Endocrine Abstracts

Introduction: Assessment of hypoglycaemia in neonates is challenging due to limited blood availability, and lower glucose readings compared to adults. ISO criteria used to assess the accuracy of glucometers are based on adult needs where values below 4.2 mmol/l require intervention. National neonatal guidelines define an action threshold for hypoglycaemia as a glucose below 2mmol/l. We currently use a Radiometer ABL-735 blood gas analyser which uses 35 microlitres of blood. We trialled the Nova Statstrip glucometer (NSG), which uses 1.2microlitres of blood, and internally corrects for haematocrit.

Aims: To compare the accuracy of the NSG against the Radiometer and establish if the NSG could be a reliable alternative for gluco-analysis on the unit.

Method: We reviewed paired glucose readings from the NSG and Radiometer, collected prospectively between December 2008 and February 2009 from babies on the unit. We performed statistical tests to assess the accuracy and precision of the NSG compared to the Radiometer.

Results and Conclusion: We obtained 730 validated paired values. 166 had a value less than 4.2 mmol/l on the Radiometer, 10 below 2 mmol/l and 2 below 1 mmol/l. 98.80% of NSG values less than 4.2 mmol/l and 97.70% of values greater than 4.2 mmol/l met the ISO criteria. Bland-Altman and linear regression analysis showed good correlation between the readings (r-squared =0.8259). An Error Grid showed that most infants would be appropriately managed as per our local guidelines when Radiometer values were below 2 mmol/l. The NSG performed well on statistical analysis compared to the Radiometer, and performed well across a wide range of haematocrits. We recommend its use on our unit.