The isCGM (FreeStyle Libre “flash”) is approved in a number of countries for use in adults, but has had minimal study in pediatric groups and no reports of its use during ketoacidosis. This study in pediatric patients sought to compare paired samples from this isCGM to the Accucheck Aviva glucose monitor under conditions of varying glucose levels and varying degrees of ketoacidosis.

Twelve hyperglycemic T1D patients, ages 6-16 years, were studied with a minimum of 6 paired comparisons in each condition: 3 with ketoacidosis (pH 7.2-7.3) and one with severe ketoacidosis (pH <7.2), and all others nonketotic. Up to a 15% measurement difference between BG and tissue glucose measurements for BG > 100 mg/dL, or an absolute difference of 15 mg/dL for BG < 100 mg/dL, were considered within target range. Blood glucose levels ranged from 35 to 556 mg/dl (1.94-30.89 mmol/L) and pH levels 6.9 to 7.6. Mean absolute relative differences (MARD, isCGM vs. Accucheck paired measurements) were: 426 non-ketoacidotic with BG >100 mg/dL: MARD of 11.77% with 69.2% of values within a 15% difference; 96 non-ketoacidotic with BG < 100 mg/dL: MARD of 12.4% with 78.1% within 15%; 17 ketoacidotic with BG>100 mg/dL: MARD of 12.42% with 70.6% within 15%; and 6 severely ketoacidotic: MARD of 22.68% with 16.7% within 15%, constituting a major difference with severe ketoacidosis. Clarke Error Grid Analysis for results with non-ketoacidotic values with BG <=79 mg/dL: 93% in Zone A and 7% in Zone B; ketoacidotic values: 94.11% in Zone A and 5.88% in Zone B; and severely ketoacidotic values: 50% in Zone A and 50% in Zone B. Of the 545 paired comparisons obtained, only 17 were obtained during ketosis and 6 during severe ketoacidosis. These results show that this “flash” isCGM and BG measurements determined by Accucheck Aviva are highly concordant, however, the major differences observed during severe ketoacidosis urge caution in interpreting the blood glucose levels under this condition.


G.A. Bou: None. M. Wurm: None. B.C. Hansen: None. K. Schwab: None.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at