Discrepancies between methods of continuous glucose monitoring in key metrics of glucose control in children with type 1 diabetes

OBJECTIVE: We aimed to compare glycemic control and variability parameters obtained from paired records of real-time continuous glucose monitoring (RT-CGM) and flash glucose monitoring (FGM). METHODS: Ten Polish boys and 11 girls aged 15.3 ± 2.1 years with type 1 diabetes for 7.7 ± 4.5 years and glycated hemoglobin 7.35 ± 0.7% (57 ± 5 mmol/mol) were recruited between August 2017 and June 2018 and equipped with devices for RT-CGM (iPro2 system with Enlite electrodes) and FGM (FreeStyle Libre) for 1 week. Afterwards, Glyculator 2.0 software was used to calculate and compare key metrics of glycemic control listed in the International Consensus on Use of Continuous Glucose Monitoring, with distinction into all record/night-time/day-time blocks when appropriate. RESULTS: Agreement between the two systems' measurements across patients ranged from poor (R2  = .39) to nearly perfect (R2  = .97). Significant differences between RT-CGM and FGM were observed in five important metrics: coefficient of variation (median difference: -4.12% [25%-75%: -7.50% to -2.96%], P = .0001), low blood glucose index (-0.88 [-1.88 to -0.18], P = .0004), % of time below 70 mg/dL (3.9 mmol/L) (-4.77 [-8.39 to -1.19], P = .0015) and 54 mg/dL (3 mmol/L) (-1.33 [-4.07 to 0.00], P = .0033) and primary time in range (TIR) 70-180 mg/dL (8.58 [1.31 to 12.66], P = .0006). CONCLUSIONS: RT-CGM and FGM differ in their estimates of clinically important indices of glycemic control. Therefore, such metrics cannot be directly compared between people using different systems. Our result necessitates system-specific guidelines and targets if TIR and glycemic variability are to be used as an endpoint in clinical trials.