Optimising fully closed-loop insulin delivery

Hybrid closed-loop systems are increasingly being implemented in clinical care for people with type 1 diabetes.

These systems improve glucose control and have quality-of-life benefits, but still require the user to calculate insulin doses at mealtimes which can be challenging and burdensome.

A fully-automated closed-loop system not requiring meal announcements would significantly reduce the burden of diabetes management and would increase uptake of this technology. It would also reduce workload for healthcare professionals supporting people using closed-loop systems.

We have developed a fully closed-loop system which has been shown to be safe and effective for people with type 2 diabetes.

A feasibility study in people with type 1 diabetes showed safety and efficacy when compared to non-automated therapy, but glucose outcomes with fully closed-loop were below recommended targets. Notably, the clinical efficacy of fully closed-loop therapy was highly variable between participants.

Our research will analyse data from existing closed-loop datasets to identify factors which differentiate ‘responders’ and ‘non-responders’ to closed-loop therapy. This will be used to optimise the fully-automated closed-loop algorithm for people with type 1 diabetes.

The performance of the fully closed-loop controller will be evaluated using a computer simulator of a virtual population with type 1 diabetes.