The Epidemic Severity Index: Estimating Relative Local Severity of Novel Disease Outbreaks

Determining the severity of a novel pathogen in the early stages is difficult in the absence of reliable data. The pattern of outbreaks for COVID-19 across the globe have differed markedly above and below 30{degrees}N latitudes, suggesting very different levels of severity, but countries worldwide have implemented the same lockdown strategies. Existing methods for estimating severity appear not to have been useful in informing strategic decisions, possibly due to mismatches between the data required and those available, overly sophisticated methods with undesirable biases, or perhaps confusion and uncertainly generated by the wide range of estimates these methods produced early on. The Epidemic Severity Index (ESI) is a simple, robust method for estimating the local severity of novel epidemic outbreaks using early and widely-available data and that does not depend on any estimated values. ESI allows rapid, meaningful comparisons across territories that can be tracked as the outbreaks unfold. The ESI quantifies severity relative to a parameterised baseline rather than attempting to estimate values for infection fatality rates, case fatality rates or transmission rates. The relative nature of the ESI sidesteps any problems of confidence associated with absolute rate estimation methods and offers immediate practical strategic value.