The vaccine in question is built on the prefusion F protein, the conformation of the viral surface glycoprotein that RSV displays when it is about to fuse with a host cell. This prefusion form exposes the key antigenic sites that elicit the most potent neutralizing antibodies, which is why it has become the preferred scaffold for both RSV vaccines and antibody therapeutics. When administered during pregnancy, typically between 24 and 36 weeks of gestation, the vaccine stimulates the mother to produce antibodies that cross the placenta, arming the newborn with passive protection during the most vulnerable first months of life. The strategy has proven effective, but it has also been shadowed by safety concerns that have shaped the entire regulatory debate around maternal RSV immunization.

Those concerns are not trivial. In earlier clinical testing, some trials of maternal RSV vaccines were halted after interim data suggested an excess of preterm deliveries among vaccinated mothers. A previous rapid review pooling all tested maternal RSV vaccines, including candidates that never reached the market, reported an association with preterm birth. Because prematurity remains the leading cause of death in children under five worldwide, and because survivors face elevated odds of lifelong complications ranging from respiratory disease to neurodevelopmental impairment, even a modest safety signal demands rigorous scrutiny. At the same time, RSV itself kills, and the alternative preventive option, the monoclonal antibody nirsevimab, is expensive enough that maternal vaccination remains the primary strategy in much of the world. Policymakers therefore need a clear answer about where the true risk lies.

To provide that answer, a Finnish research team conducted a systematic review and meta-analysis following the PRISMA reporting guidelines. The authors searched PubMed, Scopus, and Web of Science on August 1, 2025, with no language or date restrictions, and screened the resulting records in Covidence. Their inclusion criteria were deliberately narrow: they considered only studies of the market-approved prefusion F vaccines, RSVPreF and RSVPreF3, given to pregnant individuals, and only studies reporting preterm birth, defined as any delivery before 37 weeks and zero days of gestation. They excluded trials of the non-approved vaccine formulations precisely because those candidates had already been withdrawn over preterm birth signals and could confound the safety picture of the products actually in clinical use.

The screening process winnowed 404 identified studies down to just seven that met all criteria. Three were randomized controlled trials, together encompassing 12,833 births, and all were multinational, conducted across as many as 24 countries. Four were observational studies, retrospective or prospective, covering 4,245 births, and all four came from the United States. Vaccination typically occurred between gestational weeks 24 and 36. Where birthweight was reported, in four of the seven studies, the means were comparable between vaccine and control groups, ranging from 3.2 to 3.34 kilograms in vaccinated pregnancies against 3.15 to 3.42 kilograms in controls. Risk of bias was low in two of the randomized trials and raised some concerns in one, while the observational studies were rated moderate in three cases and serious in one, assessed with the Cochrane RoB 2.0 and ROBINS-I tools respectively.

The pooled results tell a story of two evidence streams in tension. In the randomized trials alone, the meta-analysis found a statistically significant increase in preterm birth risk among vaccinated mothers, with a relative risk of 1.26 and a 95 percent confidence interval of 1.08 to 1.46, and notably zero heterogeneity across trials. Yet when the researchers turned to the observational data, the signal vanished and even reversed direction: the relative risk was 0.78, with a confidence interval spanning 0.46 to 1.32 and substantial heterogeneity of 63 percent. Combining all seven studies yielded no overall association between maternal RSV vaccination and preterm birth, with a relative risk of 0.96 and a wide confidence interval of 0.68 to 1.38, accompanied by 80 percent heterogeneity. The certainty of this evidence was rated as low using the GRADE framework, downgraded for risk of bias and imprecision, since the interval includes both meaningful benefit and meaningful harm.

The authors attribute the discordance between trial and observational findings primarily to differences in the lower gestational age threshold for vaccination. The randomized trials enrolled women from 24 or 28 weeks of pregnancy, whereas every observational study administered the vaccine only from 32 weeks onward. Because the risk of spontaneous preterm delivery naturally concentrates in the window shortly after vaccination in earlier-gestation cohorts, a lower enrollment threshold is more likely to capture deliveries that occur soon after immunization, a pattern that can be misread as vaccine-caused when the two events are merely close in time. The geographic restriction of the observational data to the United States, compared with the multinational scope of the trials, adds a second layer of difference in populations, healthcare systems, and coding practices that may further explain the divergence.

On the secondary outcome, the analysis rested on only two studies reporting neonatal intensive care unit admissions, covering 3,620 births. The pooled estimate actually favored vaccination, with a relative risk of 0.74, meaning roughly 24 fewer NICU admissions per 1,000 infants, but the confidence interval of 0.34 to 1.61 was so wide that no firm conclusion is possible. The certainty of evidence here was rated very low, reflecting the observational origin of the data, small sample, high heterogeneity of 92 percent, and risk of bias. Birthweight and other neonatal morbidities were prespecified as additional outcomes, but inconsistent reporting across the included studies prevented formal meta-analysis, underscoring how early the evidence base for neonatal safety outcomes beyond preterm birth remains.

These findings arrive at a consequential regulatory moment. The Global Advisory Committee on Vaccine Safety has concluded that, despite the preterm birth concerns, the benefits of maternal RSV vaccination outweighed the risks in 98 percent of simulations when the vaccine is given between 27 and 36 weeks of gestation. On that basis, the WHO Strategic Advisory Group of Experts issued recommendations in September 2024 that were subsequently endorsed by the WHO itself. Real-world effectiveness data from the 2024-25 season in England and Scotland have already shown highly favorable outcomes, confirming that the vaccine works when deployed at population scale. The authors note, however, that hybrid immunization strategies deserve continued consideration, because nirsevimab offers superior protection, and the preterm birth signal observed within the randomized trials has not been conclusively resolved.

The review’s authors are candid about its limits. With only seven studies, publication bias could not be properly assessed; ongoing trial registries were not searched; and rare vaccine-related adverse events cannot be detected without large-scale nationwide surveillance. The subgroup signal in randomized trials, a 26 percent relative increase in preterm birth with tight confidence intervals, stands as the single most important caveat in an otherwise reassuring picture, and the low-to-very-low certainty ratings mean the true effect could plausibly fall on either side of harm or benefit. What the analysis establishes is that the totality of current evidence on the approved prefusion F vaccines does not confirm an overall association with preterm birth, and that NICU admissions were numerically lower among vaccinated infants. As maternal RSV vaccination scales up globally, the authors argue, continued pharmacovigilance focused on gestational outcomes, alongside studies of broader neonatal morbidity, is not optional but essential. The vaccine’s promise for protecting newborns from one of childhood’s deadliest pathogens is real; so is the responsibility to keep watching.

Subject of Research: Safety of maternal RSV prefusion F vaccination regarding preterm birth and neonatal intensive care admission

Article Title: RSV‐Pre‐F Vaccination During Pregnancy and Neonatal Outcomes—A Systematic Review and Meta‐Analysis

Article References: Leskinen, N., Haapanen, M., & Kuitunen, I. (2026). RSV‐Pre‐F Vaccination During Pregnancy and Neonatal Outcomes—A Systematic Review and Meta‐Analysis. Immunity, Inflammation and Disease, 14(9), Article e70510. https://doi.org/10.1002/iid3.70510

Image Credits: AI Generated

DOI: 10.1002/iid3.70510

Keywords: RSV, maternal vaccination, preterm birth, prefusion F vaccine, meta-analysis, systematic review, neonatal outcomes, NICU admission, nirsevimab, vaccine safety, randomized controlled trials, public health