A big first this month for next generation coronavirus vaccine. The first clinical trials for a US-developed pancoronavirus vaccine have been under…
Digging Into the Evidence on Covid Vaccines in Pregnancy
Decision-making about drugs in pregnancy can be very difficult. Our illogical aversion to clinical trials in pregnancy means people making particularly complex decisions usually have little specific information about efficacy for them when a drug first becomes available. Plus, efficacy trials aren’t big enough to assess the uncommon and rare pregnancy outcomes we want to be reassured about: For that we have to wait until the drugs are in wide use. At that point, we’re swamped with a complicated research and surveillance landscape. All this makes vaccination in pregnancy an easy target for misinformation campaigns. And that’s happening now in the US.
In an interview, the FDA Commissioner dismissed a large body of evidence on the safety and efficacy of Covid vaccination in pregnancy because it wasn’t from randomized trials. He claimed that the randomized trial in pregnancy was “closed without any explanation.” (False.) Soon after, the Department of Health and Human Services circulated a document to justify their retreat from support for Covid vaccination in healthy pregnant women, misrepresenting studies to make claims of harm. I discuss those studies below, and summarize the large body of evidence that has accumulated in people, mostly for mRNA vaccines.
Bottom line: Studies with over 1.5 million Covid-vaccinated pregnant participants show vaccines are effective in pregnancy, and reduce some poor pregnancy outcomes. Extensive surveillance and study – particularly of the mRNA vaccines – has not established safety concerns for pregnancy.
I’ve broken this post into 3 sections, with the first including those recent claims from the US:
Safety surveillance
Safety surveillance is critical after drugs and vaccines are authorized. Major drug regulators require manufacturers to have plans for ongoing studies, and they monitor other research and other regulators’ findings as well. There are several types of non-randomized research of the outcomes of vaccination in pregnancy – including pregnancy outcome registries, and studies with comparison groups, like case-control studies. The range of data available is covered in the systematic review section below.
This section begins with conclusions from Europe’s drug regulator, which publishes the most detail, and includes an update from the US.
European Medicines Agency
By January 2022, the EMA had concluded that there was “growing evidence indicating that mRNA COVID-19 vaccines do not cause pregnancy complications for expectant mothers and their babies.” Monitoring of individual vaccines in pregnancy, they reported, is undertaken by their pharmacovigilance committee:
“The companies are required to provide regular updates and conduct studies to monitor the safety and effectiveness of their vaccines as used by the public. Authorities also conduct additional studies to monitor safety and effectiveness of the vaccines, including their use in pregnancy. These measures allow regulators to swiftly assess data emerging from a range of different sources and take appropriate regulatory action to protect public health if needed.”
EMA’s conclusion in 2022 included references to 12 journal publications: All of them are included in the systematic review below.
Safety updates on individual vaccines:
BNT/Pfizer vaccine (Comirnaty): Safety updates are here. The most recent posted there covered data up to December 2023, including a small amount of data from clinical trials, and analysis of the pregnancy and fetal/neonatal data from serious adverse event reporting. EMA’s conclusion: “There were no safety signals regarding use in pregnant/lactating women that emerged from the review of these cases or the medical literature.”
Moderna vaccine (Spikevax): Safety updates are here. The most recent posted again covered data up to December 2023. EMA’s conclusion: Analysis of the data “did not raise any new safety concern… The benefit-risk evaluation remains positive.”
Hipra vaccine (Bimervax): Safety updates are here. The most recent posted covered up to September 2023, and there was as yet no clinical data to report.
Another 2 vaccines have been approved in Europe more recently: from Novavax (Nuvaxovid) and Arcturus (Kostaive). There are no safety updates for these yet.
USA
A brief summary of safety was included in the CDC’s vaccine committee (ACIP) meeting on June 25, 2025 (see slides 28 and 29). They report that their own surveillance studies include over 68,000 pregnancies, and show no increased risk of serious poor outcomes, including pregnancy loss. This is supported by reference to 4 of their journal publications: All 4 are included in the systematic review below.
What about those other 4 studies cited as justification for a new government policy of not supporting Covid vaccination for healthy pregnant women? The document is online here. An expert was quoted in a news report on this document describing it as disinformation – on pregnancy, I agree with that characterization. (I didn’t check the rest.) Here’s what that report says about 3 of 4 of those cited studies:
“In two instances, the HHS memo makes claims that are actively refuted by the papers it cites to back them up. Both papers support the safety and effectiveness of covid vaccines for pregnant women. The HHS document says that another paper it cites found ‘an increase in placental blood clotting in pregnant mothers who took the vaccine.’ But the paper doesn’t contain any reference to placental blood clots or to pregnant women.”
I agree with this. The 2 cited studies that conclude the opposite of what the document claims are Velez 2024 and Dick 2022a. (Note: Dick 2022a is inaccurately cited in the document – one of that lead author’s papers is cited by surname and one by first name). The paper that does not mention pregnancy at all is Faksova 2024.
The 4th study cited is Dick 2022b, and it isn’t properly represented. This is an analysis of births at one hospital [retrospective cohort study]. A large number of patients were excluded (22%), mostly because of missing data (including timing of vaccination), and/or the patient having been vaccinated before pregnancy, or having had Covid. The outcomes they were studying were preterm birth and babies that were smaller than expected at birth.
In that study, there was no difference in the rate of preterm birth between the vaccinated and unvaccinated women. The researchers concluded that the subgroup vaccinated in the second trimester had a higher rate of preterm birth, which they said needed further study. This was a secondary subgroup analysis. Given the comparatively small number of women known to have been vaccinated in the second trimester, the large amount of missing data/excluded is a particular concern as well.
The Dick 2022b study is one of 67 investigating the risk of preterm birth after mRNA vaccination included in the systematic review in the section below: 42 studies investigate other vaccine types.
Randomized trials
I mentioned in my intro that the FDA Commissioner had publicly claimed that the randomized trial in pregnancy was “closed without any explanation,” and that this was false. He was certainly referring to the large planned trial for the BNT/Pfizer vaccine: The fact that it was terminated is brought up in anti-vaccine literature as a reason to be suspicious.
There is actually no drama there: It’s a common situation. Once the vaccine was widely available and recommended for pregnancy, the manufacturers report that it was too hard to recruit participants to a placebo-controlled trial. The original plan was to start recruiting up to 4,000 participants in the US in February 2021 – which was soon after the vaccine became available in the US. That size phase 2/3 trial was primarily intended to study adverse reactions to the vaccine and immune responses, to see if they were similar to those from trials of non-pregnant people. The trial would also have gathered useful data on common adverse events.
After a year, Pfizer amended the trial protocol, announcing that the trial was terminated “due to enrollment challenges into a placebo-controlled trial.” Planned analyses were adapted to account for the much smaller number of participants: 726 participants had been enrolled. Results were posted at ClinicalTrials.gov, including follow-up of infants to 6 months of age. Those results are included in the systematic review below in this post.
A similar fate befell the only other randomized trial of a Covid vaccine in pregnant participants that I’m aware of – for the single-shot Janssen/Johnson & Johnson viral vector vaccine. That trial was originally planned to start in the US in August 2021, aiming to recruit around 824 pregnant participants. The amended protocol for this trial reported that it also struggled with recruitment. Changes had been made to try to increase enrollment over a 10-month period, but the trial was abandoned with only 51 participants recruited. Results were also posted at ClinicalTrials.gov, but don’t appear to have been included in the systematic review below. (I’ll let the authors know.)
This situation is, unfortunately, typical for an epidemic. A review before the Covid pandemic found that of 84 emergency vaccine trials, only 8 protocols included pregnant women – and only 3 of those were trials that were completed and reported. Those authors concluded: “Pregnant people are underrepresented in vaccine clinical trials conducted during outbreaks, resulting in underreporting of pregnancy-related outcomes and a lack of protection for pregnant people and neonates from infectious diseases.” Authors of another study, this time of phase 3 and 4 vaccine trials between 2018 and 2023, found similar results: Only 6% of 400 trials included were for, or open to, pregnant participants.
Systematic reviews
To find the best data I could for this question, I looked for good quality recent systematic reviews. Good systematic reviews are research projects using very thorough processes to try to reduce bias in finding, selecting, and evaluating studies and their results. I sifted through the systematic reviews on recency and methodological quality criteria. (Notes on this process are below this post.)
The best recent review I found is the Safe in Pregnancy living systematic review of randomized and non-randomized studies maintained by an international consortium, funded by the Gates Foundation. The team searches for new eligible studies every 2 weeks. A dashboard shows the most recent analyses. It was updated while I was working on this post, and I included the analyses from June 25, 2025. At that point, they had analyzed 289 of the studies they had found (some are still in a queue for analysis), including 1,521,849 vaccinated pregnant participants (plus unvaccinated participants in control or other comparison groups).
There are some reporting issues with this dashboard – for example, some studies are missing study design categorization in the main table, the PRISMA flowchart reporting article screening isn’t complete, excluded studies aren’t listed, and neither are the studies waiting to be added to the analyses. (There are also some earlier journal publications for this review, listed in my notes below.)
There were 2 other reviews that made it onto my short list. One of those analyzed 8 case-control studies that compared mRNA vaccines with non-mRNA vaccines (Oliveira 2024). The authors concluded that mRNA vaccination was associated with a lower incidence of fetal or neonatal death (OR 0.16; 95% CI 0.08-0.33).
The other is the systematic review of 67 studies that was used by the journalist interviewing the FDA commissioner I mentioned in the intro (Fernández-García 2024). That included 723,988 vaccinated pregnant participants (plus unvaccinated participants in control or other comparison groups) from a search in January 2023. This review was not just older than the Safe in Pregnancy living systematic review. It had other drawbacks – particularly a non-reproducible and more limited search for studies – so I have not used it in this post.
Now for the results from the Safe in Pregnancy living systematic review. The major conclusions are, firstly, that Covid vaccines are similarly effective in pregnancy or not. This is summarized in a recent journal publication from this living systematic review: Ballivian 2025. And secondly, that no safety concerns have been established for mothers, fetuses, or babies. That was summarized in another earlier publication: Buekens 2024. Those basics add up to some benefits. For that data, let’s go to the recent meta-analyses.
It’s critical to keep in mind that although 289 studies is a lot, there are massive differences among these studies, and they address different issues, typically measured in different ways. Therefore, only a small proportion of the data may be suitable for any single meta-analysis. (See my post 5 Tips for Understanding Data in Meta-Analyses for more on this.) Add to that the inherent weaknesses in many of the types of studies, and the certainty in results is low to very low. They get even more uncertain if you want to drill down to particular vaccines or Covid variants, for example. Here are some outcomes to give you an indication of the effects of mRNA vaccines – they are drawn from the meta-analyses in the dashboards, and the calculations of number of studies and participants are mine:
- Reduction in symptomatic Covid-19 (Omicron) in mothers: Vaccine effectiveness 57.90% (CI: 24.54; 76.48) – based on 3 studies with 15,094 vaccinated participants.
- Reduction in severe Covid-19 or hospitalization for Covid (Omicron) in mothers: Vaccine effectiveness 71.16% (CI: 58.84; 79.79) – based on 3 studies with 15,094 vaccinated participants.
- Reduction in severe Covid-19 or hospitalization for Covid (Omicron) in infants: Vaccine effectiveness 63.41% (CI: 32.95-80.03) – based on 3 studies with 45,468 vaccinated participants.
- Reduction in preterm birth: Relative risk 0.87 (CI: 0.79; 0.96) – based on 9 studies with 305,835 vaccinated participants.
- No increase in miscarriage after vaccination: Odds ratio 0.91 (CI: 0.74-1.13) – based on 3 studies with 22,346 vaccinated participants.
- No increase in admission of baby to neonatal intensive care: Relative risk 0.95 (CI: 0.86; 1.04) – based on 5 studies with 88,509 vaccinated participants.
Finally, a tip to using the dashboard: If you find it hard to filter data, try a different browser. Vivaldi was OK, but Safari was much easier – it was too hard using it in Firefox.
Citation for Safe in Pregnancy: Tulane University, Institute for Clinical Effectiveness and Health Policy, London School of Hygiene and Tropical Medicine, University of Washington. COVID-19 Vaccines for Pregnant Persons: A Living Systematic Review and Meta-Analysis. https://www.safeinpregnancy.org/ [Accessed June 26, 2025]
You can keep up with my work at my newsletter, Living With Evidence. And I’m active on Mastodon: @hildabast@mastodon.online and less so on Bluesky (hildabast.bsky.social).
~~~~
All my Absolutely Maybe posts on Covid vaccines, beginning from March 2020, are tagged here.
All previous Covid-19 posts at Absolutely Maybe
My posts at The Atlantic and at WIRED.
Disclosures: My interest in Covid-19 vaccine trials began as a person worried about the virus, as my son was immunocompromised: I have no financial or professional interest in the vaccines. I have worked for an institute of the NIH in the past, but not one working on vaccines. More about me.
The cartoon is my own (CC BY-NC-ND license). (More cartoons at Statistically Funny.)
Notes on identifying and selecting systematic reviews
I used Epistemonikos to identify systematic reviews, limiting results to systematic reviews published from 2023, with the search terms: title:(covid-19 OR SARS-CoV-2) AND title:(pregnan* OR maternal) AND title:(vaccin*) on June 20, 2025. I found 17 systematic reviews. In addition, my regular general screening for studies on Covid vaccine identified a systematic on safety in populations of interest that included pregnancy.
I narrowed down the group by excluding reviews which did not have a search date in 2023 or later (11 reviews [a]). On quality grounds, I then excluded 2 reviews [b] which did not meet all these criteria: pre-registered protocol; search of multiple databases; showing a PRISMA flowchart; assessing the quality of included studies.
The 5 remaining reviews included 3 publications from a living systematic review project. I included the data from an update of that review on the project dashboard on June 25, 2025: This included analysis of 289 of the 299 studies found in total. The final pool of systematic reviews:
- Living systematic review, 4 versions: Buekens 2024; Ciapponi 2024; Ballivian 2025; and June 2025 update.
- Fernández-García 2024.
- Oliveira 2024.
Excluded systematic reviews:
[a] 11 reviews with search dates earlier than 2023:
PMIDs: 36007670; 36444098; 36670389; 36794075; 36800343; 36896895; 37454153; 37605626; 38432215; 40472669; and (doi) 10.2174/0126667975269133231117062158.
[b] 2 reviews for methodological reasons: PMID: 38291854; 39202263.