A novel corrective model based on red blood cells indices and haemolysis index enables accurate unhaemolysed potassium determination in haemolysed samples – Hemokalc project
Year: 2025
Type: article
Abstract: OBJECTIVES: Haemolysis is a major preanalytical issue that affects potassium measurements, often leading to sample rejection and delayed clinical management. This study proposes a novel corrective model for accurate unhaemolysed potassium prediction. METHODS: Blood samples from 14 healthy volunteers were used to prepare a range of haemolysates via freeze-thaw method. First, the relationship between potassium variation and haemolysis variation (ΔK/ΔHI) was studied both individually and globally to assess inter-individual variability. Then, to achieve a more personalised unhaemolysed potassium prediction, a novel corrective model was developed based on: potassium levels in paired unhaemolysed and gradually haemolysed samples, measured haemolysis index, mean corpuscular haemoglobin concentration, mean corpuscular volume and intraerythrocytic potassium level. The bias between true and model-predicted unhaemolysed potassium values was calculated and compared to the reference change value (RCV%). RESULTS: Global data showed a strong correlation between ΔK and ΔHI (Pearson r=0.97, p<0.0001), following a linear relationship: ΔK=0.33*ΔHI (p<0.0001). However, individual data revealed substantial inter-individual variation (min ΔK=0.23*ΔHI and max ΔK=0.39*ΔHI). The correction model achieved 100 % accuracy for the 116 prepared samples, with predicted unhaemolysed potassium values falling within a ± 10 % bias range (mean ± standard deviation of bias = -0.5 ± 2.8 %). CONCLUSIONS: We propose a novel, reliable, and cost-effective corrective model to predict unhaemolysed potassium from haemolysed samples. Compared with previously published models, the integration of red blood cells indices allows for a more personalised, patient-centred approach with high efficiency.
Authors: Charles R. Lefèvre, Bérénice Vigier, Mathilde Favalelli, Jordan Garnier, Alexandre Scanff, Maxime Pawlowski, Nicolas Collet, Claude Bendavid
Institutions: Inserm, Institut National de Recherche pour l'Agriculture, l'Alimentation et l'Environnement, Centre Hospitalier Universitaire de Rennes, Nutrition, métabolismes et cancer, Université de Rennes, École des Hautes Études en Santé Publique, Institut de Recherche en Santé, Environnement et Travail, CIC Rennes
Language: English
Subfield: Physiology
Field: Medicine
Domain: Health Sciences
Open Access status: green
OpenAlex ID: W4413052225