RESUMO
BACKGROUND: Rapid correction of dysnatremias can result in neurological complications. Therefore, various formulas are available to predict changes in plasma sodium concentration ([Na+]) after treatment, but these have been shown to be inaccurate. This could be explained by sodium acumulation in skin and muscle tissue, which is not explicitly considered in these formulas. We assessed the association between clinical and biochemical factors related to tissue sodium accumulation and the discrepancy between predicted and measured plasma [Na+]. METHODS: We used data from an intensive care unit (ICU) cohort with complete data on sodium, potassium, and water balance. The predicted plasma [Na+] was calculated using the Barsoum-Levine (BL) and the Nguyen-Kurtz (NK) formula. We calculated the discrepancy between predicted and measured plasma sodium and fitted a linear mixed-effect model to investigate its association with factors related to tissue sodium accumulation. RESULTS: We included 594 ICU days of sixty-three patients in our analysis. The mean plasma [Na+] at baseline was 147±6 mmol/L. The median (IQR) discrepancy between predicted and measured plasma [Na+] was 3.14 mmol/L (1.48, 5.55) and 3.53 mmol/L (1.81, 6.44) for the BL and NK formulas, respectively. For both formulas, estimated total body water (p=0.027), initial plasma [Na+] (p<0.001) and plasma [Na+] change (p<0.001) were associated with the discrepancy between predicted and measured plasma [Na+]. CONCLUSION: In this ICU cohort, initial plasma [Na+], total body water, and plasma [Na+] changes, all factors that are related to tissue sodium accumulation, were associated with the inaccurateness of plasma [Na+] prediction.