Evaluation of continuous flash glucose monitoring in a pediatric ICU setting.

Glucose monitoring is of great importance among patients in intensive care units (ICU). The purpose of this study is to assess the performance of a new flash glucose monitoring (FGM) system in a pediatric ICU setting. Sixteen consecutive patients admitted in pediatric ICU aged > 4 years, expected length stay > 2 days and with no medication or existing diagnosis affecting glucose metabolism were enrolled. FreeStyle Libre sensor was applied to the upper arm of the patients (8 boys). FGM measurements were compared to 3 "references": arterial blood gas analysis, capillary blood analysis and biochemical serum analysis. Mean age of patients was 8.03 ± 2.91 years. Sensors remained in situ for a median of 9.71 ± 5.35 days. Removal of the sensor was mainly attributed to the completion of the predefine life-span of the sensor or discharge of the patient from the ICU. We compared 711 pairs of measurements between the sensor and other glucose measurement methods. Glucose values from the sensor were consistently lower with mean absolute relative difference (MARD) being 28.34%, 25.11% and 18.99% compared to the blood gas analyzer, capillary blood glucose meter, and biochemical serum analysis, respectively, but a wide interindividual variability. Significant linear correlations between age and MARD values were observed. Surveillance error grid (SEG) analysis showed 92.04%, 94.67% and 95.52% of the readings in the none or slight risk zone respectively. FreeStyle Libre is well tolerated although not adequately accurate with a tendency to underestimate glucose levels in critically ill pediatric patients.