This is a randomized crossover study testing the efficacy of the Fault Detection algorithms using the Zone MPC algorithm and DiAs artificial pancreas platform in adult patients with type 1 diabetes. The trial will last for 6 weeks for each individual subject, with three weeks using the AP algorithm and three weeks using sensor augmented pump in a randomized order
Investigational Device: Artificial Pancreas System with Fault Detection Algorithms: Roche Accu-Check Spirit Combo Insulin Pump, Dexcom G4P System with Share, Diabetes Assistant (DiAs) on Android phone, DiAs Web Monitoring (DWM) \- referred to as Remote Monitoring Server, MPC control algorithm, Health Monitoring System (HMS) algorithm. Sensor and infusion set fault detection algorithms will be applied offline with data obtained from server and notifications will be sent to the clinician. Control Arm: Sensor-augmented insulin pump therapy: Subject will use their personal insulin pump and Dexcom G4P System with Share. Primary Objective: To determine the efficacy of the fault detection algorithm. The primary outcome is based on the amount of time the sensor glucose is \>250 mg/dL in the 4 hours preceding detection of the infusion set failure during sensor augmented pump therapy vs. closed-loop control with fault detection alerts. Secondary Objectives: To determine the effectiveness of the sensor fault detection algorithm. To determine the efficacy of the Zone MPC controller by evaluating glycemic outcomes Number of Subjects: There will be 20 subjects recruited: 10 at Stanford and 10 at Denver (up to 36 subjects will be enrolled to reach 20 subjects completing the study) Diagnosis and Main Inclusion Criteria: Adult subjects between 18 and 55 years of age inclusive, diagnosed with type 1 diabetes. Trial Design: This outpatient study will be conducted over 6 weeks as shown in the figure below. The 6-week period will consist of two 2-week blocks of prolonged infusion set wear with a 1-week sensor run-in period preceding each block. In each block, subjects will wear an infusion set for up to 7 days. A new infusion set will be inserted at the start of each week in the block. Following enrollment procedures, subjects will be randomized at a ratio of 1:1 to either use the AP system with fault detection algorithms (intervention) or sensor-augmented pump therapy (control) in the first block.
Study Type
INTERVENTIONAL
Allocation
RANDOMIZED
Purpose
TREATMENT
Masking
NONE
Enrollment
20
The AP system using fault detection algorithms will determine whether insulin infusion problems are occurring and may prevent severe hyperglycemia due to its predictive nature
Amount of time sensor glucose levels are >250 mg/dl
Amount of time sensor glucose levels are \>250 mg/dl in the control arm versus the experimental arm
Time frame: 4 hours after insulin infusion set failure
Effectiveness of sensor fault detection algorithm as defined by % of sensor failures caught by the system
Effectiveness of sensor fault detection algorithm as defined by % of sensor failures caught by the system
Time frame: During 2 week intervention period
Mean sensor glucose values
mean sensor glucose values
Time frame: 2 week intervention period versus 2 week control period
Percent of time in range between 70-180 mg/dl
Percent of time in range between 70-180 mg/dl
Time frame: 2 week intervention period versus 2 week control period
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.