COVID-19 mRNA Vaccines in Pregnancy Cut Hospitalization Risk in Four-Country Study

When COVID-19 vaccines first rolled out, pregnant women were largely left out of the original clinical trials, leaving them and their doctors to make vaccination decisions without direct evidence from randomized studies. Now, one of the most rigorous real-world analyses to date has filled part of that gap. Drawing on health records from more than 60,000 pregnant women across the United Kingdom, Sweden, Spain, and Norway, researchers have emulated two hypothetical clinical trials to measure how well mRNA COVID-19 vaccines actually work when given during pregnancy. The verdict: vaccination substantially lowers the risk of COVID-19-related hospitalization, while its protection against infection itself is real but considerably more modest.

The study, published in Nature Communications, was led by Núria Mercadé-Besora and Berta Raventós, who share first authorship, with senior supervision from Talita Duarte-Salles of IDIAPJGol in Barcelona and Hedvig Marie Egeland Nordeng of the University of Oslo. The team used a technique known as target trial emulation, a statistical framework that allows observational health data to be analyzed as if it came from a randomized controlled trial. Rather than simply comparing vaccinated women with unvaccinated women at the end of a study period, the method specifies a hypothetical trial design in advance, including eligibility criteria, treatment strategies, follow-up windows, and outcome definitions, and then reconstructs that trial within routine healthcare databases. This approach is designed to minimize the biases that plague conventional observational vaccine studies, such as confounding by indication and immortal time bias, where people appear to accrue follow-up time before they could plausibly have been exposed to the intervention.

The researchers emulated two distinct trials. The first compared initiation of a primary two-dose mRNA vaccination series with remaining unvaccinated. The second compared receipt of a first booster dose, a third dose overall, with remaining unboosted among women who had already completed their primary series. Crucially, the analysis used weekly matching: every week during pregnancy, women who received a first or booster dose were matched to comparable pregnant women who had not yet been vaccinated, with matching based on characteristics such as age, calendar time, prior COVID-19 history, and other health factors recorded in each country’s data. This weekly matching structure mimics the randomization point of a real trial, ensuring that vaccinated and unvaccinated women were compared at equivalent moments in their pregnancies and in the pandemic timeline.

The data came from four national and regional health databases, each mapped to a common data model that allowed identical analyses to be run in every country. In total, 62,061 pregnant women were included in the primary vaccination trial and 41,096 in the booster trial. The outcomes examined were three tiers of COVID-19 severity: documented infection, hospitalization with COVID-19, and admission to an intensive care unit. Combining results across the four databases through meta-analysis, the team found that primary vaccination during pregnancy had a vaccine effectiveness of 26.2 percent against infection, with a 95 percent confidence interval ranging from 16.2 to 35.1 percent. Against hospitalization, however, the effectiveness was markedly higher at 55.6 percent, with a confidence interval from 44.1 to 64.8 percent.

For the booster dose, the picture was somewhat different. Vaccine effectiveness against infection was estimated at 33.6 percent, with a confidence interval from 1.6 to 56.6 percent, and against hospitalization at 34.1 percent, with a confidence interval from 0.6 to 56.4 percent. These intervals are wide, meaning the booster results are consistent with a broad range of possible effects, from nearly zero to moderately strong protection. The researchers note that ICU admissions were too rare in the data to analyze at all, which is itself an encouraging signal about the overall severity of COVID-19 among vaccinated pregnant women in these populations, even if it prevents a precise estimate of the most extreme benefit.

The contrast between protection against infection and protection against hospitalization is a familiar pattern in vaccine research, but it carries particular weight in pregnancy. mRNA vaccines work by instructing cells to produce the SARS-CoV-2 spike protein, prompting an immune response that includes neutralizing antibodies and T cells. Antibody levels wane over months, and viral variants with spike mutations can partially evade neutralization, which erodes protection against infection over time. Protection against severe disease, by contrast, is more durable because it relies not only on preventing viral entry but also on the immune system’s ability to respond rapidly once infection occurs, blunting viral replication before it causes tissue damage. The finding that roughly half of COVID-19 hospitalizations among pregnant women could be averted by primary vaccination therefore aligns with the broader immunological understanding of how mRNA vaccines perform.

The heterogeneity across the four countries is another technically important aspect of the findings. The researchers report that protection against infection was more modest and heterogeneous across databases, meaning the effectiveness estimates differed from country to country. Such variation can arise from differences in testing practices, vaccination timing relative to dominant variants, background infection rates, the proportion of the population with prior infection, and how each health system records diagnoses and hospitalizations. Meta-analysis is designed to account for this variability by pooling estimates while quantifying the spread between them, but it also means a single headline number should be read as an average across different pandemic contexts rather than a universal constant. The hospitalization benefit was more consistent, suggesting that the severe-disease protection conferred by vaccination is robust across these different healthcare settings.

Why does this matter for pregnant women specifically? Pregnancy is associated with altered immune function and physiological changes that increase the risk of severe COVID-19, and infected pregnant women face higher risks of complications compared with non-pregnant peers of the same age. Yet because pregnant women were excluded from the original vaccine trials, early real-world evidence on effectiveness in this group remained limited, as the authors themselves state. Previous studies, including a national observational study in Israel on a third dose during pregnancy, have suggested benefit, but large multi-country analyses with rigorous trial emulation designs have been scarce. By harmonizing data across four European countries and applying a consistent methodological framework, this study provides some of the strongest observational evidence available that mRNA vaccination during pregnancy works as intended against severe outcomes.

The study also carries a human dimension. The authors acknowledge the significant scientific contributions of Professor Angela Lupattelli, who passed away during the course of the research. Lupattelli played an important role in the study’s design and contributed substantially to the handling of Norwegian data, and the team’s tribute underscores the collaborative, multi-year effort required to bring together health records from four countries under a single analytical framework. The work drew on infrastructure developed for the OMOP common data model, including a pregnancy extension table that allowed researchers to identify pregnancies and gestational ages consistently across databases, a technical achievement that made the weekly matching design feasible.

For clinicians and public health authorities, the findings offer a nuanced message. Primary mRNA vaccination during pregnancy is associated with a lower risk of COVID-19-related hospitalization, an outcome of central concern given the elevated severity risks in pregnancy. Protection against infection is more modest and varies across settings, which means vaccination should not be framed as a guarantee against catching the virus, but rather as a substantial shield against its most serious consequences. Booster effectiveness estimates are less certain, with wide confidence intervals that leave open questions about the magnitude of additional benefit from a third dose during pregnancy. As variants continue to evolve and antibody protection wanes, the study’s emulation framework provides a template for ongoing monitoring, allowing researchers to update effectiveness estimates as new doses, new variants, and new pandemic phases emerge, all without waiting for the randomized trials that pregnant women were never part of in the first place.

Subject of Research: Target trial emulation of effectiveness of COVID-19 mRNA vaccination during pregnancy in four European countries

Article Title: Target trial emulation of effectiveness of COVID-19 mRNA vaccination during pregnancy in four European countries

Article References: Mercadé-Besora, N., Raventós, B., Trinh, N. T., Li, H., Hayati, S., Giuliodori, A., Burkard, T., Burn, E., Català, M., Prieto-Alhambra, D., Morales, D. R., Nyberg, F., Marie Egeland Nordeng, H., & Duarte-Salles, T. (2026). Target trial emulation of effectiveness of COVID-19 mRNA vaccination during pregnancy in four European countries. Nature Communications. https://doi.org/10.1038/s41467-026-78469-x

Image Credits: AI Generated

DOI: 10.1038/s41467-026-78469-x

Keywords: COVID-19, mRNA vaccines, pregnancy, vaccine effectiveness, target trial emulation, epidemiology, Europe, Nature Communications, hospitalization, public health, Target, trial