Shane Walker, Senior Principal Analyst, Healthcare Technology
Digital health was a continuing trend at CES 2018 with 26% of exhibitors having an offering within biometrics, health and biotech, sports and fitness, or wearables. Looking at just companies related to digital health, more than 18% of exhibitors were related to consumer-side healthcare. IHS Markit forecasts the global market for digital health will grow from $15.6 billion in 2017 to $21.6 billion during 2021 at a CAGR of 8.5%, rapidly outpacing the traditional medical equipment markets. While much of the show was an evolution from previous years, including sleep monitoring, embedded sensors in clothing, and wrist-worn activity trackers, a few items incorporated broader technology trends such as machine learning, blockchain, medical-related wearables, and home-based diagnostics.
Digital Health includes:
- Consumer Medical – Self monitoring medical solutions, hardware, software and services, that are markets directly to the consumer.
- Sports and Fitness – Self monitoring solutions, hardware, software and services, that cater to sports and fitness use cases.
- Virtual Healthcare – Clinical services pushed through virtual channels that support digital engagement and remote monitoring.
After much work on using machine learning for population health management, hospital revenue cycle management, ICU monitoring, and diagnostic analytics, now the application of machine learning to digital health is emerging. An example of this from CES 2018 is Selvas AI and their Selvy Checkup AI disease prediction service. The Korea-based company is building on 20 years of human-machine interaction technology development (handwriting, voice, optical character recognition, and big data analytics) with a second entry into the healthcare space. Selvy Checkup is intended to be a medical service that predicts the probability of incidence for six common cancers using the company’s machine learning platform. Prior to Selvy Checkup, Selvas AI had launched Accuniq, a body composition analyzer that also draws on machine learning to anticipate user habits and make recommendations.
As part of the IHS Markit Digital Health Intelligence Service, the latest report on artificial intelligence (AI) in digital health notes that the added resources of chat bots, computer vision, machine learning, and natural language processing will enhance foresight, improve decision making, and enable automation. In the short term, AI in digital health is expected to add scalability to virtual healthcare with health bots for primary care and with resources to automate and improve management of remote patients. Unique health bot encounters are expected to pass the one billion mark in 2021 (in Canada, China, France, Germany, Japan the United Kingdom, and the United States) and separately, machine learning is forecast to enable the remote monitoring of more than one million patients by the same time. Despite the increasing uptake, there are substantive limitations to AI systems for data qualification, contextual awareness, and workflow integration.
On the patient side, individuals are increasingly empowered to make informed health decisions via AI, both in terms of personal medicine and fitness/nutritional purposes. This is facilitated by virtual health assistants and applications delivering personalized feedback, sometimes in real time. Voice is the next big play in health bots or virtual health assistants as natural voice with appropriate permutations allow for a human touch.
At CES 2018, there was an exploration of blockchain beyond currency, including many contractual events and authentication models. Within the healthcare sector, blockchain was noted to potentially affect health record transfers, legally binding automated contracts for payments, and the tracking and prevention of counterfeit drug sales.
For example, a startup company called Bowhead Health has plans to implement blockchain into its personalized healthcare platform as of Q2 2018. Bowhead’s platform will incorporate a connected device that dispenses nutraceuticals and refills automatically. A user can monitor their wellness through the Bowhead mobile app and store the information with the same cryptographic encryption as Ethereum wallets. Users can also provide their anonymous data to researchers through blockchain and receive Anonymized Health Tokens ($AHT) in return. Each research request will be viewed by Bowhead users individually, so each request can be evaluated, and then accepted or declined. Bowhead expects the platform to eventually include genetic information from users, along with data stemming from other monitoring and diagnostic devices in the home. Among these early diagnostic capabilities is a test for nutrient and hormone imbalances from blood and saliva samples. A combination of six supplements and pharmaceuticals are dispensed and recommendations are also available by certified doctors and wellness professionals.
Wearables for Health
With more than 580 wearables exhibitors, highlighting all the interesting developments in the world of wearables is a difficult task. However, the following are a few notables related to healthcare.
- The OrCam MyEye 2.0 is an artificial vision device that reads printed and digital text aloud from any surface, and recognizes faces and products including UPC codes. Designed for the blind, partially sighted, and for those with reading difficulties, the glasses-mounted camera follows hand gestures to direct its line of sight and recognize what the user wishes to identify. OrCam was founded by the team from Mobileye, the collision avoidance system and autonomous driving developer.
- The Resound LiNX 3D 2.4 GHz hearing aid with binaural directionality announced a new feature, ReSound Assist, that enables users to remotely connect with their hearing professional through the ReSound Smart 3D app. This allows adjustments to be made to the hearing aid without an office visit, which typically requires several follow up visits to optimize the hearing aid settings.
- Suunto’s Movesense all-in-one sensor (accelerometer, gyroscope, magnetometer, temperature, heart rate, and ECG) has now been accessed by 400 developers, including many exhibiting at CES such as Ain1.ai, Runteq, Supa, and TriMix. Ain1 is a Finnish startup focused on development of a lightweight platform for coaching and rehabilitation.
- Samsung Gear, paired with a smartphone camera, the Family Hub refrigerator, the Samsung Bixby voice assistant, and the Samsung Health app, provides an advanced level of health coordination for activity tracking, nutritional information, calories burned and allotted, and meal recommendation.
- During early 2017, World Media & Technology Corp. announced that their research in non-invasive, continuous blood glucose estimation had been found to be sufficiently accurate to allow the company to proceed with integration of the algorithm in the Helo platform utilizing a photoplethysmograph (PPG) sensor. The application has yet to be integrated, but is expected to be available during this year. The Helo device currently features monitoring of blood pressure, ECG, heart rate, respiratory rate, heart rate variability, step count, and calorie burn. The device and platform will continue to be considered non-medical.
- Wearables for enterprise—Valencell continues to find more partners for its biometric sensor technology, including those focused on the market for in-ear heart rate monitoring, also known as hearables. But the company is also working on enterprise applications such as integration with construction hardhats, VR for pain management, and driver fatigue monitoring. The National Institute of Standards and Technology (NIST) is also researching wearable technology for enterprise applications, such as the manufacturing workplace. This research has involved incorporating nontoxic gold wires into flexible plastic film that can be used for health monitoring. Gold does not corrode, but it will crack when bent, so the research is focused on developing a porous membrane that enables conductivity as well as flexibility. The long-term goal is to enable the tracking of hundreds of real-time physiological changes, including potassium levels in sweat, and sugars and proteins in the blood stream, which can highlight the onset of disease before they are detected by physicians.
Clinical Care Continues to Move to the Home
While the CPAP is already a home-based clinical device, Philips Respironics has introduced a smaller form factor travel CPAP system called DreamStation Go. DreamStation Go features 1/3 fewer components than other travel machines and is half the size of the System One and DreamStation platforms. Two other examples of clinical care moving into the home are EarlySense and TytoCare.
- Israel-based EarlySense, the patient monitoring platform with a contact-free sensor that sits under the mattress, continues its move into the consumer space with the fertility monitor Percept. The clinically-proven sensor does not require urine samples or any wearable, which makes it a convenient way of tracking and analyzing ovulation, menstrual cycle, sleep patterns, and relaxation levels. The sensor takes the body’s signals from sleep (heart rate, breathing rate, and movement) and differentiates them using proprietary signal analysis. From this data ovulation dates are predicted and AI improves prediction accuracy over time.
- Another Israel-based company, TytoCare, provides several FDA Class 1 devices, including connected otoscope, camera, thermometer, and stethoscope for home-based health monitoring. The devices are used as part of the TytoHome telehealth platform. Examinations can be conducted in real time or the data can be stored in the cloud to be shared during a scheduled video conference with a physician.
Expanding Digital Health Beyond Early Adopters
The consensus from most players in the digital health market is that consumers are interested in sharing their personal information when it improves their health outcomes or empowers them to better care for themselves. However, there have been cases when sharing data is problematic or where the process is too complicated or non-intuitive, and this may lead to a potentially negative experience. This can create a barrier to reaching an audience beyond early adopters.
Early digital health adopters tend to be very proactive, well-engaged, and healthy. The challenge is reaching the large population of people managing chronic conditions, who may be skeptical about how the data will be used. This is a challenge worth undertaking, and one that will drive uptake of digital health devices and services.
In the near term there are mixed signals in the US market for digital health, which is one of the largest markets for both devices and services. The FDA plans to maintain a focus on advancing digital health, including expediting innovative technology into the market. Digital health should also prove to be a valuable tool within bundled payment programs, particularly for post-acute care. That said, the Centers for Medicare & Medicaid Services (CMS) has ended two of these mandatory bundled payment programs for cardiac and hip replacement, along with reducing the number of regions included in the mandatory joint replacement model. However, this setback in the move toward value-based care may be temporary, as CMS is still working toward a long-term goal of shifting volume to value.