Short-Term Ambulatory EEG

CHAPTER 7

SHORT-TERM AMBULATORY EEG

JASON L. SIEGEL, MD and WILLIAM O. TATUM, IV, DO

INTRODUCTION

Ambulatory EEG (aEEG) is an essential tool in the management of patients with paroxysmal neurological disorders. Identifying both ictal and interictal abnormalities has direct effects on treatment, but the question for aEEG remains, “How long should the patient undergo the procedure?” Prolonging aEEG incurs higher cost, more patient inconvenience, and may be unnecessary, depending on the type of information the clinician is trying to obtain. Though aEEG can be performed for several days, short-term ambulatory EEG (ST-aEEG) can serve as a cost effective and accurate test.

We define ST-aEEG as an aEEG that lasts 24 hours or less. The benefit of a ST-aEEG is that it captures at least one night of sleep during prolonged aEEG recording. The earliest aEEG technology was limited by cassette tape recording, with no method to reformat or filter after the electroencephalography (EEG) had been recorded. There were also limitations in battery capacity, and therefore it was initially implemented for use over a 24-hour time period (1,2). Since then, these limitations have improved as technology has advanced, and ST-aEEG has more advanced applications and practical uses for the clinician.

RATIONALE FOR 24-HOUR MONITORING

Neurologists consider many factors when deciding on the most appropriate type of EEG for their patient. Important factors include diagnostic yield and duration of the recording time, but patient availability, environment, artifact burden, cost, accessibility, and reimbursement also merit consideration (3) (Table 7.1).

TABLE 7.1 Relative Comparison of EEG Methods Used in the Evaluation of Paroxysmal Episodes

EEG techniques vary, and the “best” test depends on what question the neurologist is trying to answer. The gold-standard method of event diagnosis, classification, and characterization is inpatient video-EEG monitoring. This method, however, is expensive, takes patients out of the natural environment in which they normally have events, and may not be accessible due to local expertise, insurance coverage, or geography. The most common EEG method is the 20- to 30-minute routine scalp EEG obtained as an outpatient. While readily accessible, inexpensive, easy to perform, and quickly interpretable, it has a lower yield when compared with long-term EEG monitoring. Furthermore, seizures are rarely captured during an outpatient routine scalp EEG recording. The added benefit from ST-aEEG has been demonstrated to provide a significantly higher yield of capturing a seizure during aEEG monitoring compared with a routine scalp EEG performed as an outpatient (4).

In order for ST-aEEG to be useful, there must be an abnormality that is readily detected in less than 24 hours. Patients with paroxysmal events that occur during sleep and those with frequent daily events are the best candidates for ST-aEEG (Figure 7.1). Other practical reasons for ST-aEEG include assessment of persistent epileptiform discharges (EDs) prior to considering a trial of weaning antiseizure drugs (ASDs) from patients with prolonged seizure freedom. One practical concern involves addressing the EEG prior to considering a trial of tapering ASDs before determining if a patient is an acceptable risk to operate a motor vehicle and be released to driving.

FIGURE 7.1 Repetitive right temporal sharp waves present on CAA-EEG during sleep.

The yield of 24-hour continuous EEG (cEEG) monitoring varies depending on the patient population. In critically ill patients, cEEG monitoring found that seizures were detected within the first 24 hours of cEEG monitoring in 88% of all patients who would eventually have seizures (5). Additional hospital-based studies have shown that within 30 minutes, EEG in the intensive care unit (ICU) identifies 2% to 3% of patients who had seizures, with 18% to 34% of patients having epileptic discharges (6). Extending the EEG to 16 to 24 hours, 14% of patients had new or additional EDs and 6% of patients had new or additional seizures (6). These patients are critically ill and may have multiple reasons to have provoked seizures, limiting the generalizability of these studies to patients in the home, ambulatory setting. Similar to critically ill patients, however, the yield of identifying seizures is greatest for ambulatory patients within the first 24 hours of recording (7).

IDENTIFICATION OF SEIZURES

The earliest studies of aEEG were conducted with four channels without video capabilities. Despite using limited channels during aEEG, some studies found a high concordance with respect to recording seizures (1). Though ST-aEEG can reliably identify electrographic seizures (8,9), if used routinely to identify ictal events it has a low probability to detect an event in a given 24-hour period. When looked at over a 5-day period, one aEEG study found that 58% of typical ictal events (seizure or nonepileptic attack [NEA]) occurred within 24 hours, which improved to 78% after 72 hours and reached 100% by 96 hours (8). Still, 58% of ST-aEEG demonstrated efficacy in the majority of patients.

In patients without a known diagnosis of epilepsy, events occur less frequently. In this population, 8.5% to 17.4% had epileptiform abnormalities or seizures within 24 hours (4,9,10). There was an even lower yield (5.1%) in aEEGs when they were ordered by nonneurologists. Also of concern was that not all clinical events that were suspicious for seizures had EEG changes (4,9,10). The low yield raises a question of effectiveness when ordering an ST-aEEG as a diagnostic tool in the evaluation process of an undiagnosed patient.

Children pose unique difficulties in completing even a 24-hour aEEG due to movement and intolerance of wearing the EEG. In children with clinically definite seizures, a 24-hour aEEG identified seizures in 55% of epilepsy patients with a clinical event, and 95.5% of these events had some detectable ictal EEG changes (11). Though the electrographic evidence of ictal activity had high concordance with the clinical episodes, the efficacy of the overall utility for patients with infrequent events is dubious if only 55% had conclusive results after an ST-aEEG. However, absence seizures recur multiple times daily and may be so brief that the patient is unable to quantify them and remain unaware of them (7,11). Quantifying seizures through ST-aEEG in this case may be helpful in optimizing ASD therapy.

Observing electrographic ictal changes on EEG is the gold standard in differentiating epileptic seizures from NEAs. Most NEAs occur during daytime, waking hours. The frequency of NEAs is similar compared to epileptic seizure events. Only about half of NEAs are identified within 24 hours, which improves to nearly 100% by 96 hours (7,11).

Despite the low occurrence of seizures over a given 24-hour period, ST-aEEG may play a significant clinical role in being able to identify subclinical focal seizures when patients report that they are seizure-free. In one large study of 502 patients evaluated with computer-assisted aEEG (CAA-EEG), about 23.4% of events were unidentified by the patient via push-button activation or bystander observation, though they were detected by the computer algorithm utilized for seizure detection (10). Thus, for quantification, CAA-EEG demonstrated significant benefit for seizures without awareness (Case Report on page 124).

An ST-aEEG may also be used to classify the type of epilepsy. Seizures have a focal or generalized onset suspect based on interictal or ictal abnormalities identified on ST-aEEG (Figure 7.2). This is essential for providing electroclinical characterization of focal seizures for treatment. Patients with epilepsy will have a range of seizure frequencies. Rarely, patients have an event within a 24-hour period though this is highly dependent on the seizure type and epilepsy syndrome. Therefore, ST-aEEG may be an effective tool in diagnosing “spells” or “events,” classifying seizures for treatment, and quantifying seizures (such as absence and myoclonic seizures).

FIGURE 7.2 CAA-EEG with an event of “staring” identified by a caretaker during an absence seizure.

IDENTIFICATION OF INTERICTAL EDs

The presence of interictal epileptiform discharges (IEDs) in the aEEG serves as an indirect biomarker for seizures and is the hallmark of epilepsy. Neurologists may be able to arrive at conclusions about patients’ events by identifying IEDs on ST-aEEG. Recognizing normal variations, benign variants, and artifact is the first crucial step to distinguish abnormal IEDs given the ability of some waveforms on EEG to mimic IEDs (12). Often confused with abnormal IEDs, normal fluctuations of background activity during light sleep and benign variants (such as wicket spikes) may prove to be a challenge when interpreting EEG (13). In addition, the presence of a “spike” due to artifact may be misinterpreted as abnormal and lead to misdiagnosis and mistreatment (14). After excluding normal waveforms and artifact on aEEG, overall the concordance with routine scalp EEG in detecting epileptiform abnormality is about 77%: 79% of focal IEDs and 100% of generalized IEDs (1).

As opposed to the low yield of recording seizures during a 24-hour period of ST-aEEG monitoring, recording IEDs occurs more frequently. The mean latency to detection of the first IEDs on ST-aEEG is under 6 hours. Up to approximately one-half of patients (range 24%–45.5%) will have an appearance of IEDs within 20 minutes (15), where a wide range of recovery has been found within 24 hours ranging from 33% to 85% (7,15,16). Patients who reported paroxysmal events at least once a week have been demonstrated to have a higher yield of diagnostic cEEGs compared with those who have more infrequent episodes (16). Generalized epilepsies have shorter latency for recording an IED compared with those with focal epilepsy, with the median latency for recovering focal IEDs occurring at a mean of about 4.5 hours, and the latency of recovering generalized IEDs within 1 hour (7).

Only gold members can continue reading. Log In or Register to continue