A SEIR-like model with a time-dependent contagion factor describes the dynamics of the Covid-19 pandemic

We show how a simple deterministic epidemic model without spatial structure can reproduce the evolution of confirmed Covid-19 case numbers in diverse countries and Brazilian states through use of a time-dependent contagion factor, beta(t). One expects that this function provides a link between the growth rate and mitigation policies. The model inserts a state A (presymptomatic) between states E (exposed) and I (infected) in the usual SEIR model, as well as distinguishing between confirmed and unconfirmed infected. With transition rates fixed at literature values, we vary the four free parameters in beta(t) to obtain a good description of time series of the cumulative number of confirmed cases. We then analyze the relation between changes in the contagion factor, as inferred from the time- series analysis, and mobility indexes based on cell-phone data.