Comparison of two methods of measurement of whole blood glucose in the neonatal period.

The purpose of this study was to compare the performance and accuracy of the HemoCue B-Glucose photometer system and reagent strip tests used in conjunction with reflectance photometry against a reference plasma glucose method. One hundred consecutive babies admitted to the neonatal unit over a 6-month period were enrolled in the study. Each baby had a heelprick capillary glucose measured by HemoCue and reagent strip tests. At the same time venous plasma glucose and haematocrit were measured. The mean difference between the reagent strip test and plasma glucose was significantly less than the corresponding value for the HemoCue (0.015 +/- 1.41 vs 0.837 +/- 1.565 mmol l-1, mean +/- SD); however, the agreement limits between both methods and plasma glucose were wide. No significant effect of haematocrit was detected on either method. The HemoCue photometer does not offer any advantage over the widely used reagent strip tests in the neonatal period. However, the limits of agreement of both methods compared with plasma glucose are too wide to be clinically acceptable in the neonatal period.