800-P: Real-World Use of Do-It-Yourself Artificial Pancreas Systems in Adults with Type 1 Diabetes

Abstract

Background and Aims: Members of the diabetes community have started developing their own Do-It-Yourself Artificial Pancreas Systems (DIY APS) , also known as Open-Source Automated Insulin Delivery (AID) Systems and use the online hashtag “#WeAreNotWaiting” to promote the development of it. DIY APS systems inclued OpenAPS, Loop and AndroidAPS. The aim of the study was to evaluate retrospectively glycemic outcomes with the use of DIY APS in adults with T1D. Methods: Subjects with T1D started using open-source AID systems (AndroidAPS with Insight insulin pump and Dexcom G6 or FreeStyle Libre 2 CGM and Loop with Minimed Paradigm and Dexcom G6) . Glycemic outcomes at baseline on open-loop and after 12 months on DIY APS were compared including time in different glucose ranges, management indicator (GMI) , mean glucosa levels, glycemic variability and Patient Glycaemic Status (PGS) . Results: Patients with T1D users open-source AID systems (22 AndroidAPS and 1 Loop) were included (age 38.4±years, 56.5% females, diabetes duration 20.8±10.6 years and capillary HbA1c 6.6±0.5%) . 18 patients previous insulin pump users with a use 3.5±4 years, 5 patients previously treated with multidose and CGM use 2.6±2.1 years. GMI decreased from 6,6±0.6% to 6±0.5% (p<0.001) . At 12 months, time in range (TIR) 70-180 mg/dL increased from 69.9±11.5% at baseline to 87.5±7.8% (p<0.001) , time in hyperglycemia >180 and >250 mg/dL were reduced from 23.5±11,5% to 9.6±7.5% and from 5.1% (2-9) to 0.6% (0-3) , respectively (p<0.001) and time in hypoglycemia <70 and <54 mg/dL decreased from 5±3.6% to 2.8±2.1% (p=0.006) and from 1.3±1.5% to 0.4±0.5% (p=0.014) . Improvements in glycemic variability from 35.7±6% to 30.3±6.2% (p=0.001) . At 12 months, PGS and sensor glucose were reduced from 54.4±36% to 23±19.6% and from 144.7±18.3% to 127±14% (p<0.001) . Conclusions: Open-source AID systems is safe and improved outcomes glycemic after 12 months of use in a real-world clinical setting in adults with T1D. Disclosure S.Amuedo: None. M.Antequera: None. S.Azriel: None.