635-P: A Simplified Approach for Evaluating and Visualizing CGM Data in People with Diabetes

Introduction: The Ambulatory Glucose Profile (AGP) visualizes and summarizes complex CGM data and contains multiple and often difficult to interpret measures. Our objective was to create a simple visual summary of the data that can provide insights for CGM users. Methods: Repeated measure correlation was used to quantify the association among AGP metrics. Then, multilevel principal components analysis (PCA) was applied to determine the key metrics that explained the most variation within the data. We then created a 3-D measure using the relevant parameters and designed a 2-D visualization which interprets the 3-D measures and exhibits the trajectory of a patient’s glucose status over time. Results: Based on the correlation analysis, mean glucose, time in range (TIR) and time above range are strongly correlated. Two pairs, TIR and SD, time below range (TBR) and CV, are moderately correlated. In the PCA study, 3 components explained over 95% of the data variation. Considering the correlations and mPCA, we determined that TIR, CV, and TBR were sufficient to describe a patient’s status. Then, we constructed a 2-D framework to visualize the status changes over time. The figure shows a CGM user’s data over 4 weeks demonstrating improvement in the 3 key metrics over time. Conclusions: This novel data visualization based on the key metrics may assist CGM users and their clinicians in assessing overall progress with glucose control over time. Disclosure S. Liu: None. M. Shomali: Employee; Self; WellDoc, Inc. A. Kumbara: Employee; Self; WellDoc, Inc. A. K. Iyer: Employee; Self; WellDoc, Inc. M. Peeples: Employee; Self; WellDoc, Inc. M. A. Dugas: None. K. Crowley: None. G. Gao: None.