You can download a lot of different apps for your smartphone. In fact, from the Apple App Store alone, the number is now over 900,000.
In the not-so-distant-future — among those hundreds of thousands of apps — we may be instructing our patients to download their pancreas app.
Okay, that is a bit facetious.
To be more accurate, we could be instructing our patients to download a controller app for an implantable artificial pancreas, or for that matter, any number of implanted medical devices.
There are currently at least three active clinical trials utilizing smartphones in some capacity as part of closed-loop insulin delivery systems. These studies have important implications for the use of smartphones in other implanted medical devices down the road.
In these trials, the smartphone is paired with at least a sensor and a delivery system. These highly-controlled trials are generally in the safety and concept validation phase; for example, patients in one trial are being closely monitored in an out-of-hospital setting overnight (a hotel room) by a clinical team operating remotely (a different room at the same hotel).
In one trial by researchers at UC-Santa Barbara and UVA, patients use the smartphone to monitor sensor data, insulin administration data, and enter carbohydrate intake. In addition, users may adjust basal/bolus dosing, though, this is not a requirement in the closed-loop system. In the initial studies, patients started in an open-loop operation mode using their home pump settings before converting to closed-loop operation. As such, in this study, the system itself will only act to modify user-input settings if the blood glucose falls outside of preset parameters (1).
In another trial being run by the JDRF and Legacy Health System, the smartphone itself is being used to calculate insulin or glucagon doses based on data from continuous glucose monitors (2). In a third study supported by UVA along with JDRF, the Sansum Research Institute, and University Hospital (Montpelier), the smarpthone is again used as the controller device to process continuous glucose measurements and adjust insulin dosing (3). Both studies also use Android operating systems.
It’s worth noting that the above-noted clinical trials utilize Android devices in which the “phone” functionality has been disabled, essentially using the device only for its wireless connectivity and its user interface. This is doubtlessly a security precaution given the high risks associated with either intentional or unintentional system failure.
There is, however, ample reason to be hesitant about integration of smartphones with the operation of implanted medical devices. There have been warnings in the past that implanted cardiac devices were vulnerable to hacking and modification.
More recently, the FDA issued a warning over cyber-security concerns around implanted medical devices. And at cyber-security conferences over the past few years, hackers have demonstrated that implanted devices and their associated control systems can be subverted to deliver lethal doses of insulin or inappropriate shocks. And just over the past few days it has been reported that hackers exploiting a major security flaw in the Android OS have used apps to deliver malicious code to unsuspecting users, enabling the hacker to take control of the device.
These trials suggest applications beyond only management of diabetes. Consider, for example, that many implanted cardiac devices are already remotely monitored via proprietary networks like Medtronic’s CareLink system or Boston Scientific’s Latitude system. Data collected from these devices include parameters like AICD discharges, arrhythmia detection, lead impedance, and more.
Much like data points are transmitted to a proprietary receiver, it is not hard to surmise that instead the specific data is transmitted to a smartphone. In the appropriate settings, there may even be applications for patient-initiated modifications to the device. Similar applications could be envisioned for implanted drug pumps (e.g. Baclofen pumps), neurostimulator devices, and so on.
1. Clinical Trial Identifier: NCT01697150
2. Clinical Trial Identifier: NCT01871870
3. Clinical Trial Identifier: NCT01727817