That curve

I tweeted something today that hadn’t aged well.

Which is not to give The Economist any guff. Prediction outside of your field is difficult, especially when the entire data set covers, like, maybe five weeks.

It’s the image attached to the article, and the way it looks, rather than the particular information it conveys that I’m here to talk about now.

My question, the one which turned up the above image, to myself, was…

But first a story.

I had a physician for a while once, and he was fairly chatty, which I liked. One day, he told me that when he was in his first year of medical school, he read a study in a textbook. In this study, which is prefaced by noting that they could never do a study like this today, and not even back in the days when he was a medical student, they gathered up all the first year medical students at this one med school.

(An aside: do you know what they call the person who graduates at the bottom of the class from med school? Doctor.)

So, they took all these first years and dosed them with a standard sized dose (whatever that means) of a known cold virus. Then they tracked the symptoms. A few people had no visible symptoms, and some people had barely noticeable symptoms. A whole lot of people were sick for seven to ten days, and some people were ill longer. A few people wound up hospitalized. And all this traced onto a nice bell curve.

His point was this: you don’t really know what a virus is going to do on a case by case basis. The effect of a virus on an individual is unpredictable. But at the population level, you can know this:

A few people won’t seem to be affected. A lot of people will have the symptoms we have come to expect that virus to have, and some people will have extremely bad outcomes.

The question I had was what does the bell curve of known symptoms for COVID-19 look like, and what are those known symptoms at this point? And what does the distribution curve for extremely bad outcomes look like? Is it shifted right? Is it shifted left?

There might not be any studies on this sort of question just yet. I didn’t find any clear graphs. But here’s a study from late July that talks about recover times (worse than influenza).

So wear a mask, keep your distance, and wash your hands more often.