Evaluation of the accuracy of a microdialysis-based glucose sensor during insulin-induced hypoglycemia, its recovery, and post-hypoglycemic hyperglycemia in humans.

Background: These studies were designed to evaluate the accuracy of a microdialysis-based subcutaneous glucose sensor (GlucoDay, A. Menarini Diagnostics, Firenze, Italy) compared with a standard reference method of plasma glucose measurement during insulin-induced hypoglycemia.

Research Design And Methods: Nine subjects without diabetes were studied in eu-, hypo-, and hyperglycemia (clamp technique). The GlucoDay was calibrated against one arterialized plasma glucose measurement (Glucose Analyzer, Beckman, Brea, CA), and plasma glucose estimates every 3 min were compared with paired plasma glucose values.

Results: Accuracy of glucose estimates was not homogeneously distributed among subjects and depended on stability of the sensor's current signal during spontaneous euglycemia (R +/- -0.68). Linear regression analysis showed a good correlation between the two methods of measurement (R = 0.9), Deming regression showed the inclusion of the unit in the confidence interval of the slope (slope 0.95, 95% confidence interval 0.87-1.02), and the accuracy of the GlucoDay reached 40 +/- 15% (American Diabetes Association criteria). The mean relative difference was 6 +/- 8% in euglycemia, 13 +/- 14% during plasma glucose fall, 5 +/- 22% in the hypoglycemic plateau, and -14 +/- 16% during recovery from hypoglycemia. The Bland-Altman analysis indicated a bias of -1.9 +/- 16.6 mg/dL, whereas the Error Grid Analysis showed 94% of the Gluco- Day measurements in the acceptable zones of the grid. The time to reach the glycemic nadir was longer when measured with the GlucoDay (90 +/- 5 vs. 72.5 +/- 9 min, P < 0.05). However, absolute values of glycemic nadir, time spent in hypoglycemia, and the rate of fall of glycemia and the rate of recovery from the hypoglycemia were not statistically different.

Conclusions: GlucoDay closely monitors changes in plasma glucose before, during, and after hypoglycemia. However, these results can be achieved only if calibration of the GlucoDay is performed under conditions of sensor signal stability. Similar studies have to be performed in subjects with diabetes to validate the GlucoDay system.