Commercially developed open source helper systems that are believed to be safe and effective



Using data from real-world studies, investigators determined that commercially developed and open source automated insulin dosing systems are safe and effective.

A systematic review summarizing the real-world findings on available automated insulin dosing systems (AID) concluded that commercially developed and open source AID systems are safe and effective treatment options for people with diabetes (PwD) in different age groups and Are genders.

To the best of the investigators’ knowledge, the current systematic review is concentrating for the first time on practical studies on several AID systems. The results of the review were published in Diabetic medicine.

Unlike commercially developed systems, open source AID systems are developed by a community of PwD using freely available source codes and documentation found online.

It is estimated that over 10,000 patients are currently using these systems – including children and adolescents whose caregivers build the systems on their behalf, the researchers said. “None of the open source AID systems has yet received regulatory approval; There is no liability as with commercially developed medical devices, ”they added.

There are currently no data from randomized controlled trials (RCTs) on open source AID systems, but a study is in the works.

To conduct the review, investigators collected data from 21 studies published between 2018 and 2021: 12 on Medtronic 670G; one on Tandem Control-IQ; 1 on Diabeloop DBLG1; 2 on AndroidAPS; 1 on OpenAPS; 1 on loop; and 3 comparing different types of AID systems.

The primary endpoints for this review included percent time-in-range (TIR; 70–180 mg / dL, 3.9–10.0 mmol / L), change in TIR, and glycated hemoglobin (A1C).

Of 12 studies examining real-world use of the Medtronic 670G – the first AID system approved in the US in 2016 – 10 studies found significant improvements in TIR and 5 reported significant improvements in A1C.

In addition, real data collected over 7 weeks from 1,435 US-based PwD using the Tandem Control IQ algorithm showed a significant improvement in TIR at 3 weeks and at the end of the study of 78.2% (70.2% -85.1%) to 79.2% (70.3% -86.2%; P.<.001 without increasing the time underrun before using aid study participants had a relatively high tir compared to general population with type diabetes>

A study examining the DBLGI system among 25 users> 22 years of age showed improvements in TIR from an average (SD) 53% (16.4%) to 69.7% (P. P < .001]) without serious side effects.

Of 7 studies that analyzed the 3 open source AID systems OpenAPS, AndroidAPS and Loop, all showed a significantly reduced A1C level and an increased TIR. No increase in hypoglycemia was seen in any of the 4 studies that measured TBR.

A study evaluating device data from OpenAPS users found that the average TIR of the entire cohort of 80 people during the first 180 days was 77.5 (10.5%), with no further significant changes between days 1- 60, 61-120 and 121-180.

A significant reduction in the estimated A1C (eA1C) from 49 (14) to 44 (17) mmol / mol (6.6 [0.9%]-6.2 [0.6%]; P. P.

After performing glycemic measurements on 558 Loop users, the Loop Observational Study found that the TIR increased significantly from 67 (16%) to 73.0 (13%) and the A1C increased from 51 (11) mmol / mol after 6 months (6.8.) Decreased [1%]) to 48 (9) mmol / mol (6.5 [0.8%]) after 6 months (P. <.001>

Combined with evidence from randomized clinical trials, real evidence on AID systems and their impact on glycemic outcomes is helpful in assessing safety and efficacy, the authors concluded.


Knoll C, Peacock S, Grove M, et al. Real-world evidence of clinical outcomes from people with type 1 diabetes using open source and commercial automated insulin dosing systems: a systematic review. Diabetes med. Published online November 12, 2021. doi: 10.1111 / dme.14741



Comments are closed.