Different glucose analyzers report different glucose concentration values in term newborns

• Authors: Rina P Duke, Lehigh Valley Reilly Children's Hospital; Shasha Bai, Emory University; Joshua A Bornhorst, Mayo Clinic; Nahed O ElHassan, University of Arkansas for Medical Sciences; Jeffrey R Kaiser, Penn State Children's Hospital
• Journal Title: Frontiers in Pediatrics
• Volume: Volume 10
• Publisher: (publisher) | 2022-09-21, Pages 933508-933508
• Type of Work: Article | Final Publisher PDF
• Publisher DOI: http://doi.org/10.3389/fped.2022.933508
• Subjects: Biology, Biostatistics; Health Sciences, Pathology
• Keywords: glucose; glucose analyzer; guideline; neonatal hypoglycemia; newborn

Abstract

Background: The American Academy of Pediatrics and Pediatric Endocrine Society neonatal hypoglycemia guidelines based their glucose concentration treatment thresholds on studies that predominantly used Beckman and Yellow Springs Glucose Oxidase Analyzers. Currently, a majority (76%) of U.S. hospital laboratories utilizing glucose oxidase methodology use Vitros® Glucose Analyzers. However, a bias of ~+5% between glucose concentrations from Beckman vs. Vitros Glucose Analyzers has been reported; this could have a clinically significant effect when using published guideline treatment thresholds. Methods: To determine if there is similar instrument bias between Beckman and Vitros Analyzers in reported glucose concentrations from term newborns, we compared plasma glucose concentrations measured within the first 3 h after birth by Beckman vs. Vitros Analyzers in a total of 1,987 newborns (Beckman n = 904, Vitros n = 1,083). Data were fit using nonlinear cubic spline models between collection time and glucose concentration. Results: The non-linear patterns of initial glucose concentrations (during the first 3 h after birth) as measured by Beckman and Vitros Analyzers paralleled each other with no overlap of the fit spline curve 95% confidence intervals, with an approximate +5 mg/dL constant bias. Additionally, in method comparison studies performed in the Chemistry Laboratory on adult samples, there was a +4.2-7.4 mg/dL measured glucose bias for the Beckman vs. Vitros Analyzer. Conclusion: Glucose concentrations from term, appropriate size for gestational age newborns were about 5 mg/dL higher when measured by Beckman vs. Vitros Analyzers. Perhaps, concentrations of 45 mg/dL reported from Beckman Analyzers may be equivalent to 40 mg/dL from Vitros Analyzers. When managing neonatal hypoglycemia, it is important to know which analyzer was used and whether adjusting for potential instrument bias is necessary when following published guidelines.