COVID-19 Data-Driven SIR Models, October 2021

ABSTRACT

Introduction/ Background: COVID-19 was declared a global pandemic on March 11, 2020 by the World Health Organization. The Susceptible-Infected-Recovered (SIR) model was used in a bid to predict COVID-19. In this study, we use a data-driven SIR model to simulate the epidemic in Rwanda from March 16, 2020 to October 14, 2021. Methods: The online access of some COVID-19 data to the public has facilitated this research. The study uses publicly available data from Our World In Data (OWID). The COVID-19 reported cases are used to estimate the spreading and the recovery rates. These data-driven parameters are then recast into the basic SIR models and its simple extension, the Susceptible-Exposed-Infected-Recovered (SEIR) model. The Susceptible-Infected-Recovered (SIR) model is one of the most extensively used approaches for modeling infectious diseases. Results: The data-driven SIR model captures a single wave and single variant in some countries but has severe limitations in estimating the end of a wave or the risk of death. However, the SEIR model captures the different waves that were identified in the country but cannot be used to assess the risk of death. Also, the predictive capabilities of the SEIR model yielded better results compared with the SIR model. Impact The aim of this research is to demonstrate the inadequacy of the SIR model and its extension due to its limitations to estimate waves as well as the mortality risks. Conclusion: The public data has limitations in terms of recovered cases and exposed cases. The limitations identified for the SIR and SEIR models, which consist of not being able to estimate a wave or the risk of death, suggest the use of improved mathematical approaches to predict the outbreak of COVID-19.