Covid-19 could be detected by wearing a smartwatch that spots the infection days before symptoms occur, scientists claim.
Researchers at Stanford University in California analysed data from 31 people who had caught the infection and used Fitbits.
They found changes in heart rate, number of steps taken and sleep were evident in 80 per cent of the cases, suggesting the virus is detectable before it takes hold.
In some cases, signs of infection were clear nine days before the tell-tale symptoms of a cough, fever or loss of taste and smell started.
The researchers now believe wearables that measure health vitals could be the way out of the Covid-19 pandemic, which has killed 550,000 people globally.
And they designed an algorithm that works to spot Covid-19 infection in smartwatch wearers — but cautioned it needs fine-tuning before it is reliable.
Such a tool would be beneficial for curbing the spread of the virus because it would catch infectious people as early as possible, and limit how many people they can transmit the virus to while contagious.
Covid-19 could be detected by wearing a smartwatch (stock pictured) that spots the infection days before symptoms, scientists claim
The researchers whittled the study group down to just 31 people who all reported a positive Covid-19 diagnosis, and who all used the same smartwatch (Fitbit, pictured)
Research scientist Tejaswini Mishra led the study, which said smartwatches could be useful to detect if someone’s typical health parameters are amiss.
Technology in watches can measure heart rate and skin temperature. Some offer the added benefit of tracking sleep quality over time.
Doctors have for years known a higher resting heart rate could be a sign the body’s immune system is ramping up in response to a pathogen even in the absence of obvious symptoms.
Researchers have shown that commercial wearables could one flag a viral infection this way.
But scientists haven’t yet been able to make an effective alert system to spot any infections for any wearable tech.
To spot Covid-19, they would first need to identify exactly how the disease alters heart rate or other parameters, which would likely differ from person to person.
The researchers recruited a group of 5,322 individuals who completed surveys about their health.
They whittled the study group down to just 31 people who all reported a positive Covid-19 diagnosis, and who all used the same smartwatch (Fitbit).
They examined each person’s ‘baseline’ — their normal resting heart rate and typical fluctuations for that person.
Researchers analysed whether deviations from each person’s baseline were detected around the period of illness – 14 days prior to symptoms and seven days after.
The first finding was that 87.5 per cent of patients showed elevated heart signals compared with their previous ‘healthy window’ before or at the time of symptom onset.
In over 85 per cent of the positive cases, these signs were clear in the days prior to symptoms.
The alterations occurred three days before the patients developed a cough or fever, on average.
In some people, signals were clear nine days or earlier. But these patients were unsure of when they actually caught the virus, so the findings aren’t concrete.
At around the same time the heart rate changed, the number of steps a person took significantly decreased and sleep duration increased, suggesting fatigue.
The researchers said it was ‘interesting’ that these changes can be seen before a person actually reports symptoms.
Covid-19 was shown to change heart rate, steps and sleep in 80 per cent of cases overall.
This suggests the disease is linked to changes in physiology that can be detected through a smartwatch.
It provides hope that detecting altered parameters could help predict if a person will fall sick with Covid-19 before they are aware themselves they have it.
It could catch the carriers of the virus who don’t have symptoms yet even though they are contagious to others.
This is what governments globally are trying to do with a test and trace system, whereby people are alerted if they are suspected to have the coronavirus and therefore at risk of passing it on to others.
They are told to self isolate to prevent them going outside, mingling with others and potentially infecting many more people. It allows governments to stay one step ahead of the coronavirus.
Tests for the coronavirus are only undertaken when someone presents with symptoms, which is problematic because carriers are able to unknowingly spread the virus before they even are aware they have it.
These are heart rates for two different patients (B and C) over a number of days (each grey vertical line is a new day). The red and purple vertical dashed lines indicate the day of Covid-19 symptom onset and diagnosis, respectively. To see where heart rate elevates before symptoms, look at the top panel for each patient (the black squiggly line). The red horizontal arrow shows where the heart rate was elevated from baseline for a number of days. It also shows to spike at other times after infection, which may be due to a relapse in symptoms
Some never show symptoms at all, called asymptomatic, believed to account for as many as 50 per cent of cases.
The benefit of smart watches are that they are used by millions globally and are easily accessible.
‘There is substantial need and opportunity for population scale technology solutions for infection detection and tracking,’ Ms Mishra wrote.
‘Since most infections become apparent only upon symptom onset, the current methods of testing cannot identify presymptomatic carriers, which is a significant challenge for the implementation of early stage interventions that reduce transmission.
‘As such, accessible and inexpensive methods for early detection of COVID-19 in real-time are urgently needed.’
The team developed an algorithm to spot Covid-19 by monitoring heart rate in real time, using the data they had collected.
They tested their method on 24 Covid-19 patients — and found it worked to detect Covid-19 before or at onset of symptoms in 67 per cent of cases.
Most of the cases missed were in people who had other pre-existing conditions such as chronic respiratory diseases, which the algorithm may have failed to differentiate from Covid-19.
‘Interestingly the number of alarms increases considerably post-Covid-19 infection, suggesting lingering complications from Covid-19 illness,’ the authors wrote.
They saw that in some people, symptoms relapsed after recovery, and heart rate elevations matched this.
Other things that cause the heart rate to spike, such as medication, alcohol, stress or travelling, were picked up by the tool.
But their alarm system worked in a ‘tiered’ fashion so that it didn’t make a false alarm when someone was simply watching a scary film or exercising.
The findings have been published online and are yet to be scrutinised by other independent scientists.
But Ms Mishra and team are confident in the potential of wearables to mitigate the Covid-19 pandemic.
They added: ‘It should be noted that these wearable devices are not yet FDA-approved and our study is still modest in size.
‘It is currently unclear as to whether our approach can distinguish infections from SARS-CoV-2 from those caused by other illnesses.’