Two Way Crossover Closed Loop Study R-AP vs MPC

An artificial pancreas (AP) is a control system for automatic insulin delivery. The investigators have implemented a missed meal bolus detection algorithm for use within an AP control system. The robust R-AP system used in this protocol has been designed to handle a variety of real-world scenarios that are critical to a high-risk patient population. The investigators will test how well the new algorithm handles missed or inaccurate meal announcements. This type of algorithm may significantly improve glucose control over the standard model predictive control (MPC) closed-loop algorithm without these new algorithm features for patients with type 1 diabetes.

Participants will undergo two visits at Oregon Health and Science University (OHSU) that will evaluate missed meal bolus detection. Participants will arrive at approximately 7am for all visits, be monitored through the afternoon and discharged before dinner. During each of these intervention visits, participants will wear an Omnipod to deliver insulin and a Dexcom G6 Continuous Glucose Monitoring System (CGM) to measure glucose. The closed loop system will receive activity data through a Polar M600 watch worn by the participant. The studies will test the ability of the system to adapt to a missed meal bolus. Participants will eat self selected meals at 10 am and 2pm, both meals with a missed bolus, and these meals will be repeated across both study arms. For one missed meal bolus study, glucose will be controlled using the Robust R-AP closed-loop mode. During the other missed meal bolus study, glucose will be controlled using the MPC closed-loop mode. Participants will complete frequent glucose and ketone checks for safety.