Michael Levitt is Professor of computer science and structural biology at Stanford Medical School and winner of the 2013 Nobel Prize in chemistry.
He has been a close observer of the pandemic and the response from the outset through its movement to Europe, the U.K., and the U.S.. Last month, speaking to the Unherd podcast and youtube channel, he offered some compelling thoughts and observers, and a striking conclusion.
Q: So you noticed that the curve was less of an exponential curve than we might have feared, in those early days?
A: In some ways there was never any exponential growth from the minute I looked at it, there were never any two days that had exactly the same growth rate — and they were getting slow…of course you could have non exponential growth where every single day they’re getting more than exponential — but the growth was always sub-exponential. So that’s the first step.
Q: [In the UK] we talk endlessly about the R-rate — the reproduction rate — and apparently that began very high, maybe as high as 3, and … [we’ve now] got it down below 1 in the UK. Intuitively, if there’s a high reproduction rate, you should see that exponential curve just going up and up.
A: Well no, wait, okay. The R-0, which is very popular, is in some ways a faulty number. Let me explain why. The rate of growth doesn’t depend on R-0. It depends on R-0 and the time you are infectious. So if you are twice as long infectious and have half the R-0 you’ll get exactly the same growth rate. This is sort of intuitive, but it’s not explained, and therefore it seems to me that I would say at the present time R-0 became important because of a lot of movies — it was very popular — talked about R-0.
Epidemiologists talk about R-0 but, looking at all the mathematics, you have to specify the time infectious at the same time to have any meaning. The other problem is that R-0 decreases — we don’t know why R-0 decreases. It could be social distancing, it could be prior immunity, it could be hidden cases.
Q: You’ve been observing the shapes of these curves and how the R-0 number tends to come down and the curve tends to flatten in some kind of natural way regardless of intervention.