ABSTRACT
BACKGROUND: We have evaluated the analytical performance of Barozen H (i-SENS Inc. Seoul, Korea), which was developed for glucose testing and can be connected to hospital information network systems. METHODS: We evaluated the precision of ten glucometers based on repeatability and intermediate precision according to the guidelines of the International Standardization Organization (ISO) 15197:2011. The linearity of patient samples was in the range of 30.7-551.2 mg/dL. The accuracy of the results of Barozen H and the correlation of these results and those of the two other glucometers in comparisons with the Modular D reference device (Roche Diagnostics Ltd., Switzerland) were evaluated using 150 capillary blood and venous whole blood samples. RESULTS: The ranges for the repeatability and intermediate precision of ten Barozen H glucometers were 1.58-4.61% and 2.85-5.48%, respectively. The linearity was expressed by y=0.9681x+2.0791, and the coefficient of determination (R2) was 0.9996. When venous whole blood samples were used, the correlation coefficient (r) was 0.9914. When glucose levels were under 100 mg/dL, 95.2% of Barozen H results were within +/-15 mg/dL, and when glucose levels were 100 mg/dL or higher, 97.2% were within +/-15%. When capillary blood samples were used, 41.4% (under 100 mg/dL) and 55.4% (100 mg/dL or higher) of Barozen H results were within +/-15 mg/dL and +/-15%, respectively. CONCLUSIONS: Barozen H provided reliable results and satisfied the ISO15197:2011 criteria when venous whole blood samples were used. It is thought to be clinically useful as a hospital point-of-care glucometer.