Fitness-tracker biometrics predict COVID-19 better than temps
Temperature checks, increasingly disparaged as COVID theater, can be outperformed as an illness indicator by fitness-tracker biometrics and self-reported symptoms, according to new research.
Of particular interest were changes in an individual’s historical and contemporary resting heart rate and normal sleep duration in identifying pre-symptomatic and asymptomatic individuals. Step counts played a role in diagnosis, too.
Those who have been infected are more likely to see their heart rate increase, sometimes faster than 100 beats per minute, than to have a fever.
Combining these deviations with self-reported symptoms enabled researchers to more accurately tell the difference between volunteers who were symptomatic and COVID-19 positive and those who were not.
Almost 31,000 people from all 50 states enrolled. Of those, 3,811 reported symptoms. Fifty-four symptomatic volunteers reported testing positive; 279 were negative.
Although supported by a comparatively small sample of volunteers, the finding is significant. There is as yet no better way to combat the spread of the coronavirus than to identify infections, trace contacts and quarantine COVID-19 symptomatic and asymptomatic individuals.
The demand for one-off COVID-19 rapid antigen tests still continues to outstrip availability in the United States. And people in areas without contact restrictions should be tested on an ongoing basis until the infection rate drops dramatically and permanently because they interact with more people.
To see if a new option could be found, a biometrics app — MyDataHelps — was created by a team of researchers from the Scripps Research Translational Institute and CareEvolution, which operates HIEBus, a health information exchange technology platform.
A fitness-tracker information exchange, called Detect, or Digital Engagement and Tracking for Early Control and Treatment, was launched in March. Data collected by Fitbit, Apple’s HealthKit and Google’s Fit devices was uploaded to Detect and shared among the researchers.