Aaron Kowalski is the vice president for glucose control research at the Juvenile Diabetes Research Foundation. He also has type 1 diabetes and, like many others who have the condition, he sometimes struggles to regulate his blood sugar levels.
"I'm a scientist and I work in the field and I don't achieve perfect glucose control," he says.
In that he is not alone: Some studies have found that even highly-managed diabetes—those who test their blood sugar more than eight times a day, still spend almost 70% of their day out of normal blood ranges—an average of more than eight hours with high blood sugar levels and two hours with low.
It's not about compliance, Kowalski says. Rather, it's about limited information—and the fact that computers can think about a lot of different things faster than humans. And, as it happens, Kowalski is working to create a computer-controlled "bionic pancreas" that will help humans better manage their diabetes.
I wrote about Kowalski and JDRF's artificial pancreas project in the March HealthLeaders magazine cover story, Medical Breakthroughs That Will Change Healthcare. But there are a lot of medical breakthroughs out there—advances in nanotechnology, wireless communication, imaging, OR technology, and more—and so I wasn't able to share everything from our conversation about the future of diabetes treatment.
A solvable problem that can change lives
"Today the person with diabetes does everything themselves. We call it the open loop," he told me. Patients gather information and do their best to respond to it. But even the most sophisticated patients and tools aren't perfect, he says. Diabetes is very difficult to control, even in the most compliant patients.
JDRF has gathered world class-doctors, mathematicians, and engineers, and has begun to work with private industry to create a safe, effective artificial pancreas that reacts to changing glucose levels and automatically delivers the right amount of insulin at the right time, he says.
It builds on two already-approved devices—the insulin pump and the continuous glucose monitor (CGM). But unlike an open-loop system in which the patient is responsible for testing, reading data, and taking corrective action, the automatic closed-loop pancreas would use a control algorithm to read and interpret the information from the device and respond by dispensing insulin when needed.
Like a natural pancreas, the device works continuously taking updated readings every one to five minutes to determine which way glucose levels are going. The next step is to tie that information together with a computer or control algorithm—software that sits in the pump and interprets the glucose level reading.
One huge benefit is that the device could take over for patients when they are most vulnerable and most likely to miss a CGM alarm—while sleeping or unconscious. In the closed loop system, the system would automatically intervene.