Background and Aims Postnatal hypoglycaemia is known to increase the risk of adverse neurological outcome in infants at risk.
Adequate glucose control can only be achieved with repetitive blood sampling which means mainly repetitive pain- and stressful procedures many times daily.
Devices measuring glucose levels subcutaneously (CGMS) are widely used in adults and children suffering from diabetes. The feasibility and accuracy of such devices were never evaluated in newborn infants.
Methods In this pilot study CGMS sensors were placed in infants (GA >35+0, birthweight >2500g) at risk of hypoglycaemia in addition to routine glucose sampling (RG). The CGMS generates a glucose value every 5 minutes and needs calibration twice a day. Because of unknown glucose delay in infants we correlated RG with CGMS values at 126.96.36.199.25 and 30 minutes following RG.
Results We measured 264 RG in 20 infants and used 97 (37%) for CGMS-calibration. The strongest correlation were at 20 min with r=0.709 and corresponding MARD (mean absolute relative deviation) of 13.28%. Plotted in Clark Error Grid showed that 94.5% of the pairs were in zone A and B.
Conclusion Based on our findings it seems that CGMS is a reliable and feasible method for testing and following blood glucose levels in infants. The use of CGMS could reduce the amount of painful RG-measurements to 37% and at the same time it increases the patient’s safety through a much better glucose control.
The subcutaneous glucose delay in infants is approximately 20 minutes, which was not published before.