COVID Transmissions for 5-3-2022
Revisiting evusheld, and looking at more mature vaccine effectiveness data
Greetings from an undisclosed location in my apartment. Welcome to COVID Transmissions.
It has been 878 days since the first documented human case of COVID-19. In 878, King Alfred the Great of Wessex (the only English king to earn this descriptor) was forced into hiding and guerilla warfare against Viking invaders, and it looked like he might lose his kingdom altogether. But he managed to rally a large force and decisively defeat the Vikings later that same year. It would have been a mistake to count Alfred out when he was defeated early in 878.
Speaking of things counted out too early, today I’d like to revisit the value of the preventative antibody cocktail EVUSHELD, and then discuss a new vaccine effectiveness study that makes things look worse than they really are.
I have to admit that it has been a bit challenging for me to get back into the swing of writing this after my forced, technical-related hiatus. I’m committing to at least one issue a week for now, but carving out time is proving challenging. I hope to get back up to two issues per week soon enough. Perhaps after some major projects at my day job wrap up.
Bolded terms are linked to the running newsletter glossary.
Keep COVID Transmissions growing by sharing it! Share the newsletter, not the virus. I rely on you to help spread good information, which you can do with this button:
Now, let’s talk COVID.
Revisiting EVUSHELD
As I did last issue with remdesivir, this issue I would like to turn back to a treatment that I was skeptical about when it was first released: EVUSHELD. It is an antibody cocktail that is designed to prevent COVID-19 in people who may not be able to mount an effective response to a COVID-19 vaccine.
EVUSHELD is pretty interesting scientifically because it is a cocktail of two antibodies that were first isolated from recovered people who had had COVID-19, which were discovered at Vanderbilt University and licensed to AstraZeneca. The clinical development team at AstraZeneca optimized these two antibodies and prepared them for clinical delivery. This is the application of thoughtful science to the blunt tool of convalescent serum, used earlier in the pandemic—powerful antibodies were isolated, improved, and combined to yield something that does a better job than the previous generation of options.
The only problem is that as soon as EVUSHELD was authorized under an EUA, a preprint suggested that it might have less potency against the Omicron variant. I was concerned, when I first wrote about this months ago, that it would mean this tool was now compromised and it would be less effective.
Well, I’m pretty glad to say I appear to have been more worried than I needed to be. EVUSHELD appears to be quite effective in preventing COVID-19, at least in a mouse model: https://www.biorxiv.org/content/10.1101/2022.03.17.484787v1
How is this possible? Well, we have to keep in mind the old adage, “If you can dodge a wrench, you can dodge a ball.” Sometimes a particular level of protection is sufficient, but not necessary, to satisfy your desired conditions. Under these circumstances, it may be possible to lower your protection level a little bit and still be protected from the disease outcomes that you’re hoping to avoid. Sure, perhaps the EVUSHELD-mediated neutralization of Omicron-variant viruses was lower than vs previous variants, but that doesn’t mean it wasn’t enough to offer protection. Again, at least in the model used in the preprint linked above.
This is reassuring, and it leaves me confident that we continue to expand our repertoire of quality defensive measures against COVID-19. EVUSHELD may not be perfect—and having some more human data with Omicron would be nice—but in combination with other measures like vaccination, masking, and strategic use of antivirals in the situations where all of that fails, it helps us build a firewall against the most serious outcomes from this virus.
Is vaccine effectiveness against hospitalization falling?
I was recently alerted to the following paper in The Lancet, which suggested that vaccine effectiveness vs hospitalization, at least with the Pfizer mRNA vaccine, had fallen to 55% with the Omicron surge beginning in December 2021 through February 2022: https://www.thelancet.com/journals/lanres/article/PIIS2213-2600(22)00101-1/fulltext
I want to begin by saying that if all other things are equal, then this would suggest that hospitals all over should have been absolutely stuffed with patients, because a whole lot of people in that study’s geography got COVID-19 during that period, and the vaccinated ones were not hospitalized in droves. Therefore, we have to conclude that not all other things are equal. So, what changed?
This is an important opportunity to learn a little bit about what “vaccine efficacy” really is. Vaccine efficacy/effectiveness1 (VE) is very easy to misinterpret. As a percentage, it’s easy to think that a VE of 90% means that 90% of the time you’ll be fine when exposed to COVID-19. As it turns out, that’s not what it means.
All VE numbers are based on a reduction in relative, not absolute, risk. Here’s how it’s calculated:
So what we look at is how much the risk of a disease event (in the case of our Lancet study, hospitalization for COVID-19) goes down in the vaccinated population relative to the unvaccinated population. If there are 2 cases in the unvaccinated population and 1 in the vaccinated population, you have a VE of 0.50, or 50%.
Now this is where things get interesting. Back in December 2020, you heard some pretty eye-popping numbers for vaccine efficacy against hospitalization, pushing 95% or greater. But the thing is, the placebo group in that clinical trial was a very different set of people than your typical unvaccinated population today.
The immunological history of your unvaccinated group makes a difference, because the risk in the unvaccinated arm matters. Back in December 2020, far fewer people had experience with SARS-CoV-2, and far fewer were immune. The placebo group was closer to a population that was “naive” to COVID-19, and it was carefully selected to be balanced in terms of risk factors with the active, vaccinated group. Membership in these groups was randomly assigned.
In 2022, in this Lancet paper, we have a very different situation. The unvaccinated people who went into the Omicron wave, there, had already survived two years of a deadly pandemic. They were people who had chosen not to get vaccinated, as well, whereas the vaccinated group were of course those who chose to become vaccinated.
These two situations mean that it was more likely for at-risk people to be depleted from the unvaccinated group relative to the vaccinated one. This is quite evident in the demographic tables in the Lancet paper, with many comorbidities being meaningfully more frequent in vaccinated populations.
There are two things driving it, as I spelled out before, but I want to be very explicit: if the people with risky comorbidities remained unvaccinated, they were more likely to die, so that would reduce their frequency among the unvaccinated. If they got vaccinated, they became less likely to die, but they were no longer counted in the unvaccinated group, likewise depleting their numbers there.
The end result is that the unvaccinated group became more likely to be resilient to COVID-19 not because of anything inherent, but simply because the people at the greatest risk had been removed from the group. It becomes a trick of statistics. And at the same time, that unvaccinated group likely did get at least some immunity from prior infections, since they had been unprotected during prior waves as well. These pieces of information taken together mean that the risk of hospitalization due to COVID-19 in unvaccinated people has changed over time.
That’s easy to misinterpret, so I want to emphasize that it is because a lot of people died or removed themselves from the unvaccinated population by getting vaccinated, leaving a highly artificial group who either had low risks or preexisting immunity from having gotten sick with COVID-19 (and experienced all the potential damage that entails).
So, when I said “if all other things are equal,” that is what I think isn’t equal here. The event rate of COVID-19 hospitalization has gone down overall in the entire population, which means there are fewer events to prevent. If there are fewer events to prevent, then a low rate of hospitalization among the vaccinated will look like less of a relative reduction.
If hospitals had been truly overrun with vaccinated people during the BA.1 Omicron wave, I might have come to a different conclusion, but since they weren’t, I’m left thinking that this Lancet paper just shows how the overall susceptibility of the population to COVID-19 has changed over time, which impacts relative risk calculations. Under those conditions, the 55% VE is actually rather impressive. Even though many unvaccinated comparator people had already been infected and thus had some protection, the vaccinated people were still more than half as likely to get hospitalized with COVID-19. If nothing else, that leaves me pretty convinced that vaccination is the way you want to be immunized to COVID-19—not infection.
Part of science is identifying and correcting errors. If you find a mistake, please tell me about it.
Though I can’t correct the emailed version after it has been sent, I do update the online post of the newsletter every time a mistake is brought to my attention.
No corrections since last issue.
What am I doing to cope with the pandemic? This:
Drying salami
Living as a kosher person, it can be hard to get access to certain things that other people take for granted, like dried salami, which is usually made with pork. Kosher salami made from beef exists, prepared in the traditional fermentation method, but it usually hasn’t been dried and is instead soft and wet.
A friend recently pointed out that you can just dry your own salami at home! It’s easy. Most packages have their own string to hang in your kitchen—or wherever—and then you can just let time and ambient conditions take their course. It’s worth doing this in a relatively climate-stable environment where the temperature doesn’t get too high, by the way.
Here’s where mine started out:
And here’s where they are now, about a week and a half later:
I imagine there are some folks for whom this sort of thing is incredibly routine—otherwise they wouldn’t keep putting strings for hanging on these products in the store—but for me, it’s a real innovation. I’m excited to enjoy the final product once it’s fully dried out in somewhere from a few days to another week from now.
Last issue sparked some really interesting discussion about antivirals, and also about progress towards approval of a pediatric vaccine. I recommend going back and reading those discussions!
With this space today, though, I want to mention that I’ve seen an uptick in the number of low-prominence antivaccine activists who decide to subscribe so they can comment here with propaganda and lies. This is delightful to me, for two reasons. One, it says to me that by being here, spreading scientific truth as I am doing, I am making these grifters feel threatened, and they think I am a big enough target that it’s worth coming after me. Great.
Two, I ban them permanently within minutes of their commenting, and it takes me a matter of seconds—far less time than it takes them to craft their lies.
Honest scientific questions about vaccination will always be welcome in the comments section of this newsletter—even if skeptical! Science is made better by skepticism, and it’s a testament to the quality of vaccine science that it is fully capable of standing up to good-faith inquiry. But propaganda and lies? That’s another matter, and I won’t allow those sorts of things to have a home here.
You might have some questions or comments! Join the conversation, and what you say will impact what I talk about in the next issue. You can also email me if you have a comment that you don’t want to share with the whole group, or if you are unable to comment due to a paywall.
If you liked today’s issue, please consider becoming a paid subscriber and/or sharing this newsletter with everyone you know.
Please know that I deeply appreciate having you as readers, and I’m very glad that if we must be on this pandemic journey, at least we’re on it together.
Always,
JS
Efficacy vs effectiveness is a distinction meant to help us tell the difference between real-world studies vs clinical trials. Efficacy is for clinical trial results, representing more idealized experimental conditions with greater scrutiny. Effectiveness is based on real-world data, and so is a little more practical but a lot messier from an experimental rigor point of view.
In further news on my participation in the Pfizer omicron-specific trial, at my three-month visit last week following my booster that was either the omicron vaccine or the original, I was unblinded, and offered another booster, which was definitely the omicron-specific vaccine. I had been in the control group, and had received the original vaccine. I'll be giving them some more blood tomorrow, one week after my omicron-specific booster. This is the first time they've wanted blood so soon after the shot. Previously, it has always been one-month after the shot. The staff mentioned that there are a bunch of sub-studies they're keeping track of.
Thanks for the lesson on statistics and the skew in the Lancet article! However, I have to say that I thought the point of the Lancet article was to produce data on how the efficacy of the vaccines waned over time. Given that fact I thought 50% was pretty good given the variable efficacy of the flu shot from year to year!