Amanda Cagle is a doctor of philosophy, not medicine, who has been successful in working with an online community to build an artificial pancreas for her daughter, Magnolia, 7, who was diagnosed with Type I diabetes when she was three years old.
The Artificial Pancreas is a system of devices that closely mimic the glucose regulating function of a healthy pancreas. It includes a continuous glucose meter with a transmitter, and an insulin pump controlled from Amanda’s phone watch. The Artificial Pancreas not only controls the diabetes better, it’s preventing thousands of finger pricks and shots and allows mother and daughter to sleep through the night.
“When we left the hospital, we were given a regimen of eleven finger pricks a day minimum, at least four-to-five injections, and a nightly routine of 11 p.m., 2 a.m., and 5 a.m. wake-ups. Indefinitely,” Amanda said. “Had we kept that protocol, I’d have already taken blood from her fingers nearly 21,000 times and given her nearly 9,500 shots, plus awakened her every single night. It has been very difficult, but she’s rarely complained. Now, she’s very aware that she’s part of an experimental research project and is pretty fluent in the science behind diabetes.”
Cagle built the Artificial Pancreas using open source code developed collaboratively by an online group of innovative Type 1 patients and parents. She has been researching and experimenting for a year, and Magnolia’s been using the system exclusively since February.
Magnolia wears a continuous glucose meter. Their insurance doesn’t cover the meters, and Amanda was having a hard time keeping her in supply.
One part of the three-part continuous glucose monitoring system is the transmitter. It costs $810 every time the battery dies, which is every four months. Amanda decided to hack into the device and change the battery. However, effort after effort failed. The system always read the new batteries as dead.
“Turns out, the batteries never were dead,” she said. “Dexcom, the manufacturer, codes the device to be silent after 118 days. Researching how to work around the code, I came across Dana Lewis, a woman from Alabama, who wanted to change the volume level of her receiver, a different continuous glucose monitoring part. In trying to do that, she found that you could connect Bluetooth devices to some models of insulin pumps, and a medical movement, #OpenAPS was born.”
OpenAPS, Open Artificial Pancreas System, is a patient-led, data-driven, do-it-yourself (DIY) project. Cagle said about 400 people around the world are using this system, all with phenomenal results.
It works like this: Every five minutes, blood sugar data runs through an algorithm more sophisticated than any even nearing FDA trial. The algorithm considers carbs on board, active insulin, duration of insulin, sensitivity factors, physical activity, upcoming meal times, hormonal changes, and other data that otherwise can’t be accurately calculated. This math is done constantly, 24/7, and a dosing decision is made and sent to a re-engineered insulin pump via a homemade Bluetooth radio device.
“Essentially, it is a closed loop system,” Cagle said. “The pump doses or withholds insulin automatically, largely eliminating lows and aggressively treating highs hours before expected spikes. The OpenAPS community has worked tirelessly to make this DIY project available to everyone. It’s created by a community of parents, patients and volunteers, who continue to improve the build and share it with others.”
Diabetes is the leading cause of blindness, amputation, and kidney disease in adults, and the 6th leading cause of death globally, costing billions a year.
Low blood glucose levels come with shaking, confusion, dizziness, possible loss of consciousness, and even death. Cagle said this DIY system greatly reduces, and potentially, could eliminate, complications of and many costs associated with diabetes.
“Magnolia used to experience lows every single day, often four times or more,” she said. “That’s nearing 5,500 times thus far. Since building the Artificial Pancreas, we’ve had essentially none. It’s astonishing. The open source algorithm predicts possible lows up to three hours in advance, withholds basal insulin automatically, and alarms the patient, family or nurse in plenty of time to take preventive action.”
Being able to sleep all night is another major advantage. They no longer have to get up at 11 p.m., 2 a.m., and 5 a.m.
“Obviously that’s exhausting, for both of us,” she said. “We’ve gone years without regular completion of sleep cycles. Plus, making a potentially lethal dosing decision at 2 a.m. is always a danger, and doing this for years on end isn’t a sustainable expectation of caregivers.”
Cagle said dosing for Type 1 requires advanced algebra and consideration of not just food intake and glycemic index, but also insulin on board, sensitivities, previous and future activity levels, hormonal changes, stresses, illness, etc. – plenty of room for human error.
“The open source algorithm takes all of this into account, does the math every five minutes, 24/7, and doses or withholds accordingly,” she said. “I haven’t had to awaken her a single time since I finished the build. Not once, versus multiple times a night. She’s always on target.’
There are obvious major advantages for parents/caregivers.
“I used to exist in crisis mode, pretty much 24/7,” Cagle said. “Every time I’ve awakened from sleep for the past four years, it has been in a state of panic, hoping she’s alive. We’ve lost children in this community to Type 1, and I’ve met vigilant parents who’ve shared stories of waking to find their child comatose, or worse. It’s intense. But, that fear is largely gone. I trust she’ll be steady through the night, not fall out during the day, or accumulate highs and fluctuations that will later rob her of vision and mobility. The weight this system lifts is really immeasurable.”
It’s a revolutionary improvement. Instead of taking blood from her fingers 11 times a day, giving her five injections, and waking her frequently at night, the DIY system allows Cagle to manage most everything from an app on her phone watch.
“For example, before, if she’s at a party, I’d have to test her sugar with a meter, decide if she could enjoy the treat, give her an injection, test again in two hours, perhaps give another injection, etc.,” Cagle said. “Now, if she’s at a party, I don’t have to interact with her physically at all. I press a button on the phone that signals she’ll eat soon. Her basals increase automatically dropping her blood glucose levels in preparation. I input carbs and dose all from the phone watch, without ever needing her. An adult could, for example, dose at a dinner from a swipe on the watch, without touching anything else.”
#OpenAPS, has caught the attention of the major companies, and the attention of research universities that are supportive of patient-driven efforts.
“I’m in talks now, through a group called Open Pathways, to design and lead a study on Phase 2 diagnosis,” Cagle said. “Parents and patients build the projects, and Arizona State assigns researchers to provide support and data management. It’s phenomenal.”
Cagle said none of this should be considered medical advice. But the device is proving helpful for hundreds of people.
“This is an experimental system, but as an online community of about four hundred, we’ve clocked nearly triple the hours required of FDA trials with unanimously good results,” Cagle said. “Also, I can’t build this system for people, but I can legally, and will happily and for free, teach people how to build it for themselves. It is neither terribly difficult nor expensive.”
I need this
W
Where can I get it?
SO EXCITED FOR YOU SND FAMILY!! CONGRATULATIONS FOR HELPING MILLIONS AND DOING WHAT MEDICAL COMMUNITY WOULD OR COULD NOT!!! ❤️