Thu 2021-Nov-04

Natural vs Vaccine Immunity, Redux

Tagged: COVID / PharmaAndBiotech / Statistics

Last summer, we saw some evidence that COVID-19 vaccination conferred better immunity than ‘natural’ immunity from recovering from the disease. How has that held up in the face of new evidence?

Comparing ‘natural’ and vaccinated immuities

CDC Press Release: Vaccination is 5x better than 'natural' immunity CDC MMWR: Vaccination is 5x better than 'natural' immunity A few days ago came a study from the CDC [1] [2] which compared COVID-19 infection rates in people who were vaccinated vs people who had recovered from a previous COVID-19 infection. Any material value of the risk ratio different from 1 would be evidence that one or the other was superior.

NB: This is not just a measurement of antibody levels, as was the previous study we examined. That’s convenient to do, and fast, but is only a narrow slice through the complexity of immunity. What we have here is a full-up empirical study of humans in the wild, going about their lives — and that’s the most useful data to have so you can figure out what will happen to you, going about your life.

The view from 10,000 feet

The study examined persons older than 18 who were hospitalized for COVID-19 symptoms. They were either fully vaccinated with an mRNA vaccine in the last 3-6 months, or had already had documented COVID-19 within the last 3-6 months. About 9,000 people were evaluated – this is not a tiny study, but almost the size of the original COVID-19 vaccine clinical trials!

Result: Unvaccinated people had a risk of reinfection that was higher by 5.49 fold (95% confidence limit: 2.75 – 10.99).

Note that the risk ratio there is bounded away from 1 by the 95% confidence limit, i.e., we’re pretty sure this is a real thing and not just by chance.

So we previously thought vaccine-induced immunity was probably better. Now we have to revise our opinion: vaccine-induced immunity is more than 5 times better!

So what?

CDC Recommendation: Vaccination is 5x better than 'natural' immunity Predictably, though also responsibly, CDC administration used this as yet another reason to encourage vaccination:

“We now have additional evidence that reaffirms the importance of COVID-19 vaccines, even if you have had prior infection. This study adds more to the body of knowledge demonstrating the protection of vaccines against severe disease from COVID-19. The best way to stop COVID-19, including the emergence of variants, is with widespread COVID-19 vaccination and with disease prevention actions such as mask wearing, washing hands often, physical distancing, and staying home when sick,” said CDC Director Dr. Rochelle P. Walensky.

Isn’t it about time we start listening to this excellent advice?


Notes & References

1: CDC Media Relations, “New CDC Study: Vaccination Offers Higher Protection than Previous COVID-19 Infection”, CDC Media Releases, 2021-Oct-29.

2: CH Bozio, et al., “Laboratory-Confirmed COVID-19 Among Adults Hospitalized with COVID-19–Like Illness with Infection-Induced or mRNA Vaccine-Induced SARS-CoV-2 Immunity — Nine States, January–September 2021”, CDC Morbidity and Mortality Weekly Report (MMWR), ePub 2021-Oct-29. DOI: 10.15585/mmwr.mm7044e1

Written Thu 2021-Nov-04

Gestae Commentaria

Weekend Editor, Thu 2021-Nov-04 21:17

Interesting, this is surprising to me. Do you have thoughts about the Israeli observational study that found much better and longer-lasting protection from natural immunity?

Thoughts? Sure! Useful thoughts? Not so much…

Basically, this is a sign that we’re on the cutting edge of new knowledge here, because the experimental data is all over the map. In a few months to a few years, we’ll get it all sorted. It will turn out that studies reaching different conclusions were because of some variable we hadn’t previously appreciated, which was set differently in each study. Until then… well, things are weird and we have to cope with ambiguity.

One thing I like about this CDC study is that it’s pretty RWE (“real-world evidence”, and yes, they really do have Congressionally-mandated definitions for that!): find a bunch of people using the EMR systems, track them for a while, report what happened. It’s about what happens when humans are out in the wild, living their lives.

There are some design flaws, of course: not yet peer reviewed (at least not outside CDC), people who were vaccinated were probably told they’re safe which changed their risk-taking behavior, and so on.

If you follow the link to the Israeli study you mentioned, you’ll see that people are debating in the comments the differences between that one and the CDC one. (Though… yeah, I probably shouldn’t be reading the comments, and I backed out quickly.)

Finally, since everybody’s passing out Israeli studies (see what clever people can accomplish with true national healthcare and a unified EMR system?), check out this one from a few months ago. They showed drastically higher antibody levels from vaccination, even in the elderly, even after deconvolving with age to avoid Simpson’s paradox. I mean, yeah, it’s only antibodies; who knows what those folks’ memory B-cells are thinking?

So this is clearly an instance of Scott Alexander’s Razor:

[data] seems to have reached a level of complexity where I no longer feel comfortable having an opinion about it.

I just note that the best experts in the world are confused about apparently contradictory data, and suspect I have no leverage to tell them why any better than they can themselves.

It’s good that this is being studied, since one of the toughest questions facing us now (in my opinion) is whether there is any significant benefit from mandating vaccinations for previously infected people.

I’m less worried about that, since it seems extremely clear that the downside to getting vaxxed is pretty minimal. Use of Jacobson v Massachusetts to back mandates with legal force will make people furious. But they will get vaccinated, and thus live and not die.

I do have some background worries about people who have good medical reason they can’t be vaccinated. I have a friend who’s had Guillain-Barré syndrome, and has to consult his PCP and some immunologists before getting any vaccine. I don’t want him forced into something that will hurt him.

But for most people it’s now likely that there’s an upside to vaccination given the CDC evidence, and it’s unlikely there’s much of a downside.

Weekend Editor, Fri 2021-Nov-05 18:26

In a wonderful example of synchronicity, today the universe brought to my attention this paper in Nature.

As an example of variables we didn’t previously understand and control for, consider the leucine zipper gene LZTFL1. It has a common single-nucleotide polymorphism (SNP) rs17713054G>A. People with this variant seem to have a harder time with COVID-19, as identified in genome-wide association studies that pointed at 3p21.31, where LZTFL1 lives. It seems to work by epithelial-mesenchymal transition (EMT) in lung cells rather than immune cells.

It’s heritable, so it runs in various ethnic groups. Today’s fact is that it’s common in South Asians, i.e., Indians, Pakistanis, and so on. This partially explains why India had such a difficult time with Delta.

It’s by no means a stretch of the imagination that there might be a difference between Israelis and Americans that could explain the difference in natural immunity. (No proof, of course.)