Modelling the COVID-19 epidemics in Brasil: Parametric identification and public health measures influence

Renato Machado Cotta; Carolina Palma Naveira-Cotta; pierre magal

The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.03.31.20049130 doi: medRxiv preprint infectious cases. With uniform distributions for all five parameters, taking the previous estimates for the three first parameters, an arbitrary guesses for , and , the five estimated parameters are shown in Table 5 , together with the 95% confidence interval for each parameter. Figure 12 shows the theoretical CR(t) curve obtained with the five parameters estimation, plotted together with the 95% confidence interval bounds for this simulated evolution. One can see the marked reduction on the f(t) parameter from the estimates in Table 5, which results in the increase of the unreported to reported infectious cases, as is shown in Figure 13 .a for CR(t) and CU(t) predictions up to 150 days. Clearly, the reduction on the testing, and thus on the isolation of reported infectious individuals, leads to an impressive increase on the total number of infected individuals after 150 days (723,698 cases), including unreported (609,125) and reported cases (114,572). Figure 13 .b presents the predicted evolution of the daily reported infectious cases, which shows a peak at around t=61 days of about 2,672 reported cases.