Real-World Performance of Personal Continuous Glucose Monitors During Hospitalization

Area of Science:

• Endocrinology
• Medical Devices
• Diabetes Management

Background:

• Continuous glucose monitors (CGMs) are increasingly used for diabetes management.
• Assessing CGM accuracy in inpatient settings is crucial for patient safety and effective glycemic control.

Purpose of the Study:

• To evaluate the accuracy of personal CGMs used by patients in a hospital setting.
• To determine the correlation between CGM accuracy and laboratory glucose values, as well as vital signs.
• To compare inpatient CGM performance metrics with published literature and clinical guidelines.

Main Methods:

• Adult patients with diabetes using personal CGMs upon admission were enrolled in a noncritical care setting.
• CGM readings were time-paired with point-of-care (POC) glucometers and laboratory (Lab) blood glucose measurements.
• Accuracy was assessed using mean absolute relative difference (MARD) and Clarke Error Grids; correlation analysis was performed for vital signs and lab values.

Main Results:

• CGM-POC pairs yielded a MARD of 13.7% (Dexcom: 14.4%, Libre: 11.8%), and CGM-Lab pairs showed a MARD of 13.6% (Dexcom: 12.7%, Libre: 15.4%).
• Over 98% of CGM-POC and CGM-Lab pairs fell within acceptable zones (A and B) on Clarke Error Grids.
• No significant correlation was found between CGM accuracy and mean arterial blood pressure, hemoglobin, glomerular filtration rate, or pulse oximetry.

Conclusions:

• Personal CGMs demonstrate adequate accuracy for inpatient glucose monitoring, with limitations noted in hypoglycemia ranges.
• Inpatient CGM performance is not significantly impacted by vital signs or laboratory results.
• Observed time-in-range (TIR) and glucose variability metrics align with inpatient guidelines and are comparable to hospital-inserted sensors.