This study was the second of two studies designed to evaluate the performance of Echo's Symphony tCGM System in the critical care setting. The study was performed at Thomas Jefferson University Hospital and enrolled fifteen (15) adult patients. The skin of each patient was prepared using Prelude and a Symphony tCGM biosensor was applied to the skin site after transfer to critical care. Reference blood samples were taken from arterial line catheters at 30-minute intervals and measured on a YSI 2300 STAT Plus Glucose Analyzer. The data collected by Symphony was blinded to study subjects and Jefferson clinical staff. At the conclusion of the study period, the test sites were inspected for redness or other undesirable effects.
Continuous data from the Symphony tCGM System were compared to reference measurements from the YSI 2300 STAT Plus Glucose Analyzer. Those reference measurements were paired with the Symphony results through a data analysis algorithm. The primary statistical analytical tools used to evaluate the performance of Symphony were the Continuous Glucose-Error Grid analysis (CG-EGA) and Mean Absolute Relative Difference (MARD). The CG-EGA is a categorization of all data pairs based on the clinical significance of the accuracy. Accurate readings result in the same clinical decision when based on the CGM value versus the blood glucose value. Benign errors lead to the same clinical outcome as accurate readings even though the actual clinical decision may differ. Erroneous readings lead to clinical errors. CGM performance is measured as the sum of accurate readings and benign errors. Numerical accuracy is measured using MARD, an error calculation tool that was used to measure the absolute value of the average relative difference between Symphony and the reference measurements, on a percentage basis.