LEARNING OBJECTIVES

Introduction

Continuous electroencephalography (cEEG) allows for the real-time monitoring of cerebral function in critically ill children. Since the advent of digital electroencephalography, the use of cEEG monitoring in the cardiac and pediatric intensive care unit (ICU) has increased exponentially. Numerous studies have focused on the incidence and characteristics of electrographic seizures in critically ill children. However, cEEG can also provide important information about other aspects of cerebral function, for example, the depth of coma, the presence of focal pathology, or the prognosis after cardiac arrest. The ability to monitor cerebral function continuously and in real time is especially important in the ICU, where the neurologic examination is often limited by the presence of central nervous system dysfunction and/or by the administration of sedative and paralytic medications. Importantly, cEEG provides both temporal and spatial information. Temporal information can be used to monitor changes in cerebral function over time, while spatial information can provide clues to the presence of localized pathology, for example, focal ischemia or intracranial hemorrhage. Moreover, cEEG is noninvasive and can be performed at the bedside, thereby offering advantages over other modalities of neurologic monitoring that require invasive techniques or transport of the patient to a dedicated diagnostic suite (e.g., MRI). In addition, cEEG can provide valuable information about cerebral function across the spectrum of patient ages and conditions and has few limitations to its use.

Encephalopathy

Continuous EEG in the cardiac and pediatric ICU is frequently used to assess the degree of a patient’s encephalopathy and to trend changes over time.

The Definition of Encephalopathy

Consciousness was defined by Plum and Posner in their seminal work Diagnosis of Stupor and Coma as “the state of full awareness of the self and one’s relationship to the environment.”¹ Both primary central nervous system pathology and secondary brain dysfunction can cause encephalopathy, an alteration in a patient’s consciousness. In turn, encephalopathy can occur as either an acute or chronic disorder.

Historically, the degree of encephalopathy has been described clinically based on the patient’s level of alertness and behavioral responses during physical examination. The most famous encephalopathy grading scale based on clinical examination is that of Plum and Posner, detailed in Table 2.1.¹

Unfortunately, the terms in Table 2.1 are often used inconsistently. Therefore, a detailed description of a patient’s mental state is often the most reliable means of communicating the degree of encephalopathy to other providers. Scales that quantify a patient’s level of consciousness can also be used. The most commonly used scale for this purpose is the Glasgow Coma Scale (GCS) (Table 2.2).² The GCS was developed as a means of communicating the level of consciousness of adults with acute brain injury.² The GCS assesses eye opening, verbal response, and motor response and was first adapted for children in 1988 (Table 2.2).^2–5

Etiologies of Encephalopathy

Acute and chronic encephalopathies are common in the cardiac and pediatric ICU. They arise from a myriad of causes including, but not limited to, those presented in Table 2.3.

Standardized Terminology for the Critical Care EEG Background

The American Clinical Neurophysiology Society (ACNS) published guidelines for standardizing critical care EEG terminology and recommends using the descriptors and definitions in Table 2.4 when describing the background features of a critical care EEG.⁶

EEG Patterns in Encephalopathy

There is generally a correspondence between the degree of encephalopathy observed on physical examination and the EEG background. As a patient progresses from milder to more severe degrees of encephalopathy and vice versa, the EEG background changes accordingly (Figure 2.1). The mildest degree of encephalopathy is marked on EEG by slowing of the posterior dominant rhythm (PDR) (Figure 2.2). In normal children over the age of 8, the PDR is in the alpha (8–13 Hz) frequency range, while in younger children, the PDR is slower (<8 Hz) and increases with age (see Chapter 1). Mild encephalopathies are also marked by the intermixing of slower frequency activity with the PDR. As the encephalopathy progresses to moderate cerebral dysfunction, the posterior dominant rhythm is no longer seen, and there is a further increase in slow wave activity (Figure 2.3). There may be a loss of fast frequency activity (beta; >13 Hz) and a loss or attenuation of sleep features, including sleep spindles and K complexes. As the degree of encephalopathy progresses to severe, there is a loss of variability, state change, and reactivity (Figure 2.4). Eventually, the background becomes discontinuous (Figure 2.5) and a burst attenuation/burst suppression pattern (Figure 2.6) may ensue. The burst attenuation/burst suppression pattern is comprised of periodic bursts of polymorphic activity, often containing sharp features, separated by periods of voltage attenuation or suppression. Within the bursts, activity can be mixed in frequency and variable in morphology with slow waves, sharp waves, and spikes. The bursts may recur at regular or irregular intervals. Lower voltage bursts and longer and flatter interburst intervals correlate with worsening encephalopathy. With progressive cerebral dysfunction, the EEG may become suppressed (Figure 2.7). Generalized suppression is marked by widespread, persistent, nonreactive delta and theta activity with amplitude of less than 10 microvolts (µV). If encephalopathy progresses to brain death, the EEG is marked by electrocerebral silence (see the section Brain Death).^7–9

There is no standardized grading system for the severity of encephalopathy on EEG in either adults or children. The most commonly used grading scale in the pediatric critical care EEG literature categorizes the EEG background as normal, including sedated sleep; slow-disorganized; discontinuous-burst suppression; and attenuated-flat.¹⁰ This grading scale has been used primarily in studies of outcome after status epilepticus 38and cardiac arrest.^11–16 Other authors have developed grading scales on a per-study basis.^17–19 Our institution uses the scale shown in Table 2.5 to grade encephalopathy based on cEEG criteria, although it has not yet been validated.

GENERALIZED RHYTHMIC DELTA ACTIVITY

In addition to the above general progression of EEG changes, specific EEG patterns can be seen in the setting of encephalopathy. For example, generalized rhythmic delta activity (GRDA) can be seen in both children and adults with encephalopathy from various causes. Intermittent generalized rhythmic delta activity is typically seen in children who are awake but drowsy or mildly lethargic. It is characterized by rhythmic delta activity that recurs at irregular intervals on a background comprised of mild-to-moderate generalized theta slowing (Figure 2.8). It is typically bilaterally synchronous and attenuates with eye opening or alerting. It is most commonly seen over the frontal region; prior to the publication of the ACNS standardized critical care EEG terminology, it was referred to as frontal intermittent rhythmic delta activity (FIRDA). In children, intermittent generalized rhythmic delta activity can, alternatively, have an occipital predominance, previously referred to as occipital intermittent rhythmic delta (OIRDA). FIRDA has also been reported in association with increased intracranial pressure and deep midline lesions, and OIRDA can be seen in children with absence epilepsy. An additional form of intermittent rhythmic delta activity over the temporal region is seen in adults in association with temporal lobe epilepsy and is not commonly considered a marker of encephalopathy. Other types of generalized rhythmic delta activity are discussed in detail in Chapter 3.^8,9

REACTIVITY

An important feature of the EEG in encephalopathy is the presence or absence of reactivity, as this can serve as an additional indicator of the severity of encephalopathy. Reactivity also has prognostic implications in the setting of coma and other types of acute brain injury.^20–30 Reactivity manifests as a reproducible change in the frequency, amplitude, periodicity, rhythmicity, or other background EEG feature in response to stimulation (Figure 2.9). The ACNS recommends that testing for EEG reactivity should follow a protocol and that neurodiagnostic technologists should perform bedside testing for reactivity at least once a day.³¹ It is recommended that visual, auditory, tactile, and painful stimulation be performed;³¹ more specifically, stimulation can consist of passive eye opening and closure, a loud clap or other auditory stimulus, nostril tickle, tracheal suction, shaking of the shoulder or other light tactile stimulus, sternal rub, or painful stimulation of the nailbed, trapezius, or supraorbital nerve.³² In low voltage tracings, reactivity typically manifests as an increase in amplitude and rhythmicity, while in high voltage tracings, reactivity may manifest as a relative attenuation of the background.⁸ Alternatively, in a patient with mild encephalopathy, reactivity may manifest as a decrease in slowing.⁹ Unfortunately, interpretation of EEG reactivity may be unreliable, with interrater agreement varying between 26% and 66%.^33,34 In keeping with this, more than half of respondents to a recent survey of critical care electroencephalographers considered reactivity interpretation subjective but, nonetheless, over three-quarters of them considered reactivity testing often or always useful for prognostication after cardiac arrest.^8,9,32

Although reactivity is generally considered a positive prognostic sign, there is an exception to this in the form of stimulus-induced rhythmic, periodic, or ictal discharges (SIRPIDs).^35–37 SIRPIDs are periodic, rhythmic, or ictal-appearing 39discharges that are consistently induced by alerting stimuli.³⁵ They include periodic discharges, rhythmic delta activity, or unequivocally evolving electrographic seizures, and they can be focal or generalized.³⁵ SIRPIDs tend to be seen in patients with risk factors for a poor outcome but are not thought to be an independent predictor of a poor outcome.³⁷

FOCAL SLOWING, ATTENUATION, AND OTHER ASYMMETRIES

In addition to the above EEG patterns, which are typically seen in the setting of a diffuse encephalopathy marked by global cerebral dysfunction, encephalopathy can also result from a focal lesion, for example a large territory arterial ischemic infarct or mass lesion. In these cases, the EEG is often marked by focal abnormalities, including: focal slowing (Figure 2.10), which may be as subtle as a 1 Hz difference in the frequency of the PDR between hemispheres; focal attenuation of beta activity (defined as a persistent 35% difference in amplitude between homologous regions in the two hemispheres) (Figure 2.11) or activity of other frequencies (Figures 2.12 and 2.13); and spindle asymmetry (Figure 2.14) or asynchrony. In addition to focal cerebral pathology, patients in the intensive care unit may have focal fluid collections intervening between the cortical surface and the scalp electrode. Fluid collections may attenuate EEG activity, especially higher-frequency activity, but do not result in slowing in the absence of accompanying cortical injury. Focal enhancement of beta activity, which is typically seen as a breach rhythm in the setting of a skull defect and less commonly in the setting of focal CNS pathology, can also be seen.³⁸

SEIZURES AND RHYTHMIC AND PERIODIC PATTERNS

EEG rarely provides information as to the etiology of encephalopathy; however, there are a few notable exceptions. A particularly important clinical scenario in which the EEG provides information about the cause of encephalopathy is nonconvulsive state epilepticus (NCSE). NCSE can occur in the aftermath of convulsive status epilepticus or can arise de novo without preceding convulsive seizures. The epidemiology, risk factors, diagnosis, and management of nonconvulsive status epilepticus are discussed in detail in Chapter 5.

Generalized periodic discharges (GPDs) with triphasic morphology, formerly known as triphasic waves, can be seen in critically ill adult patients with encephalopathy (see Chapter 3). These have historically been associated with a metabolic or toxic cause of encephalopathy; however, a recent study in adults showed that GPDs without triphasic morphology had a higher association with metabolic and toxic disturbances than those with triphasic morphology and that GPDs with triphasic morphology were seen as frequently as GPDs without triphasic morphology in critically ill patients with seizures.³⁹ The epidemiology, risk factors, diagnosis, and management of rhythmic and periodic EEG patterns are discussed in detail in Chapter 3.

Burst suppression in the absence of cardiac arrest may also provide a clue as to the underlying cause of encephalopathy. Burst suppression may be rare in the absence of cardiac arrest but can be seen in the setting of overdose with medications such as baclofen, barbiturates, and bupropion.^40–43 As such, a burst suppression pattern on EEG without a clear precipitant should prompt further evaluation for a toxic ingestion.

Coma Patterns

Coma lies at the far end of the continuum of alteration of consciousness. A patient in coma should have an EEG background that lies at the severe to profound end of the spectrum of EEG patterns of encephalopathy described above. Alternatively, they may have an EEG pattern specific to coma, including alpha coma and spindle coma. These are more commonly seen in adults but can also rarely be seen in children. Patients with apparent clinical coma but a normal or only mildly abnormal EEG background should be evaluated for alternative explanations for their clinical state (e.g., locked-in syndrome, botulism or other peripheral nervous system disorder, etc.).

Alpha Coma

Alpha coma is more commonly described in adults and is marked by continuous, invariant, monomorphic 8 to 12 Hz activity. The alpha activity is typically low to moderate in amplitude and generalized, often with anterior predominance. It is not reactive to passive eye opening although may be reactive to other forms of stimulation; when accompanied by reactivity to stimulation, alpha coma is sometimes referred to as “incomplete.” Prognosis in the setting of alpha coma is dictated by the etiology of the coma, rather than the presence of alpha frequency activity per se. Alpha coma generally portends a poor prognosis after anoxic injury in adults.^44,45 However, alpha coma can also be seen in other settings, for example drug toxicity, in which case prognosis is generally good.⁴⁴ Alpha coma typically evolves within 3 to 17 days after injury into a pattern predictive of outcome;^44,46 alpha coma evolving to a pattern consistent with a mild or moderate degree of encephalopathy portends a good outcome, while alpha coma evolving to a burst suppression pattern portends a poor outcome.^44,46 The presence of reactivity in the setting of alpha coma is also predictive of outcome, although imperfectly so. In a series of 36 comatose patients in whom alpha frequency patterns were the predominant EEG pattern, 8 out of 15 patients with EEG reactivity awoke, whereas only 3 out of 19 patients without EEG reactivity awoke.⁴⁴ Patterns comprised of theta or theta-alpha activity that otherwise 40resemble alpha coma can also be seen; theta coma and theta-alpha coma have the same etiologies and outcomes as alpha coma and are considered variants of the alpha coma pattern.^45–48 Of note, the EEG in a locked-in state due to a lesion of the ventral pons appears similar to that of alpha coma; however, in the case of a locked-in state, the EEG is reactive to eye opening. Additionally, the alpha activity is predominantly posterior as compared to diffuse and anterior in anoxia.^9,49

Spindle Coma

Spindle coma, as described in adults, resembles sleep but the patient cannot be roused. The EEG is marked by a generalized slow wave pattern with persistent 9 to 14 Hz sleep spindles, often accompanied by vertex sharp waves and K-complexes. The EEG may be indistinguishable from non-REM sleep, although spindle activity is often exaggerated; in a series of 15 patients in spindle coma, spindle activity was seen in more than 75% of each 20-second epoch of a routine daytime EEG recording.⁵⁰ Each spindle has a discrete duration, and spindles are typically maximally expressed over the frontocentral regions. As in alpha coma, the EEG may or may not be reactive, and the presence of reactivity is predictive of outcome, although again imperfectly so.⁵⁰ If reactivity to noxious stimulation is seen, it should be brief, and the EEG should quickly revert to a spindle pattern.⁵⁰ This is in contrast to a patient in coma who may have spindles appear briefly or cyclically, which is much more common and can be seen in 14% to 67% of patients with posttraumatic coma.⁵⁰ In general, spindle coma has a better prognosis than other forms of coma, likely reflecting that spindle coma is typically seen in the setting of less severe cerebral damage than alpha coma and other malignant coma patterns.⁵⁰ Poor outcomes, however, do occur when spindle coma is due to an irreversible structural lesion.⁵⁰

Rhythmic Coma Patterns in Children

Rhythmic coma patterns, including alpha, theta, theta-alpha, and spindle coma have been described in children,^20,51–66 although less extensively than in adults. The electroencephalographic patterns are similar to those in adults (Figure 2.15); however, outcomes appear to be better than those in adults.⁶⁰ The frequency of the rhythmic pattern does not appear to determine outcome in pediatric patients;^60,62 it has therefore been proposed that rhythmic coma patterns in children represent a unified entity with a similar mechanism to alpha coma in adults but with a more variable expression in the developing brain.⁶⁰ As in adults, the etiology of coma is the primary determinant of outcome, and rhythmic coma patterns have been reported in pediatric patients with coma from a variety of causes, including cardiac arrest, encephalitis, traumatic brain injury, near drowning, brain tumors, stroke, metabolic derangements, sepsis, and drug toxicity.

Postanoxic Coma After Pediatric Cardiac Arrest

The EEG in postanoxic coma has received considerable attention, and numerous studies have explored prognostic features of the EEG background in relation to outcome after cardiac arrest in children. Over 10,000 children suffer from a cardiac arrest each year in the United States alone.^67–70 Survival rates range from 10% to 44%, and survivors often have poor neurologic outcomes that significantly impact their quality of life.^67–72 The ability to provide reliable prognostic information is important for counseling families and for guiding management decisions. Various clinical, laboratory, imaging, and neurophysiologic tools may be used in outcome prediction but rarely do these have perfect predictive value.^73–77 EEG is an especially appealing prognostic tool for pediatric patients as it is noninvasive and can be performed at the bedside. Older studies demonstrated poor prognosis in children in association with high voltage slowing, absence of reactivity, limited variability, epileptiform discharges, burst suppression, and periods of voltage attenuation/suppression.^20,78–86

More recently, to determine whether specific EEG features might be predictive of short-term outcome after cardiac arrest, Topjian et al.¹³ categorized the EEG background of 128 consecutive infants and children resuscitated after either in-hospital or out-of-hospital cardiac arrest as normal, slow-disorganized, discontinuous-burst suppression, or attenuated-flat. Worse EEG background categories were associated with progressively higher odds of death and unfavorable neurologic outcome (Pediatric Cerebral Performance Category [PCPC] of 3–6) at hospital discharge. However, EEG background alone was not a reliable predictor of outcome. While no patient with a normal background had an unfavorable outcome, five patients with an initial attenuated-flat background and seven patients with an initial discontinuous-burst suppression background had a favorable outcome. EEG reactivity was highly associated with EEG background category, and an absence of reactivity was associated with mortality and a worse neurologic outcome. The presence of status epilepticus was similarly associated with a worse short-term neurologic outcome but was not independently associated with mortality. Although not a reliable predictor alone, the addition of EEG data to models incorporating clinical variables resulted in modest improvements in the ability to accurately predict mortality and neurologic outcome at discharge.

41More recently, Fung et al.¹⁵ similarly categorized the EEG background of 89 consecutive infants and children resuscitated after either in-hospital or out-of-hospital cardiac arrest as normal, slow-disorganized, discontinuous-burst suppression, or attenuated-flat. Multiple additional EEG features, including frequency, symmetry, continuity, voltage, sleep features, reactivity, and variability, were also categorized. In univariate analysis, a worse EEG background category, a low level of faster frequencies, absence of continuity, lower voltage, absence of stage 2 sleep transients, absence of reactivity, absence of variability, and absence of reactivity and/or variability were associated with mortality and an unfavorable neurobehavioral outcome (PCPC of 4–6). A combinatorial model incorporating EEG background category, stage 2 sleep transients, and variability and/or reactivity optimally predicted mortality (area under the receiver operating characteristic curve [AUC] 0.82; 95% confidence interval 0.72–0.92) and an unfavorable neurobehavioral outcome (AUC 0.77; 95% confidence interval 0.67–0.86). Nonetheless, despite selecting probability cutoffs favoring specificity over sensitivity, the positive predictive value for mortality and unfavorable neurologic outcome was only 86%. This again suggests that EEG features alone are insufficient to predict unfavorable outcome. However, the addition of early EEG data to clinical data has been shown to improve prognostication accuracy by neurologists and intensivists after cardiac arrest.⁸⁷

Multiple additional studies have reported similar results. For example, Ostendorf et al.¹⁷ categorized the EEG background of 73 children resuscitated after either in-hospital or out-of-hospital cardiac arrest as normal, slow and organized (sleep architecture or posterior-dominant rhythm present), slow and disorganized, discontinuous, burst suppression, or suppression. In this study, an initial EEG obtained based on clinical indications within 12 hours of return of spontaneous circulation (ROSC) that was either normal or slow was associated with a good prognosis. Ducharme-Crevier et al.¹⁸ analyzed the EEG of 34 children resuscitated after either in- or out-of-hospital cardiac arrest and categorized sleep spindles within the first 24 hours after return of spontaneous circulation as absent, abnormal (abnormal or atypical morphology, frequency, symmetry, and scalp distribution), or normal. The presence of normal or abnormal sleep spindles was associated with a favorable neurologic outcome (PCPC Score of 1–2 or no change from baseline score 6 months post-arrest); 8/10 patients with a favorable outcome had normal or abnormal spindles during the first 24 hours after return of spontaneous circulation, whereas 2/24 patients with a poor outcome had normal or abnormal spindles during the first 24 hours after return of spontaneous circulation. Finally, Brooks et al.⁸⁸ reported an association between background suppression and poor outcome (increase in PCPC >1) at discharge for 21 children resuscitated after cardiac arrest and monitored with EEG; all patients (n = 8) with background suppression had a poor outcome.

As detailed below in the section Commonly Observed Medication Effects, medications used in the ICU, as well as drugs that are frequently associated with overdose, have a significant effect on the EEG background. It is important to take these effects into consideration when using the EEG as a prognostic tool after cardiac arrest.

Quantitative EEG (QEEG) can also be used to predict outcome after pediatric cardiac arrest.⁸⁹ QEEG uses mathematical algorithms to decompose the EEG signal into frequency and amplitude components that can then be quantified (see Chapter 9), thereby circumventing issues related to interrater reliability.⁹⁰ In a recent study assessing the ability of QEEG analysis to predict neurologic outcome after pediatric cardiac arrest, a multifactorial model of QEEG features had a positive predictive value for a favorable outcome of 0.79 and a negative predictive value for an unfavorable outcome of 0.8.⁸⁹ This suggests that, similar to the raw EEG trace, QEEG cannot be used in isolation to predict outcome after cardiac arrest.

Postanoxic Coma With Myoclonus

Postanoxic coma may be complicated by posthypoxic myoclonus (PHM) in as many as 16% to 22% of cardiac arrest survivors.^91–94 This has been extensively described in adults but only rarely in children;⁸⁸ this likely reflects an underreporting of this phenomenon as it is not uncommonly encountered after pediatric cardiac arrest (Figure 2.6c). PHM manifests as repetitive myoclonic movements of the eyelids, face, limbs, and/or trunk and may be generalized, focal, or multifocal.⁹⁵ PHM has historically been divided into two subtypes, myoclonic status epilepticus (MSE), also referred to as postanoxic myoclonus and status myoclonus,^93,96–98 and Lance-Adams syndrome (LAS).⁹⁵ Despite considerable attention in the literature, MSE is not consistently defined,⁹⁴ and descriptions of its duration, bodily distribution, and neurophysiologic correlates vary among studies.^{96,97,99–101} Moreover, there may be some overlap between MSE and LAS. In general, MSE manifests clinically as nearly continuous myoclonic jerking for at least 30 minutes that is typically generalized or multifocal.⁹⁵ It tends to onset within the first 72 hours after cardiac arrest and is thought to be subcortical in origin.^95,102 MSE can occur spontaneously or in response to stimulation.⁹⁵ The patient is usually comatose, and the mortality rate is as high as 83% to 100%.^{91–94,97,99,103,104} In a recent analysis of the semiology of MSE in 43 adult patients after cardiac arrest, Mikhaeil-Demo et al.¹⁰⁵ found that three patients with postanoxic MSE defined as persistent myoclonus for >30 minutes beginning within 3 days of arrest in a comatose patient had distal, asynchronous, and variable myoclonus; 27 patients had axial or axial and distal, asynchronous, and variable myoclonus; and 13 patients had axial, synchronous, and stereotyped myoclonus. Two of the three patients with 42distal, asynchronous, and variable myoclonus were able to follow commands at the time of hospital discharge, and 2 of the 27 patients with axial or axial and distal, asynchronous, and variable myoclonus were able to follow commands at hospital discharge. In contrast, all 13 patients with axial, synchronous, and stereotyped myoclonus died or had withdrawal of technical support. In contrast to MSE, LAS typically occurs once a patient has regained consciousness, is marked by intention and stimulus-induced myoclonus, and is thought to be due to death of Purkinje cells in the fastigial nucleus of the cerebellum.^95,106,107 It is possible that the patients with distal, asynchronous, and variable myoclonus in Mikhaeil-Demo’s study had an early onset of LAS.

EEG patterns seen in patients with MSE are variable and include burst suppression, intermittent or continuous spike-wave, generalized periodic discharges, and alpha coma.⁹⁵ In a recent analysis of the EEG features of 59 adult patients with posthypoxic myoclonus after cardiac arrest, Dhakar et al.⁹⁴ found that all patients who regained consciousness (n = 7) had a normal voltage background at all time points analyzed during the first 72 hours after ROSC (6, 12, 24, 48, and 72 hours). In contrast, the majority of patients who failed to recover consciousness had a burst suppression background during the first 24 hours after ROSC (n = 52, 59.3% of 52 at 6 hours, 78.6% at 12 hours, and 65.3% at 24 hours). There was no difference in the prevalence of electrographic seizures, periodic discharges, lateralized rhythmic delta activity, or sporadic epileptiform discharges between the two groups. Reactivity was more common in those who regained consciousness (4/7, 57.1% v. 5/52, 9.6%, p = .006). Only patients with multifocal myoclonus, defined as asynchronous involvement of >2 body parts, regained consciousness, and the EEG showed frequent, sporadic, low-amplitude midline-maximal spikes on a continuous background in 4/5 or generalized spike-wave discharges with a normal voltage background in 1/5 (two patients did not have video recordings of myoclonic movements available for review). In contrast, in patients who failed to recover consciousness, myoclonus appeared to be time-locked to highly epileptiform bursts in 24/34 patients and to blunt cortical bursts in 5/34 patients with video recordings available for review. Of the remaining patients who failed to regain consciousness, four had myoclonus that was not associated with any cortical activity and occurred during suppressed portions of the background. These are similar to the findings of an earlier study by Elmer et al.⁹³ In this study of adult survivors of cardiac arrest who underwent EEG monitoring, 69 developed early myoclonus. No patient with a suppression-burst background with high-amplitude polyspikes in lockstep with myoclonic jerks survived with a functionally favorable outcome (n = 48).⁹³ In contrast, a continuous background with narrow spike-wave discharges time-locked to myoclonus was less clearly associated with a poor prognosis (n = 8, 50% survival with all survivors having a favorable outcome defined as discharge to home or acute rehabilitation).⁹³ In contrast to MSE, LAS has been associated with spike or polyspike-wave discharges with a maximum at the vertex in only 1/3 of cases.⁹⁵ Similar to Mikhaeil-Demo et al.’s study, it is possible that the patients who recovered consciousness in the Dhakar et al. and Elmer et al. studies had an early onset of LAS. As such, prognosis in the setting of postanoxic myoclonus is nuanced and dependent on additional clinical and EEG features.

EEG Monitoring in the Detection of Cerebral Ischemia, Increased Intracranial Pressure, and Hepatic Encephalopathy

Cerebral Ischemia

In addition to providing information about the severity of acute brain injury, cEEG can aid in the identification of newly developing and ongoing cerebral ischemia, thereby providing opportunity for intervention before the development of irreversible injury (Table 2.6). The EEG reflects activity at cortical synapses, and cortical pyramidal neurons are exquisitely sensitive to reductions 43in cerebral blood flow. Changes in EEG activity are first seen when cerebral blood flow drops to approximately 25 to 35 mL/100 g/min.¹⁰⁸ Initially, there is a loss of fast activity, primarily in the 8 to 14 Hz range. Physiologically, there is a shift to anaerobic metabolism and an excess of neurotransmitter release. As cerebral blood flow continues to decrease to approximately 18 mL/100 g/min, an increase in slower frequency activity is seen, primarily in the 4 to 7 Hz range. This corresponds physiologically to an increasing degree of lactic acidosis and depletion of adenosine triphosphate (ATP). Below a cerebral blood flow of approximately 18 mL/100 g/min, there is a further increase in slowing on the EEG, primarily in the 1 to 4 Hz range. Physiologically, the sodium-potassium pumps that maintain the transmembrane ionic gradients responsible for the conductive properties of neurons begin to fail and intracellular edema ensues. This represents a state of reversible ischemia. With further decreases in cerebral blood flow below 10 to 12 mL/100 g/min, focal or diffuse attenuation of all EEG frequencies is seen in association with calcium accumulation, anoxic depolarization, cell death, and infarction.

TABLE 2.6 THE RELATIONSHIP OF CEREBRAL BLOOD FLOW TO THE EEG BACKGROUND

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

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

Templates and Treatment Algorithms Cases Basics of EEG Monitoring in Pediatric Critical Care Questions and Answers by Chapter Seizures in Critically Ill Children Convulsive and Nonconvulsive Status Epilepticus in Critically Ill Children

Stay updated, free articles. Join our Telegram channel

Join