Premières estimations de l’impact du confinement pour contrôler l’épidémie de COVID-19 en France

Article de Philippe ANGOT (I2M, Prof. Aix-Marseille Université) :


Abstract : In this Note, I present an original dynamic model of progression of Covid-19 epidemic in France, the so-called HOPE model, which remains relatively simple. Our model follows at the best four reliable indicators: the number of patients in Hospitals and in Intensive Care Units (ICU’s), the Outflux from Hospitals and the number of Deaths which are reported daily by the French Public Healthcare system. Then, we give results about the influence of the complete lockdown measures taken by the French government on March 17, 2020, initially for 15 days, then for 30 days and now until May 11, 2020 but perhaps further. We show the tremendous impact of the general lockdown on the infectious tsunami to avoid the huge natural disaster which should occur if it was not applied. Indeed, the number of deaths is found divided by the factor 120 by applying a complete lockdown of 60 days with an efficiency ratio evaluated to 75%. We discuss this impact with respect of the efficiency and/or the duration of the containment. In particular, we show that a small effort of +1% in the efficiency of the lockdown saves 600 human lives; reversely, a small relaxation of −1% in the lockdown respect costs 600 deaths more. Next, we investigate the outbreak of an uncontrolled secondary wave of infection after the lockdown. Consequently, we show that the “stop and go” strategy is probably not a reasonable and sustainable scenario but rather a real crash test for the Healthcare system. Finally, we propose the suppression strategy called “successive damping cascade” after the general lockdown which allows the efficiency ratio to go progressively to zero within several less and less controlled secondary waves.

(soumis sur le serveur HAL le 17/04/2020)

Retour en haut 

Secured By miniOrange