Inferência Bayesiana aplicada a modelos epidemiológicos compartimentais

Abstract

Over the past few years, the study of compartmental epidemiological models has intensified to study the transmission dynamics of the SARS-CoV-2 virus. In this work, we seek through Bayesian Inference techniques, more precisely the Iterative Sampling Kalman Filter (IEKF) framework, to investigate the temporal evolution of the effective reproduction number Rt of the COVID-19 pandemic in the city of Fortaleza. The IEKF method assumes a Gaussian distribution for the a priori and likelihood probabilities, adjusting the a priori to a posteriori distribution using Bayes’ rule. The calculation of the reproduction number is therefore done using the SEIIR epidemiological model together with the IEKF, which allowed us to study the evolution of the dynamics of the spread of SARS-CoV-2 in the city of Fortaleza from data referring to the number of deaths. confirmed diaries made available by the Secretary of Health of the Government of the State of Ceará (SESA). Within these analyses, we inferred the temporal evolution of all the parameters of the SEIIR model as well as their respective effective reproduction number, which indicates periods of greater spread of the disease and, therefore, is an important indicator to be considered by government authorities during its analysis and decision-making in combating epidemiological crises.