Using Wearable Tech to Track a Viral Threat

10/7/20 Duke BME Magazine

The Duke research study uses data from smartphones, smartwatches and health surveys to help detect COVID-19

The CovIdentify team uses biometric data from smartwatches and smartphones to identify early signs of COVID-19 infection.
Using Wearable Tech to Track a Viral Threat

Once a tool reserved for athletes and tech-savvy exercisers, smartwatches have transformed into a popular and every-day accessory, giving users a chance to track everything from their blood oxygen levels to their workout statistics. Now, a team of engineers and clinicians at Duke University are hoping to use these ubiquitous tools for a new purpose—tracking and identifying early symptoms of the novel coronavirus.

Originally launched in April 2020, CovIdentify ( was designed to explore how data collected on smartphones, Apple Watches, Fitbits, Garmin Watches and other connected devices could help determine whether or not users have COVID-19, the disease caused by the novel coronavirus.

“By early March it was clear that there was going to be an explosion of COVID-19 cases across the United States, and we knew this was going to be a long-term health care problem,” says Jessilyn Dunn, PhD an assistant professor of biomedical engineering and one of the leaders of the project.

Jessilyn Dunn

We wanted to see if we could use the data from mobile devices and smartwatches to identify signals of early COVID-19 infection, and see if there were biomarkers or health changes that indicated if someone was more likely to develop a severe infection.

Jessilyn Dunn

In previous research, Dunn had shown that biometric data collected from wearable devices could indicate if a person was more likely to develop health issues like diabetes or cardiovascular disease, or show they had an infection. This work would be expanded in CovIdentify, where students in Dunn’s Big Ideas Lab, including PhD students Brennae Bent, Peter Cho, Karnika Singh, Will Wang, will collect biometric data about a participant, including their sleep schedules, oxygen levels, activity levels and heart rate.

“The first stage of the study was launched in the first week of April with, where participants could input their relevant demographics and medical information and agree to complete a daily survey,” says Dunn. “This is a short daily survey that would ask about social distancing and common symptoms, like nasal congestion, runny nose, cough, sore throat, headache, fever, chills and COVID-specific symptoms, like shortness of breath, nausea and loss of sense of taste and smell.”

By pairing these surveys with changes in the biometric data from the watches, the researchers could explore which symptoms would indicate common issues like allergies and a cold, and which symptoms and biometric changes were likely to indicate a COVID-19 infection.

“The biometric data helps us create a baseline to see what a person’s health usually looks like, and the ongoing data and surveys help us create a trajectory of the illness,” says Ryan Shaw, PhD, RN an associate professor of nursing and director of the Duke Mobile App Gateway . “We’ll be able to see how changes in the digital health data relate to the emergence of specific symptoms. Once this data is collected we can develop, test and refine our predictive algorithms to detect respiratory infections from the COVID-19 virus.”

In second phase of the study, which began in early June 2020, the team launched an iOS application, which can pull data from any device that syncs with the Apple HealthKit application. Beyond simplifying the sign-up and survey process, the app also expands the devices that the team can work with.

As the pandemic has played out, the CovIdentify team also noted that the virus had a disproportionate effect on underserved communities, minorities and people with chronic illnesses like diabetes and hypertension. Working with collaborators Chris Woods, MD, the associate director of the Duke Center for Applied Genomics and Precision Medicine, and Geoffrey Ginsburg, MD, PhD the director of MEDx and the Duke Center for Applied Genomics and Precision Medicine, the team developed plans to target and collect more comprehensive biometric data from these communities.

“We’re seeing that underrepresented minorities are at a much greater risk of both contracting COVID-19 and developing a more severe illness, so it’s important that we explore what biometric factors may contribute to that risk,” says Shaw. “We’re also targeting recruitment of people who have a higher risk of contracting the disease, like grocery store workers, cleaning and cafeteria staff at hospitals, and people who live in high-density housing like dorms, military barracks and nursing homes.

“We’ve procured some funding that allows us to provide devices for people who want to participate but may not be able to afford the devices as well,” says Shaw. “That being said, we’re still pursuing additional funding options so we can more comprehensively provide devices and expand the study.”

According to Dunn, this expanded data collection from the second phase of their study may improve their ability to differentiate COVID-19 from other illnesses, which will be significant as states open back up and infections potentially coincide with flu season.

“One of my lab’s goals is to arm health care professionals with tools and information to detect illness and intervene early by delivering the right treatment to the right person at the right time,” Dunn said.

If this study is successful, we’d be able to use non-invasive and accessible tools to help us control the spread of a dangerous virus, and predict when someone may need more intensive care. If we can achieve this, we may be able to help doctors save more lives.

Jessilyn Dunn