An artificial pancreas using a predictive rule-based algorithm achieves normal nocturnal concentrations of blood glucose in patients with type 1 diabetes, according to study findings.

“This study shows that the [predictive rule-based algorithm] is a new, more physiological, and highly precise controller that achieves a significant increase in overnight normoglycemia and glucose stability in patients with a previously acceptable metabolic control,” the researchers wrote.

The randomized crossover study evaluated 10 adult participants with type 1 diabetes and previously treated with continuous subcutaneous insulin infusion (CSII). All participants were aged at least 18 years, had diabetes for at least 3 years, and had received CSII treatment for more than 1 year.

The study participants spent two non-consecutive nights at the Diabetes Day Care Center, Parc Tauli Sabadell University Hospital in Barcelona, Spain. On one of the nights, the participants underwent their usual CSII therapy (open-loop night), and one night entailed the use of the predictive rule-based algorithm (closed-loop night). Each night involved the evaluation of two patients, one for each treatment approach. The closed-loop treatment duration was between 10 p.m. and 10 a.m., and included overnight control and breakfast control. Blood glucose was evaluated through samples collected every 20 minutes.

The researchers found that the percent of normoglycemic time sustained during the nocturnal period (from midnight to 8 a.m.) increased from 66.6% with open loop to 95.8% using the closed-loop algorithm (P<.05). The median time in which patients were hypoglycemic decreased from 4.2% in the open-loop night to 0% on the closed-loop algorithm night (P<.05). The open-loop night was associated with nine hypoglycemic events vs. one using the closed-loop algorithm. Post-meal glycemic deviation was not lower with the closed-loop system vs. conventional insulin pump therapy.

Source: Endocrine Today