Bench Talk

Sometimes, it seems like the Internet of Things (IoT) concept is being shoehorned into markets which don’t need it—or at least, aren’t ready for it. But there are undeniable synergies between the IoT and healthcare applications, particularly medical devices that are wearable (or even implanted under the skin). Wearable medical devices must be small, unobtrusive, power-efficient and reliable—these are all challenges that the IoT takes in its stride. However, making the human body a part of the IoT also introduces some unique new challenges.

The time patients spend with medical staff and in medical facilities is expensive. There’s great potential for cost savings if we can make better use of those human and physical resources, for example by reducing expenditure on routine tasks such as health monitoring and daily drug distribution. In 2010, the US estimated that the Medicare programme was wasting $17 billion annually by unnecessarily bringing recovering patients back in for basic procedures like check-ups and prescription refills. The Medicare losses were so significant that the government introduced rules to penalize hospitals that re-admitted patients with certain conditions, including congestive heart failure, within 30 days.

As a result of pressure to increase efficiency, we’re already beginning to see large-scale trials in which basic tasks are streamlined by a nascent healthcare IoT. Cost-saving opportunities like these show why the global healthcare IoT market could be worth almost $410 billion globally by 2022, if predictions from Grand View Research prove correct.

Every smart phone is, potentially, a connected health monitor. But, while there are numerous health-related apps for phones, in practice many of them are of questionable value, as standard mobile phone sensors are not ideal for the collection of accurate data about their users. A smart phone, however, can work effectively as an IoT gateway node to gather data from dedicated sensors and to provide a familiar user interface to control smaller devices.

First-generation wearable devices with an emphasis on health monitoring, like Fitbit and Apple Watch, already have a market value of over $2 billion. The more advanced of these devices automatically keep track of heart rate, sleep periods and physical activity. They can combine that data with user-supplied information, such as diet.

The US National Institutes of Health website lists 130 clinical trials involving the Fitbit alone. However, there is a hitch: These consumer-grade products are specifically advertised as “not a medical device,” and there are questions about the accuracy of their heart-rate monitoring.

Moving beyond consumer electronics, Netherlands-based Radboud University Medical Centre (Radboudumc), using technology from Philips, is testing wearable IoT devices to support patients with chronic obstructive pulmonary disease (COPD). COPD causes severe breathing difficulties and kills over two million people annually worldwide. In the trial, patients wear a small sensor on the chest that continuously measures their breathing and heart rate while they go about their everyday lives. The data is sent via the Internet to healthcare providers. This allows automatic alerts in emergencies and enables doctors to monitor long-term changes. It also allows the patients to get an objective view of their own condition and learn how changes in behaviour affect their health.

While constant monitoring could seem disturbing, it can also bring greater peace of mind for patients and their families. A heart patient who joined a similar remote monitoring trial in Arizona said, “It does touch you emotionally, knowing someone is watching out for you. It’s just feeling that back-up is there, knowing you have support.”

The obvious benefits of melding the IoT with healthcare suggest that we are rushing headlong towards a future in which doctors prescribe IoT devices as commonly as they prescribe medicine. However, bringing big data to the human body is not without its challenges, and the truth is that although there are countless trials underway—and some of them may expand to become standard procedure—full-scale roll-outs of medical IoT systems are still rare outside of hospitals.

The primary reason for that caution is that healthcare is literally a matter of life and death, so legal liability and ethical issues are paramount. This even forces developers to solve problems in ways that may seem counterintuitive, from an engineering point of view. For example, early implanted heart defibrillators used batteries, which patients could recharge with an inductive charger. Lawyers advised against this, because the manufacturers and medical service providers were at risk of legal challenges if the patient failed to charge the device. The result was that non-rechargeable, multi-year batteries were used. From a legal standpoint, it was cheaper to replace the whole device every few years, rather than trusting the patient to recharge it.

There are also practical challenges. Wearable devices must be comfortable for long-term use. If in contact with the skin, they must be encased in hypoallergenic materials, maintain a low operating temperature to avoid noticeable heating, and be resistant to sweat. Any device implanted under the skin must meet stringent medical standards for sterility and stability. The human body is a hostile environment for electronics, very different from a climate-controlled server room.

Biologically stable devices are well understood in the medical world, but they may be new to IoT developers with an electronics background. There are other challenges, too, which are already familiar to IoT developers: Software safety and security. Increasing functionality increases complexity, and that increases possible points of failure. Two-way network connectivity vastly increases flexibility, but also allows outside interference.

While the security issues are already familiar from other Internet-connected devices, perhaps medical IoT device designers can learn a few lessons about robustness from avionics and space hardware, where it’s standard to have duplicates of critical modules (either on standby or working in parallel), multiple paths in software and hardware to vote on decisions, and a separate, simple fail-safe monitor to oversee the whole system.

Obviously, this “strength in depth” approach puts unwelcome pressure on budgets for power, development cost, and device footprint. But it’s notable that implanted heart pacemakers already incorporate some of these features, and the most important international standard for medical device software, IEC 62304, mandates the use of independent risk control measures for safety.

While healthcare will be transformed by the IoT during the next decade, the IoT will also have to evolve to meet the unique demands of healthcare.