Biocomplexity Institute Estimates Global Importation Risk of 2019 Novel Coronavirus (COVID-19)

The 2019 Novel Coronavirus (COVID-19), which originated in Wuhan, China, in late 2019, has received significant global attention due to the rapid speed and widespread reach of the outbreak. As of early March 2020, COVID-19 has exceeded 92,000 confirmed cases in 76 countries within a span of two months. In response to this global pandemic, researchers from the University of Virginia’s Biocomplexity Institute have developed a series of visualization and analytical web applications to provide a greater understanding of the pandemic’s scope, and to help officials, healthcare systems, and policy makers worldwide, control and ultimately stop the outbreak. 

As COVID-19 began its rapid spread within mainland China in early 2020 and travel-related cases were quickly being identified in neighboring countries, researchers from the Institute’s Network Systems Science and Advanced Computing division recognized the need to characterize the risk of importation for countries throughout the world. Increased global connectivity via international trade and travel has been confirmed as a contributing factor in the rapid spread of COVID-19, with the global airline network playing a key role in the importation of emerging infectious diseases. 

To model and create an Import Risk Visualization Tool to characterize the emergence of COVID-19 globally, the Institute team started with a well-established metric known as effective distance to estimate the time of arrival (ToA) for all countries. They applied this metric on global air traffic data from the International Air Transport Association (IATA) to quantify risk of emergence for various countries through direct importation from China, and compared it against arrival times for the first 24 countries based on official reports from the World Health Organization (WHO). Using this model, the tool could then estimate ToA for all other countries. 

“In addition to quantifying arrival risk, we complemented it with a measure of a country’s vulnerability to infectious disease outbreaks (IDVI), which allows one to identify countries at risk of sustained outbreaks, and also those that may be experiencing undetected outbreaks,” said Srinivasan Venkatramanan, research scientist for the Biocomplexity Institute’s NSAAC division and co-developer of the Import Risk Visualization Tool. “We’ve shared these results with federal agencies to understand the early evolution of COVID-19’s global spread.” 

Further, Institute researchers incorporated data on airline suspensions (travel restrictions and flight cancellations to China) into the model’s estimated arrival time calculations. With the situation evolving so rapidly, the team curated the airline suspension data from multiple sources such as newspaper articles, airline websites, official airline guides, and IATA, and continues to update it frequently. 

“There is increasing evidence that the travel restrictions within China did have a significant impact on the control of COVID-19 and bought some time for other countries to prepare their response,” Venkatramanansaid. “One key result from our work was to show the impact of reactive and proactive airline suspensions on relative risks.”

As COVID-19 continues its rapid spread, the Institute team is updating the visualization tool to quantify the number of travel-related cases instead of first arrival times. According to Aniruddha Adiga, postdoctoral research associate for the Biocomplexity Institute’s NSSAC division and co-developer of the Import Risk Visualization Tool, although initially developed for global application, the team is repurposing the tool to understand travel-related case introduction and sub-national spread using other mobility datasets featuring ground transport in addition to air travel. 

“One of the central assumptions of our early work was that most travel-related cases would originate out of China,” he said. “But, recently, we’ve seen countries report cases with travel history to Iran or Italy, so we’re updating the tool to better understand travel-related cases both on the global and sub-national levels.”

The Biocomplexity Institute has extensive experience supporting global response to large-scale complex issues such as pandemics using computational and mathematical tools. These tools and techniques are crucial in supporting real-world responses that require integration of multiple large datasets along with domain expertise on the biological, clinical, epidemiological, social, and behavioral aspects of an epidemic.  

“Our team has helped support multiple epidemic and pandemic responses, ranging from 2009-H1N1, 2014-15 Ebola outbreak in West Africa, and Zika in the Americas, to name a few,” Venkatramanan said. “Being able to use models and computing platforms to support federal agencies in planning their response allows us to see the broader impact of our work in near real-time. Further, such efforts reveal new challenges which lead us towards interesting research questions to advance our respective academic disciplines.”

Several university collaborators and partners have enabled the Institute’s comprehensive response to the COVID-19 pandemic. This includes researchers from UVA’s School of Medicine, Division of Infectious Diseases and International Health, whose insights into the COVID-19 pandemic have been critical to the Institute’s modeling efforts, and the Global Infectious Diseases Institute, who awarded Institute researchers with an internal seed grant to understand the impact of migration on disease spread. 

Full details of the Institute team’s development of the Import Risk Visualization Tool are available in the medRxiv article titled, “Evaluating the impact of international airline suspensionson the early global spread of COVID-19.” For more information on all of the Biocomplexity Institute’s efforts to support planning and response for the COVID-19 outbreak, visit our Network Systems Science and Advanced Computing page.