Saturday, September 23

Wearable technology and COVID-19 – The Lancet Respiratory Medicine

A basic smartwatch or fitness tracker can be bought for less than £40. It will tell you how many steps you have walked, the calories you have expended, and the quality of your sleep. It will track your heart rate and nag you if you spend too much time sitting down. Depending on the model, there might even be monitors for blood oxygen and blood pressure. The more advanced devices are able to detect breathing disturbances. Other functions might include electrocardiograms, fall detectors, and skin temperature gauges. It is a thriving market. One in four Americans is thought to own some kind of wearable technology.

The implications for health care are enormous. “Wearables provide a nearly continuous stream of data; they monitor vital signs and help identify changes that are happening in the body in real time”, said Robert Hirten, Associate Professor of Medicine & Artificial Intelligence at the Icahn School of Medicine at Mount Sinai (New York City, NY, USA). A 2020 study involving more than 47 000 Fitbit users in the USA concluded that data on resting heart rate and sleep duration could significantly improve models predicting trends in influenza-like illnesses. In a study published last year, a machine learning model used data taken from wrist-based wearables to diagnose rhinovirus and influenza with high accuracy. It also proved capable of distinguishing between different severities of disease, 24 h before the emergence of symptoms.
“The COVID-19 pandemic has really pushed things forward”, Hirten told The Lancet Respiratory Medicine. Several studies have used wearables to predict the onset of COVID-19. Hirten co-authored a paper that found that changes in heart rate variability, captured by a smartwatch, signalled the presence of SARS-CoV-2 infection several days in advance of diagnosis. Other studies have explored the effects of vaccination on the body, which could potentially lead to individualised dosing, and how long it takes for the heart rate to return to baseline after an individual has contracted SARS-CoV-2, which could help to map the course of disease as well as identify cases of long COVID.
Researchers using wearables to detect whether the body is in the early stages of fighting infection currently tend to focus on metrics related to heart rate, which are relatively straightforward to capture. But as wearables become increasingly capable of collecting precise data on functions such as respiration rate and skin temperature, and as algorithms are honed to take into account the effect of infection on markers of activity, models can be optimised. A study of 32 individuals who contracted COVID-19 in 2020 found that 26 of them showed changes in heart rate, number of daily steps, or sleep duration. In several cases, these changes occurred more than a week before the appearance of symptoms.

All of which opens up a route to a future in which users of smartwatches and fitness trackers receive alerts letting them know that their recent biometric data are indicative of infection and perhaps they ought to consider isolating. The use of Global Positioning System (GPS) for contact tracing is already well-established. Wearables offer the opportunity to warn individuals that they might have been exposed to a virus and subsequently inform them when the exposure appears to have translated into infection.

Hirten noted that as machine learning approaches are applied to the large quantities of information extracted from wearables, they will start to tease out the subtle physiological signals that hint at the presence of an infection or other disease state. “We have a proof-of-concept that we are able to predict inflammatory events, such as respiratory viruses like SARS-CoV-2 and influenza, but the approaches are not specific to one infection or aetiology”, explained Hirten. Combining data taken from wearables with self-reported symptoms could prove productive, although users might not have the time or inclination to continually provide updated information.

“Having your own personalised baseline can help you understand your health better and raise the red flag when things move outside the baseline”, added Julia Moore Vogel, Program Director of The Participant Center, All of Us Research Program at Scripps Research (La Jolla, CA, USA). She has had long COVID for over 2 years. “I had brain fog for around 6 months after the initial infection, and it still occurs when I do too much. Daily, I have fatigue, headaches, and some chest pain”, said Vogel. Individuals with long COVID, including Vogel, commonly find that up to 72 h after overexerting themselves, they experience a flare-up of their symptoms. Carefully choosing when to be active and when to rest, a process known as pacing, can help to manage the condition.

“Pacing helps you avoid some of the lows associated with long COVID, but the problem is that you cannot always tell when you are overdoing things, and there is a lag before you feel the consequences”, said Vogel. Physical exertion is only one side of the story. The brain uses about 20% of the body’s energy supplies. “It took me a while to realise that cognitive rests are just as important as physical ones, but it is much easier to keep a handle on how much exercise you have done than it is to monitor how hard your brain has been working”, said Vogel.

The Body Battery function on Garmin smartwatches uses physical activity, sleep, and stress levels to provide a composite score from 0–100 that equates to how much energy you have left. “I have been using the body battery to manage my symptoms. I work to a threshold, and it means my symptoms are reasonably stable”, said Vogel. She is leading a planned study on whether wearables can help individuals with long COVID to implement pacing. “We still need treatments to resolve long COVID”, stressed Vogel. “But for many people, managing their symptoms through pacing can help improve their quality of life; wearables can give you quantifiable data on how to do that.”

Marielle Gross is Assistant Professor in the Center for Bioethics and Health Law at the University of Pittsburgh (Philadelphia, PA, USA). She pointed out that wearables are entangled with serious ethical considerations. “We are repurposing devices that have been designed and validated for a particular set of uses and applying them to health care, but we are not asking what it means to change the context in this way”, said Gross. Manufacturers are not typically drawn from the health-care technology community. Regulators are having to figure out how to address products that straddle the line between consumer goods and medical devices. Gross worries that we may be sleepwalking into a state of biosurveillance. After all, wearables provide data not just on where people are located but on the nuances of their physical condition.

“Are wearables genuinely improving patient health and wellbeing?”, asked Gross. She speculated that, for some people, sleep trackers might be more likely to accentuate anxiety over nocturnal habits than to improve them. “Quantification of life is not necessarily therapeutic”, said Gross. Even over the course of a single day, wearables generate vast amounts of data. There are plenty of companies that would be willing to purchase this information. Advertisers could end up knowing more about consumers than the consumers know about themselves.

“People are giving up a huge amount of sensitive, personal data—are we aware of the consequences of this? Who really benefits?”, said Gross. The data derived from wearables is often much more detailed than that contained within medical records, or obtained during clinical trials, yet the protections are far looser. Gross suggests decentralising data ownership, so that it remains with the person whose body is its source. A system analogous to that in place for online payments, which are made without the purchaser revealing all their credit card details, could be established to transmit biometric data gathered from wearables.

There are also questions over whether the advent of wearables will worsen inequalities. Users tend to be drawn from richer and better educated sections of society. Some people might be uncomfortable with navigating electronic devices that spew out endless numbers related to an array of physical functions. “The expectation is that as wearables become more advanced, and more players enter the field, prices will come down”, said Hirten. “That should help improve their availability, but we will still need to be very careful to make sure that the benefits they can provide are applied across socioeconomic groups, so that no-one gets left behind.” Vogel hopes that as the evidence base grows, there will be an unanswerable case for American health insurance companies to cover wearable devices and for public health-care systems elsewhere to routinely prescribe them.

Meanwhile, the field will continue to expand at breakneck speed. Smart facemasks, which can detect the presence of airborne viruses, might come to characterise the next pandemic. A combination of telemedicine with wearable technology could remove the need for a lot of face-to-face appointments. If a breath-borne biomarker for lung cancer were ever to be developed, a wearable could be adapted to detect it. Patients who would previously have required intensive care might be permitted to recover at home, monitored by wearables and recalled to hospital if their vital signs deteriorate. “The line between the body and the digital realm is only going to get more and more blurry”, concluded Gross.

Source link