BACKGROUND
Finger pricking is currently the common method of blood glucose measurement in patients with diabetes mellitus. However, diabetes patients have proven reluctant to regularly check their glucose profiles because of the small amount of blood that is required. Recently, a non-invasive and continuous glucose monitoring device that is based on reverse iontophoresis(GlucallTM) has been developed. In this study we wanted to evaluate the accuracy and the clinical acceptability of this new device. METHODS: The study was conducted during the period from November 2003 to January 2004 on 19 in-patients who had been admitted to Kyung Hee University Hospital. Glucose measurements using GlucallTM were performed between 10am and 4pm. The concurrent plasma glucose levels were checked hourly and they were subsequently compared with the GlucallTM data. RESULTS: The mean error(ME) of the GlucallTM measurements was -3.45+/-52.99mg/dL with a mean absolute relative error(MARE) of 20+/-15.16%. Measurements obtained by GlucallTM had a correlation coefficient of 0.784(P<0.05) with the plasma glucose levels, as was determined by linear regression analysis. This correlation was consistent regardless of the time of data collection. However, after excluding such confounding variables as age and gender, the correlation coefficient exhibited a tendency to increase. 98.9% of the results were clinically acceptable according to Clarke error grid analysis. CONCLUSION: GlucallTM does not yet have the reliability and accuracy to wholly replace the conventional methods. However, further technical advancements to reduce its shortcomings will make this device useful for the management of diabetes patients