Jacob Lund/Shutterstock

As the wearable technology market continues to expand in 2019, consumers worldwide are using their devices to help them track and improve their health. Given the surge in popularity, physicians are starting to take a closer look at how to incorporate the data collected by patients’ personal devices into their clinical workflow to provide better patient care.

One such physician is Karl A. Poterack, MD, FAMIA, medical director of applied clinical informatics at the Mayo Clinic. He notes that the healthcare market is inundated with wearable devices, with more than 400 different devices from 100 brands currently available. According to data from Futuresource Consulting, approximately 100 million wearable devices were sold worldwide in 2018, representing a growth rate of more than 18 percent from 2017. Poterack estimates that more than 50 percent of Americans use wearables to monitor their health in some capacity, including devices from Fitbit, Xiaomi, Apple, Garmin, and Samsung.

The most popular health wearables, usually a wristwatch or wristband, have multiple sensors: pedometers, accelerometers, gyroscopes, magnetometers, barometers, altimeters, GPS, and photoplethysmographs. The data is typically collected passively and continuously, including steps, energy, sleep, heart rate (HR), blood pressure, oxygen saturation, glucose, weight, and body mass index.

Vast pools of personal health data

With this wide variety of sensors, wearable manufacturers are collecting and analyzing massive amounts of data. Fitbit alone has gathered 7.5 billion nights of sleep data and 9 trillion minutes of HR data from 25.4 million active users in 87 countries, says Carolyn Walsh, vice president and head of sales and business development at Fitbit Health, which is based in San Francisco.

With access to aggregated health data collected from all parts of the world, the wearable device maker can provide useful population health snapshots. For example, Fitbit’s analysis of more than 100 billion hours of aggregated user data shows that resting HR decreases significantly after age 40 - and that the United States and Singapore have the highest resting heart rate.

While the data from wearables can empower patients to make healthy lifestyle changes, Poterack says that currently very little wearable device data is utilized clinically, as existing infrastructures do not facilitate the easy integration of patient data generated from wearables into EHRs.

However, he added that “wearables provide a tremendous volume of potentially useful physiologic data” for physicians. Because of that, the healthcare industry is looking at ways to address the existing systemic and attitudinal barriers to utilizing the data.

The industry is starting to recognize and acknowledge the considerations in obtaining, curating, storing, and retrieving data from wearable monitoring devices. Poterack says that the technical challenges of collecting and storing the data can be overcome as more scalable solutions emerge. As that happens, clinicians will be better equipped to find ways to analyze and make use of wearable device data to improve outcomes at both the individual and population level.

Chronic disease management

While many wearables on the market have overlapping functionality, some are designed for very targeted clinical use cases. Wearables that help with Type 2 diabetes management, including prescription-based continuous glucose monitoring (CGM) systems, are now available from several manufacturers. Examples include Dexcom’s G6 CGM and Abbott’s FreeStyle Libre system.

Devices that can track glucose levels day and night - without a finger-stick - and automatically relay data are poised to help patients with diabetes more easily detect patterns and trends, which can result in better diabetes management. The systems typically include a small embedded internal sensor, a reader or app that can communicate to the sensor, and software that generates shareable reports.

This category of wearables is coming to market in response to the rapid global increase in diabetes diagnoses and need for simple monitoring tools. Marc Taub, PhD, divisional vice president of product development at Abbott Laboratories, warns that the diabetes epidemic is “severe and worsening.” The World Health Organization predicts that by 2040, an estimated 642 million adults worldwide will have diabetes, up from 425 million people today with the disease.

Taub says that while these patients should monitor their glucose levels four times per day, “the reality is that because of the difficulty of finger prick tests and available self-monitoring options, patients only check levels on average 1.6 times per day, which can put them at risk for an adverse event.”

Abbott’s CGM wearable, FreeStyle Libre, uses a sensor typically embedded in a patient’s upper arm that takes glucose readings from the patient’s interstitial fluid (ISF). While blood glucose readings from a finger-stick tend to be about 5-10 minutes ahead of ISF readings, an ISF reading taken from the embedded sensor means the patient can frequently self-monitor, and most importantly, avoid finger-stick tests. The sensor filament is approximately 0.4 mm thick, and the patient swipes a reader over the sensor for a one-second scan. Each scan provides the wearer a current glucose reading, a trend arrow, and an eight-hour history.

The FreeStyle Libre 14-day reader is FDA approved and is only available with a prescription. Abbott estimates that most commercially insured patients will pay between $40 to $75 per month for the FreeStyle Libre 14-day sensors.

Data sharing and analytics

Because the CGM systems regularly collect glucose readings, they could have value for larger clinics monitoring multiple patients as well as small or independent medical practices. The FreeStyle Libre system generates multiple reports per patient, including a snapshot, glucose pattern insights, daily log, daily patterns, and a glucose summary. The reporting software is free to download and use by both patients and physicians. Patient can share PDFs of the reports, or they can share all the data with their physician by linking a LibreView account to their healthcare practice.

Taub added that future embedded sensors could measure the quality of sleep, HR, and other health-related information. “That data will be collected and processed with connected devices in a way that is simple, with minimal human intervention,” he says. “The sensor data will support improved clinical decision support systems that use artificial intelligence (AI) and machine learning for data analysis. Having these regular monitoring systems in place will result in more efficient disease management and better outcomes.”

Pain management and rehabilitation

Data analytics are also playing a role in medical and therapeutic applications using virtual reality (VR). Boston-based VRHealth has developed immersive environments that use Oculus VR headsets to take a patient through specific movements and tasks. By monitoring patient performance and reactions during each VR session, the platform analyzes session data and, with the help of AI algorithms, builds a results portal for clinicians.

VRHealth has developed 10 different VR environments and is being used at Hoag Medical Group in Los Angeles, Beth Israel Deaconess Medical Center in Boston, and Stanford Health Care in Palo Alto, Calif.

Currently, the company has VR environments designed for upper extremity and fully body rehabilitation, cervical spine range of motion assessments, and motor cognitive training. Other existing applications from VRHealth include environments used for pain distraction, neck training, memory span testing, and a cognitive behavior therapy (CBT) application for hot flashes. The hot flash environment includes an AI therapist named “Luna” for psychological assistance and self-management of symptoms.

“Virtual reality environments and analysis takes the value of wearable data to a whole new level,” says Maya Bein-Nachal, vice president of operations at VRHealth. “By controlling the virtual environment for a stroke patient, a clinician can closely monitor recovery progress and quantify the patient’s interactions. The physician can challenge the patient’s body and mind by inserting different tasks and movements into their virtual environments.”

Remote patient monitoring for health intervention

Eran Ofir, CEO and co-founder of Somatix, has developed another targeted use for wearables: behavior tracking via gesture detection. By detecting simple hand gestures in real time via a single wristband wearable, the Somatix platform can deliver actionable insights to caregivers and increase treatment adherence for patients. “People constantly use their hands, and there is a wealth of health data and information to be found in these daily gestures,” Ofir says.

The Somatix platform compiles movement data from a Somatix wearable and/or with a variety of different existing devices, including a smartwatch, smartband accelerometer, and gyroscope sensor. The platform uses predictive analytics and cloud-based machine learning to passively monitor and remotely track massive volumes of detected gesture data coming from the wearable device. By doing so, the platform can facilitate CBT-driven health intervention.

“Gesture detection can identify a person’s activity and differentiate [between] a person brushing their teeth [and] someone shaving,” Ofir says. “By identifying micro-movements and patterns, a caregiver can know if an elderly person has taken their medication or possibly fallen. A healthcare provider or health coach could also know if a patient who should not be smoking is having a cigarette.

The Somatix digital health platform currently has two applications: SafeBeing and SmokeBeat. SafeBeing elderly remote patient monitoring is designed to help caregivers identify signs of irregularities in an elderly person’s routine. SmokeBeat is used for passive smoking monitoring. By monitoring gestures, SmokeBeat can issue alerts and enables variable combinations of personalized CBT incentives for improved adherence with a clinician’s prescribed cessation therapy.

Improving outcomes at both individual and population levels

With the growing popularity and functionality of wearables, they are delivering a steady stream - and in some cases a deluge - of information to patients and their healthcare providers. With their ability to monitor critical health readings, such as glucose levels, and to stimulate the senses through applications, such as virtual reality, wearables and smart sensors are destined to take on a more valuable role in diagnostics, rehabilitation, lifestyle and behavior change, and preventive care.

With more platforms and form factors, including easy-to-apply adhesive sensors, expected to emerge this year, physicians will have the capability to consistently monitor critical patient stats with real-time data from a wearable that can send an alert at the onset of a problem.

As this happens, wearables will take on more clinical value. “They have the potential to produce significant advances in individual health assessment, population health, and even healthcare delivery itself,” Poterack says.

Melanie McMullen is an experienced tech journalist and founder of BaySide Media in Oakland, Calif. She is the former editor-in-chief of Internet Magazine and LAN Magazine. Melanie has written hundreds of articles on smart technology, data analytics, sensors, mobile apps, and the Internet of Things (IoT). She can be reached at melanie@baysidemedia.com.