Predicting the Size and Duration of the COVID-19 Pandemic

This article explores the ongoing COVID-19 pandemic, asking how long it might last. Focusing on Bahrain, which has a finite population of 1.7M, it aimed to predict the size and duration of the pandemic, which is key information for administering public health policy. We compare the predictions made by numerical solutions of variations of the Kermack-McKendrick SIR epidemic model and Tsallis-Tirnakli model with the curve-fitting solution of the Bass model of product adoption. The results reiterate the complex and difficult nature of estimating parameters, and how this can lead to initial predictions that are far from reality. The Tsallis-Tirnakli and Bass models yield more realistic results using data-driven approaches but greatly differ in their predictions. The study discusses possible sources for predictive inaccuracies, as identified during our predictions for Bahrain, the United States, and the world. We conclude that additional factors such as variations in social network structure, public health policy, and differences in population and population density are major sources of inaccuracies in estimating size and duration.

Automated summary

This article examines the potential duration of the ongoing COVID-19 pandemic in Bahrain through numerical solutions of various epidemic models and curve-fitting solutions. It highlights the complexity of estimating parameters and how this can lead to inaccurate predictions, while discussing potential sources of inaccuracies such as social network structure, public health policy, and population differences. The study concludes that additional factors beyond numerical models play a significant role in predicting the size and duration of a pandemic.