Ever since the first smartwatches came to market, there has been significant interest in how they can be used to improve healthcare. There has also been skepticism about how valuable these wearables can be, including concerns from Stanford researchers that data gathered from wearables was “way off the mark” just a few years ago.
However, the latest release of the Apple Watch should put that debate to bed. Among the many new features made available to the public, one that grabbed the attention of many in the healthcare industry, was the addition of an ability to detect atrial fibrillation (AFib) by a self-performed electrocardiogram (ECG) – an early warning sign of potential heart disease for users.
Even before the ECG function was announced, commentators had been asking what other features we might see in future versions, such as blood sugar tracking capabilities, and how this might tie into the broader demand for personalized medicine as people look to take greater control of their own health.
As Elsevier’s VP of Life Science Solutions Services Tim Hoctor put it:
With the popularity of wearable tech and the hype created by global tech companies such as Apple and Google come vastly increased expectations among the general public. By generating and collecting health data, users believe that their data can be put to use to aid research and benefit patients.
However, the real question is whether the data from wearables can go beyond helping us as individuals and benefit society by assisting in unlocking scientific discoveries such as new drugs or treatments. If that is to happen, it means overcoming two key barriers: a lack of collaboration and deploying the right platforms to manage the vast amounts of data being generated.
The path to greater data
At an individual level, we’ve already seen huge leaps in the sorts of data that wearables can provide us about our health. Year on year, the accuracy of wearable devices improves. On top of that, as manufacturers continue adding more functionality, such as sleep quality tracking and glucose monitoring, the depth of the data available is far better, too.
They are also increasingly popular. More than 20 percent of Americans will use an activity tracker of some form by 2021, while 33 million are expected to be in use by 2020 in the UK. Around the world, trials are already underway for a range of specialist devices like non-invasive wearable diabetes sensors. Their popularity and increasing functionality means multiple data points from around 65 million users is coming in every day. Such a treasure trove of information could be invaluable to pharmaceutical researchers working in the field of drug discovery by narrowing down the parameters of clinical trials, provided the information could be fed back at an early stage.
Watch and learn
The problem is that the data from our wearables is largely being gathered by tech companies, not scientific researchers. And although digital giants like Apple may have some of the required skills, they’re not set up – or currently inclined – to uncover medical advances. Nor should we expect them to be, given how far away that is from their core business, Hoctor said:
The reality is that while tech companies are good at gathering data, they do not understand the true scientific implications and lack the expertise to actually uncover insights from the data they hold. It’s important that people understand that wearable tech is good at collecting simple, isolated data points – such as your heart rate during exercise or how you slept last night – but it cannot provide a clinical analysis of that data.
In short, tech companies cannot support R&D and drug discovery by themselves; they need the help of experts who are able to make sense of the data and understand clinical inferences. We are gradually seeing moves towards the use of wearable devices in clinical trials to collect results from patients, which is exactly the kind of partnership that can aid future research and improve understanding of how patients respond to treatments.
As Hoctor highlights, if the opportunity to make societal level advances from wearables data is to be exploited, the likes of Apple and Fitbit should be collaborating with partners from the scientific research sector – pharma companies, academic groups and regulators to name a few.
This call for collaboration was echoed by nonprofit organizations The Pistoia Alliance and the Michael J. Fox Foundation (MJFF) at Pistoia’s annual member conference. At the event, Sohini Chowdhury, Senior VP of Research Partnerships at MJFF, commented:
From patients to physicians, from scientists to trial sponsors – everyone wants more data. Wearable devices will be a crucial self-reported source of that data, but we know there is still a hill to climb toward field-wide integration. Now, we’re trying to close the loop and explore the value of wearables for the entire ecosystem. By pooling our resources, we can build consensus and make these reams of data accessible – which will advance our understanding of disease and testing of new therapies.
Scientists and researchers have the skills to spot trends within the data and understand scientific inferences, but without the level of integration envisaged by Chowdhury, the data simply will not be there to open up routes to new therapies or precision medicine treatments. There are significant hurdles to be overcome before such data-sharing can become routine, and it’s a matter of if rather than when it happens.
Providing a platform
Provided the tech and pharma industries do surmount these challenges, the question then becomes how to make sense of so much data. Given the amount of information that wearables produce every day, there is no question of analysis being done by human researchers. Since attempts to use generalized AI tools have demonstrably fallen flat, the only solution is to use purpose-built Artificial Intelligence (AI) and Machine Learning (ML) platforms, designed to handle clinical data and able to contextualize the information they are getting from wearable users. As Dr. Jabe Wilson, Consulting Director of Text and Data Analytics at Elsevier, outlines:
To usher in an age of AI, far greater collaboration across different disciplines and geographies is needed. It’s important to remember that while AI has great promise, it’s not simply ‘plug and play.’ The use of AI in healthcare will necessitate purpose-built platforms that are not only technologically advanced but scientifically nuanced. Such platforms will require huge volumes of accurate, varied, multi-disciplinary data, along with many years of training and algorithm-building by human ‘masters.’
The scientific nuance Dr. Wilson mentions will be key to how wearables are used in clinical trials and experiments. Researchers are already using wearables to track patients in clinical trials for diseases like Parkinson’s or heart disease. If these researchers are using an AI platform to monitor subjects and notice, for example, that some of the subjects in a particular area are experiencing particularly strong tremors, the AI needs to be smart enough to consider a range of options as to why this could be. Are the tremors happening at regular intervals? Do they happen whenever people are on the move? If so, perhaps the reason is that these subjects live in an area with cobbled streets, and driving or cycling produces lots of non-Parkinson’s related tremors. The ability to delve into the data and ask these contextualizing questions on a mass scale is the key to unlocking hugely beneficial insights.
The opportunity for medical breakthroughs thanks to mass real-world wearable data is huge. Smartwatches and other devices could provide researchers with both higher quality and higher volume data on patients than ever before – provided they are given the chance to properly interrogate it. However, there are numerous stumbling blocks that need to be addressed before that becomes a reality, including the need for far greater collaboration and the deployment of purpose-built platforms to manage the deluge of data. Without addressing these issues, wearables will remain tools that benefit individuals but won’t transcend to the societal level.