1492-PUB: In Silico Modeling of the Performance of a Novel Closed-Loop Algorithm Using the UVA/PADOVA Type 1 Diabetes Simulator

• Published in: Diabetes (New York, N.Y.) Vol. 71; no. Supplement_1
• Main Authors: DEDESHKO, ALEXEY J., FEDOTOV, VITALY, SIMIC, AMRA, THIVOLET, CHARLES, MADER, JULIA K.
• Format: Journal Article
• Language: English
• Published: New York American Diabetes Association 01.06.2022
• Subjects: Age groups; Algorithms; Body weight; Carbohydrates; Diabetes; Diabetes mellitus (insulin dependent); Insulin; Meals
• ISSN: 0012-1797; 1939-327X
• DOI: 10.2337/db22-1492-PUB

AIAPS is a novel smartphone-based closed-loop algorithm in development that can be combined with any insulin pump and continuous glucose monitor. AIAPS strives to reduce burden of diabetes by requiring only limited input and by handling unannounced meals. Prior to launching the clinical research program in-silico modelling of the most recent features was performed. The objective of this study was to evaluate the AIAPS performance using the UVA/PADOVA Type 1 Diabetes (T1D) Simulator in people with T1D. AIAPS was used in the in-silico population of 33 subjects (children, adolescents, and adults) of the UVA/PADOVA T1D Simulator. Upon initialisation, 5 min CGM data, body weight, average daily insulin dose and sex for each person was available. To model a scenario with limited information for each individual upon initialization, no data on basal insulin profiles (BIP) , carbohydrate-to-insulin ratio (CR) and insulin sensitivity factor (ISF) were put into AIAPS. Additionally, no information on meals (time or carbohydrate content) were put into AIAPS. The following scenario was modelled: 84 days of closed-loop therapy without user interaction and with variable routine conditions representing the duration of the planned clinical study. Data on glycemic control for the three age groups over the 84 days are shown in table 1. Modelling included 2-5 unannounced meals with a random schedule and carbohydrate content of 130-250 g per day. In an in-silico model the AIAPS closed-loop system exceeded current glycemic targets with a TIR >70% and a TBR <4% in all age groups without user interaction or input of BIP, CR and ISF upon initialization. The next step includes testing the AIAPS performance in a randomized controlled trial. Table 1 - Glycemic control for the three age groups. Data are mean standard deviation.