medRxiv📄 intensive care and critical care medicine

Clinical and virological characteristics of critically ill patients with influenza in France during the 2025/26 season, marked by the emergence of influenza A(H3N2) clade K

This prospective, nationwide French cohort study of 685 critically ill influenza patients found that while the emergence of the A(H3N2) subclade K during the 2025/26 season was initially associated with higher mortality in univariable analysis, this increased risk was not significant after adjusting for host-related factors such as age and clinical frailty.

de Prost, N., Bay, P., Le Goff, M., Preau, S., Guigon, A., Beloncle, F. M., Lefeuvre, C., Dartevel, A. i., Larrat, S., Coudroy, R., Deroche, L., Darreau, C., Thomin, J., Aubron, C., Tran, A., Uhel, F., Le Hingrat, Q., Tamion, F., Moisan, A., Guillon, A., Handala, L., Souweine, B., Henquell, C., Klouche, K., Tuaillon, E., Damoisel, C., Roque Afonso, A. M., Gault, E., Cappy, P., Soulier, A., Pawlotsky, J. M., Lemoine, F., Rameix Welti, M. A., Audureau, E., Fourati, S., SEVARVIR consortium,

Published 2026-02-28
📖 5 min read🧠 Deep dive
View on medRxiv ↗PDF ↗
⚕️

This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

Imagine the human body as a bustling city, and the flu virus as a gang of intruders trying to break in. Every winter, this gang shows up, but sometimes they wear a new disguise or bring a bigger weapon.

This paper is like a report card from a massive, nationwide investigation in France that looked at what happened when the flu gang attacked the city's most critical defense force: the Intensive Care Units (ICUs). The researchers wanted to know if the flu season of 2025–2026 was a "super-villain" year compared to the previous one (2024–2025), specifically because a new, tricky version of the flu (called A(H3N2) subclade K) was causing a stir around the world.

Here is the story of their findings, broken down simply:

1. The Setup: Two Seasons, One Big Question

The researchers gathered data from 685 patients across 42 French hospitals who were so sick with the flu they had to be admitted to the ICU. They compared two groups:

  • The 2024–2025 Team: The "previous year" group.
  • The 2025–2026 Team: The "new season" group, where the scary new A(H3N2) variant was the main suspect.

The Big Question: Did this new 2025–2026 flu variant kill more people or make people sicker than the flu from the year before?

2. The Twist: The "Older" Crowd

When they looked at the patients, they noticed something interesting about the 2025–2026 group. The patients were, on average, older.

  • Analogy: Think of the 2024 flu as a thief who targets young, fit people, while the 2025 flu is a thief who specifically targets the elderly neighborhood.
  • Because the new flu variant (A(H3N2)) tends to hit older people harder, the hospital was filled with more senior citizens. These patients naturally have more health issues (like high blood pressure or heart trouble) and are more "frail" (like a house with a shaky foundation).

3. The Surprise: The "Scary" Virus Wasn't Actually Scary

The researchers were worried that the new A(H3N2) variant was a "super-virus" that was inherently more deadly.

  • First Look (Unadjusted): At first glance, it looked like the A(H3N2) patients were dying more often.
  • The Deep Dive (Adjusted): But when the researchers put on their "detective glasses" and accounted for the fact that these patients were older and frailer, the picture changed completely.
  • The Verdict: Once they factored in age and health, the virus itself was not the killer. The A(H3N2) virus wasn't any more dangerous than the old flu; it just happened to attack a more vulnerable group of people.

The Metaphor: Imagine two cars crashing. Car A (the new flu) hits a fragile, old wooden house. Car B (the old flu) hits a sturdy brick house. Car A causes more damage, but not because the car is faster or stronger—it's because the house was weaker to begin with.

4. The "Genetic Fingerprint" Check

The scientists also took "DNA fingerprints" (genetic sequencing) of the viruses to see if the ones in the ICU were special "super-villains" compared to the ones just walking around in the community.

  • The Result: They found no difference. The viruses in the ICU looked exactly like the viruses in the community. There was no secret "killer gene" that made the ICU patients sicker. The severity came from the host (the patient), not the virus.

5. The Other Suspect: A(H1N1)

While A(H3N2) was the headline, the other flu type, A(H1N1), was also present. Interestingly, the patients with A(H1N1) were slightly younger and had fewer health issues, but they seemed to get sicker faster when they arrived at the hospital (lower oxygen levels). However, because they were generally healthier to start with, they survived just as well as the older A(H3N2) patients.

6. The Takeaway for Everyone

So, what does this mean for the rest of us?

  • Don't Panic about the "New" Flu: Just because a new flu variant circulates and makes headlines, it doesn't mean it's a "super-virus" designed to kill.
  • Protect the Vulnerable: The real danger lies in the people who are already weak (the elderly, those with chronic illnesses). The virus exploits their weakness, not their strength.
  • Vaccines and Meds: The study found that most patients got the antiviral drug (oseltamivir) and it worked well. However, they did find some small mutations in the H1N1 virus that might make the drug slightly less effective in the future, so scientists need to keep watching.

In a Nutshell:
The 2025–2026 flu season in France was intense, but the "new" flu wasn't a monster. It was just a regular flu that happened to knock on the doors of the most vulnerable people in the city. The tragedy wasn't the virus's strength; it was the fragility of the people it infected. The solution remains the same: protect the vulnerable, keep them vaccinated, and treat them quickly.

Get papers like this in your inbox

Personalized daily or weekly digests matching your interests. Gists or technical summaries, in your language.

Try Digest →

More like this

State-Dependent Parameter Relevance in Intensive Care: Syndrome-Specific Centroids Improve Orbit-Based Mortality Prediction from AUC 0.59 to 0.83 in 59,362 Predictions

By extending the Therapeutic Distance framework to incorporate state-dependent parameter relevance across 16 clinical syndromes in 84,176 ICU patients, this study demonstrates that syndrome-specific centroids significantly improve orbit-based mortality prediction (AUC 0.83) over established severity scores and standard machine learning models, while maintaining robustness to temporal drift and hyperparameter variations.

Basilakis, A., Duenser, M. W.2026-04-08📄 intensive care and critical care medicine

Clinician-Informed Feature Engineering Improves Machine Learning Assignment of Molecular Endotypes in the Intensive Care Unit

This study demonstrates that integrating clinician expertise into the feature engineering process for machine learning models significantly improves the accuracy, efficiency, and interpretability of assigning molecular endotypes in critically ill patients compared to purely data-driven approaches.

Sines, B. J., Hagan, R. S., Jiang, X. + 9 more2026-04-07📄 intensive care and critical care medicine

Re-evaluation Of Hypo- And Hyperoxemia In Patients With Respiratory Failure And Veno-Venous Extracorporeal Membrane Oxygenation

This retrospective study of 443 severe ARDS patients on VV-ECMO reveals a U-shaped association between mean PaO2 levels and mortality, identifying an optimal oxygenation range of 90–123 mmHg that differs from non-ECMO settings while also highlighting a significant risk associated with high FiO2.

Buenger, V., Russ, M., Hunsicker, O. + 5 more2026-04-07📄 intensive care and critical care medicine

Observation-process features are associated with larger domain shift in sepsis mortality prediction: a cross-database evaluation using MIMIC-IV and eICU-CRD

This study demonstrates that while incorporating observation-process features like measurement counts improves internal discrimination in sepsis mortality prediction models, it significantly exacerbates domain shift and degrades external calibration and transportability across different ICU databases.

Yamamoto, R., Wu, F., Sprehe, L. K. + 3 more2026-04-06📄 intensive care and critical care medicine

Therapeutic Distance: An Orbit-Based Framework for ICU Decision Support - Initial Validation in 11,627 Sepsis Patients from MIMIC-IV

This paper introduces and validates the "Therapeutic Distance" framework, an orbit-based decision support tool that analyzes 11,627 sepsis patients to identify subgroups with comparable outcomes based on clinical proximity to treatment centroids, thereby enabling individualized signal detection while acknowledging its role as a hypothesis-generating aid rather than a causal inference method.

Basilakis, A.2026-04-04📄 intensive care and critical care medicine