Fitbit Charge3 device (Fitbit Inc.)
2 Nov. 2020. Data from smart watches and fitness trackers, when combined with symptom reports, can show that the wearers likely have Covid-19 infections. These findings from research conducted by Scripps Research Translational Institute in La Jolla, California appear in the 29 October issue of the journal Nature Medicine.
A team led by Scripps Institute epidemiologist Jennifer Radin, is seeking more comprehensive and unobtrusive tools for detecting Covid-19 infections in a population. Screening populations now for Covid-19 rely on symptom questionnaires or reports of travel to regions with widespread infections. As the authors note, however, these steps can miss people with pre-symptomatic and asymptomatic infections that make up as many as 45 percent of infected individuals.
“We know that common screening practices for the coronavirus can easily miss pre-symptomatic or asymptomatic cases,” says Radin in an institute statement. “And infrequent viral tests, with often-delayed results, don’t offer the real-time insights we need to control the spread of the virus.”
The authors cite data showing fitness trackers and many smartwatches that also track a person’s health are now worn by as many as one in five people in the U.S. To determine if these devices can provide useful data for Covid-19 screening, Scripps Institute is conducting the Detect study, short for Digital Engagement and Tracking for Early Control and Treatment, to see if these data collected in near real time that can alert health authorities to viral disease outbreaks much earlier than current methods.
The Nature Medicine paper reports on data from 30,529 Detect study participants in the U.S., from 25 March to 7 June 2020. Participants wore smart watches and fitness trackers made mainly by Fitbit and Apple, providing data through the MyDataHelps app, made by health technology company CareEvolution, available in Apple iOS and Android versions. As reported by Science & Enterprise at the start of the study, participants were asked to donate their data on heart rate, activity levels, and sleep, with logs of respiratory symptoms, treatments received, and diagnostic test results.
In addition, participants reported symptoms characteristic of Covid-19 infections and other disorders, such as headaches, fatigue, fever and chills, and difficulty breathing. Among participants, 3,811 reported these or other symptoms, and 54 subsequently tested positive for Covid-19 infections. A set of symptoms emerged as better indicators of individuals testing positive for Covid-19, compared to those testing negative: stomach ache, fever and chills, fatigue, difficulty breathing, decrease in taste or smell, cough, and body aches.
The Scripps researchers combined the symptom data into a single statistical index, and also combined data collected from wearable devices — resting heart rate, sleep, and activity measures — into a different metric. Each of these composite statistical measures, symptoms and wearable data, modestly predicted Covid-19 infections among the participants. But when the team combined the symptom and wearable metrics into a single model, they found that combined metric predicted Covid-19 infections with 80 percent accuracy, more reliably than either of the separate measures.
“What’s exciting here is that we now have a validated digital signal for Covid-19,” notes Scripps Research Translational Institute founder and director Eric Topol. “The next step is to use this to prevent emerging outbreaks from spreading.” The Detect study is still recruiting participants, with the goal of enrolling some 100,000 fitness tracker and smart watch wearers, particularly health and other front-line essential workers.
More from Science & Enterprise:
- Blood Test Shown to Predict Covid-19 Symptom Severity
- Wireless Sensor Detects Multiple Covid-19 Indicators
- Study Testing Wearables Data for Covid-19 Clues
- App Tracks Public Mask Wearing for Covid-19 Evidence
- Univ Lab, Company Developing Phone-Based Virus Sensor
* * *
RSS - Posts
You must be logged in to post a comment.