Pilot Study of Intensive Care Unit Continuous Glucose Monitoring

The investigators believe that there remains a gap in implementing insulin infusions in critically ill patients to maximize the benefit and minimize adverse events like episodes of hypoglycemia. Based on the published experience with Continuous Glucose Monitor (CGM), the investigators believe that it is safe to use in critically ill patients. Furthermore, the investigators believe that in combination with a protocol with low risk for hypoglycemia at baseline, that CGM can eliminate this risk fully. In this study the investigators will: 1. Study the safety and feasibility of the continuous glucose monitor use in 20 critically ill patients for 7 days (the current maximum recommendation for sensor use). Safety data will include the rate of significant bleeding (hematoma) or infection (cellulitis) from sensor use. Feasibility data will evaluate the amount of missing glucose data over the 7-day sensor life. 2. Randomize patients treated with the current UVA intensive care insulin protocol for insulin management to the addition of \"brakes\" that reduce insulin administration based on continuous glucose monitoring data between hourly reference glucose data to prevent episodes of hypoglycemia (blood glucose \<70 mg/dl) and severe hypoglycemia (blood glucose \<50 mg/dl). This will serve as pilot data to power a larger study in the future.