Editors' ChoiceInfectious Disease

Maybe not an overreaction

See allHide authors and affiliations

Science Translational Medicine  12 Feb 2020:
Vol. 12, Issue 530, eaba9019
DOI: 10.1126/scitranslmed.aba9019


A mathematical simulation of coronavirus COVID-19 estimated the number of infections to be far worse than reported.

Starting in December 2019, a novel coronavirus, known as COVID-19, caused growing cases of atypical pneumonia in the city of Wuhan, China. The virus spread. As of 12 February 2020, a total of 45,204 confirmed cases had been reported from China and 519 confirmed cases from 27 countries worldwide. Among the reported victims, 1116 died. When facing an epidemic as such, a dangerous mistake that countries over the world can make is to rely entirely on confirmed data but underestimate the actual size of infections. Causes of underestimation include the limitation of resources, as there can be insufficient quarantine spaces, detection reagents, and medical personnel to identify infections in real time. Additionally, the lack of symptoms from the virus in its dormant state can delay confirmation. There is also a fear of data being manipulated or downplayed by officials due to economic and political concerns.

To estimate the actual number of infections, in late January Wu and coauthors used a susceptible-exposed-infectious-recovered (SEIR) model to simulate the epidemic since its onset in December 2019. Assumptions of the initial size of infection, dormant period, and the power of transmission were made based on reported data from the Chinese Center for Disease Control and Prevention. A unique feature of their SEIR model is that it simulated the transportation of infected population by flights, railroads, and shuttles through Wuhan, both domestic and international, following records on transport schedules and passenger numbers. The model also accounted for the lockdown of Wuhan since 23 January 2020. The prevalence of infections in Wuhan was then inferred based on the number of confirmed cases from countries around the world. The model estimated that the number of infections in Wuhan on 25 January 2020 had already reached 75,815, in sharp contrast to 1975 confirmed cases that were reported on 25 January. The authors warned that major cities with close transport links to Wuhan were probably sustaining a localized outbreak.

Despite opposition from Beijing and officials of the World Health Organization, who stressed concerns over the economy, by early February 2020 several countries—including Australia, Israel, Indonesia, and the United States—had imposed tight restrictions on travel to China and other countries at risk of outbreak. In the meantime, cases continued to increase exponentially in more than 20 cities with ground transport links to Wuhan. Although the accuracy of the above estimation is currently unknown, this mathematical model and similar ones might have motivated precautions that so far prevented a global disaster.

Highlighted Article

Stay Connected to Science Translational Medicine

Navigate This Article