Qualitative Study of User Experiences With Loop, an Open-Source Automated Insulin Delivery System

The “do-it-yourself” (DIY) diabetes movement has allowed people with diabetes to access fine-tuned automated insulin delivery systems for years prior to commercial systems being approved. While pump manufacturers were preparing automated insulin delivery (AID) systems for the market, more and more people with diabetes sought to have closed-loop control, leading them to Loop. This study by Sutratman and colleagues describes patient experiences using Loop as new users and experienced users through focus groups. Qualitative analyses addressed several themes that emerged from the data collected. Specific benefits of using Loop included decreased burden associated with diabetes management while glycemic targets were easier to achieve. Specifically, mental burdens, including worry, stress, and cognitive effort associated with type 1 diabetes (T1D) management, were decreased using Loop. Loop also allowed for greater discretion when managing diabetes because Loop users could bolus using their phone rather than having to retrieve a medical device. However, due to the timing of the study (enrollment ended in August 2019, before the release of Control-IQ in the United States), these comparisons may have been made against open loop therapy or earlier versions of AID systems, rather than more recently released devices.

Sleep has consistently been a challenge for people with T1D, as glycemic control overnight can be unpredictable, and continuous glucose monitors set off alarms frequently when glycemic excursions take place. Loop users described improvements in overnight blood sugar levels, improved sleep quality, and increased trust in how the Loop system would manage blood sugars overnight. One particularly unique characteristic of Loop highlighted in this article is that participants learned more about diabetes by reviewing their data. For example, Loop not only displays information about insulin on board, but also carbohydrates on board, and uses a dynamic carbohydrate absorption model to estimate carbohydrate absorption based on glycemic changes.  The “carbohydrates on board” feature is not available with commercial systems nor is it typically taught in diabetes education. There is also a breadth of data stored in Loop showing observed and predicted glucose changes and estimated times for carbohydrate absorption, and this information can be used to help individuals adjust carbohydrate absorption times when using Loop.

Although Loop users also described challenges, reporting difficulties adjusting settings, the overall feedback about the support received through the diabetes online community to help address these challenges was positive. Furthermore, participants pointed out a continued disconnect between patients and providers when using DIY systems. Physicians may still be hesitant to support Loop users or be unable to keep up with the changes with each new release. 

As additional AID systems have become commercially available after this study was conducted, it will be interesting to see how the future unfolds for Loop users. Many current Loop users may choose to transition to commercially available systems. However, the DIY Loop community may continue to be able to create new features that are beneficial to patients, paving the way for the next generation of AID systems.