BACKGROUND: The recent availability of a continuous glucose monitor offers the opportunity to match the demands of intensive diabetes management with a period of equally intensive blood glucose monitoring. The present study evaluates the performance of the MiniMed continuous glucose monitoring system (CGMS) in patients with diabetes during home use. METHODS: Performance data and demographic information were obtained from 135 patients who were (mean +/- SD) 40.5+/-14.5 years old, had an average duration of diabetes of 18.0+/-9.8 years, 50% were female, 90% were Caucasian, and 87% of whom had been diagnosed with type 1 diabetes. Patients were selected by their physician, trained on the use of the CGMS and wore the device at home for 3 days or more. The performance of the CGMS was evaluated against blood glucose measurements obtained using each patient's home blood glucose meter. Evaluation statistics included correlation, linear regression, mean difference and percent absolute difference scores, and Clarke error grid analysis. RESULTS: The CGMS values were compared to 2477 SMBG tests (r = 0.91, slope = 0.93, intercept = 14.5 mg/dL, mean absolute difference = 18.0%+/-19.8%). Clarke error grid analysis showed 96.2% of the data pairs falling within the clinically acceptable regions (zones A and B). CONCLUSIONS: These results demonstrate the agreement of the CGMS to blood glucose meter values, under conditions of home use, in patients selected by their physicians as candidates for continuous monitoring. The detailed glucose information provided by the CGMS should make successful management of diabetes more easily achieved.
BACKGROUND: The recent availability of a continuous glucose monitor offers the opportunity to match the demands of intensive diabetes management with a period of equally intensive blood glucose monitoring. The present study evaluates the performance of the MiniMed continuous glucose monitoring system (CGMS) in patients with diabetes during home use. METHODS: Performance data and demographic information were obtained from 135 patients who were (mean +/- SD) 40.5+/-14.5 years old, had an average duration of diabetes of 18.0+/-9.8 years, 50% were female, 90% were Caucasian, and 87% of whom had been diagnosed with type 1 diabetes. Patients were selected by their physician, trained on the use of the CGMS and wore the device at home for 3 days or more. The performance of the CGMS was evaluated against blood glucose measurements obtained using each patient's home blood glucose meter. Evaluation statistics included correlation, linear regression, mean difference and percent absolute difference scores, and Clarke error grid analysis. RESULTS: The CGMS values were compared to 2477 SMBG tests (r = 0.91, slope = 0.93, intercept = 14.5 mg/dL, mean absolute difference = 18.0%+/-19.8%). Clarke error grid analysis showed 96.2% of the data pairs falling within the clinically acceptable regions (zones A and B). CONCLUSIONS: These results demonstrate the agreement of the CGMS to blood glucose meter values, under conditions of home use, in patients selected by their physicians as candidates for continuous monitoring. The detailed glucose information provided by the CGMS should make successful management of diabetes more easily achieved.