Central diabetes insipidus is not a common and prognostically worse type of hypernatremia in neurointensive care.

BACKGROUND: Hypernatremia is a common sodium dysbalance in neurointensive care which is associated with worse outcome. It can be caused by central diabetes insipidus (cDI) or by other mechanisms, more often from osmotherapy and furosemide. The aim of this study was to determine the incidence of cDI and to analyse outcome as compared with other causes of hypernatremias found in neurointesive care.

METHODS: We analysed 75 hypernatremic (serum sodium, SNa+ >150 mmol/l) patients (pts) with brain diseases admitted over a period of five years to Neurologic-Neurosurgical Intensive Care Unit (NNICU). Firstly we diagnosed cDI according to measured serum and urine osmolality, eletrolyte free water clearance (EWC) and response to desmopressin acetate. The remaining hypernatremias were categorised as "non cDI". We observed Glasgow Coma Scale (GCS) on onset of hypernatremia, incidence of cerebral complications, Glasgow Outcome Scale (GOS) upon discharge from NNICU and mortality in NNICU.

RESULTS: We found cDI in 8 pts (mean SNa+ 154.8 ± 5.4 mmol/l). Most pts (67) were classified as "non cDI" hypernatremias (mean SNa+ 154.3 ± 3.6 mmol/l). There were no differences in serum sodium (p=0.682), serum osmolality (p=0.476) between the two groups, however patients with cDI indicated low urine osmolality (p=0.001) and positive EWC (p=0.049). We did not find any differences in GCS score on onset of hypernatremia (p=0.395), incidence of cerebral complications (p=0.705), GOS score upon discharge from NNICU (p=0.61) and mortality in NNICU (p=0.638). More patients in the "non cDI" group received antiedematic therapy (p=0.028) and diuretic furosemide (p=0.026). Multivariate logistic regression analysis showed that independent predictors of NNICU mortality was the highest level of serum sodium (Odds ratio, OR 1.13, per 1 mmol/l increase in maximal hypernatremia during NNICU stay, 95% confidence interval, CI 1.01-1.26, p=0.027), and GCS on admission of less than 9 (OR 2.61, 95% CI 1.41-5.44, p=0.003).

CONCLUSIONS: Central diabetes insipidus is not a frequent type of hypernatremia in neurointensive care. Prognosis is connected with serum sodium level, not with type of hypernatremia.