Evaluation of the CGMS Gold® Using a Standardized Test Brian R Hipszer, PhD, Jung-Won Park, MD, Jeffrey I Joseph, DO Results The minimum, maximum, and mean number of paired values per subject were 86, 131, and 108. The average maximum and minimum blood glucose concentrations were 183 ± 24 and 41 ± 4 mg/dL and the average range was 142 ± 21 mg/dL. The minimum, maximum, median, mean and standard deviation of the correlation coefficients are 0.3251, 0.9597, 0.7422, 0.7386 and 0.1831. Conclusions The correlation coefficient has not been widely used as a measure to assess the CGM accuracy because it is sensitive to the glycemic range under which the CGM is evaluated. However, if the test is standardized, any CGM can be evaluated and compared. Further Discussion The standardized test described here was used to evaluate accuracy of continuous glucose monitors. It has been reported that continuous interstitial glucose monitoring is inaccurate when the change in blood glucose is greater than 1.5 mg/dL/min (1) or glycemic levels are below 70 mg/dL (2). As such, this standardized test includes two periods of controlled glucose change and two periods of hypoglycemia. References 1. The Diabetes research in children network study group (DirecNet). Evaluation of factors affecting CGMS calibration. Diabetes Technology & Therapeutics. 2006;8:318-25. 2. DirecNet. Accuracy of the modified continuous glucose monitoring system (CGMS®) sensor in an outpatient setting: Results from a diabetes research in children network (DirecNet) study. Diabetes Technology & Therapeutics. 2005;7:109-14. 40 180 36 160 32 140 28 120 24 100 20 80 16 60 12 200 40 8 160 20 4 120 800 0 0 40 60 120 180 240 300 360 420 480 540 600 0 0 8 16 24 32 40 Time (min) 200 40 180 36 160 32 140 28 120 24 100 20 80 16 60 12 200 40 8 160 20 4 120 800 0 0 40 60 120 180 240 300 360 420 480 540 600 0 0 8 16 24 32 40 Time (min) 200 40 180 36 160 32 140 28 120 24 100 20 80 16 60 12 200 40 8 160 20 4 120 800 0 0 40 60 120 180 240 300 360 420 480 540 600 0 0 Sensor Current (nA) 200 Sensor Current (nA) Paired reference/sensor data were collected in twelve subjects. In each subject, the blood glucose was manipulated in a predefined manner using intravenous infusions of insulin and dextrose. Each reference value was the average of two plasma glucose concentrations measured simultaneously on both channels of the YSI 2300 Stat Plus (Yellow Springs, OH) using a single venous blood sample. Blood samples were obtained at alternating intervals of two and five minutes. The sensor’s output current from the CGMS Gold (Medtronic Diabetes, Northridge, CA) was recorded every five minutes. Spline interpolation was used to pair each reference value in time to a corresponding sensor value using Matlab (Mathworks, Natick, MA). Any data pairs recorded in the 60 minutes after sensor insertion were not analyzed. The correlation coefficient was computed for each subject. R 0.3251 0.9548 0.7372 0.7077 0.8431 0.7472 0.8537 0.6391 0.5530 0.6474 0.9597 0.8951 8 16 24 32 40 Time (min) Sensor Current (nA) Methods n 123 131 119 124 112 107 93 99 86 91 104 110 Reference Glucose (mg/dL) We investigated the accuracy of the Medtronic CGMS Gold® under a standardized glycemic pattern. Subject 1 2 3 4 5 6 7 8 9 10 11 12 Reference Glucose (mg/dL) Objective Reference Glucose (mg/dL) Thomas Jefferson University, Philadelphia, PA 19107, 2. Drexel University, Philadelphia, PA 19104