As wearable technology become cheaper and more widely used, it's time to begin conversations about confidentiality and how best to use the data to improve health, writes Adam Drury in the HSJ.
Over the past two years there has been a significant rise in wearable devices that support personal fitness and health. Sales of fitness trackers (from market leaders such as FitBit, Jawbone and Garmin in the UK) have risen exponentially in the last 12 months. For example, Fitbit sold 21 million units globally in 2015 and had revenues just short of $2bn. Adoption levels for wearables are still growing, with a compound annual growth rate of 25 per cent expected over the next five years.
Costs of such devices have fallen significantly in the last 12 months, and new models are regularly being released. There is an increased investment in the underpinning 'cloud based' wellness platforms to help manufacturers differentiate themselves through user experience (such as the LifeLog platform from Sony, which records general lifestyle as well as health information).
Fitness tracking data (such as number of steps, amount of sleep, amount of exercise, calories burnt and heart rate information) is starting to improve in quality, and the range of data that can be captured is increasing as the quality of sensors improves, the cost goes down, devices get smaller and battery life improves.
Furthermore, fitness and health sensors are also being integrated into other items from some of the industry leader players - such as the Apple Watch, Samsung Gear S2 and Google Fit - meaning that fitness tracking, practicality, fashion and mobile technology are starting to merge. As a result, the healthcare sector is clearly seen by technology corporates as an area for major growth in future.
The range and use of apps that support fitness and health is increasing. More than 30 per cent of mobile owners use fitness apps, and usage has increased 60 per cent year on year since 2014. Over 100,000 health and fitness apps are now available for download.
Market leaders such as RunKeeper, Strava, MyFitnessPal, MapMyFitness, and Endomondo dominate the market through bundling their apps with a range of fitness tracking devices, while others have bespoke, proprietary apps to support the device (e.g. Sony fitness trackers predominantly work with their own 'Lifelong' application). With so many vendors competing in the personal health data space, there has to date been no real effort to establish a common set of interoperability standards for storing health and fitness data from tracking devices to enable access from a range of different apps for secondary analysis. There is a rise in the number of Internet connected devices that can store and transmit health related data
The Internet of Things ("IoT") is transforming the ability of common devices to store and transmit data to a phone or over a Wi-Fi connection. This has traditionally been in the realm of disease management devices (such as blood sugar trackers and medication adherence management systems) but is now moving more into consumer health. One great example now hitting the mass market are IoT scales which can measure weight, body fat index, different types of body fat, resting heart rate etc., and store information in an app or the 'cloud'. Companies such as Withings, who were one of the first manufacturers of digital health devices, and have recently been bought out by Nokia, see such technology as the next step on from wearable devices to provide a wider range of information to the consumer at an affordable price.
However there has also been a rise in the range of personal health devices which can give health information on your phone - For example: AliveCor offer a £90 device (available from Amazon) that attaches to the back of your mobile phone and takes a mobile ECG trace that can be seen on the phone screen. This is designed for patients with (or at risk) of Atrial Fibrillation (AF) Wireless Blood pressure monitors have become widely available, with prices from £90 for products from companies such as Withings.Both disease and personal health management segments have become cheaper, smarter, easier to use, and more prevalent amongst the general public.
As these devices proliferate, there is a range of immediate and longer term potential benefits associated with their use: Better tracking and measurement of health indicators (e.g. daily exercise, sleep quantity and quality, blood sugar measurement). People taking increased ownership of, and accountability for, their personal health and wellbeing (e.g. public consciously tracking their sitting time, patients tracking their sugar intake). Routine and predictive indicators to maintain and improve health (e.g. notifications for when to get up and walk, when to take insulin).
One major issue now facing users is that many of the fitness trackers and IoT devices store information in their own proprietary apps or databases, meaning that it is difficult to connect / integrate all of your personal fitness and health information into one place. For example: The information from my IoT scales is stored in a different app and a different format to my fitness tracker
It is difficult to export data from apps to do analysis and 'cause and effect' testing to understand (for example) how my weight impacts my fitness (and vice versa). Partnerships are starting to be formed between device manufacturers and apps developers to enable data sharing (e.g. Withings partnership with Nike, MyFitnessPal, and Runkeeper), but such developments do not integrate data but share a copy of the data from one app with another.
This situation is not helped by major market players such as Apple, Samsung and Google developing their own toolkits for their devices which are specific to their own brand of hardware and therefore not truly interoperable. Another key issue is the provenance of the information generated from such devices by primary and secondary care clinicians. For example:
What value will clinicians place on user recorded and/or smart device generated information in the overall clinical decision making process? Will they make their own assessments and measurements anyway? Even if a clinician did want to use the information, how would they access it and would it be stored as part of the main clinical record, as an adjunct, or through a link (with the patient's permission) to a personal health record? Recent partnerships between Epic, one of the US' leading Electronic Patient Record providers, and Apple regarding integrating information from the Apple Watch into the EPR record is a welcome development in this area.
Privacy concerns are seen as one of the major blockers to the widespread use of wearables data in the health domain. Addressing issues such as 'Who has access to my information, and for what purpose?' will need to be resolved before the use of personal health data becomes a more common part of the clinical decision making process.
So what are the next steps / ways forward?
First, further progress will require an agreed set of international standards (that are non-supplier proprietary) to enable: Standardised recording of health data in apps and their back end databases;
Interoperability between devices (using existing standards such as Bluetooth, Wifi, NFC and HyperCat); and Open APIs to enable apps to pull data from many sources and development of more sophisticated analytical tools.
Second, there is a need for open debate between clinicians - and between clinicians and their patients - on how they could use information generated from such devices as part of the care process, so patients can understand what to expect when a clinician accesses information from personal health apps and devices. This should take the form of challenging both clinicians and patients to use the technology to its full potential, and may lead to re-designing care pathways to make best use of user-generated data, as well as to changes to training to help clinicians deal with the increasing range of data that will be available to them.
There is a need to establish the case for wider use of devices and encouraging the use of them: there is little comprehensive evidence for the benefit and use of these devices, making it difficult for clinicians to advise their use or even prescribe them. There is a need for real-world evidence to be gathered and analysed on the impact of these devices, with health outcomes and cost impacts estimated. There should also be an open conversation on which types of devices have the largest impact and on which types of patients - targeted outreach, as opposed to a blanket approach, might be more impactful for chronic disease management. Models of how core clinical record information and personally generated information are considered in totality are needed - systems at the moment don't support the bringing together of these two worlds of data together.
In conclusion, the integration of consumer health devices into the mainstream of long-term condition management has huge potential, but is an opportunity which we risk missing if a number of key questions and issues are not addressed. Doing so will require the health professionals and, most importantly of all, the patients, to recognise the potential benefits and lead the debate, ensuring that technology and healthcare practice can be brought together in the most effective way possible to improve outcomes and reduce service costs.