Chasing John Snow: data analytics in the COVID-19 era
Jesse Pietz, Scott McCoy, Joseph Wilck
Abstract
During the first half of 2020, the lives of people around the world abruptly changed due to COVID-19. Data visualisations and models related to the spread of the disease became ubiquitous. In this paper, we survey 25 different data analytics dashboards, highlight the modelling approach taken by each, and develop a multi-attribute utility theory model to assess their effectiveness in communicating key features that explain the spread of infectious disease. We show that the dashboards that feature dimensions that span the categories associated with compartmental epidemiology models tend to be relatively robust data visualisations, and we highlight that information systems need to be improved to include data on actions to reduce the spread of the disease. We analyse the actions taken by countries around the world and show that when governments employ strict measures early, particularly those that enforce social distancing and include widespread testing and comprehensive contact tracing, they are more likely to experience better outcomes. Recommendations for how countries should respond in future pandemics are detailed.