A Study to Evaluate a Multiple Model Probabilistic Predictive Controller (MMPPC) for Closed Loop Insulin Delivery

Type 1 Diabetes Mellitus Clinical Trial

Official title: A Study to Evaluate a Multiple Model Probabilistic Predictive Controller (MMPPC) for Closed Loop Insulin Delivery

Status Withdrawn

Clinical Trial Summary

You are invited to participate in a research study for the development of an artificial pancreas. An artificial pancreas uses a program which takes information from a continuous blood glucose monitor and uses that information to tell an insulin infusion pump how much insulin to deliver. The primary purpose of this study is to gain experience with insulin delivery algorithms or programs program (algorithm) provides the best regulation of glucose levels so that there are no severe low blood glucose reactions and blood glucose levels are generally between 70 to 180 mg/dl.

Clinical Trial Description

You are invited to participate in a research study for the development of an artificial pancreas. An artificial pancreas uses a program which takes information from a continuous blood glucose monitor and uses that information to tell an insulin infusion pump how much insulin to deliver. The primary purpose of this study is to gain experience with insulin delivery algorithms or programs program (algorithm) provides the best regulation of glucose levels so that there are no severe low blood glucose reactions and blood glucose levels are generally between 70 to 180 mg/dl. If the system is working properly, you would not need to enter the amount of food you were eating, give an insulin bolus, or change your basal rates while wearing the device. You would need to periodically check to be sure the continuous glucose sensor was functioning properly and you would need to respond to alarms that might occur if your blood glucose was too high, too low, or the glucose sensor or pump were not working well. In addition to this the investigators will also gain experience with insulin delivery algorithms to minimize the number of glucose readings which are above or below target. It is our intention to modify the algorithms during these studies.

In this study the investigators plan to use a commercially available insulin infusion pump (OmniPod) to deliver lispro (Humalog) insulin. The investigators will use Navigator continuous glucose sensors both to monitor glucose levels (sensor 1) and provide the glucose concentration for the closed loop algorithm (sensor 2). The signal from the second Navigator will be sent by serial cable to a computer which will be at the patient's bedside. A control algorithm will reside on the computer, and the amount of insulin to be delivered will be transmitted to the OmniPod Personal Device Manager which will then send a radiofrequency (rf) signal to the Omnipod pump residing on the subject. A health care provider will be in attendance and monitoring discrete blood glucose levels (YSI, GlucoScout, or HemoCue measurements) at least every 30 minutes. While the Navigator and Omnipod are commercially available they will be used in this study as part of an investigational system while you are in the hospital, an investigational system is one that is not approved for use by the FDA. ;

Study Design

Intervention Model: Single Group Assignment, Masking: Open Label, Primary Purpose: Treatment

Related Conditions & MeSH terms

• Diabetes Mellitus
• Diabetes Mellitus, Type 1
• Type 1 Diabetes Mellitus

• NCT number: NCT01492062
• Study type: Interventional
• Source: Stanford University
• Status: Withdrawn
• Phase: N/A
• Start date: February 2011
• Completion date: November 2013