Estimating the epidemic growth dynamics within the first week

FIORITI, V.; ROSELLI, I.; CHINNICI, M.; ARBORE, A.; SIGISMONDI, N.

Information about the early growth of infectious outbreaks are indispensable to estimate the epidemic spreading. A large number of mathematical tools have been developed to this end, facing as much large number of different dynamic evolutions, ranging from sub-linear to super-exponential growth. Of course, the crucial point is that we do not have enough data during the initial outbreak phase to make reliable inferences. Here we propose a methodology to estimate the epidemic growth dynamics from the infected cumulative data of just a week, provided a surveillance system is available over the whole territory. The methodology, based on the Newcomb-Benford Law, is applied to Italian covid 19 case-study. Results show that it is possible to discriminate the epidemic dynamics using the first seven data points collected over fifty Italian cities. Moreover, the form of the most probable approximating function of the growth, within a six weeks epidemic scenario, is identified.