

Back on July 9, I was able to run this comparison, in a post called It’s Not Just Like the Flu….. 

I realize that was a million cases ago. Time flies! Still, here’s where we stood on July 9:

3 million people had contracted COVID-19, vs 35 million who contracted the flu last year. SO. There were ten times as many flu cases in a season as C19 in its first six months in the US.  The flu was ten times more prevalent.

Meanwhile, 114,000 people had died from COVID-19, versus 34,000 who died from the flu last year.

In other words, COVID’s one tenth as many cases had resulted in three times as many deaths. Big difference.

Hard to argue that COVID is less deadly because more contagious.

The WSJ authors, though, weren’t counting confirmed COVID cases. They were thinking about the current speculation that (at the high end) there may have been 10 to 13 times more COVID cases in the US than were ever tested and confirmed. 

So what if there were? If you used that reckoning, you’d have about the same number of COVID cases and flu cases. And you’d still have three times as many deaths.

I ran it again a couple of days ago, with more recent numbers, 4,200,000 confirmed cases of COVID. Multiply it by ten to play alon with the argument that only about one tenth of all cases are confirmed (lots of cases are either asymptomatic or too mild to get tested, or we were out of tests, or….).  Suppose we’ve really had  42M cases, only a tenth of which were ever confirmed. On the day we hit 4,200,000 confirmed cases, we also hit 146,584 deaths. So now we are up to four times as many deaths as a whole flu season.

It just doesn’t compute. Such a weird argument.