Cheryl Rofer

Dr. Fauci is warning us that it may take more than we thought to get to herd immunity. Now that anti-science Donald Trump is on the way out and Fauci is advising President-elect Joe Biden, he’s telling us what he thinks.

A year ago, we knew next to nothing about SARS-CoV-2. Since then, experts have bootstrapped us all the way to effective vaccines. The bootstrapping started from what we know about other coronaviruses and pandemics in general. It’s a matter of informed guesses, testing them against each other and observations, modifying them, and testing again.

Because the math of epidemiology is similar to the math of chemical kinetics, I’ve been following the modeling. It is also bootstrapping, guessing parameters, testing them, and modifying them. It’s a set of multiple parameters (an unknown number of them) being fitted to data that has serious limitations. The estimates get better as we get more data. That’s what Fauci is saying. Our first guess for herd immunity was around 70%. With almost a year of data, it looks like that could be as high as 90%.

This ties in with what I’ve been thinking, but I want to let the people who are expert in epidemiology and virology lead. I can see what they are doing, but they have knowledge that they’ve acquired through experience that I don’t have. Here’s my version of what’s behind Fauci’s warning.

We don’t know R₀, the inherent rate of spread of SARS-CoV-2, and we won’t know for some time. Pulling it out of the data is a statistical operation and needs a lot of data. Additionally, the data are lumpy – that’s what people are talking about in “superspreaders.” The lumpiness is measured by another parameter, k. We don’t know k either.

We can measure the rate of spread in a particular situation. I’ll call that R, but it sometimes is designated R_T and other names. I have seen confusion of R and R₀, even from modelers, who should know better. R₀ does not vary, but R does. However, the estimate of R₀ changes with time, as more data comes in. Because the level of immunization depends on R₀, that changes with time too. That’s what Fauci is saying.

New Mexico has recently gone from an R of about 1.3 to an R of about 0.86. That’s R, not R₀. R above 1 means the number of cases is going up; less than 1, that the number is going down. The reason R has decreased for New Mexico has very little if anything to do with the virus itself or R₀. It’s because people are acting more responsibly and staying home, wearing masks, practicing hygiene. Here’s a site that estimates up to date values of R for all the states, although they call it R_t.

In order to back out R₀ from that, we need to know how much of that decrease is due to staying home, how much to wearing masks, and so on. It’s hard to get at those numbers. Modelers estimate them, try them out in the models against observation, modify them, and try again.

And k. Don’t forget k. Lumpy data can help or hinder the analysis. What I see so far is that it seems to be more of a hindrance.

Fauci is probably being conservative – it’s better to predict a more difficult situation so that people can be relieved when it’s not that bad – but there are many indications that the first guesses of 70% immunization for herd immunity are too low. As percent immunization goes up, estimates of R₀ and k will improve.

SARS-CoV-2 won’t be easy to control. But we control measles, and it needs 95% herd immunity. A year ago, nobody – nobody! – was immune to SARS-CoV-2. That’s why it has swept the world. For the diseases we’re familiar with, most people have some immunity, whether old folks who had the measles when they were kids or residual immunities to the flu carried over poorly every year. Better times are coming.

Cross-posted to Nuclear Diner

Dr. Fauci’s WarningPost + Comments (137)