First Outpatient Evaluation of a Tubeless Automated Insulin Delivery System with Customizable Glucose Targets in Children and Adults with Type 1 Diabetes

Abstract

Background: The objective of this study was to assess the safety and effectiveness of the first commercial configuration of a tubeless automated insulin delivery system, Omnipod^® 5, in children (6-13.9 years) and adults (14-70 years) with type 1 diabetes (T1D) in an outpatient setting. Materials and Methods: This was a single-arm, multicenter, prospective clinical study. Data were collected over a 14-day standard therapy (ST) phase followed by a 14-day hybrid closed-loop (HCL) phase, where participants (n = 36) spent 72 h at each of three prespecified glucose targets (130, 140, and 150 mg/dL, 9 days total) then 5 days with free choice of glucose targets (110-150 mg/dL) using the Omnipod 5. Remote safety monitoring alerts were enabled during the HCL phase. Primary endpoints were difference in time in range (TIR) (70-180 mg/dL) between ST and HCL phases and proportion of participants reporting serious device-related adverse events. Results: Mean TIR was significantly higher among children in the free-choice period overall (64.9% ± 12.2%, P < 0.01) and when using a 110 mg/dL target (71.2% ± 10.2%, P < 0.01), a 130 mg/dL target (61.5% ± 7.7%, P < 0.01), and a 140 mg/dL target (64.8% ± 11.6%, P < 0.01), and among adults using a 130 mg/dL target (75.1% ± 11.6%, P < 0.05), compared to the ST phase (children: 51.0% ± 13.3% and adults: 65.6% ± 15.7%). There were no serious device-related adverse events reported during the HCL phase, nor were there episodes of severe hypoglycemia or diabetic ketoacidosis. Conclusion: The Omnipod 5 System was safe and effective when used at glucose targets from 110 to 150 mg/dL for 14 days at home in children and adults with T1D.

Keywords: Artificial pancreas; Automated insulin delivery; Closed-loop; Omnipod; Tubeless insulin pump.

FIG. 1.

Components of the commercially intended Omnipod 5 Automated Insulin Delivery System. From left to right: (1) tubeless insulin pump (Pod) containing the automated insulin delivery algorithm (52 × 39 × 14.5 mm without adhesive, 26 g when empty); (2) Omnipod 5 application (app), pictured running on provided locked-down smartphone handheld device (144 × 67.6 × 12.4 mm, 165 g); (3) interoperable CGM (Dexcom G6, see Dexcom product documentation for additional information). The Pod is a lightweight, waterproof (IP28), self-adhesive insulin pump that delivers insulin through an automatically inserted cannula. The automated insulin delivery algorithm is built into the Pod, which receives glucose measurements every 5 min directly from the on-body CGM. The algorithm then commands the Pod to deliver microboluses every five minutes based on current and projected glucose values, with the goal of achieving and maintaining a set target glucose value. The user interacts with the system through the Omnipod 5 app, which communicates with the Pod through Bluetooth^® wireless technology. Actions performed using the app include: complete initial setup, activate and deactivate Pods, start Automated Mode, use the bolus calculator, deliver insulin boluses, enable the HypoProtect feature, view insulin delivery and CGM history, respond to system alerts and alarms, check Pod status, and adjust settable parameters. The app home screen (pictured) prominently displays the current CGM value and trend, as well as the amount of insulin on board, information about the last bolus, and a link to view the CGM history graph. The bolus calculator is accessed using the circular icon near the bottom of the screen. Since the algorithm runs on the Pod, and the Pod and CGM are both worn on-body and communicate directly, automated insulin delivery can continue uninterrupted even if the handheld device containing the app is not nearby. Copyrighted image used with permission. © 2020 Insulet Corporation. All rights reserved. CGM, continuous glucose monitor.

FIG. 2.

Interquartile plot of median sensor glucose profile over 24-h for (A) adults (age 14–70 years) during the HCL free-choice period overall (n = 18); (B) adults using the 110 mg/dL glucose target during the HCL free-choice period (n = 12); (C) children (age 6–13.9 years) during the HCL free-choice period overall (n = 18); and (D) children using the 110 mg/dL glucose target during the HCL free-choice period (n = 11), each compared to the median sensor glucose profile per age group from the 14-day ST phase. No prespecified glucose targets were used during the ST phase, as the ST was managed according to the patient's personal diabetes treatment goals. The data are presented as median (line) and interquartile range (shaded area) of sensor glucose per time of day across all participants and days. The target range of 70–180 mg/dL is indicated by black dashed lines, and the target glucose (if applicable) is indicated by a solid black line. HCL, hybrid closed-loop; ST, standard therapy.

FIG. 2.

Interquartile plot of median sensor glucose profile over 24-h for (A) adults (age 14–70 years) during the HCL free-choice period overall (n = 18); (B) adults using the 110 mg/dL glucose target during the HCL free-choice period (n = 12); (C) children (age 6–13.9 years) during the HCL free-choice period overall (n = 18); and (D) children using the 110 mg/dL glucose target during the HCL free-choice period (n = 11), each compared to the median sensor glucose profile per age group from the 14-day ST phase. No prespecified glucose targets were used during the ST phase, as the ST was managed according to the patient's personal diabetes treatment goals. The data are presented as median (line) and interquartile range (shaded area) of sensor glucose per time of day across all participants and days. The target range of 70–180 mg/dL is indicated by black dashed lines, and the target glucose (if applicable) is indicated by a solid black line. HCL, hybrid closed-loop; ST, standard therapy.