Predicting Glycated Hemoglobin Through Continuous Glucose Monitoring in Real-Life Conditions: Improved Estimation...

Area of Science:

• Endocrinology
• Biomedical Engineering
• Data Science

Background:

• Continuous glucose monitoring (CGM) enhances glycemic control in diabetes management.
• CGM data, when analyzed effectively, can provide reliable estimates of key clinical markers such as glycated hemoglobin (HbA1c).
• The REALISM-T1D study contributes to advancing HbA1c estimation techniques for clinical application.

Purpose of the Study:

• To evaluate advanced data analysis techniques for improving HbA1c estimation from CGM data.
• To compare the accuracy of improved HbA1c estimation methods against traditional methods and laboratory assays.
• To assess the potential of these enhanced methods as a complement or alternative to conventional HbA1c testing.

Main Methods:

• Acquisition of CGM data from 27 adults with type-1 diabetes using Dexcom G6 sensors over 120 days.
• Performance of standard glycated hemoglobin (HbA1c) laboratory assays at follow-up visits to serve as the gold standard.
• Analysis of CGM data incorporating smart interpolation for missing values and optimized interstitial glucose (IG) weighting functions for MIG to HbA1c regression.

Main Results:

• Improved HbA1c estimation quality was observed using advanced interpolation and IG weighting functions compared to methods relying solely on mean interstitial glucose (MIG).
• Bland-Altman plots demonstrated a reduction in variance, indicating enhanced accuracy against the gold standard.
• Statistical analysis (Wilcoxon signed-rank test) confirmed a significant improvement (P = .0179) in bias-compensated mean squared error over conventional MIG-based methods.

Conclusions:

• Enhanced HbA1c estimation methods provide superior prediction quality compared to MIG-alone approaches, offering timely feedback to clinicians.
• These refined methods address existing discrepancies in literature and show promise for clinical utility.
• Further development could position these techniques as a viable alternative or adjunct to standard HbA1c assays.