Intensive care unit (ICU) management of aneurysmal subarachnoid hemorrhage (SAH) can predispose patients to medical complications, which are closely associated with the outcome (1). Specifically, both hypernatremia and hyponatremia commonly occur during ICU management of patients with SAH (1–17). Hypernatremia is associated with unfavorable neurologic outcomes (2, 3, 16) and mortality (14), whereas hyponatremia is associated with poor neurologic outcomes (12), vasospasm and delayed ischemic neurologic deficits (6, 7, 9, 10, 13), and a prolonged ICU stay after SAH (10, 12).
However, available literature did not specify the target range of serum sodium levels in SAH patients during ICU management. Furthermore, the definitions of hyponatremia and hypernatremia varied in all studies from less than 125 mmol/L (13) to 135 mmol/L (2, 6–13, 15, 18) for hyponatremia, and more than 143 mmol/L (3) to 155 mmol/L (14) for hypernatremia. Likewise, current guidelines for the management of SAH do not mention the specific target values of serum sodium levels (19, 20).
The purposes of this study were to examine the association of abnormal serum sodium levels with unfavorable neurologic outcomes and to identify the target range of serum sodium in patients with SAH.
Study design and setting
The Kagawa University Hospital is a 613-bed teaching institution with an eight-bed neurointensivist-managed ICU in Japan. This single-center, retrospective, case-control study was performed by a review of medical records and was approved by the institutional review board of University Hospital. It was conducted in accordance with the ethical standards established in the 1964 Declaration of Helsinki and its later amendments. The IRB waived the requirement for patient consent due to the retrospective nature of the study.
Study participants and inclusion criteria
We included all patients ≥18 years old who were consecutively hospitalized for a confirmed diagnosis of SAH and had at least five arterial sodium measurements between January 1, 2009 and December 31, 2015. Patients who were given only comfort care within 24 h of admission were excluded.
General management of SAH in the ICU
The management carried out was in line with Guidelines for the Management of Aneurysmal Subarachnoid Hemorrhage from the American Heart Association/American Stroke Association (20). In addition to general intensive care, all patients were monitored for clinical deterioration or cerebral infarction arising due to delayed cerebral ischemia (8). Fluid management was targeted at maintaining euvolemia. Prophylactic hemodynamic support, hemodilution, and administration of mannitol or hypertonic saline solution were not performed.
Control of serum sodium levels in SAH patients in the ICU
Analysis of arterial blood gas and sodium levels was routinely performed in the ICU on admission, after intubation and neurologic intervention by coiling or clipping, and every 6 h. Additional measurements were carried out as needed, by critical care physicians. Treatment of hyponatremia generally comprised the administration of oral sodium chloride (NaCl) tablets, with or without fludrocortisone. Maintenance fluids were mainly provided as 0.9% NaCl solution. For the management of hypernatremia, the intravascular volume status was initially assessed. Desmopressin was considered when the serum sodium level was >150 mmol/L due to central diabetes insipidus (DI).
The following data were collected: age, sex, Hunt and Kosnik (H&K) grade, treatment modality (coil or clip), the number of serum sodium measurements, maximum and minimum serum sodium levels during ICU stay, modified Rankin scale (mRS) upon hospital discharge, length of ICU stay, length of hospital stay, and hospital mortality.
The primary outcome was the association of maximum and minimum serum sodium levels measured in the ICU with the incidence of an unfavorable neurologic outcome, which was assessed by the mRS upon hospital discharge (21). The mRS is a measure of global disability. Its seven outcome categories are as follows: no symptoms at all; no significant disability; slightly disability; moderate disability; moderately severe disability; severe disability; and death. We evaluated all patients with SAH using mRS at hospital discharge in real time and listed their medical records. The neurologic outcome was defined as unfavorable when the mRS score was 3 to 6 and as favorable when the mRS score was 0 to 2. The secondary outcomes included identification of the cut-off levels of maximum and minimum serum sodium measured at the ICU and the association with an unfavorable neurologic outcome.
Demographic factors and baseline characteristics were summarized using descriptive statistics. The groups were compared using the Student t test or the Mann–Whitney U test, as deemed appropriate. Categorical comparisons were drawn using the Fisher exact test. Univariate and stepwise multiple regression analyses were conducted to explore the independent factors that predicted unfavorable neurologic outcomes in the study population.
Receiver operating characteristic (ROC) curve and the area under the curve (AUC) were estimated based on the Mann–Whitney U test. For each continuous variable, the cut-off value that gave the best combination of sensitivity and specificity was identified.
The comparison of baseline characteristics between dysnatremia (−) and dysnatremia (+) in mild damage (H&K grades I–II) and severe damage (H&K grades III–V) groups was made.
Two prediction models for unfavorable neurologic outcomes were also provided and their superiority was compared to confirm the impact of dysnatremia regardless of the severity of the actual brain injury. Model 1 was comprised of age (>65 years), gender (male), Hunt & Kosnik grade (I–II or III–V), and treatment modality (coil or clip). Model 2 was comprised of factors in Model 1 plus dysnatremia (presence). Each score was determined based on β-coefficients from logistic regression analysis. We calculated the AUC of two models and compered them.
Alterations of serum sodium levels in patients with episodes of either hypernatremia or hyponatremia alone and with both were examined. Statistical analyses were performed using JMP statistical software (version 11; SAS Institute, Cary, NC). A two-sided P value of < 0.05 was considered statistically significant for all analyses.
Baseline characteristics of the study population
There were 131 patients (mean age, 62.2 years; 37 were men) with 4,811 serum sodium measurements included in this study. Unfavorable neurologic outcomes were observed in 45% of these patients. Comparison of the clinical characteristics between the unfavorable and favorable groups is shown in Table 1. On univariate analysis, the groups significantly differed in age (P < 0.01), H&K grade (P < 0.01), number of serum sodium measurements (P < 0.01), maximum serum sodium level (P < 0.01), and minimum serum sodium level (P = 0.03), during their ICU stay.
Predictors of unfavorable neurologic outcomes in patients with SAH
On multiple regression analysis (Table 2), unfavorable neurologic outcomes were significantly associated with age, H&K grade, maximum serum sodium levels at the ICU (OR, 1.18; 95% CI, 1.05–1.35; P < 0.01) and minimum serum sodium levels at the ICU (OR, 0.88; 95% CI, 0.77–0.99; P = 0.048).
ROC curve analysis
The ROC curves of unfavorable neurologic outcomes according to the maximum and minimum serum sodium levels during the ICU stay were constructed. The respective AUCs, sensitivities, and specificities for prediction of unfavorable neurologic outcomes are shown in Table 3. The cut-off values of serum sodium levels were obtained at a maximum of 145 mmol/L and a minimum of 132 mmol/L. Based on these values, a hyponatremia episode was defined as the condition when the serum sodium level ≤132 mmol/L at any point during the first 2 weeks in the ICU, whereas a hypernatremia episode was defined as the condition when the serum sodium level ≥145 mmol/L at any point during the first 2 weeks in the ICU.
Association between episodes of hyponatremia and/or hypernatremia and the incidence of an unfavorable neurologic outcome
Patients with episodes of hyponatremia and hypernatremia accounted for 88.2% of unfavorable neurologic outcomes, whereas those with normal sodium levels accounted for only 15.6% of unfavorable neurologic outcomes (Fig. 1).
The comparison of baseline characteristics between dysnatremia (−) and dysnatremia (+) in mild damage (H&K grades I–II) and severe damage (H&K grades III–V) groups
According to the cut-off values of the serum sodium levels obtained by the ROC curve analysis, patients with a maximum of 145 mmol/L and/or a minimum of 132 mmol/L were defined as the dysnatremia group.
In the mild damage group, the incidence of unfavorable neurologic outcomes was relatively higher in patients with dysnatremia than those without dysnatremia (29.6% vs. 9.4%, P = 0.09) (Supplemental Table 1a, http://links.lww.com/SHK/A588). In the severe damage group, the incidence of unfavorable outcomes was significantly higher in patients with dysnatremia than in those without dysnatremia (74.6% vs. 30.8%, P < 0.01) (Supplemental Table 1b, http://links.lww.com/SHK/A588). Therefore, in this small number of study patients, patients with dysnatremia developed unfavorable neurologic outcomes in both mild and severe damage groups.
The comparison of two prediction models for unfavorable neurologic outcomes
Comparing two models revealed that model 2 with dysnatremia developed higher AUC compared with model 1 (0.86 vs.0.82) (Supplemental Table 2a, b, and c, http://links.lww.com/SHK/A589).
Alteration of the serum sodium levels in patients with episodes of hyponatremia and/or hypernatremia during the first 2 weeks in the ICU
Patients with both hypernatremia and hyponatremia episodes initially had increased serum sodium levels, which peaked during the acute phase (days 1–2). Subsequently, serum sodium levels decreased and reached a nadir in the late phase (days 7–12) (Fig. 2).
Patients with episodes of hypernatremia alone initially had increased serum sodium levels, with the peak during the acute phase (days 2–4); whereas patients with episodes of hyponatremia alone manifested with gradually decreasing serum sodium levels before reaching a nadir in the late phase (days 6–12) (Fig. 3).
Alteration of serum sodium levels in H&K grades I–II and III–V
In patients with H&K grades III–V, serum sodium levels increased in the acute phase (days 1–2) and gradually decreased before reaching a nadir on days 6–12. In patients with H&K grades I–II, serum sodium levels gradually decreased before reaching a nadir on days 6 to 12 (Fig. 4).
In the current study, we demonstrated that both maximum and minimum serum sodium levels during ICU management of SAH patients were significantly associated with unfavorable neurologic outcomes. According to the two additional analyses (Supplemental Table 1, http://links.lww.com/SHK/A588 and 2, http://links.lww.com/SHK/A589), dysnatremia is confirmed to be an independent risk factor for unfavorable neurologic outcomes regardless of severity of the actual brain injury.
A particular strength of this study was that the serum sodium levels were measured every 6 h in the ICU; therefore, more precise alterations of serum sodium levels were obtained. On the other hand, previous reports have measured serum sodium levels either daily (7, 9–11, 17); every 3 days after SAH (2); at three time points (1, 2, and 5 days after SAH) (3); or every 8 h after SAH for 5 days (16). Moreover, in this study, we were able to determine the threshold for both hyponatremia and hypernatremia; therefore, the target sodium levels during ICU management of SAH patients could be determined.
Cut-off values for hypernatremia and hyponatremia
Several studies showed that hypernatremia was independently associated with poor neurologic outcomes (2, 16, 17, 22). In these reports, hypernatremia was defined as serum sodium levels >145 mmol/L, which was similar to our identified cut-off level. On the other hand, Wartenberg et al. (1) reported that hypernatremia, which was defined as a serum sodium level >150 mmol/L, was not an independent predictor of poor neurologic outcome. This different result might be due to a different study population, in which more than 20% of patients did not receive any intervention for the SAH aneurysm (1).
Hyponatremia, which is usually defined as a serum sodium level <135 mmol/L or <130 mmol/L, is more common than hypernatremia in patients with SAH (2, 23). Although hyponatremia was reported to have certain effects on the various outcomes of SAH, such as length of the hospital stay and development of vasospasm (4), several studies demonstrated that hyponatremia was not associated with unfavorable neurologic outcomes (1, 2, 17, 24). In 3 of 4 studies (2, 17, 24), hyponatremia was defined as a condition with a serum sodium level <135 mmol/L, which was higher than the cut-off level (132 mmol/L) that we obtained in this study. Therefore, we believe that hyponatremia was not determined as a prognostic factor in SAH (2, 7, 23). Wartenberg et al. (1) defined hyponatremia as a condition with <130 mmol/L serum sodium levels and also reported that it was not significantly associated with unfavorable neurologic outcomes; however, the patient cohort included in the study appeared to be different, as mentioned above.
Timing of hypernatremia/hyponatremia episodes after SAH
In our current study, episodes of hypernatremia alone and both hyponatremia and hypernatremia were observed in the acute phase (within the first 4 days), whereas episodes of hyponatremia alone and both hyponatremia and hypernatremia were observed in the late phase (days 6–12). Consistent with our current data, the results of Qureshi et al. (2) demonstrated that a higher proportion of patients developed hypernatremia within 3 days, when compared with patients who developed hypernatremia on day 6 or day 9 . On the other hand, Pierrakos et al. (5) reported that the median time from the onset of SAH to the development of hyponatremia (<135mEq/L) due to the syndrome of inappropriate secretion of antidiuretic hormone (SIADH) was 4 days. Kao et al. (8) reported that over 50% of patients with hyponatremia developed a nadir of serum sodium level at 7 days after SAH. Most reports, including the current study, showed that hyponatremia occurred in the late phase; however, the nadir of hyponatremia may vary due to several causes. Indeed, Sherlock et al. (18) reported that over 40% of hyponatremia episodes developed in the acute phase (within 3 days) of SAH.
Pathophysiology of hypernatremia/hyponatremia after SAH
Several factors, such as hypovolemia (11), inappropriate fluid therapy (11), SIADH (11), glucocorticoid deficiency (11), and cerebral salt wasting syndrome (CSWS) (8), can induce hyponatremia in SAH patients. On the other hand, central DI arising due to a dysfunction of the hypothalamic nuclei (16, 25) is considered a main cause of hypernatremia in SAH. Therefore, dysregulation of serum sodium levels in SAH cannot be concluded by one simple explanation. Based on the current data from patients with H&K grades III to V and who developed both hypernatremia and hyponatremia episodes, we speculate that in severe cases of SAH, direct damage of the hypothalamic nuclei might have initially induced hypernatremia (14), which directly caused further brain damage by increasing extracellular fluid osmolality (2). Subsequently, a severely damaged brain and stress might have induced CSWS in the majority of hyponatremia with severe cases (8). On the other hand, in our patients with H&K grades I to II and mild SAH, minimal damage to the supraoptic and paraventricular nuclei might have caused SIADH-induced hyponatremia (11, 14).
Taking into consideration the several patterns of serum sodium level alterations, we propose that severe cases of SAH should be carefully monitored for the development of hypernatremia within 4 days after admission and of hyponatremia in the late phase (days 6–12). On the other hand, mild cases of SAH should be monitored for the development of hyponatremia in the late phase (days 6–12). Because the causes of hyponatremia are multifactorial, proper diagnosis prior to intervention is mandatory for appropriate treatment.
This study had several limitations. First, there was the potential selection bias because it was a retrospective cohort study conducted at a single center. Moreover, uncontrolled confounding factors may have existed. Second, the neurologic outcomes of patients after discharge were not followed up. Third, we could definitively not exclude the possibility that differences in patient population, such as comorbidities, could have a varying impact on neurologic outcomes. Fourth, the impact of interventions to manage serum sodium levels (i.e., the outcome of patients whose dysnatremia status was corrected with interventions) could not be evaluated due to the retrospective nature of the study; therefore, a further study with interventions will be required. Fifth, the mechanism of dysnatremia in patients with SAH could not be examined. Finally, the sample size in this study was relatively small.
Both the maximum and minimum sodium levels during the ICU management of SAH patients were significantly associated with unfavorable neurologic outcomes. The cut-off values for the maximum and minimum serum sodium levels were determined as 145 mmol/L and 132 mmol/L, respectively.
The authors are grateful to all physicians and nurses at the study site for their crucial contribution to the successful completion of this study.
1. Wartenberg KE, Schmidt JM, Claassen J, Temes RE, Frontera JA, Ostapkovich N, Parra A, Connolly ES, Mayer SA. Impact of medical complications on outcome after subarachnoid hemorrhage
. Crit Care Med
2. Qureshi AI, Suri MF, Sung GY, Straw RN, Yahia AM, Saad M, Guterman LR, Hopkins LN. Prognostic significance of hypernatremia and hyponatremia among patients with aneurysmal subarachnoid hemorrhage
3. Fisher LA, Ko N, Miss J, Tung PP, Kopelnik A, Banki NM, Gardner D, Smith WS, Lawton MT, Zaroff JG. Hypernatremia predicts adverse cardiovascular and neurological outcomes after SAH. Neurocrit Care
4. Mapa B, Taylor BE, Appelboom G, Bruce EM, Claassen J, Connolly ES Jr. Impact of hyponatremia on morbidity, mortality, and complications after aneurysmal subarachnoid hemorrhage
: a systematic review. World Neurosurg
5. Pierrakos C, Taccone FS, Decaux G, Vincent JL, Brimioulle S. Urea for treatment of acute SIADH in patients with subarachnoid hemorrhage
: a single-center experience. Ann Intensive Care
6. Chandy D, Sy R, Aronow WS, Lee WN, Maguire G, Murali R. Hyponatremia and cerebrovascular spasm in aneurysmal subarachnoid hemorrhage
. Neurol India
7. Kubo Y, Ogasawara K, Kakino S, Kashimura H, Yoshida K, Ogawa A. Cerebrospinal fluid adrenomedullin concentration correlates with hyponatremia and delayed ischemic neurological deficits after subarachnoid hemorrhage
. Cerebrovasc Dis
8. Kao L, Al-Lawati Z, Vavao J, Steinberg GK, Katznelson L. Prevalence and clinical demographics of cerebral salt wasting in patients with aneurysmal subarachnoid hemorrhage
9. Nakagawa I, Kurokawa S, Nakase H. Hyponatremia is predictable in patients with aneurysmal subarachnoid hemorrhage
—clinical significance of serum atrial natriuretic peptide. Acta Neurochir (Wien)
10. Vrsajkov V, Javanovic G, Stanisavljevic S, Uvelin A, Vrsajkov JP. Clinical and predictive significance of hyponatremia after aneurysmal subarachnoid hemorrhage
. Balkan Med J
11. Hannon MJ, Behan LA, O’Brien MM, Tormey W, Ball SG, Javadpour M, Sherlock M, Thompson CJ. Hyponatremia following mild/moderate subarachnoid hemorrhage
is due to SIAD and glucocorticoid deficiency and not cerebral salt wasting. J Clin Endocrinol Metab
12. Saramma P, Menon RG, Srivastava A, Sarma PS. Hyponatremia after aneurysmal subarachnoid hemorrhage
: implications and outcomes. J Neurosci Rural Pract
13. Maimaitili A, Maimaitili M, Rexidan A, Lu J, Ajimu K, Cheng X, Luo K, Sailike D, Liu Y, Kaheerman K, et al. Pituitary hormone level changes and hypxonatremia in aneurysmal subarachnoid hemorrhage
. Exp Ther Med
14. Takaku A, Shindo K, Tanaka S, Mori T, Suzuki J. Fluid and electrolyte disturbances in patients with intracranial aneurysms. Surg Neurol
15. Kurokawa Y, Uede T, Ishiguro M, Honda O, Honmou O, Kato T, Wanibuchi M. Pathogenesis of hyponatremia following subarachnoid hemorrhage
due to ruptured cerebral aneurysm. Surg Neurol
16. Beseoglu K, Etminan N, Steiger HJ, Hanggi D. The relation of early hypernatremia with clinical outcome in patients suffering from aneurysmal subarachnoid hemorrhage
. Clin Neurol Neurosurg
17. Bales J, Cho S, Tran TK, Korab GA, Khandelwal N, Spiekerman CF, Joffe AM. The effect of hyponatremia and sodium variability on outcomes in adults with aneurysmal subarachnoid hemorrhage
. World Neurosurg
18. Sherlock M, Agha A, Thompson CJ. Aneurysmal subarachnoid hemorrhage
. N Engl J Med
19. Bederson JB, Connolly ES Jr, Batjer HH, Dacey RG, Dion JE, Diringer MN, Duldner JE Jr, Harbaugh RE, Patel AB, Rosenwasser RH. Guidelines for the management of aneurysmal subarachnoid hemorrhage
: a statement for healthcare professionals from a special writing group of the Stroke Council, American Heart Association. Stroke
20. Connolly ES Jr, Rabinstein AA, Carhuapoma JR, Derdeyn CP, Dion J, Higashida RT, Hoh BL, Kirkness CJ, Naidech AM, Ogilvy CS, et al. Guidelines for the management of aneurysmal subarachnoid hemorrhage
: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke
21. Banks JL, Marotta CA. Outcomes validity and reliability of the modified Rankin scale: implications for stroke clinical trials: a literature review and synthesis. Stroke
22. Alimohamadi M, Saghafinia M, Alikhani F, Danial Z, Shirani M, Amirjamshidi A. Impact of electrolyte imbalances on the outcome of aneurysmal subarachnoid hemorrhage
: a prospective study. Asian J Neurosurg
23. Spatenkova V, Bradac O, de Lacy P, Skrabalek P, Suchomel P. Dysnatraemia is frequently a poor prognostic indicator in patients with acute subarachnoid hemorrhage
having targeted sodium management. J Neurosurg Sci
24. Zheng B, Qiu Y, Jin H, Wang L, Chen X, Shi C, Zhao S. A predictive value of hyponatremia for poor outcome and cerebral infarction in high-grade aneurysmal subarachnoid haemorrhage patients. J Neurol Neurosurg Psychiatry
25. Crompton MR. Hypothalamic lesions following the rupture of cerebral berry aneurysms. Brain