Data, so the saying goes, is the new gold – and when it comes to medicine, it’s hard to disagree. If nothing else, that’s clear from the statistics, with work by Verified Market Reports finding that the global market for big data analytics is already worth over $30bn – a figure that could double before the end of the decade. It’s a similar story, meanwhile, when it comes to the biggest industry players. In December 2023, for instance, J&J announced it was hiring some 6,000 data science and digital specialists, coupled with a swanky new research centre near San Francisco. Similar work is happening across the Atlantic, with AstraZeneca investing $250m in data-driven cancer research.

Nor is this enthusiasm hard to understand. Providing crucial insights up and down the pharma supply chain – and offering lab technicians and family doctors opportunities to understand patients and sharpen treatment – data is transforming medicine from a bewildering range of directions. But for that to happen,  of course, the information first has to be secured. Enter wearables. Encompassing watches and phones, they can gather reams of data on a subject’s daily routine before beaming it back to a technician many miles away. That’s echoed by yet more sophisticated biosensors, capable of measuring everything from blood glucose levels to muscle contractions.

It goes without saying that for patients and clinicals both, the pivot towards wearables could be transformative. All the same, it’d be wrong to suggest that researchers can just sit back and expect the findings to roll in. For one thing, distinguishing useful data from mere noise is challenging, especially for a profession weaned on strict clinical trial protocols. That’s echoed by making patients themselves comfortable with new technologies, even as technical and privacy challenges make themselves felt too. Step back, though, and the revolutionary force of such devices feels impossible to ignore – especially when you appreciate the newest generation of quantum sensors gradually entering the space.

Heart through sleeves

It’s hard to overstate the enthusiasm for wearables in contemporary medical life. As so often, the numbers here are revealing, with work by Statista finding that the global medical wearable devices sector could reach $83bn by 2026, up from barely $20bn in 2021. That’s echoed in specific markets too: the US alone could see the industry enjoy a CAGR of 14.6% through the end of the decade. At the same time, this energy is being taken up by particular companies. Eli Lilly and Merck are just two of the US pharma giants to invest heavily in wearable technology, while the European Commission is spearheading similar work from its headquarters in Brussels.

Speak to the experts, at any rate, and the boom quickly makes sense. For starters, suggests Dr Can Dincer, a sensors expert at the FIT Freiburg Center for Interactive Materials and Bioinspired Technologies, and the University of Freiburg, consider just how these devices make data collection – especially compared to the situation a few decades ago. Imagine you’re living in a rural location, Dincer says hypothetically, and you “need to drive two hours every week” to get something looked at for a trial or checkup. That’s not ideal – nor is the need for nurses to leave their clinic to monitor the vital signs of patients on the other side of town. But if the US still devotes some $400m a year to home medical visits for infants and children alone, Dincer says that wearable devices are quickly making many trips redundant. Dr Stefano Canali, a philosopher of science at the Politecnico di Milano, agrees, suggesting that “remote passive monitoring” is becoming an increasingly common idea across medical life.

Of course, none of this would matter if the devices themselves were unreliable. Fortunately – at least when it comes to pure data collection – that isn’t generally true. That’s apparent if you happen to own an iPhone: you may not conceive of it as a medical device, yet it faithfully registers the number of steps you do each day. When it comes to more specialised technology, meanwhile, sensors are powerful, versatile machines. Some, explains Dincer, focus on capturing physical signs: heart rates or body temperatures are two obvious examples. Chemical sensors, conversely, can be targeted to measure electrolytes. And while one-off data points are certainly useful, Canali says that what really makes these machines so special is their ability to “holistically” track the body – for instance, by helping us understand the relationship between heart rate and glucose levels.

Sensor applications

The practical opportunities here are unsurprisingly vast. At the University of Manchester, for example, a team of scientists is exploring how wearable sensors could provide insights into the vital signs of cancer patients. Other research is even more focused. At MIT, tests on mice suggest that the right sensor could detect fibrosis with an accuracy of 86%. That’s echoed by advances in so-called quantum sensors. Though both Dincer and Canali suggest that the technology is at an early stage of development – “I think that they are not really ready for a transition from the lab to real applications,” says Dincer – that could soon change. Collecting data at the atomic level, these devices could be a hundred million times more sensitive than traditional alternatives.

Some sensors focus on capturing physical signs, while chemical sensors, for example, can be targeted to measure electrolytes. Image Credit:

The number of new data science and digital hires recently announced by J&J.
Wall Street Journal

Even so, the path to a sensor-centric future is far from guaranteed. One involves the technology itself. Quite aside from the limited spread of quantum sensors, there are plenty of more prosaic examples here too. Though GPS can do a reasonable job of calculating your daily step count, for instance, distinguishing hiking and dancing from morning strolls is rather trickier. That reflects broader technical limitations. Though they’re good at detecting specific variables, after all, sensors struggle with causation. To explain what he means, Dincer offers the example of diagnosing Covid-19. Imagine, he says, that a sensor detects that a subject has a high temperature. The problem with such an approach, the academic stresses, is that sensors alone can’t tell whether such symptoms are caused by the coronavirus or just a regular cold.

The amount the US devotes to home medical visits for infants and children each year.

Canali, for his part, emphasises the challenges sensors pose to medical philosophy. Especially for trial conveners – but also, to an extent, for regular doctors – clinical evidence is valid if gathered in formal settings. But placed on flesh-and-blood subjects going about their daily lives, wearable sensors are by their nature informal, with none of the double-blind tests or placebo controls inherent to other spheres of medical research. Altogether, stresses Canali, that can make some researchers “struggle” with the concept.

Wearing thin

Combined with related problems of getting patients themselves to employ wearable devices properly – especially around taboo issues like alcohol consumption, some subjects may artificially decrease their intake, distorting the data sensors receive – and it’s no wonder that both Canali and Dincer are unsure about the future of wearable sensors. Notwithstanding the sector’s headline growth, that’s reflected in practice. In February 2024, to give one example, the FDA warned against smartwatches that claim to measure blood sugar levels without piercing the skin.

The size of the global wearable devices sector by 2026.

Underlying these technical questions, meanwhile, is the question of data management. Beyond the usual worries around cybersecurity, which could soon be solved by relying on blockchain and other innovations, Canali fears that tech giants may hoard the data their sensors produce. “In some cases,” he says, “scientists have problems accessing the data themselves – even in research settings – because the data collected by technologies are proprietary algorithms from Google or something like that. I think the discussion on protection and privacy is crucial, and should also align with discussions of who’s the owner of the data, who keeps the data – and where the data is kept.”

Fortunately, there are signs the sector is taking these concerns to heart. In the EU, for instance, the European Health Data Space is a new initiative aimed, it says, at supporting “individuals to take control of their own health data” while also creating a shared set of rules around how that can occur. For its part, the FDA is publishing similar guidance.

Apart from the obvious strengths these devices offer generally, there are other causes for optimism too. As an example, Dincer envisages a world where sensors are gamified, and users are encouraged to boast on social media if they keep their calorie or drink intake to healthy levels. That dovetails, Dincer adds, with the pressures of an ageing population. As he explains, using sensors to understand how we get old “may allow the users to improve their lifestyle, leading to subsequent healthier ageing”. Given the broad potential of wearable sensors, such an eventuality wouldn’t be surprising.