Mathematical models are a valuable tool for better understanding the spread of the flu and predicting its future course. Using epidemiological data, they enable the analysis of complex dynamics influenced by multiple factors (behavior, vacations, circulating influenza viruses, seasonality, etc.).

Since mainland France entered the epidemic phase, the Pasteur Institute and Santé publique France have been providing weekly forecasts on the dynamics of the flu epidemic at the national and regional levels. These forecasts help predict the epidemic’s trajectory over the next four weeks and the timing of the peak.

These tools aim to help health authorities and healthcare professionals better anticipate healthcare needs and adapt prevention measures. Aside from adopting preventive measures (wearing masks, handwashing, ventilating rooms), annual flu vaccination remains the most effective way to protect oneself.

Like any modeling, these forecasts have limitations, particularly related to the assumptions on which they are based and external factors (behaviors, influenza viruses, etc.).

How to interpret the forecasts?

The forecasts are presented in two formats:

• Panel A: Projected trend over 4 weeks (blue line and white dots for the median, blue area for the 95% confidence interval). Recent data are shown in black, and curves from previous seasons serve as a reference.

• Panel B: Probability of the peak occurring each week (colored bars; darker bars indicate a higher probability). The dotted line indicates the last week of available data.

The Pasteur Institute will provide a dedicated section where users can access this research, conducted by the Mathematical Modeling of Infectious Diseases unit, led by Simon Cauchemez. This information will also be available on the Santé publique France website.
The authors of these models are Juliette Paireau^a,b, and Simon Cauchemez^a.

• Download the full forecast (PDF) from 12/18/25

• Santé publique France Flu Special Report

• Link to the article in the Journal de la recherche