For people with diabetes, an artificial pancreas is the Holy Grail. Health Tech that automatically monitors blood glucose levels and administers the correct type and amounts of insulin could make diabetics’ lives much simpler. When Medtronic’s iPro2 received FDA approval for use in the U.S. in late 2016, that was a very big deal indeed. Of course, people who have and live with dangerous, potentially life-threatening diseases aren’t always enthusiastic about jumping in line to be among the first to adopt new technology, regardless of the promises. After the Medtronic news, I asked an acquaintance with Type 1 Diabetes (T1D) if he was excited about the new development and if he talked with his doctor about it. His answer, “Yeah, it could be great, but right now I’m comfortable with testing my blood sugar levels and adjusting insulin levels on my own. I want to give it a while before replacing that with a brand new technology I’m not sure I can trust, regardless of the FDA. And my doctor agrees.”
A recently completed and reported multi-site clinical trial with a new type of artificial pancreas system may give people with T1D more confidence in the artificial pancreas concept. Researchers from Harvard’s John A. Paulson School of Engineering and Applied Science, the William Sansum Diabetes Center in Santa Barbara, California, the Mayo Clinic in Rochester, Minnesota, the University of Virginia’s Center for Diabetes Technology, and the University of Padova in Padua, Italy collaborated in a 12-week trial with more than 60,000 hours of combined use. The test technology hardware included a Dexcom G4P CGM (continuous glucose monitor) and a Roche Accu-Check insulin pump. The key to the study, published in Diabetes Care, was a smartphone app with two vital functions: health monitoring and predictive control. Thirty patients with T1D in California, Virginia, and Minnesota participated in the trial. The CGM and the insulin pump were placed under the skin and controlled by the smartphone app via Bluetooth. The software used a model-predictive control (MPC) algorithm that based insulin injection levels on variables including meals consumed, physical activity, sleep, stress, and metabolism. The software learned as time passed how the individual patients’ glucose levels varied based on the monitored variables. The result was a program tailored for the individual based on behavior and real-time biometrics, rather than using fixed dosages based solely on glucose levels.
Expert clinicians and the patients reviewed the MPC algorithm-based adaptations during the test period. Approximately 10% of the recommended adaptations were manually overridden. At the end of the 12-week period, the conclusion was the MPC software yielded significantly reduced HbA1c levels and time in hypoglycemia. The results of the study were heartening, particularly since it showed measurable improvements with test subjects who were already disciplined and adept at reading their glucose levels and administering insulin based on conventional methods. This was an uncontrolled study, meaning there was no control group for comparison. The study must be replicated and validated by additional research and testing. As artificial pancreas technology progresses and matures, however, the time draws closer when physicians and patients alike will trust and use some form of the lifestyle altering Health Tech.