Preparedness for Viral Outbreaks: Making Accurate Predictions from the Outset

Abstract

An epidemic of an unknown virus introduces extreme uncertainty and can, as we found in 2020, easily lead to panic reactions. Only data subjected to mathematical analysis—statistical analysis above all else at the outset—can give a quick understanding of the seriousness of the situation and provide the means for immediate contingency planning. To understand the Covid-19 crisis, we used Extreme Value Theory (EVT) techniques that extract information from the tails of distributions, where the information about ‘catastrophic events’ resides. We describe how the information so obtained gave us the predictive power essential for contingency planning very early in the Covid-19 pandemic. We also describe the epidemiological model we developed, which extends the initial statistical work and gives a complete toolkit for predicting the behaviour of a viral outbreak in time to be effective in dealing with it. Finally, we talk briefly about the importance of a readily accessible worldwide population and disease database to aid in the preparation for future emergencies.