Safety and Efficacy of The iLet Bionic Pancreas System (algorithm 2.0), a new automated insulin delivery (AID) system, in people with type 1 diabetes

Automated insulin delivery (AID) is a new therapy that aims to improve outcomes for people with diabetes. AID links an insulin pump and continuous glucose monitoring (CGM) to a math program (algorithm) that automatically adjusts insulin delivery to try and keep blood glucose levels in a normal range. The purpose of this study is to test how well a new AID system can safely and effectively control blood glucose. The iLet Bionic Pancreas (BP) System using the current insulin dosing algorithm 1.0 is commercially available in the USA. The new AID system has modified the iLet BP System to include a new insulin dosing algorithm (2.0). The version 2.0 algorithm is entirely new and uses different underlying principles to achieve a similar user experience. The system is based on the commercially available iLet Bionic Pancreas (BP) System (which uses algorithm 1.0), using an insulin infusion pump containing a new algorithm (2.0), a Dexcom G6 continuous glucose monitor (CGM), and a smartphone with the iLet App installed. This study consists of a single-arm test-run period or periods for initial qualification of algorithm 2.0, followed by a prospective random-order, two-period crossover trial, assessing outcomes over 4 weeks of using the BP with each of the two insulin dosing algorithms (1.0 and 2.0). Standard Therapy will involve the collection of baseline CGM data for 14 days. Participants will wear a blinded Dexcom G6 CGM sensor for 14 days. If they are a current CGM user and agree to share their last 14 days of CGM data, they are able to bypass Standard Therapy. Once baseline CGM data have been collected, participants will use the BP system running algorithm 2.0 for 2 weeks to evaluate its safety and performance. After the test-run is completed, the participants will return to their usual diabetes management regimen and the results will be analysed. If the minimum safety and efficacy standards are not met, or if they are met but the analysis of the results indicates changes to the algorithm could improve performance, algorithm 2.0 will be modified and a second test-run period will be performed. Each test-run period will be up to 4 weeks (two weeks of Standard Therapy to collect baseline CGM data followed by two weeks use of the BP system running algorithm 2.0). During the crossover trial, participants will use the BP for a 4-week period using the 1.0 algorithm and the other 4-week period using the 2.0 algorithm, in random order. During the first 3 weeks of each 4-week study period, participants will be encouraged to use the qualitative meal announcement feature (meals announced as Breakfast, Lunch, or Dinner, and Usual for me, More, or Less). Meals are announced by tapping the Meal Announcement icon at the bottom of the home screen. The meal type is then selected as "Breakfast", "Lunch", or "Dinner". The carbohydrate amount will default to "Usual for me", reflecting the usual amount of carbohydrates typically eaten at that meal. This amount can be changed to "More" or "Less", depending on the carbohydrate content of the meal. During the last week of each 4-week study period, the participant will use the BP in fully closed-loop mode, where they will be asked to continue using the BP without meal announcements. The intervention remains exactly the same when meals are announced and not announced. The study will compare how well each of the two algorithms (1.0 and 2.0) can manage glucose levels following meal announcements and without meal announcements. Participants eat their usual diet throughout the study. Although this is a random order crossover trial, no 'wash out' period is required between treatments. This is due to the half life of rapid acting insulin being so short that any effect has gone within 4 hours of changing treatments. Participants will be asked to complete the following self-reported surveys and questionnaires: Hypoglycaemia survey - every day during the use of the BP in the Test-Run period, and every other day during the use of the BP in the Crossover Trial. iLet and Algorithm Experience Questionnaire - at the end of the Test-Run period, and after 3-weeks and 4-weeks use of the BP with algorithm 1.0 and algorithm 2.0 during the Crossover Trial. Diabetes Constraints Scale - at the start and end of using the BP during the Test-Run period, and at the following time points during the Crossover Trial : start of BP use, and after 3-weeks and 4-weeks use of the BP with algorithm 1.0 and algorithm 2.0.