Experimental smartwatch COVID-19 detection improving - study

Smartwatch alerting systems for early detection of COVID-19 infection are coming closer to reality, researchers reported on Monday in Nature Medicine.

They tested their new system, developed with open-source software, in 2,155 wearers of Fitbit, Apple Watch, Garmin watches or other devices.

Ultimately, 84 of the volunteers were diagnosed with coronavirus infections - including 14 of 18 people without symptoms.

Overall, the researchers' algorithms generated alerts in 67 (80%) of the infected individuals, on average three days before symptoms began.

"This is the first time, to our knowledge, that asymptomatic detection has been shown for COVID-19," they said.

Presently, the system mainly depends on measurements of wearers' resting heart rate, said study leader Michael Snyder of Stanford University School of Medicine in California.

Going forward, he said he hopes watch manufacturers will be able to provide other types of highly accurate physiologic data.

"Many stressors can trigger the alerting," Snyder said.

"Most of these are easy to spot - travel, excessive alcohol, even work or other types of stress, so the user knows to ignore the alerts."

When watches can report other health data such as heart rate variability, respiration rate, skin temperature, and oxygen levels, it will become easier to distinguish the COVID-19 cases from other non-COVID-19 events, researchers said. "Right now we are running this as a research study," Snyder said. "But soon we hope that FDA approved devices will dominate this area." -- Reuters