Overestimation of hypoglycemia in infants with a high hematocrit

Background: In neonates, hypoglycemia is an emergency condition requiring urgent treatment. Therefore, rapid and reliable blood glucose measurements are necessary. However, this step has been proven difficult because of both analytical and preanalytical variables. In our children's hospital, we incidentally observed cases of hypoglycemia that were not in line with the clinical picture of the infants. Remarkably, most of these infants had a high hematocrit.

Methods: Glucose concentrations were determined in blood samples from healthy participants that were collected in Li-heparin capillary and pediatric tubes. The effect of hematocrit on glucose consumption over time was studied by artificially increasing sample hematocrits. To study the effect of sample cooling, glucose concentrations were followed over time in samples stored at room temperature and on ice.

Results: In all samples, glucose concentrations declined with time. This effect was most dramatic [up to 18 mg/dL (1 mmol/L) in the first 30 min] in samples with high hematocrits and collected in capillary tubes. Cooling of samples clearly reduced glucose consumption; however, this was not evident in the first 30 min.

Conclusions: Overestimation of hypoglycemia in infants must be considered if samples are not centrifuged or are not analyzed immediately after sampling. The extent of overestimation is more pronounced in samples with a high hematocrit, collected in capillary tubes. Cooling of samples does not prevent glucose consumption in vitro during the first 30 min. These results emphasize that, for glucose analysis, prompt handling of samples of newborns with a high hematocrit is necessary.