Serum sodium measurements in postoperative critically ill patients: The difference between direct and indirect methods

Abstract

BackgroundDerangement of serum sodium (Na) concentration is common and associated with adverse outcomes in postoperative patients. Therefore, accuracy of Na measurements should be relevant. Several studies had suggested that Na concentration measured by using central laboratory machine had been lower than the actual Na concentration in patients with hyper-proteinemia. Although hypo-proteinemia is common in postoperative patients, validation of Na measurements in this cohort is still poor. MethodsIn this retrospective study, we included patients who underwent an elective operation for esophageal cancer or for head and neck cancer and who required postoperative intensive care for >72 h. We obtained Na concentrations using indirect (blood central laboratory test using serum) and direct method (arterial blood gas analyzer using whole blood) simultaneously. We assess the difference of these paired Na concentrations, and determined its association with serum protein levels. ResultsThe average Na concentrations measured by the two methods were significantly different (indirect:139.0 mmol/L, direct:137.2 mmol/L, p<0.001). The incidences of hyponatremia (<135 mmol/L) and hypernatremia (>145 mmol/L) indicated by the two methods were significantly different (hyponatremia; indirect:11.4%, direct:22.0%, p<0.001, hypernatremia; indirect:2.8%, direct:1.0%, p = 0.005). The average serum protein levels concentration was 4.5 g/dL, and it was weakly associated with the difference in Na concentrations measured by the two methods. ConclusionThe results suggest that the derangement of Na concentration might be underestimated when only one method is used for measurement of Na concentration in postoperative patients with hypo-proteinemia.