Reproducibility and reliability of hypoglycaemic episodes recorded with Continuous Glucose Monitoring System (CGMS) in daily life

AIM: Continuous glucose monitoring may reveal episodes of unrecognized hypoglycaemia. We evaluated reproducibility and reliability of hypoglycaemic episodes recorded in daily life by the Medtronic MiniMed Continuous Glucose Monitoring System (CGMS).

METHODS: Twenty-nine adult patients with Type 1 diabetes underwent 6 days of continuous subcutaneous glucose monitoring, applying one CGMS on each side of the abdomen. Blood glucose was measured by HemoCue B-Glucose Analyzers six times daily and two different 4-point calibration sets were generated (set A and B). Using these calibration sets, CGMS raw data were recalibrated generating four different CGMS data sets [left-A (left side of abdomen, calibration set A), left-B, right-A and right-B]. Agreement between CGMS data sets was evaluated during hypoglycaemic events, comparing CGMS readings = 2.2 mmol/l with nadir values from corresponding CGMS data sets. CGMS readings were also compared with independent self-monitored blood glucose (SMBG) values.

RESULTS: With hypoglycaemia (CGMS readings = 2.2 mmol/l) in calibration set left-A, values below 3.5 mmol/l were present in 99% (95% CI: 95-100%) of samples in left-B, 91% (95% CI: 84-96%) of samples in right-A, and 90% (95% CI: 83-95%) of samples in right B. In 84% of these episodes (95% CI: 59-96%) independent SMBG values were below 3.5 mmol/l. Difference in duration was observed with a median difference of 20 min; (left-A vs. right-B).

CONCLUSION: Hypoglycaemic episodes recorded by CGMS are reproducible and agreement with independent SMBG values is acceptable for retrospective recording of hypoglycaemic events with CGMS.