Tomorrow (Friday, 12/3): AMA here, 3 p.m. EST (note new time for this week). Subscribers only — subscribe now to get notified and join in!
At times, the COVID-19 pandemic has seemed a master class in how not to do public health messaging. The guidance on masking, for example, has lurched back and forth, leaving the public confused and with limited trust. The booster messaging is the latest example and, to me, among the most extreme. Some people “may” get a booster; others “should.” The “may” and “should” groups have shifted over time, in many cases seemingly without much reason.
Omicron has made this worse. The message that (a) the new variant may show some vaccine evasion and (b) the solution is to definitely get a booster is … confusing. The vaccines work less well, so I should get another one? This seems puzzling, although there is good science behind it. But public health messaging is not explaining that science.
This last point is key. The messaging around COVID has rarely explained why we change guidance, and that erodes trust. Experts say different things (see, e.g., this article), and without knowing more context, it is difficult to see why there is disagreement. So today I’m going to discuss boosters. I’m going to start at the beginning: why there is uncertainty about the number of shots in general, and what boosters might do. Only following that will I talk through the decision to get one. And, yes, I’ll touch on both Omicron and natural immunity.
The TL;DR: Very likely you should get a booster, and for sure if you are older or at all at higher risk. Also, thanks to Westyn Branch-Elliman for talking this through with me over the weekend, but don’t blame her if you don’t like what I say.
Step back: How do we decide vaccine dosing?
There is a lot of variation in dosing schedules for non-COVID vaccines. Your kids get two measles shots. They get five doses of the Tdap vaccine; hepatitis B is three shots. Most adults get only one flu vaccine, but some older adults will get two in a typical year.
The variation in schedules reflects an attempt to maximize immunity development at necessary times. Changes in dosing schedule often respond to new information. For example: at some point, it became clear that babies were at risk for pertussis in the first few weeks of life, before they could get vaccinated themselves. In response, Tdap boosters in pregnancy were introduced; this induced antibody development in the pregnant person, which was passed to the infant, lowering pertussis risks.
Developing optimal vaccine dosing takes time. You need a lot of data to figure out that one dose of measles vaccine is 93% effective but two is 97%. It should not be remotely surprising to us, then, that we are still figuring out the optimal dosing schedule for the COVID-19 vaccines. There are actually many open questions. What is the optimal spacing between the first and second dose? Should younger people have only one dose? Should people who have already had COVID have only one dose?
Central among these questions is whether the standard dosing series should be three doses for everyone, three doses for some people, four doses for some and three for others, and on and on. Does the need for additional doses interact with the length of time between dose 1 and dose 2? We do not have perfect answers to these questions right now, and we will not for some time. We will eventually, just as we have detailed dosing schedules for other vaccines. But not this week.
We’ll make our current choices about boosters with this background uncertainty, and that’s just the reality of the moment. It is important to be clear, though, that this uncertainty is expected and it is different from saying there is uncertainty about overall vaccine efficacy or safety. We are very certain that COVID vaccines protect against serious illness and death and reduce infection risk. We are very certain that the vaccines are safe, and do not contain tracking microchips (just to take on one claim). We can be certain about these things while also being somewhat unsure about the exact right dosing schedule.
Step back: The goal of boosters and their effects
The goal of vaccines is to prevent serious illness and death. An additional benefit of vaccines is that they lower the overall risk of infection, which also lowers the risk of spread to others.
For older and more vulnerable individuals, the standard two-dose vaccine series shows waning effectiveness against the primary goal. In this group, we are seeing increasing evidence of infections leading to hospitalization and, in some cases, death. The risks remain much lower than among unvaccinated people, but they are elevated relative to the trial data and to what we saw early on.
For younger people, who are at lower risk of hospitalization and death in the first place, the concerns about waning protection are less significant because a mild breakthrough case is less likely to turn serious.
A simple way to understand it may be to say that over time the vaccine protection against any infection has waned (the “benefit” described above gets lower). For individuals with less-robust immune systems, this affects the vaccine’s success against the primary goal, because even a mild infection can be serious. For younger people, that’s less true.
The reason that immunity wanes in this way is that, over time, circulating antibodies disappear. You’re not left unprotected, however. Targeted T-cells and B-cells retain the memory of how to produce antibodies. These components of the immune system are also more flexible, making them much more likely to be effective against variants. But this part of the immune system takes a few days to ramp up; antibodies work faster. The result is that a breakthrough case would be more common, although in the vast majority of cases it would be mild, because the immune system is ready.
For some people with limited immune response, a third dose of the vaccine may be required to generate immunity at all. For most people, though, the intent of a booster is to increase the production of antibodies, generating the ability for a faster response to viral exposure. The expected result is fewer breakthrough cases. This is, in fact, precisely what we are seeing. The U.K. has rolled out boosters by age group, and the FT has the best graphs (hat tip: John Burn-Murdoch). Graphs below — hospitalizations and then cases.
These graphs show compelling evidence that boosters are working, against cases overall and against hospitalizations for older individuals. Yesterday, new data from Israel confirmed this on cases, in particular.
Bottom line: in the current setting, it seems clear that a booster will lower the risk of a case and, if you’re more vulnerable, serious illness.
How long will the additional protection last? Will we need more boosters? When? These answers are not clear yet. This is part of that uncertainty we cannot avoid.
Does Omicron change this?
We are still waiting to learn more about the new variant. Based on what we know about other strains, there is reason to believe that T-cell and B-cell immunity will still be effective (if perhaps slightly diminished). Antibodies, however, may be less effective (may be! we still do not know for sure!) because they are a less close match to the Omicron spike protein. This would increase the risk of a breakthrough case.
Although it sounds confusing, having more antibodies in this situation is helpful. Part of the less-than-perfect match can be addressed by having a high antibody volume. This means that the success of a booster in terms of preventing a breakthrough case may be higher with a new variant. It is also worth noting that with a new, more transmissible variant, case rates are likely going to be higher, making a breakthrough more likely.
Does breakthrough COVID count as a booster?
For reasons that are somewhat unclear, there is a lot of resistance to discussing infection-induced (sometimes called “natural”) immunity to COVID-19. My sense is that a big piece of this is a desire to encourage vaccinations among individuals who have had COVID-19 and haven’t yet had a first vaccine dose. Evidence out of Israel and elsewhere suggests that at least one dose provides better protection than infection-induced immunity alone.
This is a bit separate from the question of breakthrough COVID, though. If you are double-vaccinated and have had COVID after vaccination, should you get another booster? In a real sense, you’ve already had a booster. There is little reason to think there would be a downside to an additional shot, but the upside is very likely more limited. (This is consistent with this new pre-print, which shows very durable protection from two doses + COVID).
So … should I get one?
In the end, despite the messy messaging, I think it’s not that complicated.
You should get a booster if you want to lower your risk of getting a breakthrough COVID infection in the next few months.
There are many reasons this might be beneficial. One is if you’re at risk for serious illness from such an infection. A second is if you’re worried about infecting others (either in general, or specific others like unvaccinated kids or immune-compromised people). A third is if it would be practically difficult to have COVID (like, you’d miss work or school). Or maybe you just do not want to have COVID.
On the flip side: Are there reasons not to get a booster? Only limited ones. We are not seeing red-flag adverse effects from it. To the extent that there are theoretical concerns, they are largely around young men and the risk of myocarditis (which is typically mild and treatable). A healthy 19-year-old man with limited high-risk contacts is probably the most complicated type of person to consider for a booster; the benefits are fairly small and the risk of side effects is small but larger than for other groups. To the 35-year-old men who write to ask me if they should be worried about myocarditis, I will say that, for better or worse, you’re basically too old.
The other concern raised about boosters is vaccine equity. Global vaccine equity is important, but more boosters in the U.S. is only a tiny, tiny factor. Last-mile delivery issues, vaccine resistance, cost — these are all much more central.
Overall, I think a booster is a good idea for nearly everyone and crucial for anyone with any serious risk factors. (Yes, I had one.) The arguments are not as strong as the arguments for a first or second dose; not even close. I can see why reasonable people differ. But if you do not want to get breakthrough COVID, especially heading into the winter, get a booster.
Does it matter which one?
Not really, no. If you’ve had the J&J vaccine, you should get an mRNA booster. If you had one of the mRNA, you could get that again, or the other one. There is speculation that maybe we’ll all ultimately want to do boosters with some other vaccine type, but this is uncertain. Which is the word of the day.
I want to come back to where I started, about the messaging. It’s not that the message currently being sent is wrong. The advice that we are getting — on boosters or anything else — is usually based on the best knowledge at the moment. The problem is that it’s too often delivered with an air of certainty that the knowledge doesn’t support.
This would be fine if people were amnesiac robots. If everyone woke up in the morning, forgot everything from the day before, checked the current guidance, and seamlessly took up that guidance, then the strategy of changing the guidance every day would be great. It would let you incorporate even small changes in knowledge in a seamless way.
But people are not amnesiac robots, and our trust in guidance is based on consistency and understanding. In this world, we need messaging with more nuance, messaging that does a better job of explaining why the guidance is what it is and why it might change.
You should very likely get a booster. Hopefully now you can do so while understanding a bit more about why.