Coronavirus: Michael Levitt Follow-up

This is a quick follow-up to my recent blog post about Michael Levitt's projections on COVID-19. Now Michael Levitt is an American-British-Israeli biophysicist who won the 2013 Nobel prize for chemistry. So he is definitely far smarter and intelligent then me. But I have issues with his projections. For example, I suspect he just got lucky with his China projections. I first came across his projections based on this Jerusalem Post (Mar 20) where he was quoted with the following:

In Wuhan, where the virus first emerged, the whole population theoretically was at risk of becoming infected, but only 3% were.

The Diamond Princess cruise ship represented the worst-case scenario in terms of disease spread, as the close confines of the ship offered optimal conditions for the virus to be passed among those aboard . . . “Those are extremely comfortable conditions for the virus and still, only 20% were infected. It is a lot, but pretty similar to the infection rate of the common flu,” Levitt said.

Hmm, well here's an LA Times (Aug 11) article regarding San Quentin State Prison, which I'd argue is similar to the Diamond Princess in terms of people being confined to a limited space:

COVID-19 spread unchecked across California’s oldest prison in ways that stunned public health experts, despite efforts to control the disease. As of Monday, there had been more than 2,200 cases and 25 deaths, among a population of more than 3,260 people. On Sunday, a guard became one of the latest to die.

That means more than two-thirds of the prison’s population has been infected, said Dr. George Rutherford, epidemiologist and infectious diseases expert at UC San Francisco.

And though new cases have slowed, they are still occurring — with 60 reported in the last two weeks — suggesting herd immunity has not yet been achieved.

San Quentin’s death toll translates to a mortality rate of about 767 people dying out of every 100,000 persons.

Note that the LA Times is using a mortality rate when they're talking 0.8%. The CFR % would be 25 deaths over 2,200 cases or 1.1%. The CDC is using a CFR % of 0.4%. I would assume that San Quentin State Prison age population is somewhat evenly distributed above the age of say 25. Meaning I doubt they're all over the age of 75 and therefore have a much higher chance of dying from COVID-19. So I think that a CFR % that is nearly triple that of the CDC is something to consider, especially since it is likely that a few others will die.

But the big thing I'd like to point out is the infection rate of 67.5% and growing (2,200 cases out of 3,260 people) is far higher than the 3% or 20% that Michael Levitt used/is using to come up with his forecasts. Are the forecasts he is using to calculate the spread of the virus across various countries skewed by the fact that his assumptions about the risk of getting infected is far too low?