Dysnatremia, its correction, and mortality in patients undergoing continuous renal replacement therapy: a prospective observational study

BACKGROUND: Although dysnatremia has been reported to be correlated with mortality risk, this issue remains unresolved in patients undergoing continuous renal replacement therapy (CRRT). Furthermore, it has not been determined whether change in or correction of sodium is related to mortality risk in this subset. METHODS: A total of 569 patients were prospectively enrolled at the start of CRRT between May 2010 and September 2013. The patients were divided into 5 groups: normonatremia (135–145 mmol/L), mild hyponatremia (131.1–134.9 mmol/L), moderate to severe hyponatremia (115.4–131.0 mmol/L), mild hypernatremia (145.1–148.4 mmol/L), and moderate to severe hypernatremia (148.5–166.0 mmol/L). The non-linear relationship between sodium and mortality was initially explored. Subsequently, the odds ratios (ORs) for 30-day mortality were calculated after adjustment of multiple covariates. RESULTS: The relationship between baseline sodium and mortality was U-shaped. The mild hyponatremia, moderate to severe hyponatremia, and moderate to severe hypernatremia groups had greater ORs for mortality (1.65, 1.91, and 2.32, respectively) than the normonatremia group (all P values < 0.05). However, later sodium levels (24 and 72 h after CRRT) did not predict 30-day mortality. Furthermore, the changes in sodium over 24 or 72 h, including the appropriate correction of dysnatremia, did not show any relationships with mortality, irrespective of baseline sodium level. CONCLUSIONS: Sodium level at the start of CRRT was a strong predictor of mortality. However, changes in sodium level and the degree of sodium correction were not associated with the mortality risk in the patients with CRRT. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1186/s12882-015-0215-1) contains supplementary material, which is available to authorized users.