Recombinant flu vaccine cut nursing-home outbreaks in one season, not the next

Influenza tears through nursing homes, where frail, older residents live in close quarters and a single outbreak can strain an entire facility. A large pragmatic trial set out to test whether switching a nursing home from a standard flu shot to a recombinant one would reduce outbreaks, and the answer it returned is a qualified yes: fewer confirmed outbreaks in a season with real flu circulation, and no measurable difference in a season when flu barely appeared [s1]. The trial was an investigator-initiated study funded by a vaccine manufacturer, Sanofi Pasteur, a funding source worth noting even though the design was pragmatic and the outcome was facility-level [s1].

Most flu vaccines are grown in eggs, a process that can introduce mutations blunting the match to circulating strains. A recombinant vaccine (RIV4) is made without eggs, using a protein-expression system, and delivers a higher dose of the key antigen — a plausible reason it might protect older recipients better [s2]. Whether that biological rationale translates into fewer outbreaks in real facilities is what this trial tried to measure directly, rather than relying on antibody levels [s1].

How the trial was run

The study was a pragmatic, cluster-randomised clinical trial: whole nursing homes, not individual residents, were randomly assigned to give either the recombinant quadrivalent vaccine (RIV4) or a standard-dose egg-based inactivated quadrivalent vaccine (IIV4) to their residents and staff [s1]. It ran across two influenza seasons, enrolling 965 facilities in 2019-2020 and 1023 facilities in 2020-2021 [s1]. Eligible homes were Medicare-certified, had at least 50 long-stay residents, and housed high proportions of adults aged 65 and older [s1].

The outcome was facility-reported influenza outbreaks, tracked through monthly logs and split into two definitions: a "confirmed" outbreak meant one or more laboratory-confirmed cases, while a "suspected" outbreak meant two or more cases of influenza-like illness [s1]. Counting whole outbreaks, rather than individual infections, is what makes this a test of facility-level benefit — the thing an administrator actually wants to prevent.

What it found

The two seasons told sharply different stories, and the difference is the point. In 2019-2020, a season with moderate influenza activity, recombinant-assigned facilities had fewer outbreaks by both measures. Suspected outbreaks occurred in 148 (30.6%) of the recombinant homes versus 178 (37.0%) of the standard-vaccine homes, and laboratory-confirmed outbreaks in 177 (36.6%) versus 214 (44.5%) [s1]. The adjusted relative risk for confirmed outbreaks was 0.82 (95% CI 0.71-0.96) — a roughly 18% reduction, with a confidence interval that stayed below 1.0 and so reached statistical significance [s1].

The second season erased the signal. In 2020-2021, when pandemic precautions suppressed flu almost everywhere, confirmed outbreaks were rare and evenly split — 16 in recombinant homes versus 15 in standard-vaccine homes — with no significant difference [s1]. Covid-19 outbreak risk, checked as a comparison, did not differ between the groups in either season, as expected for a flu vaccine [s1].

Why the two seasons matter

A vaccine can only prevent outbreaks of a virus that is circulating. The 2020-2021 null is therefore not evidence that the recombinant vaccine failed; it is evidence that there was almost nothing to prevent that winter [s1]. Reading the two seasons together, the fair summary is that the benefit appeared when flu was around and vanished when flu was absent — which is exactly how a real vaccine effect should behave, and a useful guard against over-reading either season alone.

The caveats are real. This was a pragmatic trial measuring outbreaks reported by facilities, not laboratory-verified infections in every resident, so the outcome carries the noise of real-world reporting [s1]. A cluster-randomised design, in which whole homes rather than individuals are assigned, also means the analysis rests on far fewer independent units than the resident headcount suggests, which is part of why the confidence interval is as wide as it is [s1]. The effect size in the one informative season was modest, and the confidence interval reached close to 1.0 [s1]. And the manufacturer funding, while disclosed and investigator-initiated, belongs in any reader's weighing of the result [s1]. A companion analysis comparing recombinant and egg-based vaccination in the same nursing-home setting adds context but does not settle the size of the benefit [s2].

What it means

For nursing homes deciding which flu vaccine to stock, the trial offers modest, real-world support for the recombinant option during seasons with meaningful flu circulation, alongside an honest reminder that the benefit depends on how much flu actually shows up [s1]. The finding is a facility-level signal, not a promise to any individual resident. What to watch next is whether the pattern holds across more typical seasons and in trials not funded by a vaccine maker.

Sources

• [s1] Recombinant Influenza Vaccine and Influenza Outbreaks in US Nursing Homes: Results From a Pragmatic Cluster-Randomized Clinical Trial. Open Forum Infectious Diseases, 27 August 2026. https://doi.org/10.1093/ofid/ofag522
• [s2] Recombinant vs Egg-Based Quadrivalent Influenza Vaccination for Nursing Home Residents. JAMA Network Open, 2 January 2025. https://doi.org/10.1001/jamanetworkopen.2024.52677