CREATE Trial Data Add to Support of Open-Source AID Use in T1D

Findings presented at the 82nd Scientific Sessions of the American Diabetes Association support the safety and efficacy of open-source automated insulin delivery (AID) systems among children and adults living with type 1 diabetes (T1D).

Results of the 24-week, multicenter, randomized controlled CREATE trial, which compared open-source automated insulin delivery (AID) systems with sensor-augmented pump therapy (SAPT) show, show AID diabetes technology to be safe and effective for children and adults living with type 1 diabetes.

Among the 48 children aged 7 to 15 years and the 49 adults aged 16 to 70 years, most had little to none AID system experience prior to the study.

“We compared open-source AID (using the OpenAPS algorithm from a version of AndroidAPS implemented in a smartphone with the DANA-i insulin pump and Dexcom G6 continuous glucose monitor) to SAPT,” the authors wrote. Their findings were released during the recent 82nd Scientific Sessions of the American Diabetes Association, held virtually and in New Orleans. The primary study outcome was change in percent of time spent in the target sensor glucose range (TIR), which for this study was defined as 3.9 to 10 mmol/L, from study run-in to the last 2 weeks of the trial.

An overall 14% difference was seen for 24-hour TIR, favoring the AID group. Looking at each age group randomized to AID, the mean (SD) TIR for the adults was 74.5% (11.9%) (Δ+, 9.6% [11.8%] from run-in; P < .001) and for the children, 67.5% (11.5%) (Δ+, 9.9 [14.9%] from run-in; P < .001). These improvements from baseline were immediate and sustained, the study authors noted. At run-in, the TIR was 64.7% (12.9%) for the adults and 57.4% (10.6%) for the children.

Similar results were not seen in the group randomized to SAPT use. For adults in this study arm, the mean TIR at the end of the study was 56.5% (14.2%), and for children, 52.5% (17.5%). There was no change when comparing findings with the run-in numbers for this group, which were 60.3% (15.6%) for the adults and 55.1% (12.6%) for the children.

In addition, 4 times as many participants in the AID group vs the SAPT group met the above 70% threshold for TIR as set forth by international guidelines: 60% vs 15%, respectively.

Nighttime results continued the trend of more TIR seen in the AID group:

  • AID group from run-in to study end:
    • Adults: 62.6% (18.1%) to 85.2% (12.7%)
    • Children: 59.1% (16.5%) to 76.8% (15.8%)
  • SAPT group from run-in to study end:
    • Adults: 62.8% (20.9%) to 53.5% (20.1%)
    • Children: 57.0% (18.9%) to 57.2% (21.4%)

“While open-source AID is not yet approved by the FDA, we know people worldwide are using it to manage their diabetes,” said Martin de Bock, PhD, FRACP, pediatric endocrinologist at the University of Otago, Christchurch, New Zealand, and lead investigator of the study, in a statement. “Our findings demonstrate that this is a safe and effective technology and adds to the body of evidence supporting use of open-source AID for improving glycemic outcomes.”

There were no instances of severe hypoglycemia or diabetic ketoacidosis in either study arm and no increase in overall hypoglycemia.

Reference

Burnside MJ, Lewis DM, Crocket H, et al.The CREATE Trial: Randomized Clinical Trial Comparing Open-Source Automated Insulin Delivery with Sensor Augmented Pump Therapy in Type 1 Diabetes. Presented at: 82nd Scientific Sessions of the American Diabetes Association; June 3-7, 2022; New Orleans, Louisiana, and Virtual. Abstract 286-OR. Accessed June 8, 2022. https://eppro02.ativ.me/src/EventPilot/php/express/web/page.php?page=IntHtml&project=ADA22&id=251