The present work is a statistical analysis of the COVID-19 pandemic. As the number of cases worldwide overtakes one million, data reveals closed outbreaks in Hubei and South Korea, with a new slight increase in the number of infected people in the latter. Both of these countries have reached a plateau in the number of Total Confirmed Cases per Million (TCCpM) residents, suggesting a trend to be followed by other affected regions. Using Hubei’s data as a basis of analysis, we have studied the spreading rate of COVID-19 and modelled the epidemic center for 10 European countries. We have also given the final TCCpM curves for Italy and Lombardia. The introduction of the α-factor allows us to analyse the different stages of the outbreak, compare the European countries amongst each other, and, finally, to confront the initial phase of the disease between Europe and South America.

Methods

By dividing the TCCpM curves in multiple sections spanning short time frames we were able to fit each section to a linear model. By pairing then the angular coefficient (α factor) of each section to the total number of confirmed infections at the center of the corresponding time interval, we have analysed how the spreading rate of Covid-19 changes as more people are infected. Also, by modelling the TCCpM curves with an asymmetrical time integral of a Normal Distribution, we were able to study, by fitting progressively larger data ensembles, how the fitting parameters change as more data becomes available.

Findings

The data analysis shows that the spreading rate of COVID-19 increases similarly for all countries in its early stage, but changes as the number of TCCpM in each country grows. Regarding the modelling of the TCCpM curves, we have found that the fitting parameters oscillate with time before reaching constant values. The estimation of such values allows the determination of better parameters for the model, which in turn leads to more trustworthy forecasts on the pandemic development.

Interpretation

The analysis of the oscillating fitting parameters allows an early prediction of the TCC, epidemic center and standard deviation of the outbreak. The α factor and the recovered over confirmed cases ratio can be used to understand the pandemic development in each country and to compare the protective measures taken by local authorities and their impact on the spreading of the disease.

Funding

CNPq (grant number 2018/303911) and Fapesp (grant numebr 2019/06382-9).

	Year	Citations

References