Efficacy of bedside optic nerve sheath diameter measurement in differentiating provoked seizure from unprovoked seizure in the emergency department

Abstract

BACKGROUND:

Emergency departments (EDs) are typically the first medical contact for seizure patients, and early diagnosis and treatment is primarily the responsibility of emergency physicians.

OBJECTIVES:

Demonstrate the efficacy of bedside ocular ultrasonography for optic nerve sheath diameter (ONSD) measurement in differentiating provoked seizure from unprovoked seizure in the ED.

DESIGN:

Prospective observational study

SETTINGS:

Tertiary care hospital

PATIENTS AND METHODS:

Patients presenting to the ED with seizure were divided into two groups according to medical history, physical examination, laboratory results, cranial computed tomography findings and electroencephalography results. Patients with seizures that did not have a specific cause (unprovoked) were compared with patients who had seizures caused by underlying pathology (provoked). The measurement of the ONSD was taken at the bedside within 30 minutes of arrival. The study compared the ONSD values, age, sex, type of seizure, and Glasgow Coma Score between the two groups.

MAIN OUTCOME MEASURE:

Efficacy of ONSD to distinguish between provoked and unprovoked seizures.

SAMPLE SIZE:

210 patients

RESULTS:

One hundred and fourteen (54.3%) patients were in the provoked seizure group and 96 (45.7%) were in the unprovoked seizure group. The ONSD measurements were significantly higher in the provoked seizure group compared with the unprovoked seizure group (median 6.1 mm vs. 5.2 mm, P<.001). The cut-off value of ONSD higher than 5.61 was significantly associated with the prediction of the provoked seizure (P<.001). The area under the curve value was 0.882 (95% CI: 0.830-0.922) with a sensitivity of 86.5 and specificity of 78.9%.

CONCLUSIONS:

Bedside ONSD measurement by means of ocular ultrasound is an effective method for differentiating provoked seizure from unprovoked seizure.

LIMITATIONS:

Statistical significance of age on ONSD and exclusion of pediatric patients.

CONFLICT OF INTEREST:

None.

INTRODUCTION

Seizures are defined as episodes of temporary neurological alteration brought about by over-excitation and hypersynchronization of neuronal activity.^1,2 Seizures are responsible for an estimated 1.6 million ED visits annually, and approximately 400 000 patients seek medical help for new-onset seizures in EDs.³ It is important to identify the cause, both to make a correct diagnosis and provide appropriate treatment, as well as to avoid unnecessary treatment of patients deemed to be at low risk of recurrence.⁴

Seizures can be grouped into two major categories: provoked and unprovoked. Acute provoked seizures, which are also called acute symptomatic seizures, are the first occurrence of a seizure in a given patient, usually caused by a neurological or systemic event triggering the seizure activity.⁵ This type of seizure occurs due to a variety of conditions, most commonly including cerebrovascular disease, infections, trauma, drug toxicity, life-threatening metabolic derangements, autoimmune/inflammatory conditions, and neoplastic diseases.⁶ Unprovoked seizures occur in the absence of any apparent trigger or acute cause and tend to be recurring in character. The most well-known example is epilepsy, which may be accompanied by medical and psychiatric illnesses.^1,2

It is essential to take a comprehensive history, perform a thorough physical examination and to order EEG and a brain imaging studies to make a distinction between patients with acute symptomatic seizures and those with the unprovoked variety.⁴ An emergency physician responsible for the management of a patient with a first seizure episode should primarily decide whether an underlying serious condition has caused the event.¹ Nevertheless, there are no studies evaluating the effectiveness of emergency physicians in differentiating between acute provoked and unprovoked seizures and thus making important decisions about ordering brain imaging and laboratory studies.⁶

Elevated intracranial pressure (ICP) may indirectly serve as a means of identifying the type of seizure. If left untreated, elevated ICP can accompany head trauma, central nervous system (CNS) tumors, hydrocephalus, hepatic encephalopathy, and impaired CNS venous outflow.⁷ Recently, a method that employs ONSD measurement with bedside ocular ultrasonography to quantify ICP has drawn widespread attention.^8,9 The rationale of measuring ONSD is based on the fact that the optic nerve is a direct continuation of the CNS, and thus is wrapped within the subarachnoid space, and directly reflects intracranial pressure. ONSD measurement by ultrasonography is noninvasive, easy-to-use, portable, and fast.¹⁰

The relationship between seizures and ICP has not been clearly identified and still remains controversial. But on the basis of the Monro-Kellie hypothesis and the molecular mechanisms of ICP and ictal activity, increased blood flow during seizures as a result of depletion of energy reserves are issues in the pathophysiology of ICP elevation during seizures.^11,12 Several reports have documented ICP elevation in patients with seizures and ICP elevation appears to be independent of the type of seizure (partial, generalized, convulsive, or non-convulsive); however, these studies on increased ICP have been mostly done with invasive methods.^13–15 To our knowledge there is no study evaluating ONSD in patients with seizures in the literature. In this study, we aimed to assess the efficacy of bedside ONSD measurement in differentiating provoked seizure from un-provoked seizure.

PATIENTS AND METHODS

This single-center, observational study was conducted prospectively between January 2018 and January 2021 at the University of Health Sciences Antalya Training and Research Hospital. The protocol of the study was approved by the ethics committee of the hospital. Patient consent was obtained from the patients who were eligible for the study. The study included patients older than 18 years who had an active seizure in the ED or who were admitted in the postictal period. Exclusion criteria were age under 18 years, eye and orbital diseases, optic nerve neoplasm, orbital mass, history of neurosurgical procedures, patients admitted with trauma and patients who refused to participate in the study. Patients were grouped by whether the seizure was provoked or unprovoked. Provoked seizures had an underlying cause, while there was no acute cause in the unprovoked patients. Demographic characteristics, seizure type (focal seizure, generalized tonic-clonic seizure, non-convulsive status epilepticus (NCSE) or status epilepticus (SE), metabolic profile, toxicological screen, imagining findings and patient outcomes were recorded.

All patients were examined with bedside ultrasonography per study protocol after they underwent a primary clinical assessment performed by the primary emergency physician. All ultrasonographic examinations were performed by four emergency physicians, who were practicing in the ED and had performed ultrasonography for a minimum of 4 years; all had certificates for performing both basic ultrasonography and advanced-level ultrasonography examinations including ONSD measurement. The principal investigator had 10 years, and the other three had at least 3 years of post-qualification experience. Sonographic examinations were performed within 30 minutes after ED presentation on patients who met the inclusion criteria and before the administration of antiseizure medications. The patient assumed a supine position and closed his/her eyelids, onto which a small amount of gel was spread. The optic disk was then spotted using a 7.5 MHz linear ultrasonographic probe (Mindray Medical, Xinsight, Germany). Both eyes were initially scanned in the vertical and horizontal planes through the eyelid, followed by the visualization of the optic disk. As described previously, ONSD was measured in the transverse and sagittal planes at a site located 3 mm proximally from the optic disk, by quantifying the width of the hypoechoic area as a surrogate marker of the optic nerve sheath (Figure 1). The measurements were performed in the coronal and longitudinal plane and two measurements were obtained for each plane. ONSD was recorded in millimeters (mm) and the median of the ONSD was calculated.

Figure 1. ONSD Measurement technique: transverse measurement of optic nerve at a site 3 mm proximal from the optic disk, with closed eyelids.

Initial cranial imaging studies were ordered independently by the first emergency physician who was responsible for the patient according to the indication that he/she determined within 15 to 20 minutes after ONSD measurements and they were reported by an independent radiologist who was unaware of the ONSD measurements. All patients consulted with the neurology department, and the final diagnosis was confirmed.

IBM SPSS Statistics for Windows, version 26.0 (IBM Corp., Armonk, N.Y., USA) was used for statistical analysis. For descriptive statistics, the mean and standard deviation are used to present continuous data with a normal distribution. The median with minimum-maximum values are used to present continuous variables without a normal distribution. Numbers and percentages are shown for categorical variables. The Shapiro-Wilk, Kolmogorov-Smirnov, and Anderson-Darling tests were used to analyze the distribution of the numerical variables. The Mann-Whitney U test was used to compare two independent groups where numerical variables were not normally distributed. A nonparametric covariance analysis using the “sm.ancova” package in R version 4.2.2 software (https://www.R-project.org/) was used to compare the effect of age on ONSD differences between the groups. The Pearson chi-square and Fisher exact tests were used to compare the differences between categorical variables in 2×2 tables. The Fisher Freeman Halton test was used in RxC tables. The receiver operating characteristic (ROC) analysis was done using the DeLong method with the Youden index to determine the ONSD cut-off value that predicts the development of unprovoked seizures. The area under the curve (AUC) and the corresponding 95% confidence interval (CI) were calculated. Level of significance (P) was set at less than or equal to .05.

RESULTS

Of the 210 patients who presented to the ED with seizure, 114 (54.3%) patients had unprovoked seizure and 96 (45.7%) patients had provoked seizure. The mean age of patients with provoked seizures was significantly greater than those with unprovoked seizures (60.3 [19.4] years vs. 49.8 [20.7] years, respectively, P<.001) (Table 1). Age also had an effect on the median ONSD difference between the groups (P<.001) as indicated by the graph of the regression model of age on ONSD (Figure 2). The two groups had similar sex distribution (P=.700). The unprovoked seizure patients had more neurological/psychiatric comorbidities (P=.001). Epilepsy was the most common disorder (79.4%) among patients with unprovoked seizures, while provoked seizures were most commonly related to Ischemic stroke (36.1%) and intracranial mass (26.2%). The differences in the rates of patients with epilepsy, intracranial mass, and Ischemic stroke were significant between the two groups (P<.001). There was a significant difference in the frequency of seizure types between the groups (P=.035). The proportion of patients with status epilepticus was significantly higher in patients with provoked than unprovoked seizures (P=.035). The median GCS was significantly lower in Group 2 than in Group 1 (13.0 vs. 15.0, P<.001). In patients with provoked seizures, neurological, metabolic, and toxic etiological factors were detected in 72 (75.0%), 18 (18.8%), and 6 (6.2%) patients (Table 2). Ischemic stroke was the most frequent neurological disease. Among the metabolic causes, hyponatremia was more common (66.7%) than other metabolic conditions.

Table 1.

Demographic and clinical characteristics of the patients with unprovoked and provoked seizures.

	Unprovoked seizure (n=114)	Provoked seizure (n=96)	P
Age (years)	47.0 (19.0-93.0)	63.5 (18.0-90.0)	<.001
Gender
Male	60 (52.6)	54 (56.2)	.700
Female	54 (47.4)	42 (43.8)
Neurological/psychiatric comorbidity	97 (85.1)	61 (63.5)	.001
Comorbidities
Epilepsy	77 (79.4)	10 (16.4)	<.001
Intracranial mass	0	16 (26.2)	<.001
Cerebrovascular disease	8 (8.2)	22 (36.1)	<.001
Intracranial bleeding	0	1 (1.6)	.210
Delirium	3 (3.1)	5 (8.2)	.306
Psychiatric diseases	3 (3.1)	2 (3.3)	.999
CNS infections	3 (3.1)	2 (3.3)	.999
Dementia-Parkinsonism	3 (3.1)	3 (4.9)	.999
Type of seizure
Focal	29 (25.4)	18 (18.8)	.035
Generalized	68 (59.6)	48 (50.0)
Non-convulsive	13 (11.4)	20 (20.8)
Status epilepticus	4 (3.5)	10 (10.4)
Glasgow Coma Scale score	15.0 (5.0-15.0)	13.0 (3.0-15.0)	<.001

Data are n (%) or median (minimum-maximum).

Figure 2. Nonparametric regression curves of age on optic nerve sheath diameter (1: unprovoked seizure, 2: provoked seizure).

Table 2.

Etiological factors for patients with provoked seizures (n=96).

Neurological etiologies	72 (75.0)
Spontaneous intracranial hemorrhage	13 (18.1)
Ischemic stroke	26 (36.1)
CNS infection	10 (13.9)
Intracranial mass/tumor	16 (22.2)
Hypertensive encephalopathy	1 (1.4)
Arteriovenous malformation	1 (1.4)
Cerebral venous sinus thrombosis	3 (4.2)
Posterior reversible encephalopathy syndrome	2 (2.8)
Metabolic etiologies	18 (18.8)
Hyponatremia	12 (66.7)
Hypoglycemia	2 (11.1)
Diabetic ketoacidosis	3 (16.7)
Hepatic encephalopathy	1 (5.6)
Toxic etiologies	6 (6.2)
Substance use	3 (50.0)
Alcohol withdrawal syndrome	1 (16.7)
Methanol intoxication	1 (16.7)
Lithium intoxication	1 (16.7)

Data are n (%).

The unprovoked seizure group had significantly higher lactate levels than the provoked seizure group (median 4.2 vs. 2.3, P<.001) (Table 3). Other laboratory parameters were similar between the groups. Pathological findings on cranial CT were present in 64 patients (30.5%). In decreasing order, spontaneous intracranial hemorrhage (ICH), intracranial mass/tumor, cerebral infarct, and cerebral edema were detected in the provoked seizure group (Table 3). The ONSD measurements were significantly higher in the provoked seizure group than in the unprovoked seizure group (median 6.1 mm vs. 5.2 mm, P>.001).

Table 3.

Comparison of laboratory and imaging parameters between groups.

	Unprovoked seizure (n=114)	Provoked seizure (n=96)	P
Sodium (mEqL)	138.0 (131.0-146.0)	137.0 (112.0-142.0)	.070
Glucose (mg/dL)	92.0 (71.0-411.0)	91.0 (40.0-550.0)	.694
Lactate (mmol/L)	2.3 (0.8-21.4)	4.2 (1.2-10.0)	<.001
pH	7.3 (6.8-7.5)	7.3 (7.1-7.5)	.570
pO2 (torr)	50.0 (31.0-99.0)	50.0 (23.0-99.0)	.565
pCO2 (mmol/L)	40.0 (21.0-76.0)	40.0 (23.0-56.0)	.335
HCO3 (mEq/L)	23.0 (9.0-29.0)	23.0 (13.0-28.0)	.055
Computed tomography
Normal	114 (100)	32 (33.3)	<.001
Pathological	0	64 (66.7)
Cerebral infarct	0	15 (23.4)
Intracranial hemorrhage	0	20 (31.2)
Intracranial tumor	0	16 (25.0)
Cerebral edema	0	13 (20.3)
Optic nerve sheath diameter (mm)	5.2 (4.5-6.7)	6.1 (5.0-6.8)	<.001

Data are n (%) or median (minimum-maximum).

A ROC curve indicated that provoked seizure could be predicted with a high predictive power when ONSD was greater than the cut-off value of 5.6 (P<.001) (Figure 3). This value had an area under the curve (AUC) of 0.882 (95% CI: 0.830-0.922), a sensitivity of 86.5% and a specificity of 78.9% for the differentiation of provoked and unprovoked seizure (Table 4). The overall accuracy for this model was 82.4%.

Figure 3. Receiver operating characteristics curve of optic nerve sheath diameter measurements as predictor provoked seizures.

Table 4.

The predictive power of the optic nerve sheath diameter for differentiating provoked seizure from unprovoked seizure.

	Provoked Seizure	Unprovoked Seizure	Sensitivity (95% CI)	Specificity (95% CI)	Accuracy	PPV	NPV
Optic nerve sheath diameter
High (>5.61 mm) (n=107)	83 (86.5)	24 (21.1)	86.5% (78.0-92.6)	78.9% (70.3-86.0)	82.4%	77.6%	87.4%
Low (≤5.61 mm) (n=103)	13 (13.5)	90 (78.9)

Data are n (%).

The rate of discharge from the ED was significantly higher in the unprovoked than the provoked seizure patients (75.4% vs 17.7%; P<.001). The two groups had similar mortality rates (P=.208). Patients with provoked seizures were hospitalized more often (P<.001) (Table 5).

Table 5.

Outcomes by seizure groups.

	Unprovoked seizure (n=114)	Provoked seizure (n=96)	P
Current status
Discharge from ED	86 (75.4)	17 (17.7)	<.001
Hospitalization in clinical wards	24 (21.1)	50 (52.1)
Transfer to an ICU	4 (3.5)	27 (28.1)
Mortality	0	2 (2.1)	.208

Data are n (%).

DISCUSSION

Previous research has examined the ability of increased optic nerve sheath diameter (ONSD) to predict increased intracranial pressure (ICP) and a possible connection between seizures and ICP elevation has been established. However, the potential use of ONSD in distinguishing between provoked and unprovoked seizures has not yet been explored.^11,15 Our study supports previous findings that measuring ONSD can be an effective way to evaluate ICP in intracranial disorders. However, our main focus was not to repeat this finding, but to investigate the predictive value of ONSD in identifying the type of seizure, which has not been previously studied.

Measuring ONSD involves some degree of variation between individuals and researchers, making it challenging to establish a clear cut-off point that differentiates a normal and abnormal ONSD, and to develop a general equation linking ONSD and ICP. There are conflicting results in the literature about the role of operator in measuring ONSD by ultrasonography; some studies have found that the technique is effective and accurate, even when used by inexperienced operators,¹⁶ while others have suggested that even when used by skilled operators, ultrasound measurement of ONSD may not be a reliable tool, despite good inter-rater reliability.¹⁷ Despite these challenges, our study found a significant difference in median ONSD values between the study groups (P<.001). A cut-off value of 5.61 mm or higher for ONSD had a sensitivity of 86.5% and a specificity of 78.9% for identifying provoked seizures.^18,19 This suggests that a higher value for ONSD increases the likelihood of provoked seizures and can aid in the differential diagnosis for emergency physicians.

The severity and associated risks of seizure activity are related to the formation and spread of discharges, which may not be easily identified through standard visual inspection and pattern classification.¹⁸ Therefore, it is crucial for emergency physicians to accurately predict and appropriately treat such serious clinical events in the emergency department. Some observational studies have shown that seizures negatively impact quality of life and increase mortality compared to the general population. The severity of seizures can vary and is dependent on various factors, such as seizure type, underlying pathology, comorbid conditions, and the timing and accuracy of the treatment plan.^20,21 In our study, the ONSD value increased from simple seizures to complicated seizures, which suggests that ONSD can be used to predict seizure activity caused by potentially serious underlying conditions or serious, potentially fatal seizure types such as status epilepticus; these findings are consistent with previous literature.^12,22,23 These findings support the idea that ONSD measurement could be used to diagnose nonconvulsive status epilepticus and to evaluate the response to therapies aimed at reducing ICP in comatose patients admitted to the emergency department.

There is limited information in the literature to guide the decision-making process for admitting patients with a first seizure episode.¹ It has been proposed that abnormal imaging signs, long-lasting or recurring episodes, or failure to fully stabilize a patient may lead physicians to admit patients to manage medical emergencies and perform a neurological evaluation.^24–26 In our study, there was a significant difference between the patients discharged from the emergency department and those admitted to the hospital. This finding supports the clinical use of ONSD in patients with seizures as a prognostic and predictive tool in the emergency department. Another important point to note is the statistical significance of age between the groups on ONSD in our study. Although there are conflicting results about the effects of age on ONSD in the literature,^26,27 the statistical significance of age on ONSD in our regression analysis and exclusion of children are limitations of our study and should be validated by further research.

In conclusion, measuring ONSD using ultrasonography may provide a noninvasive, inexpensive, and potentially useful tool for determining ICP increase and cerebral disorders in patients with seizures. However, there are conflicting results about the dependence of the technique on operator experience.²⁸ As ONSD increases, the likelihood of abnormal brain imaging study results and worse clinical outcomes also increases. A baseline ONSD measurement can be used to differentiate between provoked and un-provoked seizures among patients presenting to the emergency department with seizures. The conflicting results about operator dependence should be further examined by comparing ultrasonography measurements with standard measurements of ONSD using brain CT or MRI in future studies.

Funding Statement

Funding: None.