Séminaire Univers / |
| « Analysis of French COVID-19 hospital data with a SEAFHCDRO model » |
|
Gary Mamon |
|
The French daily hospital data with general admissions, admissions to critical care, deaths and releases, allows to objectively fit models of the growth of the COVID-19 pandemic in France. After an introduction to simple SIR evolutionary models, I will show that the data invalidates these models. I will present 7-, 8- and 9-phase models, which appear to be the simplest required to model the hospital data. These models involve up to 9 ratios of timescales over branching fractions, as well as the R_0 factors before and after the national lockdown and a normalization. I will show the results of Bayesian modeling for France considered as a single region, as well as considering 8 French départements, which are assumed to be genetically equivalent (hence same ratios) but with their own R_0 factors and normalizations. These results show that the national lockdown was a success, suggesting that well over one million deaths would have occurred in France without it. I will also show predictions for the fraction of the population that will have become immunized to the current strain of SARS-2, as well as for future evolution past the partial lifting of the lockdown this May 11.
|
|
mardi 5 mai 2020 - 10:30 Webinaire Institut d'Astrophysique de Paris |
| Pages web du séminaire / Seminar's webpage |