Simplicity is a key strength of the package: it only takes a few lines of code to run and compare different types of models! Dev by Nathanael Hoze, with contributions from Marga Pons, @jessmetcalf.bsky.social, @mtwhite.bsky.social, @hsalje.bsky.social.

Rsero is useful to determine optimal designs for serosurveys depending on context. For example, should you target the general population, children <10y.o. or <20y.o.? This will depend on the expected annual probability of infection (FOI)...

You can also use Rsero to evaluate the impact of interventions, here mass drug administration on Trachoma prevalence in Nepal.

Rsero can assess how infection risk may vary with individual characteristics, e.g. gender and location.

Rsero can help you determine if your data are more consistent with a scenario of endemic ‘constant’ circulation, or with the occurrence of 1, 2 or 3 outbreaks in the last 50 years. It can estimate the year when these outbreaks occurred.

Thrilled to attend Institute Pasteur Dakar Epidemics Forum, celebrating the 100th anniversary of the Institute! Impressed by the numerous contributions to public health in the last century and exciting developments in recent years.

Fantastic away days of @pasteur.fr Global Health Department in beautiful Chantilly! Exciting scientific discussions on research projects in Cameroon, Madagascar, Cambodia, Brazil, Algeria, Senegal… also talking about mental health and how to better support staff.

Join our 1-week course on Infectious Disease Mathematical Modelling at Institut Pasteur in beautiful Paris, April 7-11 2025! Deadline for applications: 7 December 2024.

More information at www.pasteur.fr/en/education...

The blue curve shows the fit to the data. Red one shows the scenario without treatment. We estimate the campaign prevented about 6k hosp. for RSV broncholitis. It’s about 1 hosp. prevented per 40 doses administered. Efficacy was estimated at 73% (61-84).

To evaluate this impact, we developed a math model that captured key properties of RSV (e.g. waning immunity, severity profile by age), of the French pop (e.g. contact patterns by age) and of the campaign (e.g. doses administered over time) and was calibrated to the data.

In infants >2m (no/little treatment), peak amplitude in 2023-24 was similar to that in previous season. In contrast, peak in infants aged 0-2m (targeted for treatment) was 47% lower in 2023-24 than in 2022-2023. This suggests treatment had a strong impact.

Two days of great scientific discussions for the French infectious disease modelling community (supported by anrs-mie). So nice to see this community grow!

We evaluated it using seasonal flu in France. We found that, compared to standard methods (PF), accuracy of mPF_v2 was improved for forecasts 1 to 4 weeks ahead, peak timing and size. This approach allows combining strengths from statistical and mechanistic models.