a b s t r a c t

Introduction: The aim of this study was to investigate the relationship between the optic nerve sheath diameter measured on non-contrast head computed tomography and the diagnosis and prognosis of sponta- neous subarachnoid hemorrhage on emergency department (ED) patients.

Method: We used a matched control group of patients with the same age and gender who were diagnosed in the ED with spontaneous SAH and who admitted to the ED with headache. Four emergency medicine attending phy- sicians made the ONSD measurements. For measurements, the 3-mm posterior location where the optic nerve enters the eyeball was used.

Results: This study was done with 61 spontaneous SAHs with an equal number of control patients. The median ONSD for control and spontaneous SAH groups was 5.76 [interquartile range (IQR): 0.96] mm and 6.72 (IQR: 1.42) mm, respectively (p b 0.001). The area under the receiver operating characteristic curve was determined as 0.791 (confidence interval 95% 0.710-0.872). At an ONSD threshold value of 6.1 mm, the sensitivity and spec- ificity of SAH was 72%. There was no significant relationship between ONSD and in-hospital mortality in sponta- neous SAH patients (p N 0.05). The intra-class Correlation coefficients for inter and intra-rater reliability were 0.84 and 0.95, respectively.

Conclusion: In patients with spontaneous SAH, the ONSD measured in the orbital sections of a Head CT.is strongly correlated with a SAH diagnosis. Assessment of ONSD in head CTs taken with spontaneous SAH suspicion may contribute to the diagnoses of spontaneous SAH.

(C) 2017

Introduction

Headache is a frequent and high-risk complaint in emergency de- partments (ED). Many pathologies that threaten life and should be diag- nosed at an ED present with headache. Headache is the most common symptom of subarachnoid hemorrhage (SAH). The headache classically starts suddenly and peaks quickly. In aneurysmal SAH, sudden intracra- nial pressure increases due to hemorrhage within the first minute. The interruption of blood flow is due to global ischemia and cytotoxic edema. These are the early pathophysiological changes, which lead to headache in the SAH [1-3]. Increased intracranial pressure (over 20 mm Hg) was detected in N 50% of aneurysmal SAH cases [4].

The most commonly used and most easily accessible examination for the diagnosis of these pathologies is non-contrast head computed

* Corresponding author at: Izmir Tepecik Research and Educational Hospital, Department of Emergency Medicine, Izmir, Turkey.

E-mail address: [email protected] (M. Yesilaras).

tomography (CT). Although this examination provides much valuable information for spontaneous SAH and stroke, it does not exclude these diagnoses 100% [5,6]. The diagnostic value is much lower for conditions such as venous sinus thrombosis and idiopathic intracranial hyperten- sion. Sensitivity of non-contrast head CT for SAH was reported as 92.9% to 98.7% [7,8].

The optic nerve sheath is a continuation of the dura mater, which also surrounds the brain. When the intracranial pressure increases, the Optic nerve sheath diameter also increases. This measurement can be done by ultrasound at the 3 mm posterior location of the eyeball [9]. The optic nerve sheath diameter is usually not assessed during ex- amination of a head CT. The ONSD expands in proportion to intracranial pressure [10]. Thus, it may also be an indicator for mortality. The ONSD measurement is a non-invasive method that can assess intracranial pressure [11]. Different thresholds have been defined in different popu- lations to indicate an increase in intracranial pressure [12-14]. In pa- tients with spontaneous SAH, high intracranial pressure and no response to treatment are associated with mortality [15]. Many studies

http://dx.doi.org/10.1016/j.ajem.2017.04.022

0735-6757/(C) 2017

have reported that ONSD is well correlated with intracranial pressure in patients with intracranial hemorrhage [1,16].

The aim of this study was to investigate the relationship between ONSD and diagnosis/prognosis in head CTs of patients who were diag- nosed with spontaneous SAH at ED.

Method

This single-center case-control study was conducted in a hospital with approximately 200,000 adult emergency patient admission per year that receive patient transfers from other hospitals and where emergency medical assistant training is provided. Local ethics commit- tee approval was obtained prior to the research.

Selection of patients and control group

Hospital digital data were examined, and patients aged 18 years and above who were admitted to the hospital with a spontaneous SAH diag- nosis in the ED between 2010 and 2014 were identified. The digital and written data of these patients were examined. We excluded patients who were transferred from other hospitals, who were subjected to CT after 6 or more hours of admission to the ED, with CT images which could not be reviewed over picture archiving communication systems (PACS), who had orbital trauma, cranial operation and a pathology that could increase the intracranial pressure via means other than SAH. Patients were excluded if they were transferred to another institu- tion or referred to our hospital from another institution.

The control group was selected from patients who were admitted to the ED with a headache complaint at the same time interval, who underwent a head CT and were discharged. The head CT of these pa- tients was normal, and they were discharged from the ED with im- proved symptoms. All control patients were ranked using a random number table. After this random sequencing, the first patient with the same age and gender for each of the cases was selected as the control patient. In the absence of a control patient of the same age, the first pa- tient in the nearest age group, young or old, was selected as the control patient. In the absence of the same gender, control patients were select- ed from the other gender.

The CTs of the control group were examined by a radiologist who confirmed that they were normal. Patients with any pathology in the head CT (arachnoid cyst, meningioma, etc.) were replaced with new control patients with the same algorithm. The consultations required for control patients and other referrals to our hospital were also exam- ined to confirm that there was no pathology.

Examination of head CTs

For the SAH group, all CTs were reviewed by an experienced radiol- ogist and rated according to the Fisher classification. All cranial CT im- ages were obtained on a “Siemens Emotion Duo 2 section/sec” device with 6-mm sections.

ONSD measurement

For ONSD measurements, orbital sections of the first head CT were used and were obtained when the patients applied to the ED. ONSDs were measured by four ED attending physician with an experience be- tween 5 and 10 years. The measurements were made digitally via PACS. Standard bone window views were used because the optic nerve sheath borders for the measurements could be better seen. The ED specialists who measured the ONSD were given 2 h of didactic and practical training on orbital anatomy and measurement technique on CT imaging. The patients in the control group and the SAH group were randomly mixed. The CTs were examined by a person outside the mea- surer and were enlarged 3-4 times so that only the orbital sections of the CT could be seen. Thus, the measurements were blinded. The

ONSDs measurements were made from 3 mm posterior to the eyeball. If measurements could not be made here, then they were made 3- 10 mm and 10 mm posterior, respectively (Fig. 1). Patients who could not be measured from both eyes were not included in the statistical analysis. In the analyses, the average value was used for measurements from two eyes, and the single measurement was used for measure- ments from one eye.

Clinical markers and outcomes for SAH

All of the SAH patients were assessed in the ED by neurosurgeons and admitted to the hospital. We recorded the Glasgow Coma Scale (GCS), the Hunt-Hess scores, possible causes of hemorrhage following hospital examinations, whether the in-hospital hemorrhage re- occurred or not, whether there was intubation requirement after admis- sion to the ED or hospital (those extubated after being intubation for op- eration were excluded), whether surgical intervention was applied or not, and patient outcome.

Sample size and statistical analysis

In this study, we included all cases that were presented after the date on which patient data were recorded electronically. The power of the data was evaluated via power analysis using www.openepi.com; the resulting power was 100%.

The data were evaluated by a histogram and the Kolmogorov- Smirnov test. We determined that the data did not have a normal distri- bution. The qualitative data were expressed by frequencies (n and %), and the quantitative data were expressed by median and interquartile range (IQR). In some cases, minimum and maximum values were also given so that the distribution could be understood. The Mann- Whitney U and Kruskal Wallis tests were used to compare quantitative data, and the chi-square test was used to compare qualitative data. Re- ceiver operating characteristic (ROC) analysis was used to determine threshold ONSD values. Statistical analyses were performed using Sta- tistical Package for the Social Sciences version 22.0.

Result

We found that 173 patients with spontaneous SAH were admitted to ED over 5 years. Sixteen patients were referred to our hospital from an- other health facility. We excluded 12 patients who were unable to ob- tain CT images, 10 patients underwent CT 6 h after the ED admission, 9 patients had additional pathologies, 3 patients were excluded from SAH diagnosis, and 22 patients were referred to other health institu- tions. All spontaneous SAH patients except one were diagnosed by CT. One patient (1.6%) was diagnosed with lumbar puncture . The ex- clusion diagram for these patients is given in Fig. 2. Two patients had bi- lateral (one SAH and one control group) and two patients had unilateral (both in the SAH group) ONSD that were not measured. This study was done with 61 spontaneous SAH patients and 61 controls. Three patients with meningioma and arachnoid cysts were excluded from the control group, and new controls were selected with the same algorithm. Fe- males accounted for 37 (60.7%) of spontaneous SAH cases and 35 (57.4) of control cases. The median age for SAH and control cases was 56 (IQR = 25, min: 24, max: 90) years and 57 (IQR = 21, min: 24, max: 83) years, respectively. The median GCS of patients with SAH was 14 (IQR = 7, min 3, max: 15). Median values of the right, left and averages of both eyes ONSD of for the control group were 5.74 mm (IQR: 0.99), 5.86 mm (IQR: 1.10), 5.76 mm (IQR: 0.96) and spontaneous SAH groups were 6.51 mm (IQR: 1.26), 6.72 mm (IQR: 1.42) and 0.672 mm (IQR: 1.42), respectively. The distribution of the ONSDs of the pa- tients for the control and SAH group was shown in the graphic (Fig. 3). This is because of artifacts caused by head and eye movements, fold- ing of the optic nerve-sheath complex on itself, or CT sections that could not visualize the optic nerve-sheath complex. Measurements were

Fig. 1. Optic nerve sheath diameter measurements made from different areas [3 mm posterior (A), 3-10 mm (B) and 10 mm posterior (C) of the eyeball] and measurement of a patient with increased optic nerve sheath diameter (D).

successful in at least one eye in 98.4% of the patients and in both eyes in 96.7% of the patients; 92% of the measurements were made 3 mm pos- terior from the eyeball, 6% from 3-10 mm posterior, and 2% from N 10 mm posterior of the eyeball. The median ONSDs and the receiver operating characteristic (ROC) curve for the SAH and control group and the areas underneath are shown in Table 1. Sensitivity and specific- ity values for ONSD measurements were calculated by ROC analysis. For cases with an average ONSD of b 4.8 mm, the SAH diagnosis exclusion prediction sensitivity is 100 (CI 95%, 93-100%) and specificity is 12 (CI 95%, 10-20%). If the average ONSD is taken as 7.6 mm, the sensitivity is 21 (CI 95%, 10-30%) and the specificity is 100 (CI 95%, 93-100%) for SAH diagnosis. When the average ONSD threshold value was taken as

6.1 mm, the sensitivity is 73 (CI 95%, 60-80%) and the specificity is 72 (CI %95, 60-80%) for diagnoses of spontaneous SAH.

We found that 23 (38%) of the spontaneous SAH cases died in the hospital. Median ONSD were 6.8 mm (IQR: 1.13) and 6.51 mm (IQR: 1.57), respectively, in patients who died and could be discharged from the hospital. (p = 0.927) ONSDs according to clinical radiological scor- ing systems and clinical course of patients for SAH group are given in Table 2. Three (4.9%) patients had recurrent in-hospital hemorrhage. Inter-rater and intra-rater reliability was calculated by four physicians with 15 cases (SAH and control group mixed) were 0.84 and 0.95, respectively.

Discussion

The ONSD of spontaneous SAH patients were measured via non- contrast head CT and were statistically higher than control patients,

Fig. 2. Patient exclusion diagram for the spontaneous subarachnoid hemorrhage group.

but this was not related to mortality. Multidetector CT in the first 6 h of suspected SAH is a common approach to rule out hemorrhage [17]. Tra- ditionally, a diagnostic test for the exclusion of SAH in patients who can- not be diagnosed by CT but whose clinical suspicion persists is LP. In a

study by Sayer et al., 204 patients were required to undergo LP to detect any spontaneous SAH in patients who were negative for CT [18]. In these cases, the measurement of the ONSD in CT may be helpful in de- termining the LP as well as the high intracranial pressure and the SAH.

Fig. 3. Graphic representation of the distribution of optic nerve sheath diameters in spontaneous subarachnoid hemorrhage and the control group.

Table 1

The optic nerve sheath diameters the area under the curve of receiver operating characteristic curve of the spontaneous subarachnoid hemorrhage and control groups.

	Spontaneous SAH group ONSD (median, IQR), mm	Control group ONSD (median, IQR), mm	p value	AUC (95% confidence interval)
Right eye	6.51 (1.26)	5.74 (0.99)	b0.001	0.763 (0.677-0.848)
Left eye	6.79 (1.65)	5.86 (1.10)	b0.001	0.764 (0.678-0.850)
Average of both eyes	6.72 (1.42)	5.76 (0.96)	b0.001	0.791 (0.710-0.872)

ONSD: optic nerve sheath diameter, AUC: area under the curve, ROC: receiver operating characteristic, SAH: subarachnoid hemorrhage, IQR: interquartile range.

The optic nerve sheath is not a standardized area in CT views. Assess- ment in this area may also contribute to the detection of increased intra- cranial pressure [12,19,20].

Most of the cranial pathologies manifest themselves with an in- crease in intracranial pressure [21]. These patients may present with headache and loss of consciousness. Most of the patients apply to the ED in the first hours. The sensitivity of non-contrast CT taken in the first 6 h is 92.9%, and this value reaches 100% after 6 h [7]. It is quite dif- ficult to know which patients are at risk in this sense. Measuring the di- ameter of the optic nerve sheath may help in this regard. In a meta- analysis published by Ohle and colleagues in 2015, they emphasized that ONSD measurements via ultrasonography is a good method to de- tect increased intracranial pressure. In this study, we determined that the median ONSD measured in a non-contrast CT was 6.72 mm (IQR: 1.42) in the spontaneous SAH group and 5.76 mm (IQR: 0.96) in the control group (p b 0.001).

In a study by Vaiman et al., all non-Traumatic cases of intracranial hemorrhage were investigated [22]. Intracranial pressure and ONSD measured on non-contrast head CT were studied in these patients. This study concluded that the ONSD of intracranial hemorrhage and SAH groups were significantly elevated in ONSD compared to normal cases.

The difference between the ONSDs of the SAH and control cases was statistically significant, and the area under the ROC curve is also 0.791 (CI 95% 0.71-0.87) for the average ONSD. The use of threshold ONSD values (4.8 mm, 7.6 mm) may be unreliable to confidently exclude diag- noses of spontaneous SAH cases because the sensitivity and specificity values have a wide confidence interval. Although the relationship be- tween the diagnosis of spontaneous SAH and the increases in ONSD is

Table 2 Distribution of spontaneous SAH patients according to their clinical and radiological grad- ing systems and clinical course.

n Median ONSD mm (interquartile range) p value

Hunt & Hess Classification

1	31	6.32 (0.95)	0.97
2	5	7.44 (1.18)
3	11	6.8 (1.03)
4	7	7.48 (1.64)
5	6	7.14 (1.85)
Fisher classification
1	1	7.99
2	8	6.57 (0.84)	0.229
3	24	6.34 (1.12)
4	27	6.91 (1.31)
Glasgow Coma Scale
14-15	32	6.33 (1.0)	0.043
13-9	13	6.82 (1.11)
8-3	15	7.22 (1.4)
Endotracheal intubation
Yes	24	6.86 (1.37)	0.407
No	26	6.52 (1.44)
Operation
Yes	35	6.82 (1.38)	0.041
No	25	6.33 (1.09)

ONSD: optic nerve sheath diameter, SAH: subarachnoid hemorrhage.

obvious, more clinical research is needed to determine the threshold values.

In an intensive care study of traumatic brain injury patients, Legrand and colleagues found a significant correlation between ONSD and GCS in patients with dead or alive (6.8 mm versus 7.8 mm, respectively, p b 0.001) [23]. We did not find a significant difference between the ONSDs of patients who lived and died in our study (6.51 mm versus 6.81 mm respectively, p = 0.927).

Limitations

In this study, it is not possible with 100% certainty that no spontane- ous SAH is present in these patients because the control group was cho- sen retrospectively. It is not possible to distinguish the optic nerve sheath from the ophthalmic artery via CT. This may have led to a higher measurement in some patients. Spontaneous SAH is not the only cause that increases ONSD. Pathological measurements require investigation of other possible causes. In this study, CT images were examined and measured at 3-4 times magnification via PACS. The ONSD has a very small diameter (b 1 cm). This condition may not be as effective and re- liable when one cannot magnify the images.

Conclusion

The ONSD measured in non-contrast head CT in spontaneous SAH cases is larger than control patients. Measuring the diameter of the optic nerve sheath via CT with spontaneous SAH pre-diagnosis can also support the diagnosis and may provide additional benefits in cases with no radiological findings. In spontaneous SAH cases, ONSD is not a prognostic indicator.

References

1. Swope R, Glover K, Gokun Y, et al. Evaluation of headache severity after aneurysmal subarachnoid hemorrhage. Interdiscip Neurosurg 2014;1(4):119-22.
2. Milenkovic Z, Babic M, Radenkovic S. Nontraumatic (spontaneous) subarachnoid hemorrhage. Facta Univ 1998;5:12-7.
3. Cahill J, Calvert JW, Zhang JH. Mechanisms of Early brain injury after subarachnoid hemorrhage. J Cereb Blood Flow Metab 2006;26(11):1341-53.
4. Heuer GG, Smith MJ, Elliott JP, et al. Relationship between intracranial pressure and other clinical variables in patients with Aneurysmal subarachnoid hemorrhage. J Neurosurg 2004;101:408-16.
5. Mark DG, Sonne DC, Jun P, Schwartz DT, et al. False-negative interpretations of cra- nial computed tomography in aneurysmal subarachnoid hemorrhage. Acad Emerg Med 2016;23(5):591-8.
6. Vermeulen MJ, Schull M. Missed diagnosis of subarachnoid hemorrhage in the emer- gency department. Stroke 2007;38(4):1216-21.
7. Perry JJ, Stiell IG, Sivilotti ML, et al. Sensitivity of computed tomography performed within six hours of onset of headache for diagnosis of subarachnoid haemorrhage: prospective cohort study. BMJ 2011;343:d4277.
8. Dubosh NM, Bellolio MF, Rabinstein AA, et al. Sensitivity of early brain computed to- mography to exclude aneurysmal subarachnoid hemorrhage: a systematic review and meta-analysis. Stroke 2016;47(3):750-5.
9. Komut E, Kozaci N, Sonmez BM, et al. Bedside sonographic measurement of optic nerve sheath diameter as a predictor of intracranial pressure in ED. Am J Emerg Med 2016;34(6):963-7.
10. Ohle R, McIsaac SM, Woo MY, et al. Sonography of the optic nerve sheath diameter for detection of raised intracranial pressure compared to computed tomography: a systematic review and meta-analysis. J Ultrasound Med 2015;34(7):1285-94.
11. Dubourg J, Javouhey E, Geeraerts T, et al. Ultrasonography of optic nerve sheath di- ameter for detection of raised intracranial pressure: a systematic review and meta- analysis. Intensive Care Med 2011;37(7):1059-68.
12. Raffiz M, Abdullah JM. Optic nerve sheath diameter measurement: a means of de- tecting raised ICP in adult traumatic and nontraumatic neurosurgical patients. Am J Emerg Med 2017;35(1):150-3.
13. Wang L, Feng L, Yao Y, et al. Optimal optic nerve sheath diameter threshold for the identification of elevated opening pressure on lumbar puncture in a Chinese popu- lation. PLoS One 2015;10(2):e0117939.
14. Bekerman I, Sigal T, Kimiagar I, et al. The quantitative evaluation of intracranial pres- sure by optic nerve sheath diameter/eye diameter CT measurement. Am J Emerg Med 2016;34(12):2336-42.
15. Heuer GG, Smith MJ, Elliott JP, et al. Relationship between intracranial pressure and other clinical variables in patients with aneurysmal subarachnoid hemorrhage. Neu- rosurgery 2004;101(3):408-16.
16. Blaivas M, Theodoro D, Sierzenski PR. Elevated intracranial pressure detected by bedside emergency ultrasonography of the optic nerve sheath. Acad Emerg Med 2003;10(4):376-81.
17. Macdonald RL, Schweizer TA. Spontaneous subarachnoid haemorrhage. Lancet 2017;389(10069):655-66.
18. Sayer D, Bloom B, Fernando K, et al. An observational study of 2248 patients present- ing with headache, suggestive of subarachnoid hemorrhage, who received lumbar punctures following normal computed tomography of the head. Acad Emerg Med 2015;22(11):1267-73.
19. Tayal VS, Neulander M, Norton HJ, et al. Emergency department sonographic mea- surement of optic nerve sheath diameter to detect findings of increased intracranial pressure in adult head injury patients. Ann Emerg Med 2007;49(4):508-14.
20. Caffery TS, Perret JN, Musso MW, et al. Optic nerve sheath diameter and lumbar puncture opening pressure in nontrauma patients suspected of elevated intracranial pressure. Am J Emerg Med 2014;32(12):1513-5.
21. Dunn LT. Raised intracranial pressure. J Neurol Neurosurg Psychiatry 2002 Sep; 73(Suppl. 1):i23-7.
22. Vaiman M, Sigal T, Kimiagar I, et al. Noninvasive assessment of the intracranial pres- sure in non-traumatic intracranial hemorrhage. Clin Neurosci 2016;34:177-81.
23. Legrand A, Jeanjean P, Delanghe F, et al. Estimation of optic nerve sheath diameter on an initial brain computed tomography scan can contribute prognostic informa- tion in traumatic brain injury patients. Crit Care 2013;17(2):R61.