Effects of Total Plasma Protein Concentration on Plasma Sodium, Potassium and Chloride Measurements by an Indirect Ion Selective Electrode Measuring System

OBJECTIVE: Ion selective electrodes (ISE) measure electrolytes by two different technologies, direct and indirect. We wished to investigate the effect of total plasma protein concentration on the plasma sodium, potassium and chloride measurements by indirect ISE compared with measurements by direct ISE. METHODS: The evaluation of the objective was performed in a tertiary care hospital on patient blood samples sent to the pathology laboratory for general chemistry analysis. To determine the degree of difference between the two ISE measurements, 195 lithium heparin plasma samples were measured and the results were separated into three groups (65 samples in each group) depending on the protein concentration (e.g. total plasma protein concentration less than, greater than and within the reference range between 62-83 g/L). The samples were analysed over a 40 day period on a Hitachi Modular ISE system - indirect ISE (Roche Australia) and Bayer Rapidlab 865 Blood Gas Analyser - direct ISE (Bayer Diagnostics, Australia). RESULTS: Using indirect ISE, low plasma protein concentrations caused a 'pseudohyper' effect in all 3 analytes and a 'pseudohypo' effect with high plasma protein concentrations. The variation in total protein concentration had the greatest effect on plasma sodium measurement. The relationship was non-linear and no accurate predictive value could be calculated for the plasma electrolytes with changes in plasma protein concentrations. CONCLUSIONS: The plasma sodium, potassium and chloride measurements are effected by changes in plasma protein concentration when measured by indirect ISE systems. Clinicians must be aware that differences exist between the ISE technologies and in border-line clinical situations, the direct ISE systems provide a more accurate estimate of plasma sodium, potassium, chloride, anion gap, osmolality and strong ion difference calculations, and should be used for clinical decision-making.