The Impact of a Standardized Sodium Protocol on Incidence and Outcome of Dysnatremias in Neurocritical Care.

BACKGROUND: Dysnatremias are common and prognostically serious in neurocritical care. We studied whether a standardized sodium protocol would improve our neurocritical care of dysnatremias.

METHODS: A 5-year prospective study of a standardized sodium protocol for 1,560 patients admitted with various brain diseases in an adult neurologic-neurosurgical intensive care unit (NNICU) was compared with a 5-year retrospective analysis of 1,440 patients without the sodium protocol. Hyponatremia was defined as serum sodium (SNa(+)) < 135 mmol/L and hypernatremia SNa(+ )> 150 mmol/L. The sodium protocol involved measuring SNa(+), serum, and urine osmolality, measured and calculated renal function parameters, fluid intake 40 mL/kg weight/day without hypotonic saline, thiazide, and desmopressin acetate in all normonatremic NNICU patients.

RESULTS: In the protocol study, hyponatremia occurred slightly less often (15.7 versus 16.3% of patients; p = 0.684), hypernatremia was significantly higher (respectively 8.5% versus 5.2% of patients; p < 0.001), and no differences were noted in hypo/hypernatremia (p = 0.483). There were no differences in the incidence of hypo-osmolal hyponatremia (respectively 3.5% versus 3.5% of patients; p = 0.987), cerebral salt wasting (CSW; respectively 1.7% versus 1.7% of patients; p = 0.883), syndrome of inappropriate secretion of antidiuretic hormone (SIADH; respectively 0.1% versus 0.3% of patients; p = 0.152), central diabetes insipidus (CDI; respectively 1.0% versus 0.6% of patients; p = 0.149). In hyponatremia there were no differences in the Glasgow Coma Scale (GCS) score upon onset of hyponatremia (p = 0.294), NNICU mortality (respectively 1.0% versus 0.4% patients; p = 0.074), and bad outcome upon discharge from NNICU (respectively 5.1% versus 6.5% of patients; p = 0.101), but in hypernatremia GCS score upon onset (p < 0.001), mortality (respectively 2.8% versus 1.0%; p < 0.001), and bad outcome from NNICU (respectively 6.7% versus 2.7% patients; p < 0.001) were significantly higher. Multivariate logistic regression analysis showed that hypernatremia, compared with hyponatremia, was a significant predictor of mortality during NNICU stay (respectively odds ratio [OR]: 1.14; p = 0.003 versus OR; 5.3; p = 0.002).

CONCLUSIONS: The standard sodium protocol lowered the frequency of SIADH, which was encountered in only one patient over 5 years. However, it did not significantly reduce the incidence and improve the outcome of hyponatremia. Hypernatremia occurred more often and had a higher mortality and worse outcome than hyponatremia, but these patients were neurologically worse upon its onset. The prospective study confirmed that CSW, SIADH, and CDI were not common in our neurocritical care.