Universal Risk Phenotype Of US Counties For Flu-like Transmission To Improve County-specific COVID-19 Incidence Forecasts

Abstract

\n The spread of a communicable disease is a complex spatio-temporal process shaped by the specific transmission mechanism, the survivability of the pathogen outside the host under harsh environmental conditions, and access to new viable hosts broadly determined by the local population characteristics, and its compliance to social distancing policies. While the key factors shaping transmission of influenza and COVID-19 are beginning to be broadly understood, making precise forecasts on case count and mortality is still difficult. Despite a diversity of approaches being used to model the COVID-19 pandemic, a single best model is yet to coalesce. In this study we introduce the concept of a universal geo-spatial risk measure, denoted as the Universal Influenza-like Transmission (UnIT) score, to quantify the risk phenotype of US counties facilitating flu-like transmission mechanisms. The UnIT score is computed as a purely information-theoretic function of past incidence data for seasonal flu epidemics, yet emerges as the dominant factor explaining observed county-specific incidence trends over a range of putative demographic and socio-economic factors for the COV-19 pandemic. The predictive ability of the UnIT score is further demonstrated via county-specific weekly case count forecasts which consistently outperform the best models in the current literature. This study demonstrates that knowledge of past epidemics may be used to chart the course of future ones, if transmission mechanisms are broadly similar, despite distinct disease processes and causative pathogens.