ABSTRACT
AIM: To compare a method of testing at alternate skin sites (AST) with that of the usual heel-stick approach (SM) for determining blood glucose levels in newborn infants. Our other aims were to compare these methods as regards their accuracy, the pain caused by the procedures, the times taken to obtain a result and the possible delay in accurate test results using AST during rapid changes in blood glucose. METHODS: One hundred and eighty-six preterm and term infants were enrolled. The blood glucose levels were determined by a standard bedside method (SM, HemoCue) and AST (Freestyle), which permitted blood samples to be taken from the arm or leg. RESULTS: The levels of blood glucose ranged between 0.6 and 8.6 mmol/l. We found a significant correlation between SM and AST (r = 0.90, p < 0.001). The coefficient of variation was similar, pain was significantly less (median pain score 3.5 vs 7.5, p < 0.01) and the time taken to obtain a result significantly shorter (mean 35 vs 111 s, p < 0.01) with AST than with SM. No significant differences were found between these methods during rapid changes in the blood glucose levels. CONCLUSION: AST, a relatively simple and painless method of determining blood glucose levels in newborn infants, is acceptably accurate and causes minimal blood loss.
