My inbox, and my Wednesday Instagram stories, are usually a good barometer of where people are in their concerns about COVID. In the past several weeks, nearly all of the questions I get are about vaccines, and specifically how to navigate what sometimes feels like confusing messaging. The White House and CDC have offered a particular set of guidance about vaccination, but in many cases this is at odds with the information people are getting from their own doctors, let alone internet comments. As usual, without a better understanding of why these recommendations differ, it’s hard to sort through them.

Today I want to try to help by talking through three somewhat specific questions that I think reflect where many of the readers of this newsletter are.

First: bivalent boosters. Assuming you’re fully vaccinated already, what is the value of getting the bivalent booster?

Second: vaccines in pregnancy. Is there a best time to get boosted?

Third: vaccines for the under-5 set. We’re now a few months out. Have we learned anything more?

Bivalent boosters

TL;DR: The most important point here is that it is an extremely good idea for anyone who is older or immunocompromised to get the booster. If you’re over 50 or (especially) over 65 — booster!

To refresh your memory… The mRNA COVID vaccines (Pfizer and Moderna) work by inducing your body to produce antibodies to the spike protein of the SARS-CoV-2 virus. The vaccines were developed to target the original viral sequence (sometimes called the “wild type” variant). The initial vaccination protocol for adults was two doses, separated by three or four weeks.

Vaccination provides protection in at least two ways. First, your body produces antibodies to the spike protein, which are then circulating in your blood to target the actual virus. These antibodies fade out over time; the exact timing is unclear, but in general antibodies last perhaps six months. Second, in response to vaccination, your body produces T-cell memory to be able to quickly produce antibodies if necessary. This protection is much longer-lived.

Over the past couple of years of vaccination and viral evolution, we have learned three things. First, the vaccine protection against serious illness is quite robust and long-lived even with the evolution of the virus. Second, the vaccine protection against symptomatic infection is incomplete and relatively short-term. This has been especially true during Omicron. Third, the immune response of older individuals is less robust.

The combination of the last two here — the reduction in protection over time, and the vulnerability of older people — has led to the development of booster doses. A booster dose of the vaccine ramps up the immune response and generates more immediate antibodies (and more T-cell memory). It therefore provides better protection in the long term and, especially, the short term.

We have some good data on the value of booster doses overall. In these data, it has become very clear that for older or immunocompromised individuals, booster doses are crucial for lowering the risk of hospitalization and death. A new analysis out of Scotland last week showed that a booster dose reduces the risk of hospitalization by four times for people over 80; for people with underlying conditions, this could be as much as 10-fold. This data doesn’t show the same added protection for people under 60. This is likely because their baseline risk of hospitalization with initial vaccine doses is already very low.

It also appears that booster doses reduce the risk of symptomatic infection for several months; this reduction (based on a number of papers — see here, here, and here) seems to be about 50%.

The new bivalent booster is a booster dose that is reformulated to better match the circulating form of the virus. The original vaccine was developed to match the wild-type SARS-CoV-2 virus. This updated booster includes some of that original vaccine, but also includes code to induce your body to make antibodies to two forms of the Omicron variant. The hope is that this would provide better protection, against either serious illness or symptomatic infection.

We do not yet have concrete data about whether this booster provides better protection. Last week, Pfizer issued a press release saying that they saw a better antibody response to BA.4 and BA.5 with the bivalent booster. That’s encouraging, but we do not have direct evidence that this booster is better than earlier boosters in terms of performance.

So… should I get it?

If you’ve recently had COVID or had a booster shot in the past couple of months, no. Recommendations are to wait three months between doses, or between an infection and booster dose.

If you haven’t had COVID recently… For older (say, over 50 and certainly over 65) or immunocompromised individuals, absolutely yes to the booster. We are still losing 300 to 400 people a day to COVID, and nearly all of the deaths in the past several months are in individuals over 50. The vast majority are in people over 75. We know from earlier booster doses that they protect this group. If you’re in this group, please get a booster shot. If you have a parent or older relative in the group, please get them a booster shot.

For people younger than this, the value of the booster shot is a period of some protection against symptomatic COVID over the next several months. It’s difficult to know exactly how much protection; without more detail about the performance of the bivalent booster, I think it’s safe to assume it’s similar to the older version — about 50%. The booster will also ramp up protection against serious infection, but in this group, that risk is already very small after initial vaccination. Note that a lower risk of infection will also lower your risk of passing the virus on to others. These are good reasons to do it! The value is lower than in older people, but still there.

We are not seeing serious adverse reactions to the vaccine; the side effects are similar to earlier doses — sore arm, often a day or two of fatigue and fever, maybe some nausea.


During the Delta wave of the virus, unvaccinated pregnant women appeared to be at much higher risk for serious illness and death than others. However, in the more recent waves, and with vaccines, these risks are much lower. I’ve written more extensively about this here. So the first thing to say is that if you are completely unvaccinated and pregnant, it is a very good idea to get vaccinated.

Assuming you’ve had some vaccines already, there are two reasons to get boosted during pregnancy. One is that pregnancy is something of an immune-system compromise, so the value is higher than if you were not pregnant. The second is that a booster dose in the latter half of pregnancy (say, end of the second or beginning of the third trimester) can give an antibody boost to your baby for the first six months of life, before they can get vaccinated. We have seen lower hospitalization rates for infants of mothers who were vaccinated during pregnancy.

Under-5 adverse-effect data

Finally, this is a good time to review what we know so far about under-5 vaccines. The take-up of vaccines for the under-5 set has been tepid. Approximately 10% of this group, about 1.5 million children, have been vaccinated. However, that is still a lot of vaccines, and enough to start looking at realized safety data. For many parents, these early safety data are part of what they are waiting for.

To look at what has gone on with the vaccines, I pulled down data from the Vaccine Adverse Event Reporting System (VAERS) for COVID-19 vaccines for the under-5 age group. I’ve written extensively about this system when doing a similar analysis for the 5-to-11 age group, which you can read here. The short version is that this is a CDC system in which anyone can report adverse events from vaccination. It’s intended to capture a picture of safety in the real world, with millions rather than thousands of vaccines.

To begin: as mentioned, about 1.5 million children have been vaccinated. For 1,931 of that group, there are adverse events reported in the system. This is about 1 for every 800 vaccinated children. However: 831 of these events actually did not have any adverse health events (they may be incorrect timing of dose, for example). If we ignore those, 1,092 children have a reported adverse event. That’s 1 in about 1,370 children vaccinated.

Most of these cases are the adverse events we would have expected from the trials: fever, rash, injection-site pain or swelling, nausea, and diarrhea. It is important to note that the VAERS is not an ideal data set in which to evaluate these minor adverse events. Based on the trials, we know that a lot of children have fever post-vaccination and most of them will not be reported as an adverse event. On the flip side, without a comparison group, it’s hard to know whether (for example) diarrhea is more likely than it would be otherwise.

What the VAERS data is very useful for is looking at serious adverse events, like those that result in hospitalizations.

There are 28 hospitalizations reported in the data, or about 1 for every 53,000 vaccine doses. Nine of these hospitalizations are linked explicitly in the narrative to other viruses. Of the others, most are non-specific. There are four instances of seizure, which could plausibly be linked to the vaccine, and two instances of Guillain-Barré syndrome. These cases all resolved, and, again, as a share of 1.5 million children vaccinated, these events are extremely rare.

Combining these VAERS data with the fact that we haven’t seen attention or reporting on any vaccine risks for children should increase confidence in vaccine safety.

Efficacy data is less available. Serious illness is very rare in this age group, so while we would expect the vaccine to decrease the risk, that is not something we have yet seen in the data (in contrast to the 5-to-11 group, where we have enough data during Omicron to show reductions in hospitalization risk). Based on what we know, though, we would expect the vaccines to provide some protection against symptomatic illness; the trials showed 50% to 75% protection.

What if your child had COVID? In that case, it may be best to think about these vaccines as (effectively) a booster dose. They can provide additional protection against symptomatic illness and lower their risk of getting COVID this winter. And given the number of illnesses we all seem to be picking up these days, lowering the risk of this particular one seems like a good idea.