Glucose control using an advanced hybrid closed-loop system

This will be a prospective, three-stage study involving 12 pump-experienced adults with type 1 diabetes. Following a 1-week period of run-in, the first stage will take place in a supervised ‘in-hotel’ setting (5 days), followed by a second stage supervised ‘in-hospital’ (1 day) and finally a third stage in a ‘free living’ setting at home (3 weeks). Participants will be assigned the study insulin pump for a period of approximately 5 weeks. The hybrid closed-loop (HCL) insulin delivery system incorporates a proportional-integral derivative (PID) control algorithm with insulin feedback. The HCL system comprises a glucose sensor coupled with an insulin pump containing a computerised automated insulin delivery algorithm. Glucose sensor information is transmitted to the insulin pump, and the dose of insulin is calculated by the algorithm and delivered every 5 minutes to account for basal insulin requirements. Participant-initiated bolus insulin doses are still required for meals. HCL systems have been shown to be safe and effective in improving glucose control. Based upon prior experience, modifications have been made to previous iterations of the HCL system to develop an advanced-HCL system to be used as the study pump. The main modifications to the study pump include: 1) correction boluses using sensor glucose and commands by the algorithm without the need for user confirmation; 2) the ability to make therapeutic insulin dosing decisions using the sensor glucose without the need for finger-prick glucose values; 3) a more aggressive algorithm aiming to improve time-in-target glucose range. All participants will use the study pump for the study duration. Education specific to the study pump will be provided face-to-face by an experienced study doctor and diabetes nurse educator, and will be tailored to each participants’ prior knowledge. Participants will also be provided educational material and resources, including a user guide booklet for the study pump. Stage 1 – ‘In-Hotel’: Participants will live in a supervised environment for 5 days using the study pump with a study doctor and nurse present at all times. Participants’ glucose levels will be monitored remotely at all times by the study team. Three interventions will be implemented during this time including: 1) Missed meal bolus (insulin bolus unannounced) on Day 3; 2) Late meal bolus (insulin bolus will be administered 20 minutes after the meal was commenced) on Day 4; 3) Moderate intensity exercise (2 hour walk) on Day 5. All food will be provided from a standard menu with a variety of vegetarian, non-vegetarian and gluten-free meals from which participants will be able to freely choose their meals. Stage 2 – ‘In-Hospital’: Sensor miscalibration will occur in a hospital supervised setting over a 24-hour period. There will be three study doctors, four study nurses and a scientist present for this period. Participants will be randomised to either over-calibrate or under-calibrate the sensor by 40% of their true glucose value. The CL function will be turned off during this miscalibration period. Participants will continue to over or under-calibrate their sensor every 3 hours for a total of 12 hours. For safety purposes, all participants will wear a second glucose sensor which will be calibrated with their true glucose values. Following 12 hours of sensor miscalibration, the CL function will be re-activated. Frequent blood sampling will be conducted overnight using iSTAT (Abbott Point of Care, Princeton, NJ) or YSI for venous glucose levels at 30 minute intervals for 12 hours. Shorter time interval blood sampling will occur if glucose levels are <4.4mmol/L. Hypoglycaemia will be treated with usual clinical care. The frequent sampling will finish at ~8.00am the following morning. Stage 3 - Free-Living ‘At-home’: Participants will enter this stage if there have been no episodes of severe hypoglycaemia or ketoacidosis during the supervised stage of the study. If eligible to return home using the study pump, participants will be discharged home to undertake their usual activities of daily living. They will be required to upload their insulin pumps twice-weekly for review by a study doctor. After three weeks at home, they will return for their end-of-study visit and all study devices will be returned. A user-evaluation questionnaire will be completed. Participants will recommence management using their own insulin pumps with pre-existing settings.