Uncovering the Universality of COVID-19 Fatality Rates

An important and elusive characteristic of the COVID-19 pandemic is the fatality rate which allows us to understand the severity of this disease, health care needs, and the impact on large populations. To address this and other questions we present a probabilistic model to study the evolution of the COVID-19 pandemic and to correct the case fatality rates. Our model employs probabilities to estimate the time evolution of infections, recoveries, and deaths. This model discriminates asymptomatic, mild/moderate, and severe cases and allows for the estimation of undiagnosed individuals. Furthermore, we compare the model curves to official data for medium-sized cities, world metropolises, and medium-sized countries, spanning a range of populations from a few million to several million individuals. Using the undiagnosed estimates we correct the case fatality rates and find that it ranges from 0.33% ± 0.02% to 1.14% ± 0.07%. Since we applied the method to cities and countries with different characteristics, the corrected case fatality rates indicate a universality that is independent of location and other social/demographic conditions. Our results agree with sample tests and seroprevalence studies, considerably changing our understanding of the COVID-19 fatality rates. Other applications include the estimate for severe cases, ICU needs, and undiagnosed cases.