On May 27, Robert F. Kennedy Jr., Secretary of Health and Human Services (HHS) announced by video on X that COVID vaccination was no longer recommended by the Centers for Disease Control and Prevention (CDC) for healthy children and healthy pregnant women.
Vaccine schedules are fluid and subject to change based on ongoing risk/benefit assessment. Updated safety information, changes in the epidemiology of the infection, and/or change in the efficacy of the vaccine can all influence the official recommendations. So a change in recommendations is expected from time to time, and may be reasonable.
An explanation and appeal to “gold standard science”
Subsequent to the announcement on X, HHS sent a FAQ document to Congress. The justification for the change in vaccine schedule is explained:
The Secretary is committed to gold standard science as outlined in the President’s May 23, 2025 executive order. As a result, HHS is reviewing the science of risk and benefit of public health recommendations.
So far, so good. Who could be opposed to making policy based on gold standard science? This post is not to argue whether or not healthy children or healthy pregnant women should remain on the recommended schedule; I wish to assess the quality and persuasiveness of the evidence provided by HHS to support their decision.
Much of what is written below is covered elsewhere. Here is an in depth critique of the HHS FAQ to congress, including comments from authors of some of the cited papers.
Vaccination for healthy children
Per the HHS letter to congress:
…post-marketing studies have shown the vaccine to have serious adverse effects, such as an increased risk of myocarditis and pericarditis. This is highest in young males. For instance:
- Myocarditis reports in VAERS after COVID-19 vaccination in 2021 was 223 times higher than the average of all vaccines combined for the past 30 years–representing a 2500% increase.v
a. Demographic data revealed that myocarditis occurred most often in male children, with 76% of cases resulting in emergency care and hospitalization.
- A study from the UK of over 1.7 million children between the ages of 5 and 15 revealed that cases of myo and pericarditis were found exclusively in those that received the COVID-19 vaccine. vi
- A study from Japan showed that COVID-19 vaccination was significantly associated with the onset of myo and pericarditis. These occurred most often in males under 30.vii
The association between COVID vaccine and myocarditis is real. Myocarditis following COVID vaccine is rare, and usually mild, but occasionally serious. It is known that the risk is highest among boys and young men, and is greatest after the second dose.
Citations?
The first citation provided is:
The authors of this paper reviewed data from the Vaccine Event Reporting System (VAERS). They noted that reports of myocarditis after COVID-19 vaccination were 223 times higher than reporting of myocarditis after all prior vaccines. The reporting was highest for young men, after the second dose.
The VAERS database, by its very nature, is poorly suited for determining causality and quantifying risk. VAERS is intended to be an early detection system for adverse events after vaccination. Anyone can submit a report to VAERS, so the reports vary in quality, completeness, and accuracy. The low threshold for reporting to VAERS is intentional, but makes for a very noisy database.
Separating signal from noise requires diligent work. VAERS is monitored for patterns with the purpose of identifying previously unknown adverse reactions. Investigators must explore other sources of data and other analyses that are much better suited for determining if the association is real and causal. VAERS data are also unsuitable for calculating the frequency (incidence) of adverse events after vaccination.
This VAERS analysis cited in the HHS document is a dubious choice. It was published in 2024, long after COVID vaccine related myocarditis was recognized and already the subject of much more revealing research. More reliable estimates of the incidence of myocarditis, and outcomes of myocarditis after COVID vaccination had been published from analyses of numerous databases much more suitable than VAERS.
In addition, the Journal that published the article noted an “Expression of Concern”:
“The Editor and the publisher were alerted to potential issues with the research methodology and conclusions and author conflicts of interest. Sage has contacted the authors of this article on this matter, and an investigation is underway.”
This article, with alarming, but unreliable statistics was cited. Meanwhile, an extensive body of research, systematic reviews, and meta-analyses with much more robust (but less alarming) statistics was ignored.
One large meta-analysis (not cited by HHS) found the incidence of myocarditis in adolescents to be 4.5 cases per 100,000 vaccine doses overall, and 8.58 per 100,000 in males.
The next article cited by HHS is this:
Prominently displayed on the first page, directly under the Title and list of authors is this disclaimer:
This article is a preprint and has not been peer-reviewed. It reports new medical research that has yet to be evaluated and so should not be used to guide clinical practice.
HHS has used the article to inform clinical practice despite an explicit disclaimer cautioning against such use.
The third, and final publication in support of the elimination of healthy children from the vaccine schedule is this one:
This is similar to the VAERS analysis by Rose et al. It is an analysis of cases of myocarditis and pericarditis following COVID vaccination reported to the Japanese Adverse Drug Event Report (JADER) database. Similar to VAERS, this is a passive database, but it is more inclusive. VAERS collects adverse event reports after vaccination, while the JADER collects adverse events related to a broader range of pharmaceutical products. The limitations of this study are similar to those described for Rose et al. There are far more reliable data for the incidence of adverse effects after COVID vaccination.
Vaccination for pregnant women
If you would like a more pithy, more entertaining breakdown of the portion of the HHS report dealing with pregnancy, this video by Dr. Noc will do the trick.
What evidence does HHS offer to justify the change in policy?
A number of studies in pregnant women showed higher rates of fetal loss if vaccination was received before 20 weeks of pregnancy.viii Another showed statistically significant increases in preterm birth.ix Yet another study showed an increase in placental blood clotting in pregnant mothers who took the vaccine.x
Citations?
The first cited article:
This is a case-control study from Ontario Canada, comparing rates of miscarriage for vaccinated and unvaccinated women. The article is cited in support of the statement that: “…studies in pregnant women showed higher rates of fetal loss if vaccination was received before 20 weeks of pregnancy.” What the paper actually reported was no difference in miscarriages between vaccinated and unvaccinated pregnant women. In the Interpretation portion of the paper the authors stated: “The benefits of vaccinating pregnant women outweigh any potential adverse risks…”
The second cited study is this one:
This study is cited in support of the assertion that COVID vaccines had increased preterm birth. This one gets a little complicated. In the overall analysis they found no difference in preterm births between the vaccinated and the unvaccinated groups. In one of many secondary analyses they did find a small, but statistically significant excess of preterm births in the vaccinated group. It is important to recognize that this finding is an outlier. The paper was published in 2022. Subsequently, there have been larger, more powerful studies as well as meta-analyses and systematic reviews. These more powerful studies have not replicated the finding of higher preterm birth in vaccinated women. It is curious that the more powerful, but safety affirming studies were not included in the report from HHS to Congress.
The final study is the most perplexing. Per the HHS document: “Yet another study showed an increase in placental blood clotting in pregnant mothers who took the vaccine.”
This is the relevant reference:
The curious thing is the article has nothing to do with pregnancy and had no mentions of placental blood clotting. I did individual word searches for “pregnant”, “pregnancy,” “placenta,” and “placental.” The study authors have affirmed that HHS’s assertion of “placental blood clotting” is nowhere to be found in the publication.
The phantom reference
The HHS report is formatted such that citations are enumerated with lower case Roman numerals such as i, ii, iii, iv….x. Between reference viii and ix there is this un-numbered, un-cited article:
This was a retrospective cohort study examining outcomes of vaccinated and unvaccinated women receiving in-vitro fertilization. They found no difference in any outcome between the groups.
I don’t know if there was intent to cite this reference, or if it is an artifact of sloppy editing and proofreading. I found nothing in the article that would support eliminating vaccines for pregnant women.
Observations about the HHS correspondence to Congress
To the best of my knowledge this is the only available statement elucidating the rationale for change in vaccine recommendations. It is a flawed and problematic document.
- It is a very brief justification for a significant policy change
- The reasoning is based on poor-quality sources that are not well suited to identifying or quantifying risk, such as VAERS and JADER.
- A non-peer reviewed study is cited.
- There is a rogue un-numbered, un-cited reference.
- Decisions are informed by misinterpretation or misrepresentation of cited studies.
- Higher quality studies better suited to identifying and quantifying risk are not acknowledged.
- The benefits of the vaccines are not mentioned in assessing the public health implications of these decisions.
I have collaborated with trainees on numerous projects. If a first-year resident created a document like this as a first draft, I would chastise that resident for sloppiness and inaccuracy. I would seriously consider removing him/her from the project.
Conclusions
I am not qualified to opine about whether or not the CDC vaccine recommendations should include healthy children or pregnant women. I lack the mastery of the literature necessary to fully weigh the risks, benefits, and costs.
What I do know is that the document provided by HHS to members of Congress is not at all helpful in making an informed opinion. To the contrary, it is counter productive. The document is highly biased. It contains cherry-picked, poor-quality studies that exaggerate risks and ignores entire bodies of literature representing realistic, best-available risk assessment. The report misrepresented results of 2 studies resulting in amplification of undocumented risks. Benefits of vaccination were completely absent from the conversation.
Had the authors performed a comprehensive review and summary of the data, they might have been able to construct a persuasive argument in support of their policy decision. This was not the path they chose.
Health policy decisions should follow a good faith appraisal of the risks, benefits and costs of competing policies. It is clear to me that this was a favored policy in search of supporting evidence.
The casual sloppiness and disregard for accuracy demonstrated in this HHS communication is most alarming.