Evaluation of neonatal whole blood versus plasma glucose concentration by ion-selective electrode technology and comparison with two whole blood chromogen test strip methods.

OBJECTIVE: Determination of glucose concentration from whole blood samples in neonates is confounded by variable hematocrit values, sample source, user technique, and test method, which results in poor correlation with glucose values measured from plasma or serum. Recently developed ion-selective electrodes (ISE) allow measurement of glucose in the water phase of both red blood cells and plasma with a small sample size (200 microliters). STUDY
DESIGN: The purpose of this study was to compare (1) whole blood and plasma glucose measurements by the ISE method and (2) ISE glucose values with those determined by two chromogen reagent test strip colorimetric methods. Values were determined in 180 different samples obtained from 145 infants.
RESULTS: Correlation of whole blood and plasma glucose concentrations determined by the ISE method was excellent (y = 0.99x, R2 = 0.99) and no effect was seen from hematocrit values. The two chromogen test strip methods also revealed good linearity with ISE whole blood and plasma glucose values but had large confidence intervals for individual values. Sensitivity for detecting blood glucose levels < 40 mg/dl by the two test strip methods was 9 of 11 and 9 of 10 with four and one false-positive results, respectively.
CONCLUSION: These data indicate that the ISE method provides an excellent correlation of whole blood and plasma glucose measurement results, overcoming technical problems of differing hematocrit values and serum or plasma sample acquisition in neonates. Chromogen test strip methods have limited value in estimating specific glucose values, but can be useful in screening infants for hypoglycemia.