Clin Chem Lab Med. 2025 Dec 10. doi: 10.1515/cclm-2025-1484. Online ahead of print.
ABSTRACT
OBJECTIVES: Accurate determination of blood potassium levels in neonates presents significant clinical challenges, primarily attributable to technical difficulties in phlebotomy procedures and the elevated susceptibility of neonatal blood samples to hemolysis. This study aimed to develop and validate a hemoglobin-based mathematical model to correct for hemolysis-induced pseudohyperkalemia in neonatal blood samples.
METHODS: This prospective study analyzed 134 neonatal blood specimens (postnatal age ≤7 days). Controlled hemolysis induction established the potassium-hemoglobin relationship, with specimens allocated to model development (n=101) and validation (n=33) cohorts. The model utilized direct hemoglobin quantification and multivariate regression, with statistical analyses performed using SPSS 26.0.
RESULTS: Analysis of the training cohort established a neonatal-specific potassium release coefficient (ΔK+/ΔHb=0.28 ± 0.03 mmol/L per g Hb) and revealed a strong linear correlation between hemoglobin concentration and potassium elevation (Y=0.2834X, r2=0.8606, p<0.0001). Method validation showed strong agreement between constant correction and model-predicted values (p=0.887). Independent validation confirmed clinical utility, demonstrating comparable corrected and baseline potassium concentrations (p=0.0693) with excellent correlation (r=0.8845, p<0.0001). The model improved hypokalemia detection from 8.9% to 26.7 %, effectively resolving pseudo-normalization artifacts.
CONCLUSIONS: The developed hemoglobin-driven correction model provides a validated and accurate method for hemolysis interference in neonatal potassium measurement, offering significant clinical value for managing irreplaceable specimens in neonatal intensive care settings.
PMID:41364083 | DOI:10.1515/cclm-2025-1484