Accuracy of Factory-Calibrated Continuous Glucose Monitors in Critically Ill Patients Receiving Intravenous Insulin: A Prospective Clinical Trial of Two Leading Systems.
Gautam Ramesh, Emily Kobayashi, Navyaa Sharma et al.
✦ AI-curated · Sources linked
In 30 seconds
This nonrandomized, open-label, clinically blinded prospective trial assessed the accuracy of two factory-calibrated continuous glucose monitors in 40 critically ill hospitalized patients receiving continuous intravenous insulin. CGM readings were compared with laboratory serum and point-of-care glucose measurements. Mean absolute relative difference was 22.7% for Dexcom G6 Pro versus laboratory values and 25.2% for FreeStyle Libre Pro versus laboratory values; accuracy varied considerably between patients, and both systems tended to underestimate glucose. The findings suggest that factory calibration alone may not ensure adequate accuracy for inpatient use in this population.
Key findings
- Dexcom G6 Pro had a 22.7% MARD against laboratory glucose and 22.9% against point-of-care glucose.
- FreeStyle Libre Pro had a 25.2% MARD against laboratory glucose and 27.0% against point-of-care glucose.
- Both systems showed considerable inter-patient variability and tended to underestimate glucose relative to reference measurements.
- Low reference glucose values were rarely overestimated by either sensor.
Why it matters
Glucose measurements guide insulin management in critically ill patients, so sensor inaccuracies may affect interpretation of glycaemic status. The observed variability and tendency toward underestimation indicate that performance in this setting cannot be assumed from factory calibration alone. These results help frame the need to evaluate CGM readings against established glucose measurements when considering inpatient use.
What to keep in mind
The trial was nonrandomized and included 40 patients; the abstract does not report clinical outcomes associated with CGM use.