DIY artificial pancreas increases time in range vs. pump therapy in type 1 diabetes
Click Here to Manage Email Alerts
NEW ORLEANS — Adults and children with type 1 diabetes who used an open-source automated insulin delivery system significantly increased time in target glucose range compared with baseline and with sensor-augmented pump therapy.
“While open-source automated insulin delivery is not yet approved by the FDA, we know people worldwide are using it to manage their diabetes. Our findings demonstrate that this is a safe and effective technology and adds to the body of evidence supporting use of open-source automated insulin delivery for improving glycemic outcomes,” Martin de Bock, PhD, FRACP, senior lecturer and pediatric endocrinologist at the University of Otago in Christchurch, New Zealand, said in a press release.
De Bock and colleagues randomly assigned 48 children aged 7 to 15 years and 49 adults aged 16 to 70 years with type 1 diabetes to an automated insulin delivery system (21 children, 23 adults) or sensor-augmented insulin pump as a control group (27 children, 26 adults). Participants had type 1 diabetes for at least 1 year and were using an insulin pump and had HbA1c below 10.5% for the 6 months before the study. The automated system consisted of a Dexcom G6 continuous glucose monitor paired with a DANA-i insulin pump and the OpenAPS algorithm from a version of AndroidAPS used on a smartphone.
The study included a 4-week run-in period followed by the 24-week randomized clinical trial. The researchers assessed percent of time in target glucose range of 70 mg/dL to 180 mg/dL during the 24-week trial for the two insulin delivery groups.
The automated group, participants used automated insulin delivery 94% of the time for the study duration. Time in range increased immediately and was sustained throughout the trial for participants in the automated group, de Bock said during a presentation. During the last 2 weeks of the trial, the automated group spent 14% more time in range than the pump group.
“For adults, particularly at night, you can see over 30% improvement in time in range on automated insulin delivery [compared with pump therapy], so it was very effective overnight,” de Bock said.
For the automated group, mean time in range was significantly greater during the final 2 weeks compared with the run-period for both adults — a 9.6% increase for final mean time in range of 74.5% — and children — a 9.9% increase for a final mean time in range of 67.5% (P < .001 for both).
Time in range did not change significantly for the pump group. International guidelines call for targeting a time in range of more than 70%. This goal was met by 60% of participants in the automated group and 15% of the pump group.
No participants experienced severe hypoglycemia or diabetic ketoacidosis during the trial. Two participants in the automated group withdrew due to hardware issues.
“We found that people who proclaimed to be techno phobic could use the open-source automated system very effectively and had no problems. We only had two people drop out despite the relatively diverse group who enrolled in the study. This reminds us that with good support and education, automation should be offered to everyone with type 1 diabetes,” de Bock told Healio.
Collapse
Burnside MJ, et al. 286-OR. Presented at: American Diabetes Association Scientific Sessions; June 3-7, 2022; New Orleans (hybrid meeting).
Read more about
Click Here to Manage Email Alerts