medwireNews: Fully closed-loop insulin delivery results in a markedly improved time in range among inpatients with type 2 diabetes receiving noncritical care, researchers have reported.
The trial follows on from a pilot study the team published in 2016, which involved 20 patients in one hospital. This latest trial was conducted at two hospitals – one in Switzerland and one in the UK – and included 136 patients with type 2 diabetes who required insulin for blood glucose control while in hospital. Their average age was approximately 67 years and they were in hospital for a variety of reasons, most commonly sepsis (43%).
The 70 patients randomly assigned to receive closed-loop delivery did so for 15 days or until hospital discharge, whichever was shortest. The closed-loop system was set up by entering the patient’s weight and previous total daily insulin dose, and participants ate standard hospital meals and were free to eat other meals or snacks as they wished.
During closed-loop delivery, patients remained within a glucose range of 100 to 180 mg/dL for 65.8% of the time, compared with 41.5% for the 66 patients who received usual care. This was mostly due to a large reduction in the time spent in hyperglycemia, at 23.6% versus 49.5%. The number of clinically significant hyperglycemic events (>360 mg/dL) was 18 versus 41.
Speaking to medwireNews, study author Hood Thabit (Manchester University Hospitals NHS Foundation, UK) noted that ward staff may be inclined to leave blood glucose levels a little high for fear of provoking hypoglycemia, a human factor that is avoided if an algorithm is in control. And he stressed that the algorithm improved glucose control without using significantly more insulin than the staff did for usual care, at a daily median of 44.4 versus 40.2 U.
There was no difference between the groups in the time spent with glucose below 100 mg/dL, or any other hypoglycemia threshold, although the number of events of glucose below 63 mg/dL was nonsignificantly smaller with closed-loop delivery than usual care, at three versus nine.
Overnight glucose levels were within target range for 74.0% of the time in the closed-loop group, compared with 54.2% of the time in the usual care group.
“The idea is to have a system that can be used in any general ward inpatients and to have a system that is essentially automated and easy to use,” said Thabit, adding: “We wanted to make sure that it does not increase the burden on the healthcare professionals.”
Being a fully closed-loop system, staff did not need to announce meals, enter carbohydrate content, or monitor glucose levels.
“It is possible to do this in this particular patient population,” Thabit explained. “They are not type 1, so the expectation is they will still have endogenous insulin secretion, but during this time of illness due to stress, hypoglycemia, and insulin resistance, they will require extra insulin and this is where the algorithm is able to cope with those things without bolusing for meals.”
Device-related issues in the closed-loop group comprised two sensor failures and one pump-check error.
An artificial pancreas system is more expensive than usual care, but Thabit believes that the costs may balance out, because the improved glucose control will result in shorter hospital stays. Hospital stay can also be reduced by focused multidisciplinary care, as in the DICE project at Ipswich Hospital in the UK, and Thabit said it will be important to compare closed-loop delivery against approaches that “make use of what you already have, but make it work more efficiently.”
He said: “In the future, I would see that both these systems – closed-loop system and the specialist inpatient diabetes staff – can work together.” He particularly sees a role for closed-loop delivery in the care of particularly challenging patients, such as those requiring steroids.
“What we did in the first two papers was to look at the general population, to show that this system can be used in pretty much any general ward inpatient population, and the next step is to narrow it down,” said Thabit. This will include studying subgroups of patients, such as surgical patients, who may derive the greatest benefit from improved glucose control, and looking at the effect on clinical outcomes such as length of stay.
The team presented their findings this week at the ADA’s 78th Scientific Sessions in Orlando, Florida, USA, and they were simultaneously published in The New England Journal of Medicine.
medwireNews is an independent medical news service provided by Springer Healthcare. © 2018 Springer Healthcare part of the Springer Nature group