The main objective of this study is to evaluate the efficacy and safety of 24 hours of closed-loop glucose control compared with standard diabetes treatment, in patients with type 2 diabetes treated by non-insulin glucose-lowering medications. This group is being studied as it is representative of patients with glucose dysregulation.

Algorithm driven closed-loop enables automated subcutaneous delivery of insulin in response to real-time continuous glucose sensor readings. Our studies to date have assessed the safety and efficacy of closed-loop insulin delivery in patients with type 1 diabetes in a controlled setting. For patients with type 2 diabetes treated by diet or non-insulin glucose-lowering medications alone, an episode of acute illness may result in elevated glucose levels necessitating initiation of insulin replacement therapy to optimise glycaemic control. Insulin may also be required in patients with no prior history of diabetes, presenting with 'stress hyperglycaemia'. Closed-loop systems may be of benefit in such insulin-naive patients in whom the optimal dosing regimen is difficult to establish, and may provide a safer method of insulin delivery with the added benefit of continuous monitoring of glucose levels on general hospital wards, thus minimising the likelihood of hyper- and hypoglycaemic events and their known associated worse outcomes.

The study has an open-label, randomised, two-period crossover design. Participants will be randomised to undergo two 24-hour studies in a clinical research facility, during which glucose levels will be controlled by either the computer-based closed-loop algorithm (intervention) or by patients' usual non-insulin glucose-lowering diabetes treatment regimen (control). A continuous glucose sensor will be inserted on arrival for each visit. Participants will consume regular meals (matched on both visits) and carry out daily activities mimicking those occurring in an inpatient setting. Stable glucose isotopes will be administered on the two study occasions to collect data for modeling of glucose turnover around meals and during the overnight period.

Inclusion Criteria:

• Aged 18 years or older
• Type 2 diabetes for at least 1 year as defined by WHO
• Treatment with glucose-lowering medication(s) (including Exenatide) for at least 6 mths
• HbA1c between 7.0% and 10.0% inclusive (measured within past 3 mths)

Exclusion Criteria:

• Autoimmune type 1 diabetes
• Type 2 diabetes treated with insulin
• Type 2 diabetes treated with diet control alone
• Known or suspected allergy against insulin
• Proliferative retinopathy
• Current or planned pregnancy or breast feeding
• Any physical or psychological disease or medication(s) likely to interfere with the conduct of the study and interpretation of the study results, as judged by the study clinician