Getting a fuller picture of COVID-19 infections

One of the biggest challenges during the COVID-19 pandemic has been accurate testing and detection rates of the virus. In fact, most countries detected only 60% or less of all infected cases by September 2020. IIASA researchers endeavored to bridge this gap.  

Led by Alexia Fürnkranz-Prskawetz and coauthors from the Wittgenstein Centre for Demography and Global Human Capital, the study analyzed the results from 3,000 likely parameter sets of an epidemiological model. The technique was applied to the US due to the country’s regional diversity and because the US accounted for almost a quarter of all global confirmed cases and deaths by September 2020. 

The model depends on the lethality of the virus and not contagiousness, which means more contagious mutations of the virus do not affect the results. The findings suggest that non-pharmaceutical interventions, such as social distancing and masks, have been key to containing the virus and preventing future waves. 

“The number of COVID-19 infections is key for accurately monitoring pandemics. A number of challenges, however, such as varied testing policies and availability, prevent accurate detection of all cases,” explains Fürnkranz-Prskawetz. “Previous studies aimed at reaching a more accurate number by retroactively assessing infections, but they are lacking in scope and often confined to a specific population subgroup. Our research combines the reported number of COVID-19 deaths with population data and case-fatality rates to indirectly estimate the fraction of people ever infected and the fraction of people detected among the infected.”

Last edited: 17 June 2021