Glycemic Outcomes and Device Engagement Over 18 Months Following Initiation of Two Different Automated Insulin Delivery Systems in American Youth With Type 1 Diabetes

OBJECTIVE: We compared long-term, real-world glycemic outcomes and device engagement in U.S. youth with type 1 diabetes (T1D) using two automated insulin delivery (AID) systems. METHODS: We included youth (ages 6-18) with T1D from a large academic center who started on Insulet Omnipod 5 (OP5) or Tandem Control IQ (CIQ) between January 2020 and June 2024. We collected 30-day continuous glucose monitor and 30-day AID data at baseline and 3, 6, 12, and 18 months following AID initiation. We evaluated change in glycemic metrics and user behaviors (e.g., self-initiated bolusing and carbohydrate entry) by system using separate multivariable linear regression models at each time point while adjusting for baseline factors (age, sex, insulin regimen prior to AID initiation, diabetes duration, and baseline values for the respective glycemic metric). RESULTS: = 311) experienced a greater change in TIR at 3 and 6 months (+14.9% and +13.3%, respectively, vs. +10.0% and +10.2% among OP5 users), this difference waned with time (change in TIR at 18 months +9.9% with CIQ vs. +10.6% with OP5). Carbohydrate entry simultaneously decreased at the same intervals among CIQ users (199, 184, and 173 g at 3, 6, and 18 months, respectively) but not OP5 users (177, 184, and 181 g at 3, 6, and 18 months). CONCLUSIONS: Although glycemic outcomes are similar between devices over 18 months, AID user engagement changes with time and differs between systems, potentially as they "trust" their device more. Change in carbohydrate entry among CIQ users is likely a major contributor to the decline in TIR with longer duration of AID use.