The scale of the problem the vaccine addresses is staggering. In 2019 alone, RSV was estimated to cause approximately 33 million lower respiratory tract infections in children under five worldwide, resulting in 3.6 million hospitalisations and roughly 101,000 deaths. The burden falls hardest on the youngest: infants aged zero to six months accounted for an estimated 6.6 million illnesses, 1.4 million hospitalisations and 46,000 deaths in that single year. The World Health Organization estimates that nearly half of all RSV deaths in children under five occur in infants younger than six months, a window in which their airways are narrow, their immune systems immature, and their lungs still developing. RSV vaccine development has been pursued since the virus’s discovery in 1956, but the field suffered a catastrophic setback in the 1960s when an early candidate vaccine not only failed to protect but enhanced disease in vaccinated children. Only recently, with the elucidation of the prefusion F protein structure, the form of the viral surface protein against which the most potent neutralising antibodies are directed, did an effective vaccine become technically feasible.

The new review updates and extends an earlier systematic review, incorporating evidence published through June 2025. The researchers searched Medline, the Cochrane Library, Embase, ClinicalTrials.gov and the International Clinical Trials Registry Platform, screening titles, abstracts and full texts with at least two independent reviewers at each stage. They assessed the risk of bias using the Cochrane Risk of Bias 2 tool for randomised controlled trials and the ROBINS-I tool for non-randomised studies of interventions, and graded the certainty of the evidence with the GRADE framework. For the meta-analyses they employed a random-effects model, which permits the true effect size to vary across studies and yields a more conservative estimate when heterogeneity is present, calculating risk ratios and pooled effect estimates with 95 percent confidence intervals using the Cochran-Mantel-Haenszel method. The final evidence base for maternal vaccination comprised three randomised controlled trials, two evaluating the authorised RSVpreF vaccine and one testing an unauthorised Novavax nanoparticle candidate, alongside four real-world observational studies of the authorised vaccine in routine practice.

The headline findings come from the two trials of the authorised vaccine, which together enrolled 7,656 participants across 18 countries, with the largest contributions from the United States, South Africa, Argentina and Japan. When the results were pooled, maternal immunisation with RSVpreF reduced medically attended RSV-associated lower respiratory tract infection in infants by 68 percent in the first 90 days of life, with laboratory-confirmed disease falling from around 17 cases per 1,000 infants in the placebo group to roughly 6 per 1,000 among the vaccinated. For the most severe presentations, the protection was even more striking: severe medically attended RSV lower respiratory tract illness was reduced by 84 percent, translating into approximately 8 fewer cases per 1,000 infants. Both of these estimates carried a high certainty of evidence, the highest rating available under GRADE, meaning the reviewers had substantial confidence that the true effects lie close to those observed.

Protection extended to the outcome that matters most to paediatric intensive care teams. A single large trial with 7,148 participants showed that maternal vaccination reduced hospitalisation due to RSV disease by 70 percent, again with high certainty, cutting hospitalisations from about 9 per 1,000 infants to 3 per 1,000. Meanwhile, the outcomes that provoke the most anxiety among expectant parents showed reassuring, if statistically imprecise, results. There was no evidence of any difference in RSV-related mortality, all-cause mortality or admission to intensive care between vaccinated and unvaccinated groups, though these analyses were downgraded to low certainty because the events were too rare to measure reliably. Serious adverse events related to vaccination were similarly rare, with three reported in vaccinated mothers and one in the placebo group across 7,554 participants, and neither trial recorded a single infant serious adverse event attributed to the vaccine itself.

The real-world evidence reinforces the trial data. Four non-randomised studies, three using a test-negative design and one a cohort design, evaluated the vaccine’s performance under routine conditions in Argentina, the United States and the United Kingdom, collectively covering 14,003 participants. These studies, conducted at twelve general hospitals and four paediatric hospitals in Argentina plus sites in the US and UK, estimated vaccine effectiveness against RSV-related hospitalisation at 72 percent, closely mirroring the 70 percent efficacy seen in trials. The reviewers caution, however, that the certainty of this real-world evidence is very low, owing to serious risks of bias from confounding, missing data and outcome measurement, as well as imprecision from small numbers of events. Observational designs of this kind, while invaluable for showing how interventions perform outside the tightly controlled environment of clinical trials, cannot match the internal validity of randomisation.

Safety findings on the question that has shadowed maternal RSV vaccination, namely the theoretical risk of preterm birth, were also reassuring. The original Novavax nanoparticle vaccine trial had suggested a possible preterm birth signal, which contributed to that programme’s halt, but the authorised RSVpreF vaccine showed no such association. Across the observational data, which included more than 6,385 vaccinated and 6,387 unvaccinated mother-infant pairs in a single large dataset, the risk ratio for preterm birth was 1.01, essentially identical between groups. Risks of small-for-gestational-age and large-for-gestational-age births were likewise statistically indistinguishable, with risk ratios of 0.92 and 1.06 respectively. Solicited adverse events, primarily injection-site pain, swelling and redness, along with systemic reactions such as fever, fatigue, headache and muscle pain, occurred more frequently in vaccinated women, consistent with the expected reactogenicity of any vaccine, but unsolicited adverse events and adverse events of special interest occurred at comparable rates in both groups.

The biology underpinning the strategy is elegant. Administered during the third trimester of pregnancy, the vaccine stimulates the mother’s immune system to produce high-titre antibodies against the prefusion F protein. These immunoglobulin G antibodies cross the placenta via neonatal Fc receptor-mediated transport, arming the newborn with passive immunity precisely during the first six months of life, when the infant’s own antibody production is too weak to mount an effective defence. The 90-day efficacy window reported in the trials corresponds to this period of peak maternal antibody levels, and the concentration of RSV deaths in babies under six months makes even a few months of protection a powerful public health lever. It also complements an alternative prophylactic approach: long-acting monoclonal antibodies such as nirsevimab, which can be given directly to infants. The reviewers flag as a research priority the question of how maternal immunisation influences infants’ responses to monoclonal antibodies in subsequent pregnancies, and how the two strategies compare in effectiveness and cost.

The review is not without limitations. The literature search was restricted to English-language publications, potentially excluding relevant studies in other languages. Follow-up during the period of protection remains limited, leaving open questions about how long maternal antibodies persist and whether protection wanes differentially for infants born late in the RSV season, a concern sharpened by the disruption of classic winter seasonality during the COVID-19 pandemic. Data on rare outcomes such as RSV-related death and vaccine-related serious adverse events are sparse, and outcomes including the need for invasive ventilation, duration of hospitalisation and duration of ICU stay were not measured by any trial. The reviewers also note that no trials measured antibiotic use or the duration of protection, gaps they argue should inform future study design.

Despite these caveats, the authors conclude that the evidence is now strong enough to support policy decisions. Because high-certainty randomised data demonstrate reductions in medically attended RSV infection, severe infection and hospitalisation, and real-world studies corroborate those estimates, guideline developers may consider the evidence through a formal evidence-to-decision process. The team calls for large-scale post-marketing surveillance, longer follow-up periods, and studies spanning diverse geographical regions, especially low- and middle-income countries where the burden of RSV mortality is concentrated. International collaboration, strengthened surveillance networks and standardised monitoring protocols, they argue, will be essential as immunisation programmes expand. For now, the message to clinicians and expectant parents is clear: maternal RSV vaccination offers infants their first line of defence against a virus that has evaded vaccine science for nearly seven decades, and it does so with an impressive safety record and protection measured in lives kept out of hospital.