Closed-Loop Insulin Delivery May Improve Glycemic Control in Patients on Nutritional Support

By Lorraine L. Janeczko

April 18, 2019

NEW YORK (Reuters Health) - Closed-loop insulin delivery appears to improve glycemic control in hospital patients receiving nutritional support, researchers report.

"We have shown that fully closed-loop insulin delivery in inpatients receiving parenteral or enteral nutrition (or both) is safe and significantly improves glycaemic control without increasing the risk of hypoglycaemia compared with conventional insulin therapy in non-critical care," the team writes in The Lancet Diabetes and Endocrinology, online March 29.

Over roughly seven months, Dr. Roman Hovorka of the University of Cambridge and colleagues conducted an open-label randomized controlled trial in adult inpatients who were receiving enteral and/or parenteral nutrition and insulin therapy, in one medical center in the UK and one in Switzerland.

The research team inserted a cannula that delivered insulin aspart (Fiasp from Novo Nordisk), using the DANA Diabecare R pump (from SOOIL Development Co.) and a real-time continuous glucose monitoring sensor (Freestyle Navigator II from Abbott Diabetes Care), under the skin of the abdomen or arm.

Twenty-one participants received fully closed-loop insulin delivery and 22 received conventional subcutaneous insulin therapy (the glucose sensor was masked in the control group). Patients were followed for up to 15 days or to hospital discharge.

In the closed-loop group, sensor glucose remained in the 5.6-10.0 mmol/L target range 68.4% of the time, compared to 36.4% of the time in the control group (p<0.0001).

No intervention-related adverse events, and no episodes of severe hypoglycemia or hyperglycemia with ketonemia, were reported in either group. One episode of acute respiratory failure occurred in the closed-loop group, and one cardiac arrest occurred in the control group, but both were deemed unrelated to the interventions.

"The closed-loop system used commercially available components, allowing ease of use in different health-care settings and potentially allowing accelerated adoption of the system for future widespread clinical use," the authors write.

In an editorial accompanying the study, Dr. Pieter Gillard and colleagues of KU Leuven in Belgium note, "The positive feedback of participants at the end of the study indicates that closed-loop systems were received positively by most patients. Improvements in wearability, elimination of interference with drugs commonly used in hospital (eg, paracetamol or heparin), and reductions in the need for calibrations (as has been achieved with some more recently introduced glucose sensors) might further improve acceptability."

While automated closed-loop systems may help diabetes providers care more efficiently for patients, "technical issues or defects might occur, and back-up protocols will need to be activated to avoid unsafe dysregulation of glycaemic control," they caution. "In a quarter of patients, the study was suspended temporarily because of transfer to intensive care, surgery, or other procedures (eg, radiology). In these cases, back-up protocols should be in place, which might lead to an increase in workload."

Dr. Hovorka was unable to comment on the study by press time.

The study did not have commercial funding, but several authors disclosed financial relationships with Novo Nordisk and Abbott Diabetes Care. Two authors also report patents and patent applications based on closed-loop insulin therapy.


Lancet Diabetes Endocrinol 2019.