Abstract
Aims  To assess the accuracy, reliability and measurement of glycaemic control associated with the Medtronic‐MiniMed Continuous Glucose Monitoring System (CGMS) in comparison with blood glucose self‐monitoring (BGSM) in Type 1 diabetic patients.Methods  Type 1 diabetic patients (n = 18) underwent glucose monitoring by the CGMS for up to 3 days, when control was also assessed by BGSM performed eight times daily.Results  Ninety‐five per cent of paired non calibration samples were in the clinically acceptable zones of the Clarke error grid, and 97% in a consensus grid. The median bias was 0.1 mmol/l and relative bias 15%. The failure rate of sensors was 28% of those initially inserted. The CGMS detected significantly more hypoglycaemia and post‐prandial hyperglycaemia, but total duration of hyperglycaemia, blood glucose oscillations and day‐to‐day variability were similarly assessed by CGMS and BGSM.Conclusions  The CGMS has acceptable clinical accuracy. Detection by the CGMS of more hypoglycaemia and post‐prandial hyperglycaemia than BGSM promises a valuable tool for control assessment and optimization of treatment.

This publication has 17 references indexed in Scilit: