Researchers are turning to wearable technologies such as Fitbits and Apple Watches in a new bid to tackle coronavirus, using them to gather large volumes of real-time patient data to track the course of the disease.
Wearable systems work by monitoring “biomarkers” — measures such as heart rate or skin temperature — which can indicate the presence of the virus, even before symptoms are felt.
By constantly monitoring fluctuations in these metrics, they provide a valuable stream of health data that would otherwise be more difficult and laborious for researchers to capture.
This week, NHS England announced a partnership with British health tech start-up Huma, formerly Medopad, to study more than 160 patients recovering at home from Covid-19 using the company’s remote-monitoring technology.
Meanwhile, commercial devices are being used in research in the US, with Fitbit collaborating with researchers at Scripps Research and Stanford University to provide patient data from its activity monitoring devices.
‘Proactive, predictive healthcare’
The push into wearable technologies comes as scientists scramble for new solutions to curb the spread of the disease, with populations emerging from lockdowns often with little in the way of comprehensive track-and-trace programmes.
It also follows the rollout of a pioneering biometric bracelet programme in Liechtenstein in April, which began by giving one in 20 citizens a device and aims eventually to scale up to include the principality’s entire 38,000 population.
For the NHS England trial, Huma has provided pulse oximeters to detect blood oxygen saturation levels, while patients’ smartphone cameras are used to detect resting heart rate by capturing the blood flow at the surface of the skin. The data is then collated via an app to build a picture of the patient’s health for clinicians.
The result is “proactive, predictive healthcare”, said Dan Vahdat, chief executive and founder at Huma. “We can use this to quantify your disease and build personal digital health solutions.”
He added that remote patient monitoring was not limited to Covid-19: last year the start-up partnered Chinese tech conglomerate Tencent to launch a clinical trial into Parkinson’s disease, which attempted to track the onset of the condition by analysing patients’ hand movements.
In the US, Michael Snyder, a professor of genetics at Stanford University, has been analysing the potential use of wearables for public health purposes since 2014. He discovered his own Lyme disease thanks to devices he was trialling in 2015.
His lab is applying similar techniques to Covid-19, studying data from consenting wearable owners who have had the virus, been exposed to it, or are in high-risk jobs, in order to identify early warning signs.
“In three-quarters of the cases we’ve analysed, we can see people’s heart rate go up before or at the time of diagnosis,” he said.
Epidemiologist Jennifer Radin is undertaking similar research at Scripps Research in San Diego. In January, she published a paper in the Lancet, examining how Fitbit data could be used to study influenza-like viruses.
She said that such information could be used to identify emerging disease hotspots, as well as improving individuals’ treatment. “Currently when you go to your doctor, as long as you fall in this huge range, you’re considered normal,” she said. “These wearable devices can better identify what’s normal for you.”
Beyond the lab
Some researchers hope that the use of wearables in trials will set the stage for longer-term remote monitoring, in which patients would be able to track their own biomarkers in communication with their doctor, and health authorities could observe the emergence of disease indicators at a population level.
Karen Taylor, research director of the centre for health solutions at Deloitte, said she was hopeful that coronavirus would increase individuals’ interest in tracking their own health. “As soon as you have something that impacts people’s ideas of their own longevity, ways to improve it come to the fore,” she said.
Accuracy remains a major obstacle to the broader implementation of remote health tracking, said Brent Mittelstadt, a research fellow at the Oxford Internet Institute and Turing fellow at the Alan Turing Institute, arguing that models based on heart rate and sleep levels come with a risk of false positives.
The demographic and geographic skew of wearable ownership could prove troublesome in population-level tracking of the spread of diseases such as coronavirus. “That has a knock-on effect,” said Mr Mittelstadt. “You’re going to miss outbreaks in places where ownership of the devices is low.”
Jason du Preez, chief executive and co-founder of enterprise data privacy start-up Privitar, said it was vital to ensure that privacy was at the heart of any remote monitoring programme. “It is an incredibly valuable thing to be able to pull together health data,” he said, “but we have to do that in a way that maintains the trust of the data subject.” He added that after years of data scandals, consumers were wary of the potential for abuse.
Mr Mittelstadt also emphasised the importance of properly contextualising data when it comes to personalised medicine. “Otherwise, in the same way that it could lead to peace of mind, it could lead to anxiety,” he said. “Just because your blood pressure went up for 20 minutes at some point in the day.”
