Screening electroencephalograms are feasible in the emergency department

Abstract

We investigated the feasibility and quality of a wireless, four-channel screening electroencephalogram (EEG) device on patients presenting to the emergency department (ED) with a possible seizure disorder. A convenience sample was used of ED patients presenting with a preliminary diagnosis of syncope, potential partial-complex or generalized seizure disorder, head injury with prolonged symptoms or acute undiagnosed altered mental status. Study patients had a screening EEG in the ED, but the emergency physician and patient were blinded to the results of the EEG so that neither patient care nor disposition were affected by inclusion in the study. A total of 227 patients were enrolled, with a mean age of 56 years. EEG quality was acceptable, i.e. a screening interpretation was able to be provided, in 208 of 227 cases (92%). The EEG interpretation was normal in 65%, identified generalized or focal slowing in 24% and identified sub-clinical epileptiform activity in 12% of patients. Screening EEGs performed in the ED are feasible, can be acquired with acceptable quality, and may identify sub-clinical seizure activity in a significant number of patients.

Introduction

Seizure is a common presentation in the emergency department (ED), representing at least 1% of all ED visits.^1,2 However, a seizure itself is often a secondary manifestation of a broad range of aetiologies.³ The diversity of presenting complaints makes it difficult to diagnose the underlying disease.⁴ One of the primary challenges in diagnosing a seizure or seizure disorder is that on arrival many patients no longer manifest the same symptoms which prompted their presentation to the ED. In these cases, an electroencephalogram (EEG) is recommended for further evaluation.⁵

Nevertheless, EEGs are typically not performed in the ED for many reasons.⁶ The equipment is bulky, which makes it inconvenient to store in many EDs, and is relatively expensive, costing $20--40,000. Also, the time and expertise required to perform an EEG is often lacking in the ED. Finally, ED physicians are not trained to interpret EEGs, and neurologists may not be immediately available to read an EEG performed in the ED.^7,8 Therefore, patients in whom an EEG has been recommended may not receive a diagnosis and treatment for several days. Moreover, these patients are often placed on anticonvulsant medications that they may or may not need during this time.⁹

In non-ED settings, tele-EEG has been described as “a feasible, secure, timely and effective method of providing an EEG service to hospitals”.¹⁰ The experience in institutions which have adopted tele-EEG services has been positive, for both the patients and the local clinical practitioners.¹¹ However, tele-EEG has not been previously used in the ED, where patients are acutely ill and constantly on the move.

We have investigated the feasibility and quality of a wireless, four-channel screening EEG device on patients presenting to the emergency department with a possible seizure disorder.

Methods

We conducted a prospective observational study on a convenience sample of patients presenting to a suburban community hospital with a workload of 70,000 ED patients/year. The study was approved by the appropriate ethics committee.

Adult patients (age ≥ 18 years) who met the inclusion/exclusion criteria were eligible for enrolment into the study (Table 1). The inclusion/exclusion criteria were designed to encompass a broad range of chief complaints which could present as a seizure, while excluding those patients with active convulsions in whom the information provided by the screening EEG would not be clinically relevant to the emergency physician.

Table 1.

Inclusion and exclusion criteria (adult patients only).

Inclusion	Exclusion
1. Patients with known seizure and prolonged post-ictal mental status change	1. Patients actively convulsing
2. Patients with status epilepticus who have received a muscle relaxant for intubation to determine presence of subclinical seizure	2. Patients who are medically or surgically unstable
3. Patients with brief alteration of mental status of unknown origin (e.g. “spells”, syncope, “blackouts”)	3. Patients who are unable to consent, and with family members unable to provide consent for them
4. Patients with behavioural changes that may indicate non-convulsive seizures	4. Patients with head injury incompatible with use of EEG (e.g. gunshots, severe scalp abrasions)
5. Patients with acute head injury and mental status changes that may indicate non-convulsive seizures
6. Patients with previous head injury and new onset mental status changes
7. Patient with neurological examinations that may be consistent with focal or partial non-convulsive seizure (e.g. aphasia, Todd's paralysis)

After providing informed consent, patients received a screening EEG. The monitor was a portable, wireless, four-channel EEG recorder (Crystal Monitor, Cleveland Medical Devices, Inc. Cleveland, OH). A four-channel coma montage (Fp1-C3, Fp2-C4, C3-O1, C4-O2, GRND FpZ) was used to reduce electrode set-up time. The wireless telemetry allowed the patient to be moved around the ED freely while monitoring was in progress, see Figure 1. The recording process involved the placement of seven gold-cup electrodes with electroconductive paste. The EEG was performed by study coordinators and data were collected for 20 min. A data collection form was completed at the time of the EEG.

Figure 1.

Simulated patient with the portable EEG monitor attached.

After the screening EEG was completed, the data were digitally compressed in a lossless format, password encrypted and sent via email to the study neurologists. Text-paging was used to notify the study neurologists that an EEG was available for evaluation. The neurologist opened the file using the Clevemed software on a standard PC and provided an interpretation. One of three neurologists evaluated the quality of the EEG using a four-point scale (1 = poor quality/unacceptable for interpretation; 2 = fair quality/acceptable for interpretation; 3 = good quality/acceptable for interpretation; 4 = excellent quality/acceptable for interpretation). Patients were followed to obtain their discharge diagnosis (either from the ED, or in the case of an admission, from the hospital). The emergency physician and patient were blinded to the results of the EEG. Therefore, neither patient care nor disposition were affected by inclusion in the study.

The primary outcome measures were: (1) the ability to capture a 20-min screening EEG on an ED patient; and (2) the quality of the EEG as reported by the study neurologist. Secondary outcome measures included the demographic information and final disposition of the patient, as well as any presence of sub-clinical epileptogenic activity. Data were analysed using descriptive statistics and confidence intervals were calculated using the modified-Wald method, where applicable.

Results

A total of 227 patients were enrolled in the study. A screening EEG was completed, transmitted and evaluated in 225 patients, or 99% of the total. In two patients, the EEG could not be evaluated because of a software error in the monitor. Of the 227 patients, 108 (48%) were female with a mean age of 56 years (range 18–97). The racial background of the patients was: Caucasian, non-Hispanic 88%, African American 8%, Caucasian, Middle Eastern 3%, Caucasian, Hispanic 0.8% and Asian 0.4%. A past medical history of a previous seizure was reported in 30% of patients, while 15% of patients were taking anticonvulsant medications at the time of enrolment. Most patients (72%) had a head CT scan performed in the ED, and in ten of them an acute abnormality was identified.

The 227 patients were enrolled for the following reasons: 68% for a witnessed or suspected seizure disorder, 23% for syncope, 8% for altered mental status without an identifiable cause and 1% for head injury with prolonged symptoms. The final disposition of the patients from the ED was: 144 hospital admissions, 77 ED discharges, 3 ED observations, 2 transfers to a tertiary care facility and one patient discharged against medical advice.

EEG quality was acceptable, i.e. a screening interpretation was able to be provided, in 208 of 227 cases (92%) (95% CI: 87, 95), see Table 2. The screening EEG results in the 208 patients were: normal in 65%, generalized or focal slowing in 24% and identified subclinical epileptogenic foci in 12%. Of the 24 patients with active subclinical epileptogenic foci identified on the screening EEG, nine (38%) were discharged home from the ED. One patient with a known brain mass and new vasogenic oedema found on head CT was transferred to a tertiary care facility. The other 14 patients were admitted to the hospital for further work-up.

Table 2.

EEG quality.

	Number (% total)
1 = poor quality, unusable	19 (8)
2 = fair quality, acceptable	60 (26)
3 = good quality, acceptable	98 (43)
4 = excellent quality, acceptable	50 (22)
Total	227 (100)

Discussion

Emergency physicians commonly evaluate patients whose work-up requires technology outside the scope of emergency medicine, such as endoscopy or cardiac catheterization. Telephone consultation with specialists, observation medicine and admission for further testing are often used in such scenarios. Our study is the first attempt at tele-EEG screening in ED patients who have suffered a suspected seizure, i.e. use of a screening EEG performed in the ED interpreted by an off-site neurologist.

Our primary objective was to discover whether this type of screening was feasible in the ED environment of a busy community hospital. In 99% of attempts we were able to transmit a screening EEG to a study neurologist for interpretation. Only 8% of the transmitted EEGs were described as uninterpretable, mainly due to muscular artefacts and gaps in the data from interference with other ED telemetry. The identification of the best wireless frequency for telemetery in the ED requires further study.

Limitations

Our study had certain limitations. First, we did not compare our screening EEG with a standard EEG. There are many instances when a screening EEG and a conventional inpatient EEG may be discordant. For example, a screening EEG which identifies the patient to have generalized or focal slowing in the ED secondary to medications (i.e. benzodiazepines) or being post-ictal, may very well be ‘normal’ when a subsequent inpatient EEG is performed. Many studies have shown that patients who are clinically identified as having a seizure disorder may require multiple EEGs before the epileptogenic foci can be identified. Second, we did not evaluate outcomes since it was a feasibility study only. Third, the portable EEG monitor we trialled used gold-cup electrodes and electroconductive paste, which take longer to apply than stick-on EEG electrodes. Despite an average setup time of 8 min, this did not cause problems in the study. Finally, there were three neurologists who participated in the study initially. Unfortunately, one of the neurologists dropped out after two months, for non-study related reasons. Subsequently most cases were split between the remaining two neurologists, resulting in the potential for bias when ascertaining the interpretability of the EEGs.

In conclusion, screening EEGs performed in the ED are feasible. Our study shows that they are of acceptable quality and may provide additional diagnostic information in certain ED patients.

Footnotes

Acknowledgements

The study was funded by National Institutes of Health grant PA-04-006.

References

Krumholz

Grufferman

Orr

Stern

. Seizures and seizure care in an emergency department. Epilepsia 1989; 30: 175–81.

Huff

Morris

Kothari

. Emergency department management of patients with seizures: a multicenter study. Acad Emerg Med 2001; 8: 622–8.

ACEP Clinical Policies Committee; Clinical Policies Subcommittee on Seizures. Clinical policy: critical issues in the evaluation and management of adult patients presenting to the emergency department with seizures. Ann Emerg Med 2004; 43: 605–25.

Scheepers

Clough

Pickles

. The misdiagnosis of epilepsy: findings of a population study. Seizure 1998; 7: 403–6.

Krumholz

Wiebe

Gronseth

. Practice parameter: evaluating an apparent unprovoked first seizure in adults (an evidence-based review): report of the Quality Standards Subcommittee of the American Academy of Neurology and the American Epilepsy Society. Neurology 2007; 69: 1996–2007.

Quigg

Shneker

Domer

. Current practice in administration and clinical criteria of emergent EEG. J Clin Neurophysiol 2001; 18: 162–5.

Tartara

Moglia

Manni

Corbellini

. EEG findings and sleep deprivation. Eur Neurol 1980; 19: 330–4.

Roupakiotis

Gatzonis

Triantafyllou

. The usefulness of sleep and sleep deprivation as activating methods in electroencephalographic recording: contribution to a long-standing discussion. Seizure 2000; 9: 580–4.

Perucca

Carter

Vahle

Gilliam

. Adverse antiepileptic drug effects: toward a clinically and neurobiologically relevant taxonomy. Neurology 2009; 72: 1223–9.

10.

Coates

Clarke

Davison

Patterson

. Tele-EEG in the UK: a report of over 1,000 patients. J Telemed Telecare 2012; 18: 243–6.

11.

Lasierra

Alesanco

Campos

Caudevilla

Fernández

García

. Experience of a real-time tele-EEG service. Conf Proc IEEE Eng Med Biol Soc 2009; 2009: 5211–4.