Home / Resources / Articles / ADA: Closed-Loop Insulin Delivery Improves Glycemic Management in Patients Who Are Hospitalized

ADA: Closed-Loop Insulin Delivery Improves Glycemic Management in Patients Who Are Hospitalized

Jul 14, 2018

Closed-loop delivery system — the “artificial pancreas” — provides glucose control for non-critically ill patients

As the prevalence of diabetes increases, the number of patients with diabetes who are hospitalized increases as well. Maintaining blood glucose within a normal range for patients in the hospital still remains a challenge for hospital staff. This is due to acute illness, inconsistent amounts and timing of dietary intake, and medication-induced hyperglycemia. A more efficient method of glucose management is on its way with a closed-loop insulin delivery system.


An automated system consisting of an insulin pump and a continuous glucose monitor is coupled with a sensor that triggers delivery of insulin based on glucose measurements. It is also being referred to as the artificial pancreas.

This closed-loop delivery system is improving the glucose management of people with type 1 diabetes. In a two-center, randomized, open-label trial, closed-loop insulin delivery was evaluated in patients in the hospital receiving noncritical care. The primary endpoint was the percentage of time the glucose readings were within the range of 100 to 180 mg/dL up to 15 days or until patients were discharged.

In addition to glucose management, the risk of hypoglycemia was hypothesized to be less with the closed-loop insulin system than with traditional subcutaneous injections. Patients were recruited from general wards at University Hospital in Bern, Switzerland and Addenbrooke’s Hospital in Cambridge, United Kingdom. Inclusion criteria included patients age 18 years or older with inpatient hyperglycemia requiring insulin. Exclusion criteria included people with type 1 diabetes, pregnant or breastfeeding, and any medications, physical conditions, or psychological conditions that would interfere with the interpretation of the results. There were 137 patients who were randomized into two groups: 70 patients went into the study group and received closed-loop insulin delivery with the insulin pumps and continuous glucose monitors, provided by Abbott Diabetes Care, and the control group consisted of 67 patients who received traditional insulin therapy with subcutaneous injections.

Through the trial, patients were able to eat meals of their choices provided by the hospitals and were not restricted snacks or their usual activities. In both groups, glucose measurements were taken by continuous glucose monitors that were inserted into the abdomen or upper arm.

At initiation, all insulin and sulfonylurea medications were discontinued. All other diabetes medications were continued. The investigator programmed the insulin pump to deliver insulin when glucose readings indicated a need for it and also set up an alarm on the sensor to sound when glucose levels reached 63 mg/dL or lower. The control group received traditional insulin and antihyperglycemic therapies. The continuous glucose monitor was masked to the patient, clinical care staff, and the investigator. Glucose management for these patients was managed by the clinical care team, as they were allowed to modify medication therapy and manually check the patient’s glucose accordingly. Secondary outcomes included the percentage of time the glucose readings were above or below the target range. Safety end points included clinically significant hyperglycemia (>360 mg/dL) and hypoglycemia (<40mg/dL) and other serious adverse events.

The mean percentage of time that the glucose measurements were in the target range was 65.8 + 16.8% for the closed-loop group and 41.5 +16.9% for the control group with a 95% confidence interval [CI] of 18.6 to 30; p <0.001. The mean measurement for the glucose reading for the closed-loop group was 154 + 29 mg vs. 188 + 43 mg for the control group; 95% CI 23 to 47, p<0.001.

These results were observed both for overnight and daytime measurements. Hypoglycemia occurred three times in the closed-loop group and nine times in the control group. There were no adverse events such as severe hypoglycemia or hyperglycemia with ketonemia. The only adverse events that occurred with the devices were skin irritation from the adhesion and bruising at the cannula insertion sites.

From the results, it was concluded that to begin including these closed-loop systems in clinical practice, more needs to be done to assess ease of use and costs of devices. So, to conclude, the closed-loop insulin delivery provides improved glucose management without increasing the risk of hypoglycemia in patients with type 2 diabetes receiving non-critical care in hospitals.

Practice Pearls:

  • The closed-loop system will prevent hypoglycemia. The sensor in this system will alarm when levels are reaching a clinically hypoglycemic level.
  • A closed-loop insulin delivery system will provide better glycemic control than manual glycemic control by hospital staff leading to better patient outcomes.


Bally, Lia, et al. “Closed-Loop Insulin Delivery for Glycemic Control in Noncritical Care.” New England Journal of Medicine, 25 June 2018, doi:10.1056/nejmoa1805233.

Amanda Cortes, PharmD Candidate, LECOM School of Pharmacy