Ready to learn Analytics in Healthcare? Browse Analytics in Healthcare & Life Sciences courses developed by industry thought leaders and Experfy in Harvard Innovation Lab.
The divide between consumer health and fitness wearables, and medical-grade devices is slowly being bridged as technology evolves to offer advanced sensors and form factors that combine the best of two worlds. The result is vast amounts of higher quality data to feed the complex algorithms which not only deliver the personalized results so often discussed in the industry, but the ability to even preempt negative health events.
FitBit’s application in clinical trials is one example of the potential for consumer healthcare devices. FitBit is said to be used in 500 or so clinical trials currently, including trials for congestive heart failure where a continuous picture of activity is relevant to the provider. You’re even seeing its use in pharmaceutical clinical trials for neurodegenerative disorders, where the ability to remotely track motion is very impactful as to how the drug impacts a patient day-to-day, instead of only gathering this data from patient diaries. These supplementary data points on activity add value, but they have to be considered in context. In other words, the accuracy and quality of data from consumer wearables are wide ranging, and thus their use needs careful scrutiny when it comes to medical applications.
However, the growing market of medical-grade devices, designed for consumer/patient access without the need of a prescription, is enabling a new category of remote patient monitoring. This emerging category now includes blood pressure cuffs, EKG recorders, continuous temperature monitors and more, that are also connected to the network. These devices can gather the types of medical-grade data traditionally only accessible via providers during an office visit. But now they are empowering patients and consumers to self serve remotely, and share the data with any provider that they choose – enabling a new category called remote patient monitoring (RPM).
There are three key considerations when it comes to RPM and the Internet of Healthcare Things (IoHT); a world where devices remotely and continuous gather meaningful data on the state of our health before, during and after a health event.
Enough Data to Power Precision Health
The vision of IoHT is that we can gather enough data to get a full picture of an individual’s health state in order to personalize their healthcare. However, it will not only depend on an individual’s data, but collective data from the population. Wearables are the impetus behind this vision and the way to afford the collection of data outside of the clinical setting on a mass scale. As form factor and accuracy improve, adoption will grow and a plethora of data will be available and can then be targeted to the individual.
Data Quality Means Controlled Volume
As more devices enter the market and are continuously monitoring patients, the potential to end up with too much data, which can be overwhelmingly useless, is high. Machine learning and other algorithms applied to find value in the data can only do so if the volume and quality is controlled. For example, some classes of devices like continuous temperature monitoring wearables have the ability to take readings every second, but they also have the ability to allow an application to filter and control the data volume in order to make the best use of it. This focus on quality and volume, meaning not only that the data is accurate, but that it is manageable, is important in a future where devices connect mass amounts of data directly IoHT systems.
Data Context to Drive Value
Data context may be the forgotten factor in finding value in big data from wearable health devices. This is where application solution providers must ask themselves, “What issue are we trying to address?” Data context also takes into consideration the environmental factors that impact the data. Take an arthritis patient for example – motion data directly from the patient is useful in understand their daily exercise levels, but humidity may also be equally important in understanding a recent episode. On the other hand, humidity may not be so interesting for a patient healing from a broken bone. Only in the right context of what is going on around a patient can the data gathered continuously on the state of their health be truly relevant. Providing data context is an area of opportunity and one we’ll see continue to mature.
The medical-grade devices and consumer wearables markets are converging. This will change how we interact with healthcare providers and how we receive care. The internet of things analytics market is expected to reach $60 billion by 2025, with healthcare accounting for 70 percent of that market. This includes medical-grade devices that utilize new and existing sensors. When the data gathered from these devices achieves volume, quality and context, we’ll see machine learning and AI achieve the dramatic improvements in disease management, reductions in clinic and hospital visits, and the personalized prevention and treatment plans we’ve been promised.