EEG Monitoring in the Medical ICU PDF

14 EEG Monitoring in the Medical ICU Monica B. Dhakar, Stephen Hantus, and Emily J. Gilmore IN THIS CHAPTER Incidence of seizures in critically ill patients without an acute neurological illness...

14 EEG Monitoring in the Medical ICU Monica B. Dhakar, Stephen Hantus, and Emily J. Gilmore IN THIS CHAPTER Incidence of seizures in critically ill patients without an acute neurological illness Use of cEEG for evaluation of encephalopathy in the medical ICU Technical considerations for cEEG outside of the neurological ICU KEY POINTS It has been well documented that nonconvulsive seizures are common in critically ill patients with primary neurologic insults, but seizures are also frequently seen in patients with systemic illness in the absence of acute brain injury. Clinical seizures before initiation of continuous EEG (cEEG) and history of a prior neurological insult are the strongest predictors of nonconvulsive seizures in this population. Sepsis has been shown to be an independent predictor of seizures in the medical ICU, the majority of which are nonconvulsive. cEEG can also be useful for evaluating depth of encephalopathy in medically ill patients. cEEG monitoring outside of the neurological ICU can provide important diagnostic and prognostic information regarding cerebral function, yet it is underutilized in these settings. I. BACKGROUND A. Continuous EEG (cEEG) monitoring has been utilized in critically ill patients without primary neurological illness for the same indications as in those with acute brain injury Detecting nonconvulsive seizures (NCS) in patients with altered sensorium Assessing the degree of encephalopathy Copyright Springer Publishing Company. All Rights Reserved. From: Handbook of ICU EEG Monitoring, Second Edition 123 DOI: 10.1891/9780826168627.0014 124 Part II Indications Characterizing abnormal involuntary movements such as myoclonus, tremors, and clonus to determine if they are epileptic in nature B. Altered mental status outside of the neuro ICU is common and is often associated with seizures There are many etiologies of altered mental status in patients with critical illness. The exact cause is often never discovered and is multifactorial in etiology. Similar to continuous cardiac and respiratory monitoring, an increasing number of patients are being monitored with cEEG, but its use remains limited to large institu- tions with adequate EEG resources and infrastructure. Estimates of altered mental status in the ICU range from 10% to 64% of patients (1). Common etiologies of altered mental status in the medical ICU (MICU): Sepsis is the most frequent cause. Metabolic encephalopathy often stems from liver or renal dysfunction but also includes other derangements such as hyponatremia, hypernatremia, hypoxia, and hyper- and hypoglycemia. Other primary systemic diseases can present as altered mental status from secondary neurologic injury and are also associated with seizures. Posterior reversible encephalopathy syndrome (PRES), secondary to malignant hypertension, medications (cyclosporine, tacrolimus), eclamp- sia, or metabolic disturbances (hypercalcemia, uremia). Embolic cerebral infarcts or mycotic aneurysms from endocarditis Central nervous system (CNS) vasculitis from an underlying autoimmune disease such as polyarteritis nodosa, systemic lupus erythematosus, and Sjogren’s syndrome. Paraneoplastic syndromes; limbic encephalitis Seizures are commonly associated with the previously mentioned conditions with an incidence of 10% to 16% (Figure 14.1) (2–5). Therefore, cEEG monitoring should be considered in any MICU patient with persistent or unexplained alteration in con- sciousness. A vast majority of these seizures tend to be nonconvulsive without any overt signs of ongoing seizures or clinically subtle findings. History of a prior neurologic injury and clinical seizures before initiation of cEEG are the strongest predictors of NCS in this population. Anoxic brain injury secondary to cardiac arrest is a common cause of severe encephalopathy as well as seizures in medical and cardiac ICU (see Chapters 11, 12, and 24). Patients admitted to the surgical ICU, particularly transplant and postoperative patients from major abdominal surgeries, are another population at risk for altered mental status and seizures. Causes of seizures include severe metabolic derangements, toxicity from immunosuppressive therapy, antibiotics, sepsis, and postoperative complica- tions. C. cEEG can also be utilized to assess the degree of encephalopathy in critically ill patients in MICU Sepsis-associated encephalopathy (SAE) is acute brain dysfunction owing to sys- temic inflammatory response to an infection as opposed to direct effects of infec- tious process. SAE has been associated with neuronal damage, mitochondrial and endothelial injury, and disturbances in neurotransmission (6).  Chapter 14 EEG Monitoring in the Medical ICU 125 20 Prevalence of non-convulsive seizures in percentage 18 SICU 16 14 12 MICU SICU + MICU MICU 10 % of seizures % of exclusively NCS 8 6 4 2 0 Oddo et al. Gilmore et al. Kamel et al. Kurtz et al. (2) (3) (5) (4) FIGURE 14.1 Prevalence of seizures in critically ill patients admitted to medical and surgical ICUs. All the previous studies excluded patients with acute primary neurologic injury. Incidence of NCS in MICU was found to be 10% (2) and 11% (3), in SICU = 16% (4), and when combined MICU and SICU was found to be 11% (5). MICU, medical ICU; NCS, nonconvulsive seizures; SICU, surgical ICU. II. BASICS A. Incidence and predictors of NCS in critically ill patients outside of the neuro ICU setting A retrospective study by Oddo et al. evaluated the risk of seizures in the MICU of a university hospital (2). Patients with a primary neurologic injury were excluded from the study. 60% of patients were septic and 48% were comatose at the time of the cEEG. 10% of 201 patients who had cEEG monitoring experienced seizures, of which 67% were nonconvulsive. Sepsis on ICU admission was the only independent predictor of seizures (Table 14.1). TABLE 14.1 Number of Consecutive MICU Patients Who Experienced NCS MICU ADMITTING ESZs OR ESZs ONLY PDs ONLY ESZs AND DIAGNOSIS PDs PDs Sepsis, n = 120 38 (32%) 11/120 (9%) 19 (16%) 8 (7%) No sepsis, n = 81 7 (9%) 0 (0%) 5 (6%) 2 (2%) ESZ, clinically silent electrographic seizure; MICU, medial ICU; NCS, nonconvulsive seizures; PD, periodic discharge. Source: From Oddo M, Carrera E, Claassen J, et al. Continuous electroencephalography in the medical intensive care unit. Crit Care Med. 2009;37:2051–2056. 126 Part II Indications NCS or periodic discharges (PDs) were found in 22% of patients. The additional presence of PDs in these patients was associated with a higher incidence of death or severe disability at hospital discharge. Aside from the association of poor outcome, the clinical significance of PDs is controversial. Unresolved issues include whether they represent a metabolic encephalop- athy (triphasic morphology), an ictal or potentially ictal pattern, their asso- ciation with acute structural injury, and how they should be treated (or not). Despite controversy over the appropriate management of PDs, their appearance in MICU patients appears to be associated with a negative prognosis. A prospective single-center study of patients admitted to MICU with sepsis exam- ined the impact of cEEG at time of discharge and 1-year follow-up (3). In 98 patients (100 episodes of sepsis), the rate of PDs was 25% and the rate of NCS was 11%, all of which were consistent with either definite or possible nonconvulsive status epilepticus (NCSE). Prior neurological injury and clinical seizures prior to cEEG were associated with increased risk of NCSs. In this study, lack of EEG reactivity was associated with increased 1-year mor- tality whereas there was no association between mortality and PDs or NCS. Interestingly, severely sick patients (nonneuro APACHE >25 or nonneuro SOFA >8) had significantly decreased risk of PDs or NCS. It was postulated that severe brain dysfunction may preclude the development of PDs and NCS, and thus patients with an intermediate degree of encephalopathy may be at great- est risk of NCS. One year survival was 35% and no patients developed new unprovoked sei- zures among the 17 patients who had completed follow-up at 1 year. A retrospective study of patients admitted to SICU evaluated the incidence and predictors of NCS in this population (4). A total of 154 consecutive patients with persistent unexplained alteration in consciousness were studied with cEEG for at least 12 hours. Majority of patients had abdominal surgery (65%). All patients with primary neurological injury and prior neurosurgical interventions were excluded. Seizures were observed in 16% of patients, all of which were exclusively NCS including 5% with NCSE; 29% had PDs. Coma and clinical seizures prior to cEEG were independent predictors of NCS in the multivariate analysis. B. Additional uses of cEEG monitoring: Assessment of cerebral function In the neuro ICU, cEEG can be used to detect early ischemia and changes in cere- bral blood flow (CBF). Studies of carotid endarterectomy have shown that EEG is a sensitive marker of changes in cerebral perfusion, and that decreases in EEG background fre- quency correlate well with the degree of ischemia (7). cEEG has been used to detect changes in CBF predominantly in the neuro- critical care setting in patients at risk for vasospasm secondary to aneurysmal subarachnoid hemorrhage (SAH) (8). Although patients with SAH are not often managed in the MICU, the fact that cEEG changes correlate with changes in CBF from vasospasm suggests its potential value for monitoring alterations in CBF from other disease states.  Chapter 14 EEG Monitoring in the Medical ICU 127 Example: Patients with septic shock being maintained on vasopressors who are at risk for diffuse cerebral hypoperfusion. Quantitative EEG, particularly alpha-delta ratio, which is a sensitive indicator of cerebral ischemia in SAH, may be utilized to assess the CBF due to systemic etiologies. EEG can also be used for noninvasive assessment of intracranial pressure (ICP) in patients who do not have external ventriculostomy or have contraindications for invasive monitoring, and may potentially alert physicians to the possibility of impending cerebral herniation (9,10). Rise in ICP leads to decrease in cerebral perfusion pressure resulting in decreased CBF and ischemia. As mentioned previously, progressive isch- emia results in slowing of EEG frequency, which can be used as a surro- gate marker of raised ICP. A recent study evaluated the use of quantitative power spectral analysis to monitor ICP in 62 patients with various neurologic disorders (10). It was found that “pressure index” (calculated as the inverse of median frequency × delta ratio) had a significant negative correlation with ICP (r = −0/849, p <.001) indicating that EEG may have the potential to be a reliable marker of ICP. cEEG can also be used to evaluate the degree of encephalopathy and assist in identifying possible causes. EEG is a more sensitive marker of encephalopathy compared to clinical exam. EEG abnormalities in encephalopathic patients fall along a spectrum ranging from diffuse polymorphic theta or delta slowing to low-amplitude delta activity with intermittent attenuation and finally burst-suppression pattern. The sever- ity of the EEG background abnormality usually correlates with the degree of encephalopathy. A caveat is that iatrogenic causes of encephalopathy (i.e., sedative medica- tions) can look electrographically similar to organic causes (11). Therefore, it is always important to interpret EEG findings in the context of what is clinically occurring with the patient. Periodic and rhythmic patterns may be seen in these patients, particularly triphasic waves (TWs), which were previously considered a universal marker of toxic or metabolic encephalopathy, particularly hepatic encephalopathy (Figure 14.2). TWs are now classified as a subset of generalized periodic discharges (GPDs) as per recent American Clinical Neurophysiology Society (ACNS) nomenclature. This was done to remove any clinical connotation while attempting to classify these patterns. A retrospective study by Foreman et al. demonstrated that GPDs with tri- phasic morphology is just as likely to be associated with seizures as GPDs without triphasic morphology (25 % vs. 26%) (12). Another recent study found that patients with TWs may improve upon administration of IV benzodiazepine (BZD) and nonsedating antiseizure drugs (ASDs) (13). Among 64 patients with TWs, a positive (clinical and EEG) response occurred in 10 of 53 (18.9%) treated with BZDs and in 19 of 45 (42.2%) patients treated with nonsedating ASDs (immediate in 6.7%, delayed but definite in 20%, and delayed but equivocal in 15.6%). There were no differences in metabolic profiles of responders and non- responders, suggesting that some patients with TWs may have an ASD- responsive condition. 128 Part II Indications FIGURE 14.2 EEG findings in the setting of liver failure and hyperammonemia. This is an EEG from a 51-year-old man with medical history of liver cirrhosis, due to hepatitis C and alcohol abuse, and esophageal varices who was admitted with an episode of tonic stiffening of all four extremities in the setting of GI bleeding. EEG demonstrates generalized periodic discharges with triphasic morphology at 1 Hz. GI, gastrointestinal. Antibiotic-associated encephalopathy (AAE) is being increasingly recog- nized (most often seen with cephalosporins and penicillins) and is charac- terized by an encephalopathic state which may be accompanied by seizures or myoclonus, arising within days of antibiotic administration. Concurrent EEG can be very ictal appearing (ictal–interictal continuum) and can include GPDs with triphasic morphology (Figure 14.3) (14). It is important to recog- FIGURE 14.3 EEG findings in a case of cefepime neurotoxicity. A 72-year-old man with hypertension and diabetes was admitted with pneumonia and acute renal failure to the MICU. Two days after initiation of cefepime, he gradually became more somnolent, and EEG showed generalized periodic discharges at 2 to 3 Hz, which resolved after discontinuation of cefepime. MICU, medical ICU.  Chapter 14 EEG Monitoring in the Medical ICU 129 nize this entity as alternative antibiotic treatment should be considered as well as avoidance of aggressive antiseizure medication treatment. It is important to note that not all presumed alterations in mental status are cerebral in origin. Patients with inability to follow commands and a normal EEG should alert cli- nicians for the possibility of severe neuromuscular dysfunction (critical illness neuropathy/myopathy), brainstem pathology (locked-in syndrome), neuroleptic malignant syndrome, or psychiatric illness (catatonia). C. The optimal duration of monitoring of these patients has not been studied and is not defined In a large retrospective study of all critically ill patients, 90% of all seizures were detected in noncomatose patients within 24 hours and in comatose patients by 48 hours (15). In one study of 625 critically ill patients, it was found that if no seizures occurred in the first 16 hours of cEEG, the probability of recording a subsequent seizure decreases to less than 5% in patients with epileptiform discharges (16). In patients without epileptiform discharges, this 5% threshold is reached after only 2 hours. The same guidelines may be followed in patients in the MICU (17). D. Assessing response to treatment of status epilepticus with antiseizure medications and anesthetics Management of SE in the MICU should follow the same protocol as for patients with neurologic injuries and consist of using BZD as first-line agents followed by antisei- zure medications and anesthetics if required. Please refer to Chapters 30 and 31 for a review of the treatment of SE. E. Technical considerations for cEEG monitoring in the MICU Video is essential concurrent to EEG monitoring in the ICU setting, not only to iden- tify sources of artifacts but also to characterize any movements concerning for sei- zures. Video camera must be correctly repositioned when the nursing staff moves either the patient or the bedside EEG acquisition unit. Sweating and scalp breakdown are common in patients with fevers, sepsis, and prolonged systemic disease and requires frequent evaluation and consideration of scalp rest. Routine ICU care such as compressive chest therapy, suctioning, and oral care can create rhythmic artifacts that mimic electrographic seizures but are easily identified with video analysis. Electrical noise from intravenous pumps, electrical beds, dialysis machines, and other medical devices can obscure the EEG recording and require proper identifica- tion and troubleshooting by the EEG technologist. III. FURTHER CONSIDERATIONS/REMAINING QUESTIONS A. Although preliminary studies suggest that cEEG monitoring is useful for diagnosis of NCSs and evaluation of encephalopathy, more data is required to assess its utility in overall outcomes and cost effectiveness (18) At present, only a small fraction of MICU patients are selected for cEEG monitoring. Criteria and duration for monitoring MICU patients are not clearly defined. 130 Part II Indications Prospective studies on the utility of cEEG in the MICU and treatment of symptom- atic seizures are needed to address this paucity of information. B. Resources for cEEG monitoring outside of the neuro ICU are limited, despite likely benefits in this population Improving EEG monitoring efficiency would increase the impact of cEEG by being able to provide monitoring for more patients, particularly outside of the neurological ICU. Identifying high-risk patients that would benefit most from cEEG would also be useful. At present, sepsis is the most common underlying medical condition that prompts cEEG monitoring, but this still represents a very large patient popula- tion for which EEG monitoring resources are limited. C. Future prospective studies of cEEG outside of the neuro ICU will need to address the association of encephalopathy and seizures in this patient population D. Whether or not NCS/NCSE in this patient population is a marker of severity of injury and whether or not treatment of these seizures actually affects outcomes remain unanswered. Prospective, controlled studies addressing these questions are needed References 1. van den Boogaard M, Schoonhoven L, van der Hoeven JG, et al. Incidence and short-term consequences of delirium in critically ill patients: a prospective observational cohort study. Int J Nurs Stud. 2012;49(7):775–783. 2. Oddo M, Carrera E, Claassen J, et al. Continuous electroencephalography in the medical intensive care unit. Crit Care Med. 2009;37:2051–2056. 3. Gilmore EJ, Gaspard N, Choi HA, et al. Acute brain failure in severe sepsis: a prospective study in the medical intensive care unit utilizing continuous EEG monitoring. Intensive Care Med. 2015;41(4):686–694. 4. Kurtz P, Gaspard N, Wahl AS, et al. Continuous electroencephalography in a surgical intensive care unit. Intensive Care Med. 2014;40(2):228–234. 5. Kamel H, Betjemann JP, Navi BB, et al. Diagnostic yield of electroencephalography in the medical and surgical intensive care unit. Neurocrit Care. 2013;19(3):336–341. 6. Heming N, Mazeraud A, Verdonk F, et al. Neuroanatomy of sepsis-associated encephalopathy. Crit Care. 2017;21(1):65. 7. Sharbrough FW, Messick JM Jr, Sundt TM Jr. Correlation of continuous electroencephalograms with cerebral blood flow measurements during carotid endarterectomy. Stroke. 1973;4(4):674– 683. 8. Claassen J, Hirsch LJ, Kreiter KT, et al. Quantitative continuous EEG for detecting delayed cerebral ischemia in patients with poor-grade subarachnoid hemorrhage. Clin Neurophysiol. 2004;115:2699–2710. 9. Threlkeld ZD, Kottapally M, Aysenne A. Continuous qualitative electroencephalography as a noninvasive neuromonitor. Can J Neurol Sci. 2016;6(4):157–160. 10. Chen H, Wang J, Mao S, et al. A new method of intracranial pressure monitoring by EEG power spectrum analysis. Can J Neurol Sci. 2012;39(4):483–487. 11. Young GB, Bolton CF, Archibald YM, et al. The electroencephalogram in sepsis-associated encephalopathy. J Clin Neurophysiol. 1992;9(1):145–152.  Chapter 14 EEG Monitoring in the Medical ICU 131 12. Foreman B, Mahulikar A, Tadi P, et al. Generalized periodic discharges and ‘triphasic waves’: a blinded evaluation of inter-rater agreement and clinical significance. J Clin Neurophysiol. 2016;127(1):1073–1080. 13. O’Rourke D, Chen PM, Gaspard N, et al. Response rates to anticonvulsant trials in patients with triphasic-wave EEG patterns of uncertain significance. Neurocrit Care. 2016;24(2):233–239. 14. Bhattacharyya S, Darby RR, Raibagkar P, et al. Antibiotic-associated encephalopathy. Neurology. 2016;86(10):963–971. 15. Claassen J, Mayer SA, Kowalski RG, et al. Detection of electrographic seizures with continuous EEG monitoring in critically ill patients. Neurology. 2004;62:1743–1748. 16. Westover MB, Shafi MM, Bianchi MT, et al. The probability of seizures during EEG monitoring in critically ill adults. Clin Neurophysiol. 2015;126(3):463–471. 17. Herman ST, Abend NS, Bleck TP, et al. Consensus statement on continuous EEG in critically ill adults and children, Part I: indications. J Clin Neurophysiol. 2015;32(2):87–95. 18. Ney JP, van der Goes DN, Nuwer MR, et al. Continuous and routine EEG in intensive care: utilization and outcomes, United States 2005–2009. Neurology. 2013;81(23):2002–2008.