The dilemma of open-source artificial pancreas systems: do it yourself or do it with diabetologist?

Aims: To evaluate the benefits of open-source artificial pancreas systems (OpenAPS) on glycemic variability parameters in patients with type 1 diabetes. Methods: Eighteen patients receiving therapy within the past year with access to systems were evaluated. For each patient, the main indicators of glycemic variability (time in range – TIR, time above range – TAR, time below range – TBR, and glycemic variability coefficient – CV) were considered, extrapolated using ambulatory glucose profile (AGP), glycated hemoglobin, and insulin requirement. Results: Of the 18 patients (M/F: 11/7) examined, 5 came from AHCL systems, 6 used sensor-integrated systems, and 7 utilized iCGM or standalone CGM. The mean age was 37±13 years, and the mean duration of illness was 15±13 years. At baseline, patients exhibited: TIR 67±21%, TAR 29.25±12%, TBR 3±4%, CV 34.2±6, and a mean HbA1c of 6.8±1%. Each patient was fitted with a patch pump (Omnipod Dash or E-QUIL) with a CGM sensor (Dexcom G6 or G7) or iCGM (Freestyle Libre 2). An algorithm (Tidepool or Loop) compatible with the operating system in use (Android or IOS) was downloaded for each patient. Each patient counted carbohydrates. After 180 days of therapy with open-source systems, an intra-individual improvement was observed in the coefficient of variation of daily glucose (CV 30.8±6%, p=0.042), the percentage of time in range (77.3±12.5%; p=0.002), and time spent in hyperglycemia (19±14%). No significant changes were detected in HbA1c, insulin requirement, or time spent in hypoglycemia. No patient reported any algorithm malfunctions. Conclusions: This observational study provides a real-life snapshot of the effectiveness of OpenAPS systems, improving metrics compared to conventional AHCL systems in some cases. These data, though still limited, suggest the potential for greater customization of insulin pump therapy, leading to increased satisfaction among individuals with diabetes. However, it is crucial to recognize the pivotal role of the diabetologist in algorithm and device selection, transitioning from “Do It Yourself” systems to “Do It With Diabetologist” systems.