Impact of factory-calibrated Freestyle Libre System with new glucose algorithm measurement accuracy and clinical performance in children with type 1 diabetes during summer camp.

BACKGROUND: Factory-calibrated intermittently-scanned Continuous Glucose Monitoring (isCGM) device FreeStyle Libre (FSL) has recently received improvements in its glucose tracking algorithm and calibration procedures, which are claimed to have improved its accuracy. OBJECTIVE: To compare the accuracy of two generations of 14-days FSL devices (A in 2016, B in 2019) to self-monitored blood glucose measurements (SMBG) in children with type 1 diabetes in real-life conditions during a summer camp. MATERIALS AND METHODS: Two largely independent groups of youth with type 1 diabetes took part in summer camps. In 2016 they used FSL-A, in 2019 FSL-B. On scheduled days, participants performed supervised 8-point glucose profiles with FSL and SMBG. The accuracy vs SMBG was assessed with mean absolute relative difference (MARD) and clinical surveillance error grid (SEG). RESULTS: We collected 1655 FSL-SMBG measurement pairs from 78 FSL-A patients (age 13 ± 2.3 years old; HbA1c: 7.6 ± 0.8%) and 1796 from 58 in FSL-B group (age 13.8 ± 2.3 years old, HbA1c: 7.5 ± 1.1%)-in total 3451 measurements. FSL-B displayed lower MARD than FSL-A (11.3 ± 3.1% vs 13.7 ± 4.6%, P = .0003), lower SD of errors (20.2 ± 6.7 mg/dL vs 24.1 ± 9.6 mg/dL, P = .0090) but similar bias (-7.6 ± 11.8 mg/dL vs -6.5 ± 8 mg/dL, P = .5240). Both FSL-A and FSL-B showed significantly higher MARD when glycaemia was decreasing >2 mg/dL/min (FSL-A:22.3 ± 18.5%; FSL-B:17.9 ± 15.8%, P < .0001) compared with stable conditions (FSL-A: 11.4 ± 10.4%, FSL-B:10.1 ± 9.1%) and when the system could not define the glycaemic trend (FSL-A:16.5 ± 16.3%; FSL-B:15.2 ± 14.9%, P < .0001). Both generations demonstrated high percentage of A-class and B-class results in SEG (FSL-A: 96.4%, FSL-B: 97.6%) with a significant shift from B (decrease by 3.7%) to A category (increase by 3.9%) between generations (FSL-A: 16/80.4%; FSL-B:12.3/85.3%, P = .0012). CONCLUSION: FSL-B demonstrated higher accuracy when compared to FSL-A However, when glycemia is decreasing or its trend is uncertain, the verification with a glucose meter is still advisable.

Improved Estimation of Glycated Hemoglobin from Continuous Glucose Monitoring and Past Glycated Hemoglobin Data.

Background: Accurate estimation of glycated hemoglobin (HbA1c) from continuous glucose monitoring (CGM) remains challenging in clinic. We propose two statistical models and validate them in real-life conditions against the current standard, glucose management indicator (GMI). Materials and Methods: Modeling utilized routinely collected data from patients with type 1 diabetes from central Poland (eligibility criteria: age >1 year, diabetes duration >3 months, and CGM use between 01/01/2015 and 12/31/2019). CGM records were extracted from dedicated Medtronic/Abbott databases and cross-referenced with HbA1c values; 28-day periods preceding HbA1c measurement with >75% of the sensor-active time were analyzed. We developed a mixed linear regression, including glycemic variability indices and patient's ID (glucose variability-based patient specific model, GV-PS) intended for closed-group use and linear regression using patient-specific error of GMI (proportional error-based patient agnostic model, PE-PA) for general use. Models were validated with either new HbA1cs from closed-group patients or separate patient-HbA1c pool. External validation was performed with data from clinical trials. Performance metrics included bias, its 95% confidence interval (95% CI), coefficient of determination (R2), and root mean square error (RMSE). Results: We included 723 HbA1c-CGM pairs from 174 patients (mean age 9.9 ± 4.4 years and diabetes duration 3.7 ± 3.6 years). GMI yielded R2 = 0.58, with different bias between Medtronic and Abbott devices [0.120% vs. -0.152%, P < 0.0001], and overall 95% CI = -0.9% to +1%, RMSE = 0.47%. GV-PS successfully captured patient-specific variance (closed-group validation: R2 = 0.83, bias = 0.026%, 95% CI = -0.562% to 0.591%, RMSE = 0.31%). PE-PA performed similarly on new patients (R2 = 0.76, bias = -0.069%, 95% CI = -0.790% to 0.653%, RMSE = 0.37%). In external validation GMI, GV-PS, and PE-PA produced 73.8%, 87.5%, and 91.0% predictions within 0.5% (5.5 mmol/mol) from the true value. Conclusion: Constructed models performed better than GMI. PE-PA provided an accurate estimate of HbA1c with fast and straightforward implementation.