Investigators used the French National Health Data System (SNDS) to assemble a retrospective, nationwide cohort of hospitalizations for confirmed influenza coded as J09 or J10. The study window ran from October 2016 through December 2024, with the 2020–2021 seasons excluded. The authors reported 132,443 influenza-associated hospitalizations meeting inclusion criteria. The work is reported as a preprint and has not undergone peer review.
Vaccination status was determined from reimbursed community pharmacy dispensing records. The cohort contained 72,852 hospitalizations among patients recorded as vaccinated (55%) and 59,591 among unvaccinated patients (45%). The vaccinated subgroup was older on average (mean age 82.3 years) compared with the unvaccinated subgroup (mean age 79.3 years). The investigators also accounted for pneumococcal vaccination and COVID-19 vaccination in their adjustment strategy.
The study team evaluated four key endpoints: invasive mechanical ventilation (IMV), intensive care unit (ICU) admission, in-hospital death, and a predefined composite severe outcome. Entropy balancing was applied to achieve covariate balance across sociodemographic variables, chronic conditions, and prior vaccination history. Weighted regression models produced adjusted risk ratios (aRRs) for the measured outcomes. In addition, multistate survival models were used to estimate transition-specific hazard ratios (HRs) between hospitalization, ICU admission, discharge, and in-hospital death.
IMV occurred in 2,128 vaccinated hospitalizations (2.9%) and in 2,544 unvaccinated hospitalizations (4.3%). After adjustment, vaccination was associated with a lower risk of IMV (aRR 0.88; 95% CI 0.83–0.93), according to the authors. ICU admission did not differ meaningfully by vaccination status in adjusted models (aRR 0.99; 95% CI 0.96–1.01). In-hospital death was reported in 6,300 vaccinated hospitalizations (8.6%) and 4,952 unvaccinated hospitalizations (8.3%); the adjusted risk ratio for death was 0.90 (95% CI 0.87–0.93). The composite severe outcome was nominally lower among vaccinated patients (aRR 0.96; 95% CI 0.94–0.98).
The pattern of results—reduced mechanical ventilation and lower adjusted mortality without a significant reduction in ICU admission—reflects an attenuation in severity among hospitalized older adults who had received influenza vaccine, as the study team interpreted it. The reader should note that the vaccinated patients in the cohort were, on average, older and that entropy balancing was used to address measured confounding.
For the subset admitted to ICU, multistate survival modeling produced additional insights. The authors reported that vaccination was associated with a lower hazard of in-ICU death (adjusted HR 0.85; 95% CI 0.78–0.92). Vaccinated patients admitted to ICU also experienced a shorter hospital course, with a reported median time to discharge reduced by approximately two days compared with unvaccinated ICU patients. These results are presented as transition-specific effects from hospitalization through ICU and to discharge or death.
The analysis relied on pseudonymized SNDS records, and the authors state that, under French regulation, this work did not require ethics committee review or individual informed consent. The study was declared on the Health Data Hub platform (No. 24832545). Investigators noted that clinical codes, data management scripts, and analysis code are available on request from the corresponding author, and access to anonymized SNDS data requires platform approval.
Competing interests were declared: several authors are affiliated with or employed by vaccine manufacturers and one author reported research grants and personal fees from industry. The report is a preprint and has not been peer reviewed; that status is explicitly stated by the authors. The study excluded the 2020–2021 seasons; the source did not provide detailed subgroup analyses by influenza subtype, vaccine type, dose, or timing relative to symptom onset. Those specific methodological details were not reported in the abstract and would require consultation of the full manuscript or data request.
The authors conclude that influenza vaccination was associated with lower adjusted risks of IMV and in-hospital death among older adults hospitalized with influenza in France, and with shorter time to discharge among ICU patients. Because this is a preprint using administrative claims data, remaining questions include potential residual confounding, the impact of vaccine match or subtype, and whether these associations vary by vaccine formulation—questions the source did not resolve and that would benefit from peer review or additional stratified analyses.