A Cautious FDA Also Threw Out the Baby

In restricting the label for COVID-19 vaccines, the FDA is being too conservative.

An update to ​“A Vaccine for Your Mind,” February 2025.

By Mario Barro, PhD, Tim Howard, PhD, and Peter Kolchinsky, PhD

CULTURE | SCIENCE | BIOTECH

Photo by Mathurin NAPOLY / matnapo on Unsplash

June 25, 2025

In May 2025, the FDA announced guidelines for the approval of new COVID-19 vaccines. Vaccines would be approved on the basis of immunogenicity (i.e., how high the anti-COVID-19 antibodies are after vaccination) for adults over 65 years old and everyone older than 6 months with one or more risk factors for severe COVID-19 outcomes (e.g., asthma, COPD, cancer, diabetes, heart conditions, obesity, pregnancy, smoking, use of corticosteroids, a long list of kidney, lung, and liver diseases, and more). Quite a few people have at least one of these risk factors, so actually a lot of people remain eligible for a COVID-19 vaccine.

But for individuals under 65 years without risk factors, COVID-19 vaccines are no longer considered approved. Though the FDA considers pregnancy to be a risk factor that makes maternal vaccination appropriate, the CDC recently removed the COVID-19 immunization recommendation for pregnant women from the immunization schedule. This is notable since a newborn’s immune system isn’t mature enough to respond to a vaccine and so they rely on antibodies transferred from their mothers. So vaccinating pregnant women extends protection to newborns. 

The FDA will now require randomized, placebo-controlled trial data before any COVID-19 vaccine is approved for healthy people under 65. The bar for success is 30% reduction in symptomatic COVID-19 with a minimum of 6 months follow-up after immunization to ensure that benefits persist long enough to matter.

The FDA is basically saying that it’s not confident that the benefit of the vaccine outweighs the risks in otherwise healthy people under 65 and wants more data.

That doesn’t mean it was wrong for people to get vaccinated in the past, especially while we were still in the midst of a pandemic and few people had natural immunity. But now that we’re several years beyond the pandemic, the FDA is saying that it needs more data to decide on the role of vaccines for the long run. 

But does it mean that people for whom COVID-19 vaccines are no longer approved shouldn’t get the shot (or shots, as in the case of babies getting their primary immunization)? That’s a complicated question. As you’ll see from our data analysis below, we see where the FDA is coming from when we look at the benefit vs risk in people from 5 – 65, but we disagree with its risk assessment for children under 5. COVID-19 appears to still pose a high enough risk for young kids and the vaccine poses such a low risk for them that we think the FDA was overly restrictive in rescinding approval of COVID vaccines for this age group (and, in the case of the CDC, rescinding its recommendation for pregnant women to get immunized).

Hopefully our analysis will help people decide for themselves if they want to be immunized. Doctors are free to prescribe medicines off-label, which means that if you want a vaccine for yourself or your child and your doctor agrees, you should be able to get it. But it may not be as easy as just going into your local pharmacy for a shot.

Some providers may be hesitant to prescribe it without formal guidance, and others might not even stock the vaccine until there’s a clear recommendation for a broad population. Some insurers may refuse to pay for off-label shots.

For clarity, we avoid using the term booster in this document because, although the COVID-19 vaccines continue to target the same spike protein from the virus, the structure of that protein changes slightly each year as new variants emerge. Similar to the flu vaccine, these updates mean the immune response is not simply boosted, but re-primed against evolving strains. We refer to this as seasonal immunization for individuals who have previously received a COVID vaccine. For young children under 5 years who have not yet received any COVID vaccine, the primary immunization series consists of either two doses of Moderna’s vaccine (given 4 – 8 weeks apart) or three doses of Pfizer-BioNTech’s vaccine (second dose administered 3 – 8 weeks after the first, and the third dose given at least 8 weeks after the second).

So let’s delve into some data and see what makes sense for people with different risk profiles based on what we know of the virus and vaccines today.

Since the pandemic, people’s experience of COVID-19 is getting milder. Hospitalization rates have been declining across all age groups in the US due to factors such as increased population immunity through natural encounters with COVID-19 and repeated vaccinations. We don’t know if the virus is also becoming milder. It would be good to know since that tells us whether babies are facing a milder virus than the one that first came out in 2019, or whether the virus is just as bad for them and we need to protect them with vaccines until they build up immunity and get healthy enough to withstand infections. But we just don’t know yet. 

But the data below show that babies and children under 5 still have very high hospitalization rates. And given that the FDA seems most concerned with the possibility that COVID-19 vaccines cause myocarditis (at a very low rate), it’s notable that children under 5 show no evidence of this side effect. So we think that the FDA is being too conservative by not approving and recommending COVID vaccination for this population.

When we look at the 5 – 65 healthy population, we see where the FDA is coming from. COVID-19 is so much milder in these patients, and yet there is some evidence of potential myocarditis, particularly in males aged 12 – 29. 

TABLE 1:
Hospitalization rates (≥1 night stay; rate per million) have been declining in all age groups year-over-year, but children 5 years still face risk comparable to older adults:

Source: COVID-NET interactive dashboard. Note that we show actual data for October to May but, because there aren’t yet data for May-September 2024, we extrapolated 2024 – 25 data to estimate how they are trending relative to past years.

As such, we think it is reasonable for the FDA to ask for updated trial data to better understand the benefit-risk of COVID-19 vaccines in populations that now face minimal risk (e.g., healthy young adults).

But we believe the new guidelines are too narrow, leaving young children vulnerable. If it were up to us and we were deciding on how to best use vaccines, we would have approved and recommended them for healthy children under 5 years of age and to pregnant women (which protects infants under 6 months who are ineligible for vaccination due to their immature immune systems).

Our view is supported by several key points:

1. Hospitalization rates remain high in children under 5 years, especially for those younger than 2 years.
2. Risk conditions are harder to identify in young children, and most COVID-19-related hospitalizations occur in children without identified risk conditions (i.e., healthy but for their COVID-19).
3. Myocarditis has not been observed in children under 5 years of age and is extremely rare in women of childbearing age.

TABLE 2:
Most hospitalizations in young kids occur in those without recognized risk factors while 5% of hospitalized older adults do not have an underlying condition:

Source:  June 2025 ACIP meeting, CDC NHIS, UCLA Health, CDC Asthma Statistics, CDC Trends in Multiple Chronic Conditions

Before analyzing the data, it’s important to acknowledge a potential limitation: the COVID-NET case definition may lead to an overestimation of hospitalizations due to COVID. It includes individuals who tested positive for COVID within 14 days before or during hospitalization, meaning some may have a co-infection with RSV or flu or been admitted for unrelated reasons and incidentally tested positive. However, cases included in this repository are reviewed by state and local health officials, who examine medical charts for co-diagnoses, such as pneumonia and respiratory failure, which leads us to believe that these data likely support COVID as the primary cause of hospitalization. Supporting this, a recent study from England covering data from 2020 – 2023 found that ~76% of pediatric COVID hospitalizations were primarily due to the virus.

Data on COVID co-infections in children during recent seasons is sparse, but analyses of COVID cases through 2022 suggest that co-infections with RSV and flu co-infections were relatively rare. One study in the Netherlands from the 2022 – 2023 season observed that 11% of COVID cases also tested positive for another respiratory virus, most commonly rhinovirus (6.8%), seasonal coronaviruses (1.7%), and adenovirus (0.8%). One might assume that co-infections were lower in the early years after COVID-19’s emergence because flu and RSV circulation was lower than normal, but a study in New York during the 2022 resurgence of RSV and flu showed that co-infection with RSV occurred 65% less often than expected based on RSV incidence alone, and co-infection with flu occurred 85% less often than expected, indicating that co-infection with these viruses occurs in a minority of COVID-19 hospitalizations. It is also important to note that co-infections tend to be associated with more severe illness in young children. For example, children co-infection with flu and COVID had worse outcomes than those only infected with flu, including higher rates of mechanical ventilation (13% vs 4%) and BiPAP/​CPAP support (16% vs 6%). Thus, even if flu or RSV were the primary reason for a child to be hospitalized, COVID is not an innocent bystander in the co-infection scenario.

We also surveyed approximately 75 practicing emergency medicine physicians. We wanted to know whether hospitalizations reported as being due to COVID-19 actually were due to COVID-19. About 30 had not managed any COVID hospitalizations in children 5 years old this year and were excluded from the full survey. Among the remaining respondents, physicians estimated that roughly half of COVID-positive hospitalizations are primarily due to COVID (~29% of COVID-positive cases they had seen in January through May of this year were co-infections with RSV or flu). Of the hospitalizations driven by COVID, around 50% occurred in children without underlying conditions. Notably, these otherwise healthy children were still admitted to the ICU (~9%) and diagnosed with severe pneumonia (~25%). Although these rates were lower than in children with ≥1 underlying condition (~27% ICU admission and ~35% severe pneumonia) and the data are not a large enough sample size to make definitive epidemiological conclusions, the responses highlight that COVID really does seem to be a contributing factor in many or most of the pediatric hospitalizations and ICU admissions attributed to COVID.

So our survey left us believing the COVID-NET data, and so our primary analysis is based on the COVID-NET data. However, just to be extra conservative, we also redid the analysis by cutting hospitalizations and ICU admissions due to COVID-19 in half. That 50% discount is meant to help account for any co-infections or misdiagnoses that might make COVID-19 sound worse than it might be. It’s our way of saying ​“ok, maybe the reported data exaggerate the significance of COVID-19 by a factor of two”.

As you’ll see, our conclusions don’t change. Even if vaccines remain modestly effective (20 – 30% reductions in severe COVID-19 cases), COVID-19 remains a serious enough threat to children under 5 and vaccines remain safe enough that immunizing young kids and pregnant women (to protect newborns) still makes sense to us.

TABLE 3:
Using the data presented in Tables 1 and 2, we estimated the hospitalization rate due to COVID in otherwise healthy individuals (i.e., no underlying risk conditions) per million healthy individuals in each age group (≥1 night stay).

Source: COVID-NET interactive dashboard, June 2025 ACIP meeting

TABLE 4:
Using the data presented in Tables 1 and 2, we estimated the hospitalization rate in otherwise healthy individuals (i.e., no underlying risk conditions) in each age group (≥1 night stay) and present the expected number of hospitalizations that would be avoided per million healthy people immunized with a vaccine that theoretically reduces hospitalizations by 30%:

Source: COVID-NET interactive dashboard, June 2025 ACIP meeting

TABLE 5:
Using the above data from Table 4 and ICU admission rates for COVID-19 hospitalizations in each age group (range of 14 – 28%, with ~22% for children 12 years old), we estimated the expected number of ICU admissions that would be avoided in healthy individuals (i.e., completely lack risk factors) per million people immunized with a vaccine that theoretically reduces hospitalizations by 30%:

Source: June 2025 ACIP meeting (children and adult ICU admission rates)

From the analysis in Table 5, we think it is clear that, if a vaccine were even 30% effective, vaccinating pregnant women and children under 5 years of age would provide immense benefits to individuals without risk factors. Indeed, vaccinating pregnant women (since children under 6 months are not eligible to receive the vaccines) has been shown to reduce hospitalizations by 35% for infants 6 months and 54% for infants 3 months. Additionally, myocarditis has never been observed in children under 5 (see Table 6 below), so there is really only the vaccine’s benefit to consider for them. 

Alternatively, in adolescents and younger adults, we must recognize that the benefit of routine vaccination has diminished over time, and a vaccine with 30% efficacy would only prevent 2 ICU admissions per million doses in healthy individuals ages 5 – 49. We should also consider the risk of myocarditis in this group.

TABLE 6:
Myocarditis rates per million vaccine doses in 2021 – 2022 (data as of June 2022):

Source: June 2022 VRBPAC (weighted averages from data on slide 11). These data came from cases reported to the Vaccine Adverse Event Reporting System (VAERS) that were verified by the CDC through medical record review and healthcare provider interviews. We have grayed the portions of the data that are likely not even relevant anymore since these are rates of myocarditis that only occur from two shots of the vaccine in people getting immunized for the first time; that was relevant during the pandemic but at this point just about anyone over 5 years old is contemplating only a single seasonal immunization. We keep the grayed out data only for reference and not because we think those data are informative to anyone’s decision-making today.

The CDC has noted that vaccine-associated myocarditis has not been observed since the 2021 – 2022 season. However, the FDA has recently stated that in 2023 – 24 the rate of myocarditis associated with vaccination was 8 cases per million doses in individuals 6 months through 64 years, and ~25 cases per million doses in males 12 – 25 years – though, these figures are based on insurance claims data and are likely overestimates given patients can be coded with a myocarditis diagnosis before a full clinical workup (which is required for definitive diagnosis) is even completed. To be conservative, we assume that 2022 myocarditis data remain relevant through to today, in which case the benefit-risk of seasonal immunizations appears unfavorable for healthy individuals age 5 – 50 given their low rate of hospitalization due to COVID-19 and higher rate of myocarditis. Myocarditis cases generally resolve quickly, with ~80% of vaccine-associated myocarditis cases considered resolved within 3 months by the patient’s cardiologist or other healthcare provider. However, cardiac scarring has been observed by MRI, and this can result in a rare but serious arrhythmia.

FIGURE 1:
Plot of number of ICU admissions avoided per million healthy (no risk factors) people immunized (bars) and rate of myocarditis (line), by age group. It still seems prudent to vaccinate young children against COVID-19:

Figure 1. This figure shows how many otherwise healthy people would be spared from an ICU admission due to COVID-19 per million healthy people immunized (green bars) if we assume that the vaccine only cuts ICU admission risk by 20% (dark green) or 30% (light green). The red line shows the rate of myocarditis per million shots. The evidence suggests that it’s well worth vaccinating healthy pregnant women (to protect newborns) and young children, for whom myocarditis from COVID-19 vaccination is extremely rare. The calculus is less clear for older children and younger adults, making it understandable why the FDA would like more data.

In Figure 1, we calculated ICU admissions avoided based on an assumption of the vaccine being only 20% or 30% effective. These are placeholder values, but they aren’t unreasonable guesses as to how effective vaccines are likely to still be. Let’s look at recent data. An analysis for the 2024 – 25 season found that in immunocompetent adults aged 65 and older, vaccine efficacy was 32 – 53% against hospitalization and 30 – 36% against emergency department or urgent care (ED/UC) visits, depending on how long it had been since vaccination. In adults aged 18 – 64, efficacy at reducing hospitalization risk couldn’t be estimated due to low vaccination and hospitalization rates. However efficacy against ED/UC visits was 39% within 7 – 59 days post-vaccination, declined to 32% within 7 – 119 days, and dropped further to 20% within 60 – 119 days. 

For children, efficacy data against hospitalization for the current season is unavailable. Efficacy against ED/UC visits 7 – 179 days after vaccination was 79% for children aged 9 months to 4 years and 57% for those aged 5 – 17 years.

Now, remember how we promised to do an extra conservative analysis where we assumed that, due to co-infections and other incidental COVID-positive cases, COVID-19 was the primary driver for only half the hospitalizations and ICU visits blamed on the virus in the COVID-NET database? We did that in Table 7.

Even still, COVID-19 looks dangerous enough to young children that vaccinating them and pregnant women to protect newborns makes sense to us, especially when we consider how safe the vaccine is in kids and women of child-bearing age.

TABLE 7:
Using our above analysis for number of ICU admissions avoided per million healthy individuals immunized assuming a vaccine efficacy of 30% (Table 3), we applied a 50% cut to account for ICU admissions that are not driven by COVID (for all ages):

Source:  June 2025 ACIP meeting (children and adult ICU admission rates)

Table 7 shows that, in healthy kids 2 – 4 years old, immunization would only avert 1.7 ICU admissions per million in our conservative scenario. While this may seem like a modest benefit, please remember that the vaccine also prevents hospitalization. Since Table 4 showed that the vaccine would cut an estimated 16.4 hospitalizations in the healthy 2 – 4 year-old category, half that number is 8.2 kids. So even in this extra conservative scenario, we are talking about cutting 8.2 hospitalizations and 1.7 ICU admissions per million immunizations with a vaccine that doesn’t show any rate of myocarditis in kids this age. To us, that supports immunizing kids.

While skepticism around COVID vaccination and the politicization of the topic persists, particularly around benefit for children, our own conversations and surveys with infectious disease and pediatric ICU physicians suggest a strong degree of trust in the vaccine among those most familiar with the risks of both infection and vaccination. These physicians were clear with us: they have vaccinated their own children and believe the majority of their colleagues have as well. This reinforces our view that vaccinating children under 5 and pregnant women to protect newborns is prudent. Even if a vaccine is no longer formally approved for a specific group, individuals should still be able to receive it off-label if they and their physicians believe the benefit is clear.

Does rethinking COVID vaccination read through to flu vaccines?

In short, no. But given the way we have analyzed benefits and risks of COVID vaccines and shown that they make more sense for some ages than others, one might wonder why we still have universal flu vaccination. 

The short answer is that flu is very different from COVID-19 and flu shots make sense for all ages. 

We believe that flu represents a different set of challenges compared to COVID-19 which make it logical to try to vaccinate as much of the population as possible every year:

1. The risk of serious adverse events does not exist for flu vaccines in the same way as for the COVID-19 vaccines. Flu vaccines are recognized as extremely safe with a tiny risk of anaphylaxis (like any vaccine) and very low Guillain-Barré Syndrome (GBS) rate (1−2 cases per million doses). And it’s important to note that flu infection is associated with increased GBS risk, too. Meanwhile, as we saw in Table 6, COVID-19 vaccines are suspected of causing myocarditis at a rate of over 50 cases per million in adolescent males. And while we used to think that COVID-19 infection also caused myocarditis, it now looks like many of these cases were a misdiagnosis due to troponin elevations caused by general muscle damage (e.g., maybe due to strenuous coughing).
2. Children act as reservoirs for flu but not so much for COVID-19. School-age children are efficient transmitters of flu and shed virus for longer than adults, making them key drivers of flu transmission, so it makes sense to try to reduce flu in kids. Meanwhile, nasopharyngeal shedding of SARS-CoV‑2 is shorter in children than adults.
3. Flu can be deadly in individuals without underlying conditions. Though the majority of flu hospitalizations in children aged 5 – 17 are among individuals with underlying conditions (~72% 2011-12 through 2022 – 23), most hospitalized children under 5 years old did not have a risk condition. Additionally flu mortality is not closely associated with underlying conditions – e.g., in 2010-11, 49% of pediatric deaths due to flu were in children without known conditions.
4. An estimated 80% of deaths in the H1N1 pandemic occurred in people 65 years old, cementing the public need for a universal flu booster recommendation. Compare that to the COVID-19 pandemic, in which most hospitalizations and deaths occurred in older adults (note: while kids under 5 are also at high risk, there just aren’t as many of them, so the overall numbers of hospitalizations due to COVID-19 skew much older).
5. Given the above considerations, experts argued that a universal flu vaccine recommendation would broadly expand access and reduce the public health impact of flu. The first recommendation for flu vaccination came in the 1960s for adults ≥65 years old and younger individuals with underlying conditions. In 2002, the ACIP recommended that healthy children aged 6 – 23 months and their household contacts receive the shot, as evidence showed high rates of hospitalizations and complications in this age group, even if they are otherwise healthy. The recommended age gradually expanded upward to 5 years in 2006, all children 6 months to 18 years in 2008 to reduce community transmission, and finally a universal recommendation in 2010 following the H1N1 pandemic to fully expand access.

Final thoughts

We think that if the current FDA were simply against vaccines, it might have rolled back access to flu vaccines, but it hasn’t and so we think that the FDA’s position on COVID-19 immunization is rooted in its benefit-risk assessment of these specific vaccines. This gives us some hope that we won’t see the FDA start to broadly revise labels of other vaccines. 

So if healthy adults under 65 elect not to get a seasonal COVID-19 immunization, per the FDA’s labeling of the vaccines, we get it. The risks from the virus are low and the risk of myocarditis from the vaccine could be elevated enough to make one hesitate.

Still, we can’t help but conclude that the FDA was too conservative in excluding healthy children under 5 from COVID-19 vaccine labels. From what we can see, the available data suggest that it still makes sense to immunize them (and pregnant women to protect newborns). The available data suggest they face what seem like high risks from the virus and virtually no risk that we can see from the vaccine.

So we hope this analysis helps you and your doctor exercise judgment when deciding whether a COVID-19 vaccine is right for you and your kids.