Artificial Pancreas Effective in Unsupervised, At-Home Setting

Artificial Pancreas Effective in Unsupervised, At-Home Setting
Artificial Pancreas Effective in Unsupervised, At-Home Setting
Type 1 diabetes patients experienced improved glucose control and less hypoglycemia with use of an artificial pancreas vs. sensor-augmented pump therapy.

A closed-loop insulin delivery system, also known as an artificial pancreas, improved overnight glucose control and reduced hypoglycemic episodes in patients with type 1 diabetes when used in real-life conditions, according to research presented at the European Association for the Study of Diabetes (EASD) Annual Meeting.

Many studies of closed-loop insulin delivery systems have yielded positive results, but the majority have been conducted in clinical research facilities or in home settings under close supervision or with remote monitoring, Hood Thabit, MD, of the University of Cambridge Metabolic Research Laboratories and Wellcome Trust-MRC Institute of Metabolic Science in the United Kingdom, said during a presentation at EASD.

To learn more about the performance of closed-loop systems in more real-world settings, Thabit and colleagues combined data collected during free-living, unsupervised, randomized, open-label crossover studies comparing overnight closed-loop insulin delivery with sensor-augmented pump therapy.

The researchers’ analysis included 40 participants with type 1 diabetes, 24 of whom were adults recruited at three centers and 16 of whom were adolescents recruited at one center (mean age, 43 years; mean HbA1c, 8.0%).

Patients underwent training with the study devices followed by two periods of sensor-augmented pump therapy in combination with or without an overnight closed-loop system. Each period lasted 4 weeks in adults and 3 weeks in adolescents, Time spent in the target glucose range of 3.9 mmol/L to 8.0 mmol/L served as the primary outcome.

The closed-loop system was started by participants of their own volition on 866 nights, translating to 7,619 hours of closed-loop use at home without supervision or remote monitoring. The proportion of time that the sensor glucose was in target range between midnight and 8 a.m. increased by 18% with the closed-loop system vs. sensor-augmented therapy (P<.001).

Use of the closed-loop system was also linked with a significant 0.8-mmol/L decrease in mean overnight glucose (P<.001), with no difference in glucose variability. Similarly, when the closed-loop system was used, significantly less time was spent in hyperglycemia (P=.001) and in hypoglycemia, defined as below 3.9 mmol/L (P=.014).

Lower mean overnight glucose during use of the closed-loop system was effected by increased overnight insulin delivery (P<.001) without changing the total daily insulin delivery (P=.84).

“We have shown that extended home use of closed-loop without supervision is feasible in adults with type 1 diabetes. We can significantly increase time spent in target and reduce glucose overnight, and still reduce time spent in hypoglycemia, even with a low rate of hyperglycemia,” Thabit said.

“We do acknowledge that closed-loop utilization is still limited by connectivity between devices, but our hope is that this will improve and further increase the applicability and utility of closed-loop in the future.”


  1. Thabit H et al. Oral Presentation 194. Presented at: European Association for the Study of Diabetes Annual Meeting; Sept. 15-19, 2014; Vienna, Austria.