a b s t r a c t

Purpose: Central diabetes insipidus (CDI) is a marker of severe brain injury. Here we aimed to investigate the prev- alence and risk factors of CDI in cardiac arrest survivors treated with Targeted temperature management (TTM). Methods: This retrospective observational study included consecutive adult cardiac arrest survivors treated with TTM between 2008 and 2014. Central diabetes insipidus was confirmed if all of the following criteria were met: urine volume N 50 cc kg^-1 d^-1, serum osmolarity N 300 mmol/L, urine osmolarity b 300 mmol/L, and serum sodium N 145 mEq/L. The primary outcome was the incidence of CDI. Results: Of the 385 included patients, 45 (11.7%) had confirmed central CDI. Univariate analysis showed that younger age, nonwitness of collapse, nonshockable rhythm, a high incidence of asphyxia arrest, longer downtime, and lower initial core temperature were associated with CDI development. Patients with CDI had a higher incidence of Poor neurologic outcomes at discharge and higher in-hospital mortality rate (20/45 vs 76/340, P= .001) as well as 180-day mortality (44/45 vs 174/340, Pb .001). Multivariate analysis revealed that age (odds ratio [OR], 0.963; 95% confidence interval [CI], 0.942-0.984), shockable rhythm (OR, 0.077; 95% CI, 0.009-0.662), downtime (OR, 1.025; 95% CI, 1.006-1.044), and asphyxia etiology (OR, 6.815; 95% CI, 2.457-18.899) were independently associated with CDI development.

Conclusion: Central diabetes insipidus developed in 12% of cardiac arrest survivors treated with TTM, and those with CDI showed poor neurologic outcomes and high mortality rates. Younger age, nonshockable rhythm, long down- time, and asphyxia arrest were significant risk factors for development of CDI.

(C) 2016

Introduction

Although advanced postcardiac arrest care has improved the clinical outcomes of cardiac arrest survivors [1-3], approximately 15% of out-of- hospital cardiac arrest and 22% of in-hospital cardiac arrest patients sur- vive to discharge [4,5]. Brain injury is a leading cause of mortality and morbidity during postcardiac arrest care in cardiac arrest survivors [6,7]. Central diabetes insipidus (CDI), which is characterized by hypotonic polyuria, occurs from numerous conditions affecting the hypothalamic- posterior pituitary unit. Acquired CDI mostly arises from trauma or

? Funding Sources/Disclosures: The authors have no relevant financial information or potential conflicts of interest to disclose.

* Corresponding author. Department of Emergency Medicine, Chonnam National Uni- versity Medical School, 160, Baekseo-ro, Dong-gu, Gwangju, Republic of Korea. Tel.: +82 62 220 6809; fax: +82 62 228 7417.

E-mail address: [email protected] (B.K. Lee).

surgical complications [8]. A direct mechanical impact on the hypothalamic-pituitary axis or an indirect mechanism such as vas- cular damage, cerebral edema, and Elevated intracranial pressure leads to CDI in traumatic brain injury [9]. A lower Glasgow Coma Scale score, cerebral edema, and severe injury were risk factors for the development of CDI after trauma. Hence, CDI after trauma is eventually associated with higher mortality rates [10]. Central dia- betes insipidus also leads to worse clinical outcomes through dehy- dration and electrolyte imbalance. Approximately 61% of brain dead patients were reportedly diagnosed as having CDI [11]. Therefore, CDI can be a surrogate marker of severe brain injury.

Cardiac arrest producing cerebral edema can also induce CDI [12]. However, with the exception of one article and several case reports, studies concerning the incidence of CDI and the factors associated with its development following cardiac arrest are lacking [13-15]. Therefore, this study aimed to identify the prevalence of CDI after cardi- ac arrest, risk factors for its development, and its characteristics in cardi- ac arrest survivors.

http://dx.doi.org/10.1016/j.ajem.2016.04.005

0735-6757/(C) 2016

Methods

Study design and population

This retrospective observational cohort study included consecutive comatose patients who were treated with targeted temperature management after cardiac arrest at Chonnam National University Hospital, a university-affiliated hospital, between January 2008 and De- cember 2014. This study was approved by the Chonnam National Uni- versity Hospital Institutional Review Board.

Cardiac arrest patients older than 16 years who underwent TTM were included. Patients were excluded if they were transferred to an- other facility or died within 72 hours after return of spontaneous circu- lation (ROSC) or extracorporeal membrane oxygenation was applied during postcardiac arrest care.

TTM protocol

During TTM, a target temperature was maintained using either feedback-controlled endovascular catheters (COOLGARD3000 Thermal Regulation System; Alsius Corporation, Irvine, CA) or Surface cooling de- vices (Blanketrol II [Cincinnati Subzero Products, Cincinnati, OH] and Artic Sun Energy Transfer Pads [Medivance Corp, Louisville, CO, USA]). The tem- perature was monitored using an esophageal and/or rectal temperature probe. Remifentanil and midazolam were used for sedation and analgesia. A Neuromuscular blocking drug was administered on an as-needed basis to prevent shivering. Upon completion of the TTM maintenance phase, pa- tients were rewarmed at a rate of 0.25 to 0.5?C h^-1. Advanced critical care measures such as oxygenation, ventilation, glucose control, and hemody- namic optimization were provided according to standard guidelines.

Central diabetes insipidus was identified when all of the following criteria were met: urine volume N 50 cc kg^-1 d^-1, serum osmolarity N 300 mmol/L, urine osmolarity b 300 mmol/L, and serum sodium N 145 mEq/L.

Data collection and patient outcomes

The following variables were obtained for each patient: age, sex, preexisting illness, presence of a witness on collapse, first monitored rhythm, location of arrest, etiology of cardiac arrest, downtime, initial core temperature, preinduction time, induction duration, rewarming duration, CDI development, CDI onset time, lowest urine osmolarity, highest serum osmolarity, highest serum sodium concentration, highest 24-hour urine output during admission, desmopressin requirements, brain death confirmed on electroencephalography, vital status at hospi- tal discharge (alive or dead), Neurologic outcome at hospital discharge, survival time, and 180-day mortality rate. The Sequential Organ Failure Assessment score within the first 24 hours after intensive care unit ad- mission was used to assess multiple-organ dysfunction severity [16]. Neurologic outcome was assessed using the Glasgow-Pittsburgh cere- bral performance categories (CPCs) at discharge according to outcome assessment recommendations for comatose cardiac arrest patients and recorded as CPC 1 (good performance), CPC 2 (moderate disability), CPC 3 (severe disability), CPC 4 (vegetative state), and CPC 5 (brain death or death) [17]. Neurologic outcome was dichotomized as either favorable (CPCs 1 and 2) or unfavorable (CPCs 3-5). The primary out- come was the incidence of CDI.

Data analyses

Continuous variables are presented as mean +- SD or median values with interquartile ranges as appropriate. The Wilcoxon rank sum test was used to compare nonnormally distributed continuous variables. The independent t test was used to compare normally distributed con- tinuous variables. Categorical variables are presented as frequencies and percentages. Categorical variables were compared using the ?²

test or Fisher exact test as indicated. We conducted multivariate analy- sis using logistic regression for CDI development and 180-day mortality, and Cox proportional hazards regression for survival time after adjust- ment for relevant covariates. All variables with a P value less than .1 on univariate analysis were included in the multivariate logistic regres- sion model and Cox regression model. Backward selection was used to achieve the final model. Goodness of fit of the final model was evaluated using the Hosmer-Lemeshow test. Kaplan-Meier survival curves were compared between the CDI and no CDI cohorts using the log-rank test. Data were analyzed using PASW/SPSS software version 18 (IBM Inc, Chicago, IL). A 2-sided significance level of .05 was used to identify sta- tistical significance.

Results

Patient population

During the study period of January 2008 to December 2014, a total of 479 consecutive comatose cardiac arrest survivors were treated with TTM. As summarized in Fig. 1, 63 patients died or were transferred to other hospitals within 72 hours after cardiac arrest and 30 patients were treated with extracorporeal membrane oxygenation. Ultimately, 385 patients were included in this study, of whom 45 (11.7%) were di- agnosed as having CDI (Fig. 1).

The median patient age was 58.0 years (45.0-70.0 years), a witness to the cardiac arrest was present for 292 (75.8%) patients, 111 (28.8%) patients had a shockable rhythm, 204 (53.0%) patients were presumed to have a cardiac etiology, 64 (16.6%) patients collapsed from asphyxia, and 313 (81.3%) patients experienced out-of-hospital cardiac arrest (Table 1). A total of 99 (25.7%) patients had in-hospital mortality and 253 (65.7%) were discharged with a poor neurologic outcome.

Risk factors for CDI development

Baseline characteristics and univariate comparisons are summarized in Table 1. The patients with CDI were younger and had a significantly lower incidence of preexisting illness (coronary artery disease, hyper- tension, diabetes, and Renal impairment). The patients with CDI, compared with patients with non-CDI, were more likely to have a nonshockable rhythm, asphyxia etiology, longer downtime, and lower initial core temperature and less likely to have a witness to their col- lapse. In-hospital mortality and poor neurologic outcome were signifi- cantly higher in the patients with CDI.

The multivariate analysis revealed that younger age (odds ratio [OR], 0.963; 95% confidence interval [CI], 0.942-0.984; P= .001), shockable

rhythm (OR, 0.077; 95% CI, 0.009-0.662; P= .020), longer down time

(OR, 1.025; 95% CI, 1.006-1.044; P= .009), and asphyxial etiology (OR,

6.815; 95% CI, 2.457-18.899; Pb .001) were associated with CDI development.

Characteristics of patients with CDI

Of the total 45 patients, CDI developed within 24 hours after cardiac arrest in 14 (31.1%) patients and within 7 days after ROSC in the other patients (Fig. 2). The lowest urine osmolarity was 107 mmol/L (72- 162 mmol/L), the highest serum osmolarity was 338 mmol/L (327- 352 mmol/L), the highest Serum sodium concentration was 163 mEq/L (156-169 mEq/L), and the highest 24-hour urine output was 6870 mL (5720-8465 mL). Repeated nasal sprays of desmopressin were provided for treatment and median requirement of desmopressin was 8 ug (3-16 ug). Brain death was identified in 6 (13.3%) patients, and 4 patients died after Organ donation, whereas 12 (26.7%) patients were confirmed to not be brain dead and brain dead status was not determined in the

other 27 (60.0%) patients.

Fig. 1. Flowchart.

CDI and 180-day mortality

Table 2 shows the association between variables and 180-day mor- tality. After the adjustment for confounders, CDI was associated with

180-day mortality (OR, 41.977; 95% CI, 5.718-308.156; Pb .001). Using

a Cox proportional hazard regression model, we found that the presence of CDI increased the hazard of death (hazard ratio, 4.159; 95% CI, 2.828- 6.115; Pb .001).

Table 1

Patient characteristics according to development of CDI

	Total (n = 385)	CDI (n = 45)	No CDI (n = 340)	P
Age (y), median (IQR)	58.0 (45.0-70.0)	47.0 (36.5-58.0)	60.0 (47.0-70.0)		b.001
Male sex Preexisting illness Coronary artery disease	263 (68.3) 47 (12.2)	26 (57.8) 1 (2.2)	237 (69.7) 46 (13.5)		.106 .029
Heart failure	25 (7.3)	0 (0.0)	28 (8.2)		.059
Hypertension	139 (36.1)	9 (20.0)	130 (38.2)		.017
Diabetes	107 (27.8)	6 (13.3)	101 (29.7)		.021
Pulmonary disease	28 (7.3)	1 (2.2)	27 (7.9)		.228
Renal impairment	38 (9.9)	0 (0.0)	38 (11.2)		.014
Cerebrovascular accident	24 (6.2)	1 (2.2)	23 (6.8)		.336
Hepatic disease	9 (2.3)	0 (0.0)	9 (2.6)		.606
Witness of collapse First monitored rhythm Shockable	292 (75.8) 111 (28.8)	19 (42.2) 1 (2.2)	273 (80.3) 110 (32.4)		b.001 b.001
Nonshockable Location of arrest Out-of-hospital	274 (71.2) 313 (81.3)	44 (97.8) 39 (86.7)	230 (67.6) 274 (80.6)		.326
In-hospital Etiology Cardiac	72 (18.7) 204 (53.0)	6 (13.3) 7 (15.6)	66 (19.4) 197 (57.9)		b.001
Asphyxia	64 (16.6)	23 (51.1)	41 (12.1)
Other medical	86 (22.3)	9 (20.0)	77 (22.6)
Drug overdose	28 (7.3)	5 (11.1)	23 (6.8)
Drowning	3 (0.8)	1 (2.2)	2 (0.6)
Downtime (min), median (IQR)	27.0 (18.5-40.0)	35.0 (24.5-45.0)	27.0 (17.0-37.0)		.004
Preinduction time (min), median (IQR)	210.0 (150.0-300.0)	193.0 (149.0-307.5)	211.5 (150.8-300.0)		.880
Duration of induction (h), median (IQR)	2.5 (1.5-4.0)	2.0 (1.1-3.0)	2.5 (1.5-4.0)		.069
Rewarming duration (h), median (IQR)	12.0 (10.0-13.0)	12.0 (9.1-14.0)	12.0 (10.0-13.0)		.453
SOFA score	9.0 (7.0-11.0)	9.0 (6.5-11.0)	9.0 (7.0-11.0)		.262
Initial core temperature (?C), (IQR)	36.0 (35.1-36.7)	35.8 (34.2-36.3)	36.0 (35.2-36.7)		.011
Poor neurologic outcome	253 (65.7)	45 (100.0)	208 (61.2)	b0.001
In-hospital mortality	99 (25.7)	20 (44.4)	76 (22.4)	.001

Abbreviations: IQR, interquartile range; SOFA, Sequential Organ Failure Assessment.

Fig. 2. Time of CDI onset. More than 50% of cases of CDI occurred within 3 days after cardiac arrest.

The median survival (number of days) of the CDI and no CDI groups was 15 and 157 days, respectively. The 180-day mortality rates of CDI and no CDI patients were 97.8% and 51.2%, respectively (Pb .001; Fig. 3).

Discussion

In this retrospective cohort study, 11.7% of cardiac arrest survivors treated with TTM experienced CDI. Younger age, nonshockable rhythm, longer downtime, and asphyxia were associated with CDI development. Central diabetes insipidus mostly developed within 7 days after ROSC. Brain death was confirmed in 13.3% of patients with CDI, and the 180- day mortality rate was 97.8%. There were significant differences be- tween the CDI and no CDI cohorts in overall mortality at hospital dis- charge or death at 180 days. Central diabetes insipidus had a robust association with mortality.

Although various etiologies can lead to CDI, acquired CDI most com- monly develops from traumatic brain injury or transsphenoidal surgery [8]. Few studies to date have described CDI after cerebral ischemia such as ischemic stroke or cardiac arrest. Posttraumatic CDI reportedly varies at 15% to 51% [10,18,19]. Central diabetes insipidus after severe trau- matic brain injury is not an uncommon adverse event because the pitu- itary gland and its vascular supply and stalk are susceptible to trauma [20]. Hadjizacharia et al [10] reported an 11.8% incidence of CDI in pa- tients with isolated head injury from blunt trauma and an incidence of 39.5% in patients with penetrating trauma. In the current study, approx- imately 12% of patients developed CDI after cardiac arrest, similar to the rate in patients with isolated head injury from blunt trauma. Chae et al

[12] reported approximately 21% of CDI after cardiac arrest. Although it is difficult to directly compare the results, the subjects in the present study appear to be younger, have a lower rate of asphyxia etiology, and have a shorter downtime compared with those in the previous

Table 2

Multiple regression models for 180-day mortality and survival time

study [12]. We presume that those factors in combination influenced the lower incidence of CDI seen here than that in the previous study by Chae et al [12].

Central diabetes insipidus is attributed to not only direct mechanical impact to the hypothalamic-pituitary unit but also ischemia, hypoxia, and elevated intracranial pressure. Central diabetes insipidus after car- diac arrest might result from Hypoxic brain damage, cerebral edema, or elevated intracranial pressure. Low Glasgow Coma Scale score and ce- rebral edema representing severe brain injury are associated with CDI development in patients with traumatic brain injury [18]. In a study an- alyzing the development of CDI after cardiac arrest, the gray-to-white matter ratio on brain computed tomography, which reflects cerebral edema severity, remained a robust marker for the development of CDI after adjustment for confounders [12]. Therefore, we presume that the development of CDI could be attributed to the presence of cerebral edema in cardiac arrest.

In the present study, nonshockable rhythm, longer downtime, youn- ger age, and asphyxial etiology rather than cardiac etiology were associ- ated with CDI development after cardiac arrest. The previous study also demonstrated that younger age and an asphyxia etiology were indepen- dent factors for the development of CDI [12]. Nonshockable rhythm, longer downtime, and asphyxial etiology are well-known predictors of poor clinical outcome in cardiac arrest survivors [21-23]. In an experi- mental study, CDI did not develop until 90% of the hypothalamic neu- rons were destroyed [24]. Therefore, CDI development after cardiac arrest is also suggestive of severe brain injury leading to poor clinical outcomes. However, younger age was associated with CDI development in the present study, whereas younger age is known to be associated with Good neurologic outcome and survival in cardiac arrest survivors [25,26]. Although the exact reason for this finding is unclear, we pre- sume that younger patients were more prone to intracranial pressure elevations. Younger patients may be more susceptible to cerebral edema than elderly patients because the latter have greater compensa- tory reserves due to cerebral atrophy [27].

180-d mortality, OR (95% CI)

P Time to death, P

HR (95% CI)

Previous studies presenting case reports of patients with CDI after cardiac arrest described that CDI developed within 48 or 72 hours

Age 1.050 (1.028-1.072) b.001 1.022 (1.012-1.033) b.001

Preexisting illness

Diabetes 2.904 (1.508-5.592) .001 1.682 (1.244-2.276) .001

after cardiac arrest [13,15]. More than 50% of cases of CDI occurred with-

in 3 days after cardiac arrest in the present study. Chae et al [12] also re- ported that 40% of cases of CDI developed within 3 days after cardiac

Pulmonary disease Cerebrovascular accident

Witness of collapse

4.158 (1.285-13.453) .017 2.256 (1.394-3.651) .001

7.807 (1.999-30.487) .003 1.365 (0.829-2.248) .221

0.308 (0.145-0.656) .002 0.494 (0.360-0.677) b.001

arrest. Central diabetes insipidus that develops early after cardiac arrest might be a result of hypoxic brain damage during downtime, whereas CDI that develops after the recovery phase (approximately 72 hours after ROSC) might be a result of cerebral edema or elevated intracranial pressure during postcardiac arrest care. However, it is not easy to diag-

Shockable rhythm 0.302 (0.143-0.637) .002 0.342 (0.221-0.530) b.001

Etiology

Cardiac Reference

Asphyxia 2.865 (1.097-7.484) .032 1.552 (1.019-2.365) .041

Downtime 1.026 (1.008-1.044) .004 1.008 (1.001-1.015) .023

SOFA score 1.171 (1.064-1.290) .001 1.104 (1.053-1.157) b.001

CDI 59.756 (6.923-515.584) b.001 4.159 (2.828-6.115) b.001

Abbreviations: HR, hazard ratio; SOFA, Sequential Organ Failure Assessment.

nose CDI in cardiac arrest survivors because the water deprivation test, the essential examination for the differential diagnosis of CDI, cannot be performed in critically ill patients. Furthermore, the large volumes re- quired for hemodynamic stabilization at the early stage of postcardiac arrest care and the cooling used in TTM can confound cold diuresis with CDI [3,28]. Even the diagnostic performance of direct plasma mea- surement of arginine vasopressin in the differential diagnosis of CDI was

Fig. 3. Kaplan-Meier survival curves demonstrating cumulative survival over time during the 180 days after cardiac arrest in the CDI and no-CDI groups. A significant difference was seen between the 2 groups (log-rank test, Pb .001). The median survival of the CDI and no-CDI groups was 15 and 157 days, respectively.

low [29]. Therefore, we used 1 clinical and 3 laboratory findings to define CDI in the present study. The serum osmolarity and sodium con- centrations were much higher than the diagnostic threshold, whereas the urine osmolarity was much lower. Previous case reports also dem- onstrated extremely high serum sodium concentrations of 195 and 175 mEq/L [14,15].

Central diabetes insipidus was associated with a poor neurologic outcome at discharge and greater in-hospital mortality rates. Chae et al [12] also showed that CDI was an independent predictor of mortal- ity. Furthermore, CDI also showed a robust association with long-term mortality after adjustment for confounders in the present study. There was also a huge difference in median survival days between the CDI and no CDI cohorts. The greater neurologic injury severity after cardiac arrest was attributed to long-term mortality. It is not clear whether the electrolyte imbalance and water deficit derived from CDI affect mor- tality. Chae et al [12] showed that urine output was significantly differ- ent between the CPC 4 and CPC 5 groups in the CDI cohort. Therefore, it is presumed that the mortality of cardiac arrest survivors with CDI might be attributable to electrolyte imbalance and dehydration.

Desmopressin, a synthetic analog of vasopressin, is the drug of choice for CDI. Minimal dose administration is encouraged to avoid overtreatment. However, most of our patients required repeated doses of desmopressin. Despite treatment, all patients with CDI were discharged with poor neurologic outcomes. Central diabetes insipidus patients had higher in-hospital mortality rates, including 87% at 30 days. Therefore, CDI can be used a surrogate marker of severe brain in- jury after cardiac arrest. In the present study, 13% of patients with CDI were confirmed to have suffered brain death. There has been an in- crease in organ donors after cardiac arrest, and the Organ function has not differed between patients who suffered from brain death from car- diac arrest and those who suffered from brain death from other causes [30]. The American Heart Association also recommends evaluating for organ donation in brain death as class I [1]. Central diabetes insipidus development can be used as an early alert for the evaluation of and preparation for organ donation.

The present study has several limitations. First, it was retrospective and single centered, implying the need for further studies including larger sample sizes, multiple centers, and a prospective design to assess generalizability and causation rather than associations. Second, we ex- cluded patients who died within 72 hours after ROSC. This might lead to selection bias and influence the prevalence of CDI. However, its influ- ence was presumed minimal, because most cases of death within 72 hours after ROSC in cardiac arrest survivors were attributable to postcardiac arrest shock rather than the brain injury itself [31]. Third, we did not consider the effect of vasopressin administration during re- suscitation. Vasopressin has both a pressor effect and an antidiuretic ef- fect. However, because vasopressin has a half-life of 10 to 20 minutes, it

is shorter acting than desmopressin, which has a half-life of 90 to 150 min. Fourth, some CDI patients might have been missed as a result of missing data or inappropriate sample timing because we used a strict definition of CDI that included only those patients who met all 4 criteria.

Conclusions

In the current study, CDI developed in 12% of cardiac arrest survivors treated with TTM, and CDI patients showed poor neurologic outcomes and high mortality rates. Younger age, nonshockable rhythm, longer downtime, and asphyxia arrest were significant risk factors for CDI de- velopment. Further clinical studies in large cohorts are warranted.

Conflict of interest statement

The authors declare no conflicts of interest.

References

1. Callaway CW, Donnino MW, Fink EL, et al. Part 8: post-cardiac arrest care: 2015 American Heart Association guidelines update for cardiopulmonary resusciation and emergency cardiovacular care. Circulation 2015;132:S465-82.
2. Schober A, Holzer M, Hochrieser H, Posch M, Schmutz R, Metnitz P. Effect of inten- sive care after cardiac arrest on patient outcome: a database analysis. Crit Care 2014;18(2):R84.
3. Sunde K, Pytte M, Jacobsen D, et al. Implementation of a standardised treatment pro- tocol for post resuscitation care after out-of-hospital cardiac arrest. Resuscitation 2007;73:29-39.
4. Chan PS, McNally B, Tang F, Kellermann A. Recent trends in survival from out-of- hospital cardiac arrest in the United States. Circulation 2014;130:1876-82.
5. Girotra S, Nallamothu BK, Spertus JA, et al. Trends in survival after in-hospital cardiac arrest. N Engl J Med 2012;367:1912-20.
6. Dragancea I, Rundgren M, Englund E, Friberg H, Cronberg T. The influence of induced hypothermia and delayed prognostication on the mode of death after cardiac arrest. Resuscitation 2013;84:337-42.
7. Laver S, Farrow C, Turner D, Nolan J. Mode of death after admission to an intensive care unit following cardiac arrest. Intensive Care Med 2004;30:2126-8.
8. Fenske W, Allolio B. Clinical review: current state and future perspectives in the di- agnosis of diabetes insipidus: a clinical review. J Clin Endocrinol Metab 2012;97: 3426-37.
9. Yuan XQ, Wade CE. Neuroendocrine abnormalities in patients with traumatic brain injury. Front Neuroendocrinol 1991;12:209-30.
10. Hadjizacharia P, Beale EO, Inaba K, Chan LS, Demetriades D. Acute diabetes insipidus in severe head injury: a prospective study. J Am Coll Surg 2008;207:477-84.
11. Wijdicks EF, Rabinstein AA, Manno EM, Atkinson JD. Pronouncing brain death: con- temporary practice and safety of the apnea test. Neurology 2008;71:1240-4.
12. Chae MK, Lee JH, Lee TR, et al. Early central diabetes insipidus: an ominous sign in post-cardiac arrest patients. J Crit Care 2016;32:63-7.
13. Udoshi MB, Trivedi AD, Desai RC, Lichtenstein E. Diabetes insipidus following cardio- respiratory arrest. J Natl Med Assoc 1981;73:797-800.
14. Masood M, Kumar S, Asghar A, Jabbar A. An unusual case of central diabetes insipidus & hyperglycemic hyperosmolar state following cardiorespiratory arrest. BMC Res Notes 2013;6:325.
15. Choi SS, Kim WY, Kim W, Lim KS. Unexpected fatal hypernatremia after successful car- diopulmonary resuscitation with therapeutic hypothermia: a case report. J Korean Med Sci 2012;27:329-31.
16. Vincent JL, de Mendonca A, Cantraine F, et al. Use of the SOFA score to assess the in- cidence of organ dysfunction/failure in intensive care units: results of a multicenter, prospective study. Working group on “sepsis-related problems” of the European So- ciety of Intensive Care Medicine. Crit Care Med 1998;26:1793-800.
17. Booth CM, Boone RH, Tomlinson G, Detsky AS. Is this patient dead, vegetative, or se- verely neurologically impaired? Assessing outcome for comatose survivors of cardi- ac arrest. JAMA 2004;291:870-9.
18. Agha A, Thornton E, O’Kelly P, Tormey W, Phillips J, Thompson CJ. Posterior pituitary dysfunction after traumatic brain injury. J Clin Endocrinol Metab 2004;89:5987-92.
19. Hannon MJ, Crowley RK, Behan LA, et al. Acute glucocorticoid deficiency and diabe- tes insipidus are common after acute traumatic brain injury and predict mortality. J Clin Endocrinol Metab 2013;98:3229-37.
20. Harper CG, Doyle D, Adams JH, Graham DI. Analysis of abnormalities in pituitary gland in non-missile head injury: study of 100 consecutive cases. J Clin Pathol 1986;39:769-73.
21. Kim WY, Giberson TA, Uber A, Berg K, Cocchi MN, Donnino MW. Neurologic out- come in comatose patients resuscitated from out-of-hospital cardiac arrest with prolonged downtime and treated with therapeutic hypothermia. Resuscitation 2014;85:1042-6.
22. Meaney PA, Nadkarni VM, Kern KB, Indik JH, Halperin HR, Berg RA. Rhythms and outcomes of adult in-hospital cardiac arrest. Crit Care Med 2010;38:101-8.
23. Wissenberg M, Lippert FK, Folke F, et al. Association of national initiatives to improve Cardiac arrest management with rates of bystander intervention and patient survival after out-of-hospital cardiac arrest. JAMA 2013;310:1377-84.
24. Keren G, Barzilay Z, Schreiber M, Szienberg A, Aladjem M. Diabetes insipidus indicat- ing a dying brain. Crit Care Med 1982;10:798-9.
25. Terman SW, Shields TA, Hume B, Silbergleit R. The influence of age and chronic med- ical conditions on neurological outcomes in out of hospital cardiac arrest. Resuscita- tion 2015;89:169-76.
26. Winther-Jensen M, Pellis T, Kuiper M, et al. Mortality and neurological outcome in the elderly after target temperature management for out-of-hospital cardiac arrest. Resuscitation 2015;91:92-8.
27. Gjerris F, Brennum J. The cerebrospinal fluid, intracranial pressure and hernimation of the brain. In: Paulson OB, Gjerris F, Sorensen PS, editors. Clinical neurology and Neurosury. Copenhagen, Denmark: FADL’s Forlag Aktieselskab; 2004. p. 179-96.
28. Gaieski DF, Band RA, Abella BS, et al. Early goal-directed hemodynamic optimization combined with therapeutic hypothermia in comatose survivors of out-of-hospital cardiac arrest. Resuscitation 2009;80:418-24.
29. Fenske W, Quinkler M, Lorenz D, et al. Copeptin in the differential diagnosis of the polydipsia-polyuria syndrome-revisiting the direct and indirect water deprivation tests. J Clin Endocrinol Metab 2011;96:1506-15.
30. Orioles A, Morrison WE, Rossano JW, et al. An under-recognized benefit of car- diopulmonary resuscitation: organ transplantation. Crit Care Med 2013;41: 2794-9.
31. Lemiale V, Dumas F, Mongardon N, et al. Intensive care unit mortality after cardiac arrest: the relative contribution of shock and brain injury in a large cohort. Intensive Care Med 2013;39:1972-80.