The Biomedical Advanced Research and Development Authority has begun working with Evidation Health to use data from wearable devices to cut the time it takes to identify cases of COVID-19.
In a statement released Thursday, Evidation said it will analyze sleep and activity data, plus self-reported symptoms, from 300 people at risk of exposure of developing the disease with a view to creating an early warning algorithm.
- The collaboration is part of a broader push to use data from wearables to track respiratory infections. BARDA is already working with Evidation to apply the approach to influenza, while Fitbit is collaborating with Scripps Research Institute and Stanford Medicine on a similar infectious diseases effort.
The presence of presymptomatic and asymptomatic carriers of the coronavirus makes it hard to control the outbreak. Even if everyone self-isolated as soon as they developed symptoms, the virus might continue to move through communities, transmitted by people completely unaware that they are infected with the pathogen.
BARDA, part of the Department of Health and Human Services, contends it may be possible to detect presymptomatic and asymptomatic cases using patient-generated data. To test the idea, the agency is providing Evidation with $720,000. The Bill & Melinda Gates Foundation has also provided money from its $250 million COVID-19 fund.
Evidation, working with health data nonprofit 4YouandMe, will use the money to monitor healthcare workers and other first responders. BARDA believes the project may lead to computer models that use wearable and self-reported data to improve real-time predictions of the incidence of the virus and detect infections before the development of symptoms.
The concept is supported by research into the use of wearables in the monitoring of influenza. In January, researchers at Scripps published a paper in The Lancet's digital health journal about their analysis of data captured by Fitbit wearable devices. The study found a correlation between abnormal data and increases in influenza-like illness, leading the researchers to posit that wearables “could be vital to enact timely outbreak response measures to prevent further transmission” during outbreaks.
Evidation has published similar research. In one study, the company found step counts, heart rates and time spent in bed all changed one day before the onset of symptoms of influenza-like illness.
An algorithm that can detect signs of COVID-19 prior to the emergence of symptoms could potentially help to curb the rate of transmission, thereby supporting efforts to open economies without causing a surge in new infections. However, it remains unclear whether such algorithms can prospectively predict SARS-CoV-2 infections with enough accuracy and lead time to be useful in the management of the outbreak.