Background: Patient-driven innovation has led to the development of do-it-yourself artificial pancreas systems (DIYAPS) for the management of type 1 diabetes (T1D). There have been few studies comparing glycemic outcomes for DIYAPS compared to conventional sensor-augmented pump (SAP) therapy.

Methods: A retrospective double cohort study was performed. The DIY group consisted of people with T1D using DIYAPS (OpenAPS, Loop, and AndroidAPS) who contributed data to the OpenAPS Data Commons. The SAP group included age-matched SAP users whose sensor data was obtained through the Tidepool Big Data Donation Project. The primary outcome of percent of sensor glucoses (SG) in low glucose range (<70 mg/dL) was determined for both groups and compared. Secondary outcomes included mean SG, number of >=15-minute hypoglycemic episodes (<70 mg/dL) per month, and percent of SG in high glucose range (>=180 mg/dL) and in target range (70-180 mg/dL).

Results: The DIY group included 90 days of data from 74 individuals with mean age of 36 ± 15 years and 24 ± 16 years since T1D diagnosis. The SAP group had 98 individuals with mean age of 35 ± 14 years and 20 ± 13 years since diagnosis. There was no significant difference in percent of SG <70 mg/dL between the two groups (3.8% for DIY vs. 4.7% for SAP, p=0.17). There was also no difference in number of hypoglycemic episodes per month (32.9 for DIY vs. 33.4 for SAP, p=0.87) or in percent of SG <54 mg/dL (0.8% for DIY vs. 1.3% for SAP, p=0.09). Mean SG was lower for the DIY group (134.9 ± 19.1 mg/dL for DIY vs. 150.3 ± 25.7 mg/dL for SAP, p<0.0001). Percent of SG >=180 mg/dL was lower (16.9% for DIY vs. 26.1% for SAP, p<0.0001) and percent in target glucose range was greater for the DIY group (79.3% for DIY vs. 69.2% for SAP, p<0.0001).

Conclusion: Users of DIYAPS had a comparable amount of hypoglycemia to those using SAP and had less hyperglycemia. This study suggests that the glycemic benefit of DIYAPS is in reducing hyperglycemia without compromising the low occurrence of hypoglycemia.


J. Zabinsky: None. H. Howell: None. A. Ghezavati: None. D.M. Lewis: Research Support; Self; JDRF. A. Nguyen: None. J.C. Wong: Consultant; Self; pH Pharma. Research Support; Self; Dexcom, Inc.



Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at