By Glenn Craib, Technical Analyst at Jumpstart
Many people nowadays walk around with devices tracking their every move. Smartwatches and dedicated fitness trackers allow users to measure many aspects of their daily life and fitness activity. There are even cases of insurance companies offering smartwatches on a discount to customers who can demonstrate a specific level of activity.
But just how accurate are these devices, and how relevant are they to health professionals?
Recent testing by Which? has shown that the accuracy of the step counting and distance measuring of these consumer devices varied dramatically. During their testing, which simulated a marathon distance, the devices they tested ranged from underestimating the distance by up to 10 miles or overestimating by up to 6. While this testing shows that comparison of different devices readings is problematic, many users don’t use them for this purpose. They are simply looking to compare their own personal reading from one day to the next, looking to go further or quicker than before, or to hit a certain daily goal.
Initially these devices only measured step count and in some cases, location, using GPS. But, as sensor technology progressed, other features have gradually been added. Many devices now offer heart rate monitoring, which is achieved by use of photoplethysmorgophy (PPG) which uses green illumination of the skin, monitoring any reflected light. Some more recent devices, such as the series 4 iWatch from Apple, even offer medical alerts. As such, they have to be approved by the FDA in the US. These advances have resulted in dramatic improvements in accuracy, so that many devices now read within 1 – 2 beats per second of a chest mounted monitor.
So, even though consumer health trackers may have accuracy issues, they can still provide a useful guide. Some companies however, are already looking at the next level, where trackers would be to able to accurately monitor relevant patient vital signs and report the reading directly to health professionals. These devices are typically designed to be worn on the upper arm, held in place by a strap to reduce the effects of motion. Alternatively, the device can be set to track motion (steps), and only measure health data when the patient is stationary. Newer sensors and associated algorithms can measure steps (motion) and heart rate, as well as track blood pressure, temperature and oxygen levels within the blood. Sensors being developed today also include blood glucose and sweat measurement (for salt content for example). All these sensors are non-invasive, easy to use and replace when necessary.
These clinical level devices can provide a range of information from a patient at much higher measurement intervals than is typically achieved today, even within a hospital. Clinical trials are being carried out to ensure the accuracy of this new generation of devices, comparing their output with more traditional methods of measurement.
Data gathering and reporting of real-time or near real-time data is only the first stage. The next step is predicting the patients’ condition. AI and predictive algorithms can be trained in what to look for within the patients’ sensor data, in order to allow for earlier medical intervention, and also provide patients who might otherwise have required close monitoring, with more freedom of movement.
While consumer fitness tracking devices often suffer from a lack of clinical accuracy, they do have an important role in promoting general fitness. The next generation of devices, targeted at the clinical market could result in major changes to patient monitoring and care.