Evaluation of HemoCue® blood glucose analyzer for the instant diagnosis of hypoglycaemia in newborns

The aim of the study was to evaluate a portable photometer, HemoCue(R) Blood Glucose Analyzer, in the instant diagnosis of hypoglycaemia in newborns. The HemoCue is a simple, easy-to-handle photometer; with an analysis time of less than 240 s, it utilizes a modified glucose dehydrogenase method in 5 yl whole blood. The HemoCue method was compared to a hexokinase method for deproteinized whole blood in a total of 118 samples from 58 newborns. The linear regression for these samples was Y=1.19x-1.02 (range 0.7-7.2 mmol/L), r=0.90. Ten samples were less than or equal to 2.0 mmol/L with both methods and 37 samples were 12.0 mmol/L with the HemoCue method. The average difference (D) for each sample (n=118) and the standard deviation (SD) for the difference were 0.45+/-0.46 mmol/L. Blood samples with a mean value with both methods less than or equal to 2.0 mmol/L (n=20) had a D and SD of 0.71+/-0.29 mmol/L. When testing for linearity at low glucose concentrations, the HemoCue method gave significantly lower values compared to an ideal line. The HemoCue method has several advantages in the analysis of glucose in newborns: short analysis time, small sample size, and no influence from glycolysis. However, in our investigation, falsely low values occurred, especially in the low measuring range, so the HemoCue method is not suitable in the diagnosis of hypoglycaemia in newborns.