Background: Sodium concentration is a frequently used marker to discriminate between differential diagnoses or for clinical follow-up. Pseudonatremia, as a result of indirect ion-selective electrode (ISE) measurements in automated chemistry analyzers, can lead to incorrect diagnosis and treatment. We investigated whether the estimated water content, based on total protein and lipid concentrations, can be used to reduce diagnoses of pseudonatremia.
Methods: Indirect and direct ISE measurements of sodium were compared in blood samples from intensive care unit (ICU) (n = 98) and random non-ICU patients (n = 100). Differences between direct measurements using whole blood and lithium-heparin plasma were also determined. Water content, estimated by a linear combination of total protein and lipid concentrations, was used to correct indirectly measured sodium concentrations. The prevalence of pseudonatremia was evaluated in the ICU patient group.
Results: An absolute difference of 3 mmol/L was observed between direct measurements using lithium-heparin plasma and whole blood, with higher concentrations in plasma. Additionally, we observed that differences between indirect and direct measurements displayed a linear relationship with the estimated water content. The prevalence of pseudohypernatremia after indirect measurements (32%) was reduced when measurements were corrected for water content (19%).
Conclusions: In critically ill patients, sodium concentrations should be preferably measured by direct measurements. Whole blood is the preferred material for these measurements. For routine sodium analyses in other patients, correction using the estimated water content appears promising in reducing the prevalence of pseudohypernatremia by indirect measurements.