'Smart Bed' May Detect, Track COVID-19 Symptoms

Megan Brooks

June 28, 2021

Editor's note: Find the latest COVID-19 news and guidance in Medscape's Coronavirus Resource Center.

Smart-bed technology may help predict and track symptoms of COVID-19 and other respiratory infections, new research suggests.

Real-world biometric data obtained from a consumer smart-bed platform showed that worsening of COVID-19 symptoms correlated with an increase in sleep duration, breathing rate, and heart rate and a decrease in sleep quality.

The study used the Sleep Number 360 smart bed and its operating system, SleepIQ technology.

"The 360 smart bed is able to detect symptoms associated with COVID in a very unobtrusive way," investigator Gary Garcia-Molina, PhD, with Sleep Number Labs, San Jose, California, told Medscape Medical News.

The findings were presented at SLEEP 2021: 35th Annual Meeting of the Associated Professional Sleep Societies.

Promising Devices

Pathophysiologic responses to respiratory viral infection affect sleep duration and quality in addition to breathing function. "Smart" and "connected" devices that monitor biosignals over time hold promise for monitoring infectious disease.

The investigators leveraged longitudinal, biometric data captured using ballistocardiography signals from the 360 smart bed to predict, at an individual level, the presence and duration of COVID-19 symptoms.

Study participants included 122 COVID-19–positive smart bed users (mean age, 45.6 years; 40% men) and 1603 COVID-19–negative smart bed users (mean age, 49.8 years; 48% men). In both groups, mean body mass index was around 30 kg/m2.

Analysis of the biometric data showed that symptom exacerbation in the COVID-positive group was associated with a significant increase in sleep duration, respiration rate, heart rate, restful time, and motion.

Worsening of COVID-19 symptoms was also associated with a decrease in sleep quality but no apparent change in the time to fall asleep.

In contrast, no significant disruptions in sleep or cardiorespiratory metrics were evident in the COVID-negative group.

Using the data, the researchers developed a predictive model for worsening COVID-19 symptoms. The area under the receiver-operator curve (AUC) estimate was 0.84, which is similar to values reported in wearable sensors.

When the dataset was expanded beyond the initial self-reported dates of COVID symptom onset, the AUC estimate improved to 0.93.

"The duration of COVID symptoms predicted by the model aligned with the duration of symptoms reported by the user, and we were able to predict 75% of the cases within day 5 of symptoms. That's useful information that might help prevent spreading of the disease," Garcia-Molina said.

This model, he added, may be able to predict the occurrence of symptoms "associated with other conditions, like influenza or the common cold."

Digital Biomarker?

Commenting on the findings for Medscape Medical News, Adam C. Powell, PhD, president, Payer+Provider Syndicate, said that "passively observed digital biomarkers" are being used in a growing range of ways in healthcare.

"It is wonderful to see a new application related to disease detection," said Powell, who was not involved with the research.

The study "highlights a new potential digital biomarker with many potential applications ― ballistocardiography data produced by a smart-bed platform," he said. It also "suggests that a consumer smart-bed platform may be able to detect COVID symptoms."

Powell cautioned that the study had "a small sample of COVID-positive patients, has yet to be published in a peer-reviewed journal, and was coauthored by individuals with ties to a smart-bed platform company.

"Nonetheless, it is very interesting because it highlights an additional potential digital biomarker," he said.

Powell noted that there are real advantages to capturing data from a smart-bed platform. For example, users do not need to remember to wear or do something for the data to be recorded, such as with smart watches and other smart devices.

"When information can be passively captured, adherence to measurement is likely to be far better than when active participation is required," Powell said.

"There have been a number of passive digital biomarker interventions in the cardiology and behavioral health spaces. Given the increased concern about respiratory illness since the start of the COVID pandemic, there is a great need to bring new methods for detecting respiratory illness to market," he added.

The study was funded by Sleep Number Corporation. Garcia Molina and five co-authors are employees of the company. Powell has reported no relevant financial relationships.

SLEEP 2021: 35th Annual Meeting of the Associated Professional Sleep Societies: Abstract 651. Presented June 13, 2021.

For more Medscape Neurology news, join us on Facebook and Twitter.

Comments

3090D553-9492-4563-8681-AD288FA52ACE
Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as:

processing....