An RSV vaccine's second season looks weaker than its first, in three studies at once

Four papers about the same vaccine appeared in three days at the end of August: two randomised trials [s1][s2] and an editorial [s3] in NEJM Evidence on 29 August, and a Medicare cohort study [s4] with its own editorial [s5] in JAMA Internal Medicine on 31 August. Read together they answer the question of whether the bivalent RSV prefusion F vaccine prevents hospitalisation, and open a question about how long it keeps doing so.

The hospitalisation trial

The larger of the two trials was pragmatic, individually randomised and open-label, conducted in Denmark and Spain [s2]. Adults were assigned 1:1 to receive RSVpreF or no vaccine. In the first year of enrolment, participants aged 60 or older in Denmark were enrolled; in the second year, eligibility was expanded to adults 18 and older in Denmark and Spain, because fewer RSV-related hospitalisation events than anticipated had occurred in the first year [s2].

Across the 2024–25 and 2025–26 winter seasons, 513,358 participants were randomised [s2]. The incidence of RSV-related respiratory tract disease hospitalisation was 0.16 per 1,000 person-years in the vaccine group and 0.55 in the control group, an estimated vaccine effectiveness of 70.3% (95% CI 49.7–83.2) [s2]. For RSV-related lower respiratory tract disease hospitalisation the rates were 0.11 and 0.39, an effectiveness of 71.1% (45.4–85.7) [s2].

The third number is the one that keeps the result honest. For all-cause respiratory tract disease hospitalisation, the rates were 11.16 and 12.01 per 1,000 person-years — an effectiveness of 7.1% (−0.3 to 13.9) [s2]. Serious adverse events occurred in 2.0% of the vaccine group and 2.2% of the no-vaccine group [s2].

That contrast is the whole picture in two lines. RSV-specific hospitalisation is uncommon and the vaccine prevents about seven in ten of them. All-cause respiratory hospitalisation, at 12.01 per 1,000 person-years in the control group against 0.55 for RSV-related, is more than twenty times as common — and the vaccine moves it by an amount whose confidence interval touches zero [s2].

The durability trial

The second trial followed a single cohort across two seasons. Participants aged 60 or older were randomised 1:1 to RSVpreF or no vaccine during the 2024–25 winter season and followed through 2025–26, to evaluate protection across two consecutive seasons after one dose [s1]. Of 131,379 randomised, 130,991 were in the second-season intention-to-treat population [s1].

In the second season, RSV-related respiratory tract disease hospitalisation occurred at 0.11 per 1,000 person-years in the vaccine group and 0.25 in the control group — a vaccine effectiveness of 56.2%, with a 95% confidence interval of −12.4% to 84.8% [s1]. In the first season the same estimate had been 83.3% (42.9–96.9) [s1]. Across both seasons combined, rates were 0.11 and 0.37, an effectiveness of 70.6% (39.1–87.0) [s1].

There was no apparent difference in all-cause respiratory or cardiorespiratory hospitalisation in the second season among those randomised to vaccine in the first [s1]. The trial was funded by Pfizer and registered as NCT06684743 [s1].

The authors state the conclusion carefully, and it is worth copying their care: vaccine effectiveness estimates appeared lower in the second season than in the first [s1]. "Appeared" is doing necessary work. A confidence interval running from −12.4% to 84.8% is compatible with substantial waning and also compatible with no waning at all. The point estimate fell; the evidence that it fell is weak.

The real-world cohort

The JAMA Internal Medicine study looked at a population the trials largely did not reach: US nursing home residents in the first season the vaccines were available [s4].

It was a retrospective cohort of Medicare fee-for-service beneficiaries using claims from 10 September 2023 to 30 March 2024, restricted to beneficiaries aged 65 or older who resided in a US nursing home for at least one day and had continuous Medicare Part A, B and D enrolment [s4]. Effectiveness was calculated as (1 − hazard ratio) × 100% from multivariable Cox models [s4].

Among 597,430 beneficiaries — 66% female, median age 82 (IQR 75–88) — effectiveness against RSV-associated hospitalisation was 73% (95% CI 65–79) [s4]. By subgroup: 60% (35–76) in those aged 65–74, 75% (67–82) in those 75 and over, 67% (55–76) in short-stay residents of fewer than 100 cumulative days, and 80% (68–87) in long-stay residents [s4]. Effectiveness was 56% (39–69) against RSV-associated death, 59% (44–70) against severe outcomes defined as ICU admission, mechanical ventilation or death, and 66% (44–79) against RSV-associated thromboembolic events [s4].

A 73% real-world estimate against hospitalisation in nursing home residents [s4] sits close to the 70.3% from the randomised trial in a very different population [s2], which is the kind of agreement that makes both more credible.

What is and is not settled

Settled, on this evidence: the vaccine prevents RSV-specific hospitalisation in older adults, in a randomised design and in a frail real-world population, with effect sizes around 70% [s2][s4]. It also appears to reduce RSV-associated death and severe outcomes in that population, with wider intervals [s4].

Not settled: how long one dose lasts. The durability trial's second-season interval spans zero [s1]. The accompanying NEJM Evidence editorial is titled "A Hat Trick for RSV Vaccines — Prevention of Disease, Hospitalization, and Beyond" [s3], and the JAMA Internal Medicine study carries its own editorial on protecting nursing home residents [s5]. The revaccination question — whether, and when, a second dose is needed — is a policy decision that these data inform without answering.

Also not settled: the all-cause signal. A vaccine that cuts RSV hospitalisation by 70% and all-cause respiratory hospitalisation by 7.1% with an interval crossing zero [s2] is doing exactly what a pathogen-specific vaccine should do, and is not a general protection against winter respiratory admission. Both statements are true and the second one is the one usually lost.

What to watch

The durability trial's design points at its own follow-up: whether a third season separates the arms further, and whether a second dose restores first-season effectiveness. In the meantime the numbers above are seasonal estimates from two winters, one of which had fewer events than expected [s2].

This article describes trial and cohort results. It is not medical advice, and nothing here should be used to decide about vaccination.

Sources

• Durability of RSV Prefusion F Vaccine Protection across Two RSV Seasons, NEJM Evidence, 29 August 2026
• Bivalent RSV Prefusion F Vaccine to Prevent Hospitalizations in Adults, NEJM Evidence, 29 August 2026
• A Hat Trick for RSV Vaccines — Prevention of Disease, Hospitalization, and Beyond, NEJM Evidence, 29 August 2026
• Respiratory Syncytial Virus Vaccine Effectiveness in Medicare Beneficiaries Residing in Nursing Homes, JAMA Internal Medicine, 31 August 2026
• Protecting Nursing Home Residents From Respiratory Syncytial Virus, JAMA Internal Medicine, 31 August 2026