Jose Garcia-Tirado, United States of America
University of Virginia Center for Diabetes Technology (CDT)Presenter of 1 Presentation
FULLY AUTOMATED CLOSED-LOOP IN ADOLESCENTS WITH TYPE 1 DIABETES: A SAFETY AND FEASIBILITY STUDY
- Jose Garcia-Tirado, United States of America
- Mark D. DeBoer, United States of America
- Helen Myers, United States of America
- Katie Krauthause, United States of America
- Morgan Fuller, United States of America
- Jenny L. Diaz, United States of America
- John P. Corbett, United States of America
- Charlotte Barnett, United States of America
- Martha Dawson, United States of America
- Rebeca Esquivel-Zuñiga, United States of America
- Marc Breton, United States of America
Abstract
Background and Aims
Modern automated insulin delivery system (AID), relying on timely quantification of meals, have reliably improved glycemic control in people with Type 1 Diabetes (T1D) of all ages. Achieving similar control without meal information has so far proven elusive. We present a new generation closed loop control (CLC) algorithm designed for such use.
Methods
Adolescents with T1D were enrolled at the UVA Center for Diabetes Technology (CDT) in a supervised outpatient clinical trial comparing the legacy UVA CLC (USS) with our new algorithm (RCKT) during hybrid (HCL) and fully automated (FCL) use. Four 24h periods were studied with 3 meals repeated from day to day; the second and fourth dinner were unannounced. One algorithm was used for day 1-2 and the other for day 3-4, in random order. Glucose control was evaluated on standard CGM-based metric.
Results
Twenty-one adolescent participants signed consent at CDT. Eighteen completed the study. Time in range (TIR) overall was 79.1±11.4% vs 83.2±10.8 (p=0.22) with dinner bolus and 78.6±11.1% vs 84.5±7.4% (p=0.075) without, for USS and RCKT respectively. Focusing on dinner (6PM-midnight) TIR in HCL was 90.9±12.9% vs 97.8±5.7%, p=0.052, and 58.8±25.7% vs 77.0±22.6%, p=0.02, in FCL. There was no difference in time below 70mg/dL or occurrence of hypoglycemia.
Conclusions
Our new AID algorithm was shown to be feasible in adolescents with T1D, equivalent to UVA legacy algorithm in hybrid use, and led to +18% TIR during un-bolused dinners. Additional studies are needed to assess its impact over several days of use and less supervised environments.