May / June 2017

Diabetes and the Internet of Things

Scott Fotheringham, PhD

Megan was diagnosed with Type 1 diabetes in 2007 when she was five months pregnant with her first child. She began the routine, familiar to many of the 415 million diabetics worldwide, of measuring her blood glucose levels with finger pricks, followed by insulin injections. She kept a pen-and-paper log of her glucose levels and diet to help manage he disease.

“I knew keeping track of my levels led to good control of my health,” the 36-year old mother of two said. “But I always found it cumbersome. I wanted to see graphs and trends to help me make decisions.”

The situation improved in the summer of 2009 when she started to use a body-mounted insulin pump. Her programmable Animas Ping delivered an insulin injection every three minutes through a port embedded in her abdomen. In 2016 Megan received a Dexcom G4 continuous glucose monitor (CGM), a small wearable device that sends data wirelessly to her Animas Vibe pump, which keeps a record of her levels. Megan then uploads this information to Diasend, a cloud-based database. Megan, her doctor, and health care team all have access to the information, which allows them to work together to manage her disease. In future, the data will be incorporated in an electronic health record.

All of this is made possible by the Internet of Things (IoT), the network of computerized sensors embedded in medical (and other) devices that can collect, send, and receive data via the Internet. The amount of data, the speed at which it is transmitted, and its aggregation, storage, and analysis is revolutionizing the management of chronic diseases like diabetes. The IoT could lead to lower health care costs, fewer doctor visits, and the ability to amass clinical data from large populations to provide insight into treatment options.

The IoT is revolutionizing management of chronic diseases like diabetes.

Chronic disease patients can now be fitted with wearable devices similar to fitness products Fitbit and Jawbone, which track step counts, pulse, and sleep patterns. Intel and the Michael J. Fox Foundation, for example, have collaborated on a smartwatch with an app for Parkinson’s disease patients that measures tremors and communicates with users, reminding them to take medications and giving them info on disease management.1 In the ICU, glucometers, scales, and monitors that report heart rate and blood pressure have been joined by a new device that can measure core temperature and urine output in catheterized patients.2

Computers will soon be able to collect Megan’s real time data, collate historical information, then mine it for patterns to allow her to make informed choices about her life habits. Some health care start-ups, hospitals, and pharmaceutical companies have already started to do this by working with IBM Watson

Health, a cognitive computing platform:

  • Novo Nordisk, in partnership with digital and analytics company Glooko, are working with Watson to improve management and treatment options for diabetics.3
  • Medtronic makes CGMs, insulin pumps, and an app for diabetics that collect data on a user’s exercise and carbohydrate consumption. Watson analyzes the data collected by the devices and uses it to predict potential hypoglycemic events that might occur as many as three hours later. If Watson believes the meal the user is about to eat may be harmful based on his or her historical data, it can send an alert.
  • The American Diabetes Association is using Watson’s computing power to sift through more than 66 years of clinical and research data.4 By comparing an individual’s data to that of large populations, the team hopes to identify risk factors and create personalized treatment plans.

One challenge to IoT in the medical landscape is the lack of interoperability, which prevents medical devices, equipment, and databases from communicating with each other. Standardized interfaces will be necessary for this information to be useful in aggregate and to avoid it being marooned on “data islands.” 5 There are also privacy concerns and cybersecurity threats. Most ransomware attacks are already aimed at health care organizations; 6 IoT medical devices could also be vulnerable.

For Megan, the hope is that these kinks can be worked out.

“Having a CGM, I can see the data all the time, anytime,” Megan said. “It’s made a huge difference in how I relate to my disease.”