Artificial pancreas safe for hospitalized diabetics

An automatic insulin delivery system that has performed well in type 1 diabetes patients also proved safe and feasible for type 2 diabetes patients on a general hospital ward, according to a U.K. study.

The so-called artificial pancreas, or closed-loop insulin delivery system, monitors blood sugar levels and increases or decreases insulin delivery in response - approximating how a healthy pancreas would work, researchers write in The Lancet Diabetes and Endocrinology.

The system eliminates the skin-pricks and manual insulin injections that many type 2 diabetes patients currently rely on, the authors note.

The artificial pancreas "allows more responsive insulin delivery and the expectation, so far supported by clinical studies, is that health outcomes can be improved," said senior author Dr. Roman Hovorka of the University of Cambridge Metabolic Research Laboratories.

But it costs more than injections and requires patients to wear a device around the clock, Hovorka told Reuters Health by email.

For the study, the researchers enrolled 40 adults with type 2 diabetes who were receiving insulin therapy in general wards at Addenbrooke's Hospital in Cambridge. Half received closed-loop insulin delivery and half received conventional insulin injections for three days.

The artificial pancreas includes a glucose sensor inserted into the skin, which took measurements every 1 to 10 minutes and used the information to determine how much insulin to deliver.

Patients with the artificial pancreas spent about 60 percent of the three-day study period in their target blood sugar range, compared to an average 38 percent of the time in the comparison group. There were no incidents of severe high or low blood sugar in either group and no other adverse events related to the devices.

"We presently use the closed loop system in people with type 2 diabetes staying in hospital," Hovorka said. "Glucose control in hospital is often suboptimal and our aim is to improve it while people with type 2 diabetes are staying in hospital for various reasons such as treating diabetes complications."

At the moment, Hovorka and his colleagues are not planning to try the system outside the hospital, he said.

Before all people with type 2 diabetes can obtain one, "the major issue will be demonstrating cost effectiveness, through larger clinical trials, given the continual push on health care expenditure," he said. "Development of commercial systems specifically for type 2 diabetes is also a necessity."

This was a small study; a larger one may have found that the artificial pancreas helps reduce the risk of dangerously low or high blood sugar, writes Gerry Rayman of Ipswitch Hospital NHS Trust in Suffolk, U.K., in a commentary accompanying the study.

He notes that hospitalized patients are generally sicker than average type 2 diabetics outside hospitals, and whether they are hospitalized for diabetes-related reasons or not, controlling their diabetes in the hospital is an important problem to solve.

Right now, good blood sugar control is achieved in these patients just 40 percent of the time, Rayman writes.

He questions whether this test represents real-world conditions, where 85 percent of hospital admissions of diabetics are emergencies unrelated to diabetes, so these patients are not usually under the care of diabetes specialists while in the hospital.

Even for diabetes specialists, learning to insert and calibrate the glucose sensors can be tricky, and sensors sometimes fail, Rayman notes.