Measuring the Change in European and US COVID-19 Death Rates

TL;DR

This study uses a compartment ODE model to analyze COVID-19 death data in the US and Europe, revealing significant decreases in case mortality rates that are not strongly linked to common factors or interventions.

Contribution

It provides a novel quantitative analysis showing large, rapid decreases in COVID-19 mortality rates across US and European regions, challenging prior estimates.

Findings

Case mortality decreased by at least 80% in US and 90% in Europe.

Decreases occurred mainly between mid-April and mid-June.

No strong correlation found between mortality decrease and other model parameters or demographics.

Abstract

By fitting a compartment ODE model for Covid-19 propagation to cumulative case and death data for US states and European countries, we find that the case mortality rate seems to have decreased by at least 80% in most of the US and at least 90% in most of Europe. These are much larger and faster changes than reported in empirical studies, such as the 18% decrease in mortality found for the New York City hospital system from March to August 2020 (Horwitz et al, Trends in Covid-19 risk-adjusted mortality rates, J. Hosp. Med. 2020). Our reported decreases surprisingly do not have strong correlations to other model parameters (such as contact rate) or other standard state/national metrics such as population density, GDP, and median age. Almost all the decreases occurred between mid-April and mid-June, which unexpectedly corresponds to the time when many state and national lockdowns were released resulting in surges of new cases. Several plausible causes for this drop are examined, such as improvements in treatment, face mask wearing, a new virus strain, and potentially changing demographics of infected patients, but none are overwhelmingly convincing given the currently available evidence.