March 3, 2020
The 2019 Novel Coronavirus (COVID-19) is a rapidly spreading epidemic that originated in Wuhan, China, in December 2019. As of late February 2020, COVID-19 has exceeded 81,000 confirmed cases across six continents, affecting people in virtually every age group, and resulting in death in just over three percent of confirmed cases to date. The many unknowns about COVID-19 – including its precise origin, how long the disease incubates, and when during the incubation people become contagious – have made data collection, and therefore, identification of effective intervention methods challenging.
In response to this global pandemic, researchers from the University of Virginia’s Biocomplexity Institute have developed a set of visualization and analytical web applications to help provide a better understanding of the epidemic’s scope and aid in bringing the outbreak to a swift conclusion. These simulation and modeling tools integrate real-time open source data on confirmed cases of COVID-19 to help explain the complex dynamics of the disease and its spread, and ultimately provide public healthcare systems and policy makers with guidance and support for effective interventions and resource allocation.
Researchers from the Biocomplexity Institute’s Network Systems Science and Advanced Computing (NSSAC) division have created two visualization tools for COVID-19, including:
- COVID-19 Surveillance Dashboard– An interactive dashboard that integrates available historical data on confirmed cases of COVID-19 with unique features to explore temporal and spatial aspects of outbreaks simultaneously. The tool was originally released Feb. 3, 2020, and its data is updated twice daily.
- COVID-19 Cases and Clusters Outside of Mainland China Dashboard– An open-data online tool that provides a dashboard view of imported COVID-19 cases (cases in countries outside of Mainland China) through an interactive visual map. In addition to detailed information about individual cases by country, this tool provides data for case clusters, including exposure source, exposure start, hospital visit dates, recorded symptoms, and more. The tool was originally released on Feb. 13, 2020 and is updated several times daily.
While there are numerous dashboards available to support planning and response efforts for the COVID-19 outbreak, most of them focus on the latest data with a limited view of historical information. The Biocomplexity Institute’s COVID-19 Surveillance Dashboard supports all available historical data from onset of the outbreak (current data starts from Jan. 22, 2020), and enables users to view the progression through an automated time slider which demonstrates both temporal and spatial outbreak data alongside statistical data for the globe. Since its initial release on Feb. 3, the COVID-19 Surveillance Dashboard has received more than 40,000 views with users from more than 158 countries (as of March 2).
Other key features include:
- An interactive chart for cumulative and daily number;
- A visualization of all reported COVID-19 incidence data, filtered by date;
- A heatmap of selected attributes on an interactive map;
- A query tool that allows users to focus on regions of interest;
- Ability to select regions by clicking on the map;
- Ability to export subsets of the data for analysis on external tools; and,
- Mobile friendly.
Data for the Surveillance Dashboard is sourced from numerous organizations worldwide, including the World Health Organization (WHO), the Centers for Disease Control and Prevention (CDC), the European Centre for Disease Prevention and Control (ECDC), National Health Commission (NHC) of the People’s Republic of China, China’s DXY, and China’s QQ, and curated by Johns Hopkins University Center for Systems Science and Engineering (JHU CSSE) and the Biocomplexity Institute.
“The most challenging aspect of developing the dashboard has been data curation,” said Dawen Xie, research scientist for the Biocomplexity Institute’s NSAAC division and lead developer of the Surveillance Dashboard. “We recently switched to our own data curation, which enables us to publish real-time data at least twice daily, and we’re in the process of automating the entire curation process. Our overall goal is to provide the public and research community with a comprehensive yet user-friendly tool to access all COVID-19 surveillance data, customizable to meet people’s needs at different levels.”
A second tool developed by the Biocomplexity Institute is the COVID-19 Cases and Clusters Outside of Mainland China Dashboard. The comprehensive dashboard integrates multiple sources to provide detailed data for both individual cases and case clusters by country. Users can click on the interactive map to see individual case data, including:
- Number of recorded cases;
- Number known to have been imported from other countries;
- Number of symptomatic cases (and, where available, what symptoms where exhibited);
- Where the imported cases are suspected to be from (i.e. due to recent travel); and,
- The ages of those infected.
Further, users can access country-specific data on case clusters, including geographic dispersal, disease derivation (Wuhan or non-Wuhan), and case cluster timelines for travel, symptom onset, hospital admission, and disease confirmation. The dashboard also integrates linked data such as related news articles and repositories.
“A lot of people are focusing on COVID-19 end data, which is reliable and attainable, but lacking in detail,” said James Schlitt, postdoctoral research associate for the Biocomplexity Institute’s NSAAC division and lead developer of the Cases and Clusters Dashboard. “The main purpose of our dashboard is to enable investigation and analysis, so it’s distinctive in its level of detail. It integrates multiple data sources to provide individual patient accounts, such as who they are and where they got infected, as well as clustering analysis so we can better understand how the disease is transmitted and transported.”
According to Schlitt, understanding individual cases is key in identifying effective mitigation and intervention methods. “Looking at individual cases provides a way to see what is working and not working as far as interventions. You can see how cases come into a country, common transmission patterns, and see the odds of it exploding in various communities. With this data, policy makers and healthcare systems can be better equipped and prepared to deal with the outbreak.”
Data for the Cases and Clusters Dashboard is sourced from numerous organizations, including JHU CSSE, University of Washington Institute for Health Metrics and Evaluation, Fogarty International Center at the National Institutes of Health, Natural Earth, and China’s Tencent, and is curated by the Biocomplexity Institute. Since its official release on Feb. 13, 2020, the tool has received nearly 6,000 views from users largely based in the U.S., Canada, United Kingdom, and Norway as of early March.
"Our team has a long history of developing decision support tools and dashboards to support epidemic planning and response,” said Madhav Marathe, director of the Biocomplexity Institute’s NSSAC division. “We have done this in past as we supported the Ebola and Zika outbreak response. Dawen and James have done an incredible job in leading the development of these dashboards, and we have already seen growing interest in them. Our team is committed to supporting the community on problems of societal importance."
The Biocomplexity Institute is in a unique position to help solve large-scale complex problems, such as those in epidemiology, with its collaborative, multidisciplinary approach. By assembling expert teams across disciplines, the Institute aims to discover connections and identify solutions between health, information networks, security, and infrastructure.
“A lot of these issues are so complex, it’s implausible for any one narrow expertise to solve them,” Schlitt said. “We have a unique insight at the Institute and the ability to approach public health questions by applying mathematics, network science, statistics, computer science, and public health, among other domains to solve these problems. Our interdisciplinary approach is a key advantage.”
For more information on the Biocomplexity Institute’s efforts to support planning and response for the COVID-19 outbreak, visit biocomplexity.virginia.edu/nssac.