Estimation of COVID-19 cases in Brazil: mathematical modeling study

Abstract

The objective was to describe a mathematical model to estimate cases by COVID-19 in Brazil in the short term and to identify the existence of forces to inhibit the spread of SARS-CoV2. Cases confirmed by COVID-19 in Brazil from March 16 to May 31, 2020 were used to develop a mathematical equation for estimating COVID-19 cases for up to 13 future days. Correction factors were calculated to identify whether there was an inhibition of the spread of infection in the country. The relative error and Pearson's correlation were analyzed to determine the model's performance. Six models were developed (gross, S1, S2, S3, S4 and corrected). The corrected model showed good adherence to the number of estimated and observed cases. Eight correction factors were calculated, which increased during the study period. The increase in correction factors expressed the loss of the inhibition force of interventions implemented in the country for the prevention and control of infection. The proposed mathematical model proved to be viable and feasible for implementation in epidemiological surveillance services.

Keywords

• epidemiological monitoring, covid-19, estimation, mathematical modeling