Handbook of ICU EEG Monitoring

An Imprint of Springer Publishing

Handbook of

ICU EEG Monitoring SECOND EDITION

n n n n n n n n

Suzette M. LaRoche hiba aRif haideR

Handbook of

ICU EEG Monitoring

Handbook of

ICU EEG Monitoring SECOND EDITION

Editors SUZETTE M. LAROCHE, MD Medical Director Mission Health Epilepsy Center Asheville, North Carolina

HIBA ARIF HAIDER, MD Assistant Professor of Neurology Emory University School of Medicine Atlanta, Georgia

An Imprint of Springer Publishing

Visit our website at www.springerpub.com ISBN: 9780826168610 e-book ISBN: 9780826168627 Acquisitions Editor: Beth Barry Compositor: Exeter Premedia Services Pvt Ltd. Copyright © 2018 Springer Publishing Company. Demos Medical Publishing is an imprint of Springer Publishing Company, LLC. All rights reserved. This book is protected by copyright. No part of it may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the publisher. Medicine is an ever-changing science. Research and clinical experience are continually expanding our knowledge, in particular our understanding of proper treatment and drug therapy. The authors, editors, and publisher have made every effort to ensure that all information in this book is in accordance with the state of knowledge at the time of production of the book. Nevertheless, the authors, editors, and publisher are not responsible for errors or omissions or for any consequences from application of the information in this book and make no warranty, expressed or implied, with respect to the contents of the publication. Every reader should examine carefully the package inserts accompanying each drug and should carefully check whether the dosage schedules mentioned therein or the contraindications stated by the manufacturer differ from the statements made in this book. Such examination is particularly important with drugs that are either rarely used or have been newly released on the market. Library of Congress Cataloging-in-Publication Data Names: LaRoche, Suzette, editor. | Haider, Hiba Arif, editor. Title: Handbook of ICU EEG monitoring / [edited by] Suzette M. LaRoche, Hiba Arif Haider. Description: Second edition. | New York : Springer Publishing Company, [2018] | Includes bibliographical references and index. Identifiers: LCCN 2017039427| ISBN 9780826168610 | ISBN 9780826168627 (e-book) Subjects: | MESH: Electroencephalography | Intensive Care Units | Monitoring, Physiologic Classification: LCC RC386.6.E43 | NLM WL 150 | DDC 616.8/047547—dc23 LC record available at https://lccn.loc.gov/2017039427 Contact us to receive discount rates on bulk purchases. We can also customize our books to meet your needs. For more information please contact: [email protected] Printed in the United States of America.

We dedicate this edition of the Handbook of ICU EEG Monitoring to Dr. Sandra Helmers. Cancer ended Sandy’s career way too early, but not before she had the opportunity to influence generations of neurophysiologists. Both of us had the distinct benefit of Sandy’s mentorship as residents, fellows, and ultimately as colleagues. We miss her greatly, but strive to carry on her passion for education, scholarship, and extraordinary patient care. —Suzette and Hiba

Contents Contributors xi Abbreviations xvii List of Figures xxiii Preface xxxi Acknowledgments xxxiii Share Handbook of ICU EEG Monitoring, Second Edition I. TECHNICAL ISSUES Chapter 1 Chapter 2 Chapter 3 Chapter 4

Equipment for EEG Acquisition and Review 1 Susan T. Herman Electrodes and Montages 11 Jennifer L. Hopp and Chalita C. Atallah Networking, Remote Monitoring, and Data Storage 19 Joshua Andrew Ehrenberg Staffing an ICU EEG Monitoring Unit 32 Abdulrahman Alwaki, Joshua Andrew Ehrenberg, and Andres Rodriguez-Ruiz

II. INDICATIONS Chapter 5

Status Epilepticus 42 Sebastian Pollandt and Thomas P. Bleck

Chapter 6

Ischemic Stroke 52 Wendy L. Wright Subarachnoid Hemorrhage 60 Michael Mendoza and Adam Webb Intracranial Hemorrhage 68 Jonathan Elmer and Lori A. Shutter Infectious and Inflammatory Conditions 76 Olga Taraschenko and Nicolas Gaspard Traumatic Brain Injury 92 Brad J. Kolls Prognosis Following Cardiac Arrest in Adults Amy Z. Crepeau

Chapter 7 Chapter 8 Chapter 9 Chapter 10 Chapter 11

98

vii

viii

Contents

Chapter 12

Chapter 13 Chapter 14 Chapter 15

Therapeutic Hypothermia in the Neonatal and Pediatric Populations 106 Nicholas S. Abend and Courtney J. Wusthoff Prognosis in Patients Without Cardiac Arrest 116 Leslie A. Rudzinski EEG Monitoring in the Medical ICU 123 Monica B. Dhakar, Stephen Hantus, and Emily J. Gilmore EEG Monitoring in the Pediatric ICU 132 Cecil D. Hahn and William B. Gallentine

III. EEG INTERPRETATION Chapter 16 Chapter 17 Chapter 18 Chapter 19 Chapter 20 Chapter 21 Chapter 22 Chapter 23 Chapter 24 Chapter 25 Chapter 26 Chapter 27 Chapter 28 Chapter 29

Overview of Standardized Critical Care EEG Terminology Jessica W. Templer and Elizabeth E. Gerard Background Activity 152 Kevin F. Haas Lateralized Periodic Discharges 160 Jessica W. Templer and Elizabeth E. Gerard Generalized Periodic Discharges 170 Joshua Martin and Brandon Foreman Other Periodic and Rhythmic Patterns 179 Nicolas Gaspard The Ictal–Interictal Continuum 186 Suzette M. LaRoche and Valia Rodríguez Nonconvulsive Status Epilepticus 200 Hiba Arif Haider and Frank W. Drislane Neonatal Seizures and Status Epilepticus 213 Rawad Obeid and Tammy N. Tsuchida Postanoxic Encephalopathy 224 Tadeu A. Fantaneanu and Jong Woo Lee Artifacts 233 Sarah E. Schmitt Interrater Agreement for EEG Interpretation 241 Jonathan J. Halford Quantitative EEG: Basic Principles 247 Saurabh R. Sinha Quantitative EEG for Detection of Seizures 255 Hiba Arif Haider and Suzette M. LaRoche Quantitative EEG for Ischemia Detection 268 Carlos F. Muñiz, Sahar Zafar, and M. Brandon Westover

141

ix

Contents

IV. TREATMENT Chapter 30 Chapter 31 Chapter 32 Chapter 33 Chapter 34

Generalized Convulsive Status Epilepticus 281 Christa B. Swisher and Aatif M. Husain Nonconvulsive Status Epilepticus in Adults 298 Sara Hocker and Peter W. Kaplan Status Epilepticus in the Pediatric Population 310 Sarah Welsh, James Riviello, and Alexis Topjian Alternative Therapies for Refractory Status Epilepticus Emily L. Johnson and Mackenzie C. Cervenka Prophylaxis of Seizures in the ICU Population 331 Gretchen M. Brophy and Eljim P. Tesoro

320

V. ADDITIONAL CONSIDERATIONS Chapter 35 Chapter 36 Chapter 37 Chapter 38 Chapter 39 Chapter 40 Index

387

Neonatal ICU EEG Guidelines 338 Nancy McNamara and Renée Shellhaas ACNS Consensus Statement for Pediatrics and Adults 344 Susan T. Herman Billing and Coding for ICU EEG Monitoring 353 Marc R. Nuwer Report Generation and Communication With the ICU Team 359 Stephen Hantus Multimodality Monitoring 366 Carolina Barbosa Maciel, Jan Claassen, and Emily J. Gilmore EEG Monitoring in the ICU: Future Directions 381 Nicholas S. Abend and Lawrence J. Hirsch

Contributors Nicholas S. Abend, MD, Associate Professor of Neurology and Pediatrics,

Departments of Neurology and Pediatrics, The Children’s Hospital of Philadelphia, The University of Pennsylvania, Philadelphia, Pennsylvania Abdulrahman Alwaki, MD, Epilepsy Fellow, Department of Neurology, Mayo

Clinic, Rochester, Minnesota Chalita C. Atallah, MD, Visiting Instructor, Department of Neurology, University of

Maryland School of Medicine, Baltimore, Maryland Thomas P. Bleck, MD, MCCM, FNCS, Professor of Neurological Sciences, Neurosurgery, Medicine, and Anesthesiology, Rush Medical College; Director of Clinical Neurophysiology, Rush University Medical Center, Chicago, Illinois Gretchen M. Brophy, PharmD, BCPS, FCCP, FCCM, FNCS, Professor, Department of Pharmacotherapy & Outcomes Science and Neurosurgery, Virginia Commonwealth University School of Pharmacy, Medical College of Virginia Campus, Richmond, Virginia Mackenzie C. Cervenka, MD, Associate Professor, Department of Neurology, The Johns Hopkins University School of Medicine, Baltimore, Maryland Jan Claassen, MD, Department of Neurology, Columbia University Medical Center, New York-Presbyterian Hospital, New York, New York Amy Z. Crepeau, MD, Assistant Professor, Department of Neurology, Mayo Clinic,

Phoenix, Arizona Monica B. Dhakar, MD, Epilepsy Fellow, Department of Neurology, Yale School of Medicine, New Haven, Connecticut Frank W. Drislane, MD, Professor of Neurology, Harvard Comprehensive Epilepsy Program, Beth Israel Deaconness Medical Center, Boston, Massachusetts Joshua Andrew Ehrenberg, BSc, R EEG T, CNIM, Senior Product Specialist, Nihon

Kohden, Chattanooga, Tennessee Jonathan Elmer, MD, MS, Assistant Professor, Departments of Emergency Medicine and Critical Care Medicine, UPMC/University of Pittsburgh, Pittsburgh, Pennsylvania Tadeu A. Fantaneanu, MDCM, FRCPC, Assistant Professor, Department of

Medicine, Division of Neurology, The Ottawa Hospital, Ottawa, Ontario, Canada xi

xii

Contributors

Brandon Foreman, MD, Assistant Clinical Professor, Department of Neurology

and Rehabilitation Medicine, The University of Cincinnati Medical Center; Division of Neurocritical Care, University of Cincinnati Gardner Neuroscience Institute, Cincinnati, Ohio William B. Gallentine, DO, Associate Professor and Pediatric Neurologist, Division of Neurology, Department of Pediatrics, Duke Children’s Hospital and Duke University School of Medicine, Durham, North Carolina Nicolas Gaspard, MD, PhD, Associate Professor, Department of Neurology,

Université Libre de Bruxelles – Hôpital Erasme; Assistant Professor (Adjunct), Department of Neurology, Yale University, Bruxelles, Belgium Elizabeth E. Gerard, MD, Associate Professor, Department of Neurology,

Northwestern University, Chicago, Illinois Emily J. Gilmore, MD, Assistant Professor of Neurology, Yale School of Medicine;

Staff Neurointensivist, Neuroscience Intensive Care Unit, Yale New Haven Hospital, New Haven, Connecticut Kevin F. Haas, MD, PhD, Associate Professor of Neurology, Epilepsy Division, Clinical Director of Epilepsy Surgery, Vanderbilt University Medical Center, Nashville, Tennessee Cecil D. Hahn, MD, MPH, Staff Neurologist, Division of Neurology, The Hospital for Sick Children and Associate Professor, Department of Paediatrics, University of Toronto, Toronto, Ontario, Canada Hiba Arif Haider, MD, Assistant Professor of Neurology, Emory University School of Medicine, Atlanta, Georgia Jonathan J. Halford, MD, Associate Professor of Neurology, Department of

Neurology, Medical University of South Carolina, Charleston, South Carolina Stephen Hantus, MD, Clinical Assistant Professor of Medicine, Lerner College of

Medicine, Cleveland Clinic Epilepsy Center, Cleveland, Ohio Susan T. Herman, MD, Assistant Professor, Department of Neurology, Beth Israel

Deaconess Medical Center, Harvard Medical School, Boston, Massachusetts Lawrence J. Hirsch, MD, Professor of Neurology, Chief, Division of Epilepsy and EEG; Co-Director, Yale Comprehensive Epilepsy Center; Co-Director, Critical Care EEG Monitoring Program, Yale School of Medicine, New Haven, Connecticut Sara Hocker, MD, Associate Professor of Neurology, Department of Neurology,

Division of Critical Care Neurology, Mayo Clinic, Rochester, Minnesota Jennifer L. Hopp, MD, Associate Professor of Neurology, Director of The Epilepsy Division; Director, Epilepsy Monitoring Unit, University of Maryland School of Medicine, Baltimore, Maryland

Contributors

xiii

Aatif M. Husain, MD, Professor, Department of Neurology, Duke University Medical Center; Director, Neurodiagnostic Center, Veterans Affairs Medical Center, Durham, North Carolina Emily L. Johnson, MD, Assistant Professor, Department of Neurology, The Johns Hopkins University School of Medicine, Baltimore, Maryland Peter W. Kaplan, BSc (hons), MB, BS, FRCP, Director, EEG/Epilepsy, Johns

Hopkins Bayview Medical Center, Department of Neurology, Johns Hopkins University School of Medicine, Baltimore, Maryland Brad J. Kolls, MD, PhD, MMCi, Associate Professor, Department of Neurology, Division of Neurocritical Care and Stroke; Vice Chair Informatics and Teleneurology, Duke University School of Medicine, Durham, North Carolina Suzette M. LaRoche, MD, Medical Director, Mission Health Epilepsy Center,

Asheville, North Carolina Jong Woo Lee, MD, PhD, Associate Professor, Department of Neurology, Brigham

and Women’s Hospital, Boston, Massachusetts Carolina Barbosa Maciel, MD, Assistant Professor, University of Florida College of Medicine, UF Health Shands Hospital, Gainesville, Florida Joshua Martin, MD, Department of Neurology and Rehabilitation Medicine,

The University of Cincinnati Medical Center; Epilepsy Division, University of Cincinnati Gardner Neuroscience Institute, Cincinnati, Ohio Nancy McNamara, MD, Clinical Assistant Professor, Pediatric Neurology, University of Michigan, CS Mott Children’s Hospital, Ann Arbor, Michigan Michael Mendoza, MD, Fellow, Department of Neurosurgery, Emory University

School of Medicine, Atlanta, Georgia Carlos F. Muñiz, MD, Fellow, Department of Neurology, Massachusetts General Hospital, Boston, Massachusetts Marc R. Nuwer, MD, PhD, Professor and Vice Chair, Department of Neurology, David Geffen School of Medicine at UCLA; Department Head, Clinical Neurophysiology, Ronald Reagan UCLA Medical Center, Los Angeles, California Rawad Obeid, MD, Neonatal-Neurology Clinical Research Fellow, Center

for Neuroscience Research, Neurology, Children’s National Health System, Washington, DC Sebastian Pollandt, MD, Assistant Professor, Department of Neurological

Sciences, Rush University Medical Center, Chicago, Illinois

xiv

Contributors

James Riviello, MD, Associate Section Head for Epilepsy, Neurophysiology, and

Neurocritical Care, Section of Neurology and Developmental Neuroscience; Professor of Pediatrics, Department of Pediatrics, Baylor College of Medicine, Texas Children’s Hospital, Houston, Texas Valia Rodríguez, MD, PhD, Neurophysiology Lecturer, School of Life and Health Sciences, Aston University, Birmingham, UK; Professor of Clinical Neurophysiology, Cuban Neuroscience Center, Havana, Cuba Andres Rodriguez-Ruiz, MD, Assistant Professor, Department of Neurology,

Emory University School of Medicine, Atlanta, Georgia Leslie A. Rudzinski, MD, Associate Professor, Department of Neurology, Augusta University Medical Center, Augusta, Georgia Sarah E. Schmitt, MD, Associate Professor, Department of Neurology, Medical

University of South Carolina, Charleston, South Carolina Renée Shellhaas, MD, MS, Clinical Associate Professor, Pediatric Neurology,

University of Michigan, CS Mott Children’s Hospital, Ann Arbor, Michigan Lori A. Shutter, MD, Professor, Critical Care Medicine, Neurology and Neurosurgery, UPMC/University of Pittsburgh, Pittsburgh, Pennsylvania Saurabh R. Sinha, MD, PhD, Associate Professor, Department of Neurology, Duke

University Medical Center, Durham, North Carolina Christa B. Swisher, MD, Assistant Professor, Department of Neurology, Duke University School of Medicine, Durham, North Carolina Olga Taraschenko, MD, PhD, Assistant Professor, Department of Neurological

Sciences, Comprehensive Epilepsy Program, University of Nebraska Medical Center, Omaha, Nebraska Jessica W. Templer, MD, Instructor, Department of Neurology, Northwestern

University, Chicago, Illinois Eljim P. Tesoro, PharmD, BCPS, Clinical Associate Professor, College of Pharmacy; Clinical Pharmacist, Neurosciences; Director, PGY2 Residency in Critical Care, University of Illinois Hospital & Health Sciences System, Chicago, Illinois Alexis Topjian, MD, MSCE, Associate Professor, Anesthesia and Critical Care,

University of Pennsylvania Perelman School of Medicine, The Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania Tammy N. Tsuchida, MD, PhD, Associate Professor, Departments of Neurology

and Pediatrics, George Washington University School of Medicine and Health Sciences, Children’s National Health System, Washington, DC

Contributors

xv

Adam Webb, MD, Assistant Professor of Neurology and Neurosurgery,

Neurocritical Care; Medical Director, Neuroscience ICU, Marcus Stroke and Neuroscience Center, Grady Memorial Hospital; Medical Director, Performance Improvement, Emory University School of Medicine at Grady, Atlanta, Georgia Sarah Welsh, MD, Assistant Professor of Pediatrics, Pediatric Intensivist, Division of Pediatric Critical Care Medicine, Hasbro Children’s Hospital at Rhode Island Hospital, Providence, Rhode Island M. Brandon Westover, MD, PhD, Director, MGH Critical Care EEG Monitoring

Service, Department of Neurology, Massachusetts General Hospital, Boston, Massachusetts Wendy L. Wright, MD, JM, Associate Chief of Neurology and Medical Director of the Neurocritical Care Unit, Emory University Hospital Midtown; Associate Professor of Neurology and Neurosurgery, Emory University School of Medicine, Atlanta, Georgia Courtney J. Wusthoff, MD, Assistant Professor of Neurology and Neurological Sciences, Pediatrics-Neonatal and Developmental Medicine, Stanford University, Stanford, California Sahar Zafar, MD, Director, MGH/BWH/Harvard Neurocritical Care Fellowship,

Department of Neurology, Massachusetts General Hospital, Boston, Massachusetts

Abbreviations ABRET AAE ABPN ACA ACNS ACTH ADC ADEM ADP ADR aEEG AERRPS AIS AMA AMPA AS aSAH ASDA ASDs ASET ASI BI BID BIPDs B(I)RDs BP BRDs BSI BSR BZDs CA Caspr2 CBF CCEMRC CDSA CEA cEEG CFM cIV-ASDs CJD CLTM CNS

American Board of Registration of Electroencephalographic and Evoked Potential Technologists antibiotic-associated encephalopathy American Board of Psychiatry and Neurology anterior cerebral artery American Clinical Neurophysiology Society adrenocorticotropic hormone analog-to-digital converter acute disseminated encephalomyelitis absolute delta power alpha:delta ratio amplitude-integrated EEG acute encephalitis with refractory repetitive partial seizures acute ischemic strokes American Medical Association alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid active sleep aneurysmal subarachnoid hemorrhage automated seizure detection algorithms antiseizure drugs American Society of Electroneurodiagnostic Technologists absolute symmetry index bilateral-independent twice daily bilateral independent periodic discharges brief potentially ictal rhythmic discharges blood pressure brief rhythmic discharges brain symmetry index burst-suppression ratio benzodiazepines cardiac arrest contactin-associated protein-2 cerebral blood flow Critical Care EEG Monitoring Research Consortium color density spectral array carotid endarterectomy continuous electroencephalography cerebral function monitor continuous intravenous ASDs Creutzfeldt–Jakob disease Certification in Long-Term Monitoring central nervous system xvii

xviii

CPP CPR CPSE CPT CRMP CRRT CSA CSE CSF CTA CTP DAI DAR DC DCI DESC DSA DSA DTABR DWI ECI ECMO EDs EFA EMG EMUs EOG EOS EPC ES ET ETs FDA FFP FFT FIRDA FIRES FLAIR fPHT FPR FRx GABA GABA-A GABA-B GAD GCS GCSE GPD-TW GPDs GRAW GRDA

Abbreviations

cerebral perfusion pressure cardiopulmonary resuscitation complex partial status epilepticus current procedural terminology collapsing response mediator protein continuous renal replacement therapy compressed spectral array convulsive status epilepticus cerebrospinal fluid CT angiography CT perfusion diffuse axonal injury delta:alpha ratio direct current delayed cerebral ischemia devastating epileptic encephalopathy in school-age children density spectral array digital subtraction angiography delta–theta/alpha–beta ratio diffusion-weighted imaging electrocerebral inactivity extracorporeal membrane oxygenation epileptic or epileptiform discharges Epilepsy Foundation of America electromyogram epilepsy monitoring units electrooculography early onset seizures epilepsia partialis continua electrographic seizures envelope trend epileptiform transients Food and Drug Administration fresh frozen plasma fast Fourier transform frontally predominant intermittent rhythmic delta activity febrile infection–related epilepsy syndrome fluid-attenuated inversion recovery fosphenytoin false-positive rate flow-related autoregulation index gamma-aminobutyric acid gamma-aminobutyric acid receptor A gamma-aminobutyric acid receptor B glutamic acid decarboxylase Glasgow Coma Scale generalized convulsive status epilepticus generalized periodic discharges with triphasic morphology generalized periodic discharges GPDs related to anesthetic withdrawal generalized rhythmic delta activity

Abbreviations

GRDA+S GSW HIE HIPAA HSV HSV-1 IBI ICE ICH ICP IIC ILAE IM INR IOM IP IRA IT IVADs IVH IVIG JME KD KET LB LCM LCMV LDF LE LGI-1 LPDs LPR LRDA LVAD MAP MCA MD MELAS MICU µ V MMM MRS MSE NCCI NCS NCSE NDTs NIH NIHSS

xix

generalized rhythmic delta activity with superimposed sharp activity generalized spike-and-wave activity hypoxic–ischemic encephalopathy Health Insurance Portability and Accountability Act herpes simplex virus herpes simplex virus-1 interburst interval intracortical electroencephalography intracerebral hemorrhage or intracranial hemorrhage intracranial pressure ictal–interictal continuum International League Against Epilepsy intramuscular international normalized ratio intraoperative monitoring Internet protocol interrater agreement immunotherapy intravenous anesthetics drugs intraventricular hemorrhage intravenous immunoglobulin juvenile myoclonic epilepsy ketogenic diet ketamine longitudinal-bipolar lacosamide lymphocytic choriomeningitis virus laser Doppler flowmetry limbic encephalitis leucine-rich glioma inactivated-1 lateralized periodic discharges lactate/pyruvate ratios lateralized rhythmic delta activity left ventricular assist device mean arterial pressure middle cerebral artery microdialysis mitochondrial encephalopathy with lactic acidosis and stroke-like episodes medical ICU microvolts multimodality monitoring Modified Rankin Scale myoclonic status epilepticus National Correct Coding Initiative Neurocritical Care Society or nonconvulsive seizures nonconvulsive status epilepticus neurodiagnostic technologists National Institutes of Health National Institutes of Health Stroke Scale

xx

Abbreviations

NIRS NMDA NMDAR NOACs NORSE NPV NSE OR PAE PAMM PAV PbtO2 PCA PCCs pdBSI PDs PE PER PGS PHT PMA PPV PRES PRO PT PTS PTT QEEG QID QS RAI RAMPART RAS RAV RAWOD RCTs RDA rfVIIa RSE SAE SAH SBP SDH SE SEF SI-LPDs SIRPIDs SjvO2 SPECT SR

near-infrared spectroscopy N-methyl-D-aspartate N-methyl-D-aspartate receptor novel oral anticoagulants new onset refractory status epilepticus negative predictive value neuron-specific enolase odds ratio postanoxic encephalopathy postanoxic multifocal myoclonus percentage of alpha variability partial brain tissue oxygen tension posterior cerebral artery prothrombin complex concentrates pairwise-derived brain symmetry index periodic discharges plasma exchange perampanel purple glove syndrome phenytoin postmenstrual age positive predictive value posterior reversible encephalopathy syndrome propofol prothrombin time posttraumatic seizures partial thromboplastin time quantitative EEG four times a day quiet sleep relative asymmetry index Rapid Anticonvulsant Medication Prior to Arrival Trial relative asymmetry spectrogram relative alpha variability regional attenuation without delta randomized controlled trials rhythmic delta activity recombinant factor VIIa refractory status epilepticus sepsis-associated encephalopathy subarachnoid hemorrhage systolic blood pressure subdural hematoma status epilepticus spectral edge frequency stimulus-induced LPDs stimulus-induced rhythmic, periodic, or ictal discharges jugular bulb venous saturation single photon emission CT suppression ratio

Abbreviations

SREAT SRSE SSEPs SSPE STESS SV2A SW TBI TCD TDF TH TIA TICS TID TIRDA tPA TPM TWs VF VGCC VKAs VLAN VPA VPN VT WANs WBC WFNS

xxi

steroid-responsive encephalopathy associated with autoimmune thyroiditis super-refractory status epilepticus somatosensory evoked potentials subacute sclerosing panencephalitis Status Epilepticus Severity Score synaptic vesicle glycoprotein 2A spike-wave traumatic brain injury transcranial Doppler thermal dilution flowmetry therapeutic hypothermia transient ischemic attack Telephone Interview for Cognitive Status three times per day temporal intermittent rhythmic delta activity tissue plasminogen activator topiramate triphasic waves ventricular fibrillation voltage-gated calcium channel vitamin K antagonists virtual local area network valproic acid virtual private network ventricular tachycardia wide area networks white blood cell World Federation of Neurosurgical Societies

List of Figures 2

FIGURE 1.1

Components of digital ICU cEEG acquisition machines

FIGURE 2.1

Electrode options 12

FIGURE 2.2

CT artifact from metal/gold cup electrodes

FIGURE 2.3

Comparison of full vs. reduced electrode montages

FIGURE 3.1

A graphical representation of stand-alone versus integrated facility EEG networks 22

FIGURE 3.2

Remote monitoring options

FIGURE 5.1

ILAE definition of status epilepticus

FIGURE 5.2

Incidence of nonconvulsive seizures in patients undergoing cEEG 46

FIGURE 6.1

Relative attenuation without delta (RAWOD) secondary to acute ischemic stroke 56

FIGURE 7.1

Delayed cerebral ischemia (DCI) detected by QEEG techniques

FIGURE 8.1

Focal electrographic seizure in a patient with multifocal lobar hemorrhages 73

FIGURE 8.2

Ictal lateralized periodic discharges (LPDs) due to a left parietal intraparenchymal hemorrhage 73

FIGURE 8.3

Persistent delta slowing with preservation of faster frequencies in a patient with subcortical hemorrhage 74

FIGURE 9.1

MRI findings in severe pneumococcal meningitis with abscess, vasculitis, and stroke 80

FIGURE 9.2

MRI and EEG findings in herpes simplex virus type 1 (HSV-1) encephalitis

81

FIGURE 9.3

MRI findings in limbic encephalitis associated with anti–LGI-1 antibodies

84

FIGURE 9.4

MRI and EEG findings in Creutzfeldt–Jakob disease

FIGURE 9.5

High amplitude, slow frequency generalized periodic discharges (GPDs) in subacute sclerosing panencephalitis 87

FIGURE 9.6

Typical MRI and EEG features of posterior reversible encephalopathy syndrome (PRES) 88

FIGURE 9.7

Extreme delta brushes in anti-NMDAR encephalitis

14 16

25

FIGURE 10.1 Pathophysiologic mechanisms of TBI

43

64

86

90

93

FIGURE 10.2 Traumatic intracranial hemorrhage (ICH) in an 80-year-old female patient after a fall at home 94 FIGURE 10.3 NCSE in a patient with traumatic brain injury and persistent unresponsiveness 95

xxiii

xxiv

List of Figures

FIGURE 11.1 Generalized periodic discharges (GPDs) in a comatose 65-year-old on day 3 following cardiac arrest 100 FIGURE 11.2 Generalized periodic discharges (GPDs) in a patient with myoclonic status epilepticus (MSE) following cardiac arrest 101 FIGURE 11.3 EEG demonstrating intermittent myoclonic jerks following cardiac arrest

101

FIGURE 11.4 Isoelectric background, nonreactive 102 FIGURE 11.5 Burst-suppression pattern

103

FIGURE 12.1 Seizure occurrence during and following hypothermia in neonates with hypoxic– ischemic encephalopathy (HIE) 108 FIGURE 12.2 aEEG tracing demonstrating abnormally low amplitude background in a newborn receiving hypothermia for hypoxic–ischemic encephalopathy (HIE) 109 FIGURE 12.3 aEEG tracing demonstrating seizures in a newborn receiving hypothermia for hypoxic–ischemic encephalopathy (HIE) 110 FIGURE 12.4 Focal onset electrographic seizure in a child treated with therapeutic hypothermia following cardiac arrest 111 FIGURE 12.5 Seizure occurrence during and following hypothermia in children following a cardiac arrest 112 FIGURE 14.1 Prevalence of seizures in critically ill patients admitted to medical and surgical ICUs 125 FIGURE 14.2 EEG findings in the setting of liver failure and hyperammonemia FIGURE 14.3 EEG findings in a case of cefepime neurotoxicity

128

128

FIGURE 16.1 Main terms and modifiers of the ACNS Critical Care EEG Monitoring Research terminology 144 FIGURE 16.2 Schematic of ACNS Main Term 2 options FIGURE 16.3 Lateralized periodic discharges (LPDs)

145

145

FIGURE 16.4 Generalized periodic discharges (GPDs) 146 FIGURE 16.5 Lateralized rhythmic delta activity (LRDA)

146

FIGURE 17.1 Generalized desynchronized alpha and theta activity with intermixed delta activity, as seen in mild encephalopathy 154 FIGURE 17.2 Generalized high-amplitude delta activity with intermixed alpha and theta activity, as seen in moderate encephalopathy 154 FIGURE 17.3 Generalized mixture of theta and delta activity with some faster activities, as seen in moderate encephalopathy 155 FIGURE 17.4 Low-amplitude delta slowing intermixed with theta activity, nearly continuous, which is consistent with severe encephalopathy 156 FIGURE 17.5 EEG reactivity with emergence of higher amplitude generalized rhythmic delta activity 157 FIGURE 18.1 Lateralized periodic discharges (LPDs) in a patient with temporal lobe epilepsy who was admitted for breakthrough seizures 162

xxv

List of Figures FIGURE 18.2 Lateralized periodic discharges with superimposed fast activity (LPD+F) following resolution of focal motor seizures 165 FIGURE 18.3 Bilateral independent periodic discharges (BIPDs) in the setting of West Nile encephalitis and seizures 166

FIGURE 18.4 Ictal lateralized periodic discharges (LPDs) correlating to clonic movements of the right leg 167 FIGURE 19.1 Generalized periodic discharges (GPDs) following cardiac arrest 172 FIGURE 19.2 Generalized periodic discharges (GPDs) associated with myoclonus following respiratory arrest 175 FIGURE 19.3 Generalized periodic discharges (GPDs) in the setting of anesthetic withdrawal 175 FIGURE 19.4 Generalized periodic discharges (GPDs) secondary to cefepime toxicity

176

FIGURE 20.1 Bilateral independent periodic discharges plus Fast activity (BIPDs + F) in the setting of PRES 180 FIGURE 20.2 Generalized rhythmic delta activity (GRDA), frontally predominant, in a patient with acute medical problems 181 FIGURE 20.3 Lateralized rhythmic delta activity (LRDA) in association with intracranial hemorrhage and seizures 183 FIGURE 20.4 Brief (potentially ictal) rhythmic discharges, B(I)RDs

183

FIGURE 21.1 Periodic and rhythmic patterns and associated risk of seizures 188 FIGURE 21.2 Generalized periodic discharges (GPDs) with triphasic morphology and positive benzodiazepine trial 190 FIGURE 21.3 Ictal–interictal continuum (IIC) pattern secondary to anesthetic withdrawal 191 FIGURE 21.4 Ictal–interictal continuum (IIC) treatment algorithm

194

FIGURE 21.5 Bilateral independent periodic discharges (BIPDs) evolving into nonconvulsive status epilepticus (NCSE) 195 FIGURE 21.6 Lateralized periodic discharges (LPDs) with clinical correlate (ictal LPDs) 196 FIGURE 21.7 Periodic discharges (PDs) with shifting laterality secondary to baclofen toxicity 196 FIGURE 22.1 Absence SE 202 FIGURE 22.2 Secondarily generalized nonconvulsive status epilepticus (NCSE) 203 FIGURE 22.3 Focal nonconvulsive status epilepticus (NCSE) 204 FIGURE 22.4 Triphasic waves (generalized periodic discharges with triphasic morphology) 210 FIGURE 23.1 EEG features of neonatal seizures 216 FIGURE 23.2 Evolution in electrographic seizure pattern following treatment 217 FIGURE 23.3 Seizure with poorly formed spike-and-wave discharges

217

FIGURE 23.4 Burst-suppression pattern in a term infant with epileptic encephalopathy secondary to KCNQ2 mutation 218

xxvi

List of Figures

FIGURE 23.5 Runs of right temporal spike-wave discharges 220

FIGURE 23.6 Sucking artifact FIGURE 23.7 Patting artifact

219

220

FIGURE 23.8 Rocking artifact 221 FIGURE 23.9 Seizures and SE on aEEG

221

FIGURE 24.1 Burst-suppression/attenuation pattern without identical bursts 226 FIGURE 24.2 Burst-suppression pattern with identical bursts

227

FIGURE 24.3 Generalized periodic discharges, with and without triphasic morphology

228

FIGURE 24.4 Myoclonic status epilepticus 229 FIGURE 24.5 Examples of reactivity patterns FIGURE 25.1 Pulse artifact FIGURE 25.2 CPR artifact

230

235 237

FIGURE 25.3 Mechanical ventilator artifact, caused by water in ventilator tubing 238 FIGURE 25.4 Bed percussion artifact

239

FIGURE 27.1 Algorithm for calculation of amplitude-integrated EEG FIGURE 27.2 Fourier spectrum of EEG

250

251

FIGURE 27.3 Changes in Fourier spectrum with focal background slowing 252 FIGURE 28.1 Seizures on fast Fourier transform (FFT) spectrogram/color density spectral array (CDSA) 258 FIGURE 28.2 Seizures on rhythmicity spectrogram and color density spectral array (CDSA) 259 FIGURE 28.3 Example of six left hemispheric seizures on asymmetry indices FIGURE 28.4 Example of three seizures on amplitude-based trends

260

261

FIGURE 28.5 Evolution of a focal onset seizure on a quantitative EEG (QEEG) panel

264

FIGURE 28.6 Quantitative EEG (QEEG) panel demonstrating abundant right hemisphere lateralized periodic discharges (LPDs) evolving into right hemisphere seizures 266 FIGURE 29.1 The relationship between cerebral blood flow, EEG activity, and pathophysiologic cellular responses 269 FIGURE 29.2 Visual scale for grading of relative alpha variability (RAV) 272 FIGURE 29.3 Technical interpretation of asymmetry indexes and color spectrograms 273 FIGURE 29.4 Quantitative EEG (QEEG) panel showing nearly symmetric activity on monitoring day #2 following subarachnoid hemorrhage (SAH) 276 FIGURE 29.5 Quantitative EEG (QEEG) panel showing marked asymmetry on monitoring day #6 following subarachnoid hemorrhage (SAH), consistent with delayed cerebral ischemia 277

List of Figures

xxvii

FIGURE 29.6 Global decrease in alpha:delta ratio (ADR) and relative alpha variability (RAV) indicating delayed cerebral ischemia 278 FIGURE 30.1 De novo status epilepticus

284

FIGURE 30.2 Five electrographic stages of status epilepticus

286

FIGURE 30.3 A sample algorithm for treating generalized convulsive status epilepticus (GCSE) 288 FIGURE 31.1 Focal electrographic seizure arising from the right parietal–occipital region 301 FIGURE 31.2 Continuous 2- to 4-Hz generalized atypical spike-waves

302

FIGURE 31.3 Focal onset electrographic seizure maximal over the right central region 303 FIGURE 31.4 Right hemisphere lateralized periodic discharges (LPDs)

305

FIGURE 32.1 Quantitative EEG (QEEG) demonstrating refractory electrographic seizures in a 7-year-old with febrile infection–related epilepsy syndrome (FIRES) 318 FIGURE 35.1 The International 10–20 system for electrode placement (superior view) modified for neonates 340 FIGURE 35.2 Focal electrographic seizure in a term infant with right hemisphere stroke 343 FIGURE 38.1 Critical Care EEG Monitoring Research Consortium (CCEMRC) database user interface 361 FIGURE 38.2 Report generation for a continuous EEG study using the Critical Care EEG Monitoring Research Consortium (CCEMRC) database 364 FIGURE 39.1 Normal ICP waveform 371 FIGURE 39.2 Detecting ischemia with intracortical electroencephalography (ICE) 377 FIGURE 39.3 Detecting ischemia with multimodality monitoring

378

FIGURE 39.4 Detecting seizures with multimodality monitoring

379

SUPPLEMENTAL FIGURES LOCATED IN THE EBOOK FIGURE S-9.1

MRI demonstrating cerebral abscess in a patient presenting with a generalized tonic–clonic seizure

FIGURE S-11.1 Emergence of electrographic seizure from generalized periodic discharges (GPDs) FIGURE S-16.1 Generalized periodic discharges with triphasic morphology (GPD-TW) FIGURE S-16.2 Stimulus induced generalized periodic discharges with triphasic morphology (SI-GPD-TW) FIGURE S-16.3 Lateralized periodic discharges with superimposed Fast activity (LPDs+F) FIGURE S-16.4 Lateralized rhythmic delta activity with superimposed Sharp activity (LRDA+S) FIGURE S-17.1 Cyclic alternating pattern

xxviii

List of Figures

FIGURE S-18.1 Lateralized periodic discharges (LPDs)-bilateral asymmetric in a man with lung adenocarcinoma who presented with clinical seizures and was found to have multiple brain metastasis FIGURE S-18.2 Bilateral independent periodic discharges time locked with myoclonic jerks of the left leg in a patient with autoimmune encephalitis with anti-GAD antibodies FIGURE S-18.3 SPECT scan demonstrating focal hyperperfusion corresponding to the distribution of the lateralized periodic discharges (LPDs) FIGURE S-20.1 Bilateral independent periodic discharges (BIPDs) secondary to multifocal strokes FIGURE S-20.2 Generalized rhythmic delta activity (GRDA) in a 72-year-old with confusion and renal failure FIGURE S-20.3 Lateralized rhythmic delta activity (LRDA) in a man with a left frontal intraparenchymal hemorrhage and seizures FIGURE S-20.4 Brief potentially ictal rhythmic discharges (B[I]RDs) in a patient admitted for generalized convulsive status epilepticus FIGURE S-20.5 Bilateral independent periodic discharges (BIPDs) in a patient with endometrial adenocarcinoma and sepsis FIGURE S-21.1 Ictal lateralized periodic discharges (LPDs) corresponding to clonic face movements FIGURE S-21.2 Lateralized periodic discharges (LPDs) in the setting of bacterial meningitis and cefepime toxicity FIGURE S-21.3 Generalized periodic discharges (GPDs) with triphasic morphology in a patient with hepatic encephalopathy FIGURE S-22.1 Benzodiazepine trial for nonconvulsive status epilepticus (NCSE) FIGURE S-22.2 Focal electrographic seizures in a patient presenting with posterior reversible encephalopathy syndrome (PRES) FIGURE S-22.3 Lateralized periodic discharges (LPDs) evolving to focal electrographic seizures FIGURE S-22.4 Nonconvulsive status epilepticus (NCSE) manifested by ictal lateralized periodic discharges (LPDs) FIGURE S-24.1 Multifocal periodic discharges seen after rewarming following therapeutic hypothermia FIGURE S-24.2 Generalized periodic discharges (GPDs) evolving into seizure FIGURE S-24.3 Eyelid myoclonia following cardiac arrest FIGURE S-24.4 Absence of EEG reactivity FIGURE S-25.1 Ballistocardiographic, bed percussion, and EKG artifact FIGURE S-25.2 Myoclonus artifact FIGURE S-25.3 Patting artifact FIGURE S-25.4 CPR artifact on quantitative EEG (QEEG) FIGURE S-25.5 Ventilator artifact on quantitative EEG (QEEG)

List of Figures

xxix

FIGURE S-25.6 Bed percussion artifact FIGURE S-25.7 Radiofrequency artifact from cellphone use FIGURE S-25.8 Artifact from left ventricular assist device (LVAD) FIGURE S-25.9 Intermittent myogenic artifact mimicking seizures on quantitative EEG (QEEG) trends FIGURE S-28.1 Brief focal seizures not well visualized on quantitative EEG (QEEG) FIGURE S-28.2 Brief rhythmic discharges (BRDs) FIGURE S-28.3 A panel of multiple quantitative EEG (QEEG) trends aids in distinguishing seizures from artifacts FIGURE S-28.4 A novel seizure pattern on quantitative EEG (QEEG) FIGURE S-28.5 Intermittent periodic patterns mimicking seizure on quantitative EEG (QEEG) trends FIGURE S-28.6 Artifact from high-impedance electrodes appearing as paroxysmal quantitative EEG (QEEG) signal change on rhythmicity, asymmetry, and aEEG FIGURE S-31.1 Focal onset seizure in a patient with a right hemispheric stroke FIGURE S-31.2 Generalized periodic discharges (GPDs) with midline predominance FIGURE S-31.3 Generalized periodic discharges (GPDs) in the setting of cefepime neurotoxicity FIGURE S-39.1 Detecting ischemia with intracortical electroencephalography (ICE) FIGURE S-39.2 Detecting ischemia with multimodality monitoring FIGURE S-39.3 Detecting seizures with multimodality monitoring

Preface The first (and only) edition of the Handbook of ICU EEG Monitoring was published just five short years ago, arising from the need to have a concise reference manual detailing the nuts and bolts of what was uncharted territory for most neurophysiologists and neurointensivists at the time. Around the same period, the ACNS published the 2012 Standardized Critical Care EEG Terminology, and the field of ICU EEG monitoring was off and running, with application to critically ill patients of all stripes; from neonates to the elderly and from large academic medical centers to private hospitals. For neurophysiologists interested in ICU EEG, it has truly been a time of discovery. About a decade ago, a group of like-minded individuals set up the Critical Care EEG Monitoring Research Consortium (CCEMRC), which has now grown to over 50 centers in North America and Europe. The CCEMRC has fostered countless collaborative research projects that have not only increased our understanding of EEG monitoring in critically ill patients, but also highlighted what remains to be studied. The medical community involved in caring for patients with acute brain injury has increasingly recognized the significant role of continuous EEG in the detection of causes of secondary injury, such as nonconvulsive seizures and delayed cerebral ischemia. Multicenter data collection exploring the significance of rhythmic and periodic patterns continues to enlighten us, and has even led to dispelling some previously held doctrines (or at least starting to reconsider them). Continuous EEG monitoring has become a distinct subspecialty in its own right. Advances in recording and networking capabilities have paved the way for wider availability as tele-EEG services offering round the clock EEG monitoring and interpretation have become commonplace. Many clinical neurophysiology fellowship programs now incorporate ICU EEG monitoring as an essential part of their training. In fact, dedicated ICU EEG monitoring fellowships have become available and the American Board of Clinical Neurophysiology now offers a subspecialty certification in critical care EEG monitoring. Many professional societies throughout the world have also started offering education and training in this discipline at their annual meetings. The Handbook of ICU EEG Monitoring is designed to be a resource for anyone caring for critically ill patients whose care might involve continuous EEG monitoring, including neurologists, neurointensivists, neurosurgeons, nursing staff, and EEG technologists. This edition, although larger in scope, content, and physical size, broadly retains the format of its predecessor with five sections covering topics related to (1) technical issues, (2) indications, (3) interpretation (4) treatment, and (5) additional considerations. Practical considerations that aren’t typically covered elsewhere are again discussed, such as staffing models, communication and report generation, billing and coding. The section on interpretation includes new chapters on ACNS terminology, addressing periodic and rhythmic patterns with more granularity, and a review of interrater reliability of critical care EEG terminology. Treatment updates are included for the management of status epilepxxxi

xxxii

Preface

ticus, and a new chapter on alternative treatments for refractory SE has been incorporated. One of the most rapidly evolving aspects of ICU EEG involves recognition of patterns that are useful for prognostication, and is therefore discussed separately in adult and pediatric patients as well as for populations with and without cardiac arrest. Finally, we have compiled a supplementary digital atlas of figures that we hope will enhance the reader’s ability to recognize and understand clinical implications of specific EEG patterns. These supplemental figures are located in the ebook that accompanies this volume. This book would not have been possible without the collective expertise and hard work of its many contributors. We would like to thank our colleagues, including EEG technologists, neurophysiology fellows, and faculty at Emory University and Mission Hospital for their contributions and dedication to patient care, education, and research. Finally, we must thank our families, who continue to prove every day that they have our backs. Without their constant support, encouragement, and motivation, this work would not be possible. Hiba would like to thank her husband Jay and daughter Norah for their boundless support, patience, and good humor throughout the completion of this work, and also her parents for being lifelong sources of encouragement and support. Suzette would like to thank Hiba for joining her on this adventure (the second edition would not have come into existence without her) and her partner Nan who patiently offered support for this undertaking while not yet settled from the move from Atlanta to Asheville. Whether you are a veteran neurophysiologist or a trainee, we hope that this edition of the Handbook of ICU EEG Monitoring will continue to kindle your curiosity. Suzette and Hiba

Acknowledgments We acknowledge the Critical Care EEG Monitoring Research Consortium (CCEMRC) and its leadership, who had the incredible foresight to develop a multicenter infrastructure allowing for mentorship and collaborative research, which spurred the vast majority of knowledge contained within these pages. We would also like to recognize the significant contributions of our female colleagues. In a field still largely dominated by men, it is remarkable to not only have female co-editors, but also a considerable proportion of female authors who were chosen not based on gender, but on their noteworthy contributions to the field of ICU EEG monitoring.

xxxiii

Share Handbook of ICU EEG Monitoring, Second Edition

PART I: TECHNICAL ISSUES

1 Equipment for EEG Acquisition and Review Susan T. Herman

IN THIS CHAPTER

Hardware components of an ICU EEG monitoring system Computer specifications for EEG acquisition and review Software for optimization of EEG review ● ● ●

KEY POINTS

EEG acquisition and review equipment for continuous EEG (cEEG) monitoring in the ICU should meet the technical standards outlined in the American Clinical Neurophysiology Society guidelines. ICU cEEG acquisition equipment can be installed as either a fixed (wallmounted) unit or portable system. Simultaneous audio and video recording is strongly encouraged for correlation of behavioral events with underlying EEG patterns and to aid in proper identification of EEG artifacts that can be easily mistaken for electrographic seizures. Specialized hardware and software increases the utility of cEEG for monitoring at the bedside. Options include the ability to enter nursing notes, pushbuttons for seizures and other clinical events, software to integrate physiologic data (e.g., intracranial pressure, blood pressure), and quantitative EEG software for graphical display of quantitative EEG trends. ●

●

●

●

I.

BACKGROUND

A. Technical guidelines ●

The American Clinical Neurophysiology Society (ACNS) has published a consensus statement for continuous monitoring in the ICU, as well as guidelines for routine digital EEG and long-term EEG monitoring for epilepsy. ICU continuous EEG 1

2

Part I

Technical Issues

(cEEG) equipment should meet the technical standards defined in these guidelines. Relevant guidelines are available online at www.acns.org. Guideline 1: Minimum technical requirements for performing clinical electroencephalography (1). Guideline 4: Recording clinical EEG on digital media (2). Guideline 5: Minimum technical standards for pediatric electroencephalography (3). Guideline 6: Minimum technical standards for EEG recording in suspected cerebral death (4). Guideline 12: Guidelines for long-term monitoring for epilepsy (5). Consensus Statement on Continuous EEG in Critically Ill Adults and Children, Part II: Personnel, Technical Specifications, and Clinical Practice (6).

B. Components of digital ICU cEEG machines ●

Figure 1.1 is a schematic of the major components of digital EEG machines, from electrodes to display and storage. EEG signals are recorded at the scalp via electrodes, which plug into receptacles in the jackbox, or electrode box. Jackboxes are electrically or optically isolated from the power supply to prevent dangerous currents from passing through electrodes to the patient. Each jackbox input connects to input 1 of a differential amplifier. Input 2 for each amplifier is a machine reference, or common reference, electrode. The amplifier contains analog low- and high-frequency filters to exclude extraneous electrical signals.

Video Audio Amplifier 1 HFF (anti-aliasing)

LFF

Other monitoring devices

Amplifier 2 Amplifier 3 Jackbox

Amplifier 4…

ADC Analog-todigital converter

Preamplifiers

Computer Processor Memory Software QEEG

Power Calibration

Archiving media

EEG monitor display Other output DAC printer

Impedance test To network 120 V FIGURE 1.1 Components of digital ICU cEEG acquisition machines. See text for details of individual components. ADC, analog-to-digital converter; DAC, digital-to-analog converter; HFF, high-frequency filter; LFF, low-frequency filter; QEEG, quantitative electroencephalography.

Chapter 1 Equipment for EEG Acquisition and Review

3

The EEG signal is then passed to the analog-to-digital converter (ADC), which converts the analog continuous signal into discrete digital values at specified time points. The digital EEG values (and associated metadata files) are stored on the computer hard drive. EEG software is utilized to display the EEG signals on computer monitors, including postprocessing, such as montage reformatting, change in time scale and amplitude displays, and digital filtering. Video, audio, and other physiologic data streams can be synchronized with the cEEG data. EEG files are archived to removable digital media, or stored on servers or storage appliances. II.

BASICS

A. Comparison of ICU cEEG equipment to standard EEG machines ●

Table 1.1 summarizes the most important characteristics of cEEG acquisition equipment, compared with routine and video-EEG monitoring equipment (7).

TABLE 1.1 Features of Routine EEG, Epilepsy Monitoring Unit (EMU), and ICU cEEG Monitoring Equipment FEATURE

ROUTINE EEG

EMU

ICU CEEG

Physical configuration

Usually portable

Typically hard wired

Portable or hard wired

Number of EEG channels required

16 to 40

40 to ≥128

16 to 32

Sampling rate (per second)

200 to 512

200 to ≥10,000

200 to 2,000

Other physiologic inputs (optional inputs within parentheses)

EOG, EKG, EMG, respiratory effort

EOG, EKG, EMG (O2 sat)

EOG, EKG, EMG (BP, ICP, O2 sat)

Video/audio necessary

Optional

Yes

Yes

Spike and seizure detection

No

Yes

Optional

Quantitative EEG trends

No

Optional

Yes

Network connection for remote monitoring

Optional

Recommended

Yes

BP, blood pressure; EMG, electromyography; EOG, electrooculography; ICP, intracranial pressure; O2 sat, oxygen saturation.

4

Part I

Technical Issues

B. Physical configuration Fixed installation In fixed installations, computers and monitors are mounted on the walls of the ICU room, sometimes on swing arms allowing the monitors to be moved around the bed. Equipment can be mounted out of the way of bedside caregivers. Cameras and microphones can be placed in the best recording location. Fixed equipment is less likely to be damaged than portable equipment that is rolled around the hospital. If only some ICU rooms have mounted cEEG equipment, however, critically ill patients may need to be moved into a designated room equipped for monitoring. Despite several advantages, wall-mounted cEEG equipment is generally more expensive and may be underutilized. Portable equipment It can be configured on a cart or on a small-footprint pole-mount system (approximately 2.5 feet diameter base, 4 to 8 feet height [taller with camera mount]). Portable cEEG equipment has the advantage of being able to be moved to where it is needed, but it can obstruct patient care, and camera views are rarely optimal. Many portable cEEG units now have Internet protocol (IP) addressable cameras that allow remote camera control over standard network jacks, without the expense of special cabling. Portable equipment should be placed to avoid overlap of electrical cords with other equipment, and, if possible, out of the way of staff caring for the patient. Some ICU cEEG programs utilize a combination of wall-mounted equipment in high-use areas (e.g., neurological ICUs) and portable equipment for less frequently monitored areas. Increasingly, ICU EEG equipment is integrated with other ICU monitoring devices, and some systems are optimized for multimodality data collection and display. Design of EEG equipment components should take into consideration possible “rough handling” by inexperienced ICU personnel. Delicate cables with fragile connectors, nonwaterproof components (jackboxes and amplifiers), and systems with portable computer components such as laptops and tablets are especially vulnerable to accidental damage. Nonproprietary components (cameras, computer components, cables, and connectors) are preferred, as these are generally less expensive. A patient event button can be used by patients, family, or ICU personnel to mark suspected clinical events. ●

●

●

●

●

●

●

C. Security and safety features ●

●

●

Failure recovery and protection Some ICU EEG machines have a feature that allows the EEG acquisition to detect an unexpected termination of a study (such as system crash or temporary power outage) and automatically reboot and start recording again when power is restored. Uninterruptible power supplies may prevent problems during brief power fluctuations or outages. Computers should be securely fastened to the wall or pole mounts and disk enclosures should be locked to prevent loss of hard drives or other components

Chapter 1 Equipment for EEG Acquisition and Review

5

●

containing protected health information. Laptops and external hard drives should be locked and encrypted. EEG machines should have security features such as secure log-in, automatic keyboard/screen lock, and firewall and antivirus protection. Electrical safety Patients in the ICU are at increased risk for electrical shock because they often are connected to multiple electrical devices and may have indwelling devices. All ICU EEG equipment should be certified for clinical use and tested for electrical and mechanical safety by a certified technician or biomedical engineer at least annually. Portable equipment is more susceptible to damage and should be checked more frequently. Isolation transformers prevent a direct connection between the ground electrode and power line ground. Amplifiers should be electrically or optically isolated. ●

D. Jackbox In the crowded ICU head-of-the-bed environment, small jackboxes may be beneficial. Jackbox inputs are typically labeled according to the 10-20 International System or modified 10-10 System. Electrode inputs should be arrayed so that the entire jackbox and electrode pins fit into a waterproof enclosure to prevent accidental damage to the jackbox and inadvertent disconnection of electrodes. Some jackboxes include amplifiers and have wireless connectivity to the computer, internal storage, and battery power in the jackbox. Such jackboxes can continue to record EEG while the patient is disconnected from the amplifier (e.g., to go to radiology procedures). Before deciding on a wireless solution, ensure that there is no significant overlap in wireless signals from other ICU equipment.

●

●

●

●

E. Amplifier specifications (2)

●

●

●

●

The primary purpose of the amplifier is to magnify the EEG signal from a range measured in microvolts to a signal of several volts, without distortion. The amplified signal can then be further processed, recorded, or displayed. EEG machines use differential amplifiers so that common-mode signals (potentials that are the same at different recording sites and presumably artifacts) are rejected, and only differential-mode signals (potentials that are different at different recording sites and presumable of brain origin) are amplified. EEG amplifiers should have high input impedance compared to electrode impedance, so they provide minimal loading of the EEG signal being measured. Inputs: At least 16 referential EEG channel inputs are required. Thirty-two or more EEG channels are preferred for full 10-20 electrode placement, as well as additional channels for recording EKG, electrooculogram (EOG, 1–2 channels), and electromyogram (EMG). Additional channels may be needed if simultaneous intracranial EEG recording is planned. System (or machine) reference input is recommended. Ground input is recommended.

6

Part I

Technical Issues

●

Direct current (DC) input channels for connection of other physiologic monitors may be needed (thermistors, oxygen saturation monitors, intracranial pressure [ICP] monitors). Full-scale input range greater than ± 2 mV, and up to ± 5 mV is ideal. Bandpass filters 0.1 to 0.3 to 70 to 100 Hz are used for routine clinical recording. Input impedance at least 100 MΩ (up to 1 GΩ ) is recommended. ● ●

F.

ADC specifications (2) Input range ± 1 to 5 mV Sampling rate at least 256 samples per second (3 times higher than the anti-aliasing filter cutoff frequency) Resolution at least 16 bit, including sign bit Minimum amplitude resolution 0.5 μV ● ●

● ●

G. Video and audio (ACNS guideline 12) Simultaneous video and audio recording is highly recommended to allow correlation of clinical seizures and other behavioral events with EEG patterns, as well as to aid in identification of EEG artifacts (e.g., patting, chest physiotherapy, suction, ventilator artifact) (6). Video is time synchronized with the EEG data. Equipment for video recording varies greatly in picture quality and cost. Video may be color or black-and-white. Infrared cameras may be helpful in lowlight conditions. Critically ill patients are less likely to move off camera than patients in epilepsy monitoring units. Fixed wide-angle cameras therefore may be an option, but may not have adequate resolution for detection of fine motor movements. This is less of an issue with newer high-resolution cameras. Many modern cEEG units have IP-addressable cameras mounted on a pole. These allow remote pan/tilt and sometimes zoom and focus from remote locations. Standard digital video is MPEG 4, at 320 × 240 or 640 × 480 resolution. Higher resolutions are available at the expense of hard disk and server space as well as network bandwidth. Video recording size is typically 12 to 20 GB/day. Audio recordings can alert monitoring technologists to clinical episodes and allow assessment of behavior during clinical events. Systems should include an event button for the patient, family, or staff to mark events of interest.

●

●

●

●

●

H. Computer Specifications: Acquisition Machines ●

●

Computers used for EEG acquisition should have sufficient processing capability for simultaneous EEG and video acquisition, spike and seizure detection, quantitative EEG (QEEG) analysis, and network tasks. Minimum specifications Dual core processor greater than 2 GHz 4 to 8 GB RAM Discrete graphics card Network connectivity 100 mbit/sec minimum; Gigabit network interface card preferred Hard drive large enough to store at least 1 week of EEG and video data (approximately 2 GB EEG and 12–20 GB video/day = greater than 150 GB)

Chapter 1 Equipment for EEG Acquisition and Review

7

Operating system is typically Windows; version is dependent on hospital information technology specifications. ●

I.

Computer specifications: Review stations Computers used for EEG review should have sufficient processing capability for simultaneous EEG and video review, review of automated spike and seizure detection, QEEG analysis, and network tasks. Review computers may also have additional functions, such as report generation (office productivity software, voice recognition software, standardized report databases), access to hospital clinical information systems, and archiving, which may necessitate additional processing capabilities. Recommended specifications Dual or quad core processor greater than 2 GHz 8 to 16 GB RAM Discrete graphics card Network connectivity: Gigabit network interface card Hard drive large enough for installation of required software

●

●

●

J. Monitor display: Acquisition machines (2,6) Fixed and portable installations should have a monitor in the room for EEG set-up and QEEG review by ICU personnel. Screen size of at least 17” diagonal is recommended (greater than 20” preferred). Monitor resolution of at least 1,280 × 1,024 pixels is recommended.

●

● ●

K. Monitor display: Review stations (2,6) Review stations require large high-resolution monitors for accurate EEG interpretation. Screen size of at least 20” diagonal is recommended (greater than 24” preferred). Monitor resolution of at least 1,600 × 1,200 pixels is recommended (widescreen 1,920 × 1,200 pixels). More expensive monitors such as Wide Quad eXtended Graphics Array (WQXGA) with 2,560 × 1,600 pixels and Wide Quad Ultra eXtended Graphics Array (WQUXGA) with 3,840 × 2,400 pixels allow display of every stored value at a 256-Hz sampling rate. Dual monitors may be necessary for simultaneous review of EEG and video data, and especially for simultaneous review of raw EEG and QEEG trends.

●

● ●

●

●

L.

User interfaces ●

●

●

●

Keyboard and mouse remain the standard interface devices for EEG. The interface must make it easy for nurses and other ICU personnel to enter annotations, move the camera, and so on, without interfering with the ongoing EEG recording. Touchscreen monitors with large buttons for common tasks are easier for nurses to use, but are cumbersome for large amounts of text entry, and may be difficult to adequately disinfect. Security: Health Insurance Portability and Accountability Act (HIPAA) regulations require individual user log-ons for clinical systems containing protected health information. Security software should provide an audit trail for changes to EEG data. Bedside systems may include a “transparent screen lock,” which locks the computer, but allows continued viewing of EEG data, video, and QEEG.

8

Part I

Technical Issues

M. Other hardware and/or cables ●

●

●

Inputs for other physiologic monitoring devices are preferred (e.g., pulse oximetry, blood pressure, ICP, temperature, respiratory effort, brain tissue oxygenation). These data streams must be time synchronized with EEG. Some ICU EEG machines are true multifunction devices with multimodality monitoring capabilities. It is important to ensure that adequate channels are available for EEG recording, and that additional software (seizure detection, QEEG) can be installed on the device.

N. Software ●

●

●

●

Patient database A central patient and study database allows management of patient files, archiving, and report generation. Databases can synchronize information across the hospital network, so that all local machines have up-to-date clinical information. Databases may also be linked to the hospital electronic medical record to allow import of relevant patient demographic data. EEG software Should be easy to use by both EEG and ICU personnel. Some systems have separate display modes for EEG and ICU staff, with a simplified interface for ICU personnel. Essential functions include the ability to annotate an ongoing recording, as well as move the camera. Software should include a “look back” feature, which is the ability to review an already-recorded EEG without the need to interrupt ongoing recording. Many systems now include remote access to the “live” EEG session, which allows for instant messaging features as well as system adjustments from remote locations (e.g., change montage, move camera). Other useful software features: Ability to automatically stop the EEG recording at a specified interval or time, and automatically beginning a new day’s recording. Color-coded displays of 10–20 system highlighting electrodes with significant artifact so ICU staff can provide easy maintenance of affected electrodes. Automatic synchronization of common settings (montages, recording protocols) from a central location to local acquisition machines. For research purposes, software should have functions to “deidentify” EEG data and to save in an open-source EEG data format. Spike and seizure detection software Current commercially available spike and seizure detection software is not optimal for detection of common seizure patterns seen in the ICU, and little data are available on their sensitivity and false-positive rates. May produce many false alerts which prompt inappropriate treatments, and are bothersome to ICU personnel. Automated artifact identification and reduction Online (as data is acquired) artifact reduction uses a variety of source decomposition techniques (e.g., spatial filtering, principal component analysis, independent component analysis) to break EEG signals down into individual components that represent EEG and others that represent artifact. Once identified, the artifactual components are removed, and the remaining “clean” EEG

Chapter 1 Equipment for EEG Acquisition and Review

9

●

signal is recomposed. Artifact rejection is particularly important prior to QEEG analysis. QEEG software (see Chapters 27–29) (8) Allow graphical display of EEG parameters over long time periods (hours). Can be displayed at the bedside, as well as at a central monitoring station in the ICU or EEG lab. Many types of commercial QEEG software are available, and most are proprietary. The main features to consider are ease of use, how well integrated the QEEG software is with EEG acquisition software, and accuracy of event detection and artifact rejection algorithms. Some systems can be configured to send alerts and images of corresponding trends and raw EEG via email.

III.

FURTHER CONSIDERATIONS/REMAINING QUESTIONS

A. Cost The cost of computers and video recording components continues to decrease. EEG equipment has become more portable and easier to install owing to development of standard network architecture as opposed to proprietary cabling. Software (special features, seizure detection, QEEG) may add a substantial amount to the cost of ICU EEG equipment. Information technology and biomedical engineering support is necessary. Installing, configuring, maintaining, and updating ICU EEG equipment and software requires significant IT and biomedical resources. Large ICU EEG monitoring programs may require dedicated IT and biomedical personnel.

● ●

●

●

B. Multimodality monitoring ● ●

●

Device interoperability is an important issue in ICU EEG. Polygraphic and multimodality data may be clinically useful, by aiding identification of cerebral states (e.g., sleep), recognition of artifacts, and confirmation of EEG abnormalities by correlation with other physiologic changes. Techniques of multimodality recording are discussed in Chapter 39. Optimally, all physiologic patient data, as well as data from ventilators, cooling devices, and IV pumps, would be able to be incorporated into a single time-synchronized data stream.

References 1. Sinha SR, Sullivan L, Sabau D, et al. American Clinical Neurophysiology Society Guideline 1: minimum technical requirements for performing clinical electroencephalography. J Clin Neurophysiol. 2016;33(4):303–307. https://www.acns.org 2. Halford JJ, Sabau D, Drislane FW, Tsuchida TN, Sinha SR. American Clinical Neurophysiology Society Guideline 4: recording clinical EEG on digital media. J Clin Neurophysiol. 2016;33(4): 317–319. https://www.acns.org 3. Kuratani J, Pearl PL, Sullivan L, et al. American Clinical Neurophysiology Society Guideline 5: minimum technical standards for pediatric electroencephalography. J Clin Neurophysiol. 2016;33(4):320–323. https://www.acns.org 4. Stecker MM, Sabau D, Sullivan L, et al. American Clinical Neurophysiology Society Guideline 6: minimum technical standards for EEG recording in suspected cerebral death. J Clin Neurophysiol. 2016;33(4):324–327. https://www.acns.org

10

Part I

Technical Issues

5. American Clinical Neurophysiology Society. Guideline 12: guidelines for long-term monitoring for epilepsy. J Clin Neurophysiol. 2008;25(3):170–180. 6. Herman ST, Abend NS, Bleck TP, et al. Consensus statement on continuous EEG in critically ill adults and children, Part II: personnel, technical specifications, and clinical practice. J Clin Neurophysiol. 2015;32(2):96–108. https://www.acns.org 7. Kull LL, Emerson RG. Continuous EEG monitoring in the intensive care unit: technical and staffing considerations. J Clin Neurophysiol. 2005;22(2):107–118. 8. Moura LMVR, Shafi MM, Ng M, et al. Spectrogram screening of adult EEGs is sensitive and efficient. Neurology. 2014;83(1):56–64.

Additional Reading Alvarez V, Rossetti AO. Clinical use of EEG in the ICU: technical setting. J Clin Neurophysiol. 2015;32(6):481–485. Fisch BJ. Digital and analog EEG instruments: parts and functions. In: Fisch & Spehlmann’s EEG Primer: Basic Principles of Digital and Analog EEG. 3rd ed. Amsterdam: Elsevier, Ltd. 1999. 35–72. Guerit JM, Amantini A, Amodio P, et al. Consensus on the use of neurophysiological tests in the intensive care unit (ICU): Electroencephalogram (EEG), evoked potentials (EP), and electroneuromyography (ENMG). Neurophysiol Clin. 2009;39(2):71–83. Wartenberg KE, Mayer SA. Multimodal brain monitoring in the neurological intensive care unit: where does continuous EEG fit in? J Clin Neurophysiol. 2005;22(2):124–127.

2 Electrodes and Montages Jennifer L. Hopp and Chalita C. Atallah

IN THIS CHAPTER

General principles of electrode placement Electrode types (comparison of invasiveness, reusability, imaging compatibility) Commonly used montages ● ●

●

KEY POINTS

Several electrode options exist for use in ICU EEG monitoring that provide high-quality recordings with good resolution. Cost, EEG technologist time, and ease of use are key factors in the selection of EEG electrodes. The standard 10–20 system for electrode placement is still widely used although other montage options exist. ●

●

●

I.

BACKGROUND

A. Continuous EEG (cEEG) monitoring in the ICU is a continually growing and developing diagnostic modality ●

● ●

Digital EEG recording with simultaneous video recording is now widely available, and is considered the standard for diagnosis of subclinical seizures. Digital video cEEG provides good temporal and spatial resolution. Current techniques and equipment for cEEG monitoring in the ICU offer real-time review and options for remote access for data analysis.

B. There are specific considerations in the selection of EEG electrodes and montages that are unique to the ICU setting ●

Critically ill patients often undergo urgent neuroimaging procedures, but traditional gold cup electrodes are not compatible with MRI and CT and, therefore, must be removed prior to imaging. 11

12

Part I

Technical Issues

There is often the need to acquire EEG data quickly and therefore rapid electrode application is desired, but standard electrode placement by trained neurodiagnostic staff can be time consuming, especially if staff must be “called in” for the procedure. Patients may require prolonged EEG monitoring resulting in risk of skin breakdown. Infection risk is increased in critically ill patients, particularly if there are open head wounds. ●

● ●

BASICS

II.

A. EEG electrodes ●

●

Electrodes—general principles and placement Technical guidelines and consensus statements should be used for the selection, placement, and maintenance of electrodes during cEEG (1,2). Various electrode materials are available such as tin, silver, silver-silver chloride, gold, platinum, stainless steel, and conductive plastic. Silver-silver chloride and gold are most commonly used in cEEG monitoring. The 10–20 International System is the standard for electrode placement (3,4). Concurrent EKG should be recorded with the option of additional channels for electromyogram (EMG), electrooculogram (EOG), or respiratory belt to help troubleshoot artifact or correlate with clinical findings. Electrodes should be applied after they are adequately disinfected, or disposable electrodes can be used as an alternative. Conductive materials are used to adhere the electrodes and lower the impedance at the electrode–skin junction. Although choice of material may be predicated on requirements of the ICU, collodion (glue) is often considered the preferred method of electrode adherence. Electrode types—noninvasive versus invasive (Figure 2.1) Noninvasive electrodes are commonly used in centers performing cEEG. Electrode application methods for noninvasive recordings include individual electrodes as well as cap, net, and template systems. Cup electrodes are commonly used, but electrodes are also available in web/spider forms.

(B)

(C)

(A) FIGURE 2.1 Electrode options. (A) Electrode template, (B) CT compatible/MRI conditional electrodes (note, electrodes are disconnected at the yellow hub during transport for neuroimaging. The red hub, clearly marked not MRI safe, remains with the head box), (C) subdermal wire electrode. Source: Figure 2.1A from Jordan NeuroScience, Inc. Used with permission.

Chapter 2 Electrodes and Montages

13

Individual electrodes allow modification and adjustment of placement by a technician to accommodate issues such as craniotomy scars or other equipment used for cerebral monitoring. Caps, nets, or templates may be used when there are needs for rapid setup and acquisition of EEG. Electrode caps, nets, and templates can be set up by non–EEGtrained personnel. Caps are an EEG electrode application technique that includes embedded or detachable leads in a removable cap, while a net is a similar but more open system. Templates are used as a substitute for head measurement by the technician and provide a typically color-coded map for accurate electrode placement. There are reusable and disposable options which may have cost differences compared to individual electrodes. Limitations in the ICU setting are significant and include the lack of ability to modify electrode placement in settings when scalp access is paramount such as with wound care or intracranial monitoring devices. Advantages may include relative ease of use, particularly for non–EEGtrained personnel, as well as the lack of adhesive materials needed for application, but this may not be enough to warrant their use in ICU settings. One study (5) in 32 patients in an ICU setting showed that the use of an EEG template by non-EEG staff resulted in no clinically significant differences in impedance or quality of recordings compared to technologist-applied electrodes. The study also demonstrated that the template shortened time to initiation of EEG recording by 3 hours. Study limitations included the lack of evaluation of quality of long-term recordings, defined as greater than 8 hours. Invasive electrodes are an alternative to noninvasive electrodes although many are not currently recommended for routine clinical use. Subdermal needle electrodes are single-use, disposable electrodes that are used in some centers for long-term recordings in unconscious patients. Advantages include speed of application, relative lack of scalp breakdown, and stable recordings. Disadvantages to needle or wire electrodes include risk to technicians during placement, higher impedance, and potential patient discomfort. They may be more typically used for shorter length recordings. Subdermal wire electrodes are single-use, disposable Teflon-coated silver wire electrodes. Advantages include reduction of skin breakdown and better recording of EEG than disc or needle electrodes, as well as the ability for some to be CT and/or MRI compatible. Disadvantages include cost and discomfort for some patients. Pressure-placed electrodes are a newer form of invasive electrodes that provide an alternative to needle or wire electrodes. Advantages include rapid application without the need for skin prep or adhesives. There is also less risk to technicians than with needle electrodes. Disadvantages include cost and skin irritation.

14

●

●

Part I

Technical Issues

Intracranial grid/strip/depth electrodes have been used for cEEG monitoring, but are not commonly utilized in the ICU setting. Electrodes—reusable versus disposable Most centers use reusable electrodes that are cleaned and disinfected after each use. Specific measures must be taken after recordings on patients with transmissible diseases. Disposable electrodes provide an option to reduce infection risk as well as technologist’s time in cleaning electrodes, but may be associated with higher costs. Cost effectiveness must be determined for each center and may include options of utilizing both reusable and disposable electrodes for different patients in the ICU. Electrodes—imaging compatibility Many patients in the ICU undergoing cEEG require urgent and recurrent imaging with CT and MRI during their ICU stay, and imaging-compatible electrodes should be considered (6,7). Removal and reapplication of electrodes in the ICU may contribute to skin breakdown and require additional EEG technician time. Traditional gold cup EEG electrodes are not CT or MRI compatible owing to their effects on image quality and/or safety concerns. Gold cup EEG electrodes are typically considered to be safe in CT scanners, but create artifact owing to deflection of x-rays that may significantly affect the images (Figure 2.2). EEG electrodes with ferrous or magnetic material also cause susceptibility artifact on MRI images. EEG electrodes with ferrous materials or long leads are not considered to be safe in MRI owing to heating and possible physical movement of the electrodes in the scanner.

FIGURE 2.2 CT artifact from metal/gold cup electrodes. Typical “starburst” artifact pattern seen on CT with use of traditional metal/gold cup electrodes.

Chapter 2 Electrodes and Montages

15

Alternatives to traditional electrodes that are imaging compatible include CT compatible/MRI conditional electrodes, electrode caps, and invasive electrodes. Low-density, plastic, or nonmetal (carbon) electrodes can be used to avoid “starburst” artifact on CT. MRI-conditional electrodes are nonmagnetic with short electrode wires and specialized connectors. EEG and MRI technicians must be trained in appropriate techniques. Of note, not all commercially available MRI-compatible electrodes are U.S. Food and Drug Administration (FDA) approved.

B. Montages ●

●

Montages are designed on the basis of standard electrode placement and should conform to American Clinical Neurophysiology Society (ACNS) guidelines (8). The international standard for scalp electrode number and placement is the 10–20 system with 16 electrodes used at minimum. The use of fewer than 16 electrodes may result in poor quality owing to inadequate spatial sampling and difficulty in identification of artifact. A modified 10–10 system that uses additional electrodes is also accepted as a standard and is used more often for localization for epilepsy surgery, although it may also be useful for localization of frontal lobe epilepsy. The 10–10 system provides better spatial resolution than the 10–20 system, but may be less practical in the ICU setting. Limitations of the 10–10 system include increased time and effort for electrode placement, increased cost, and reduced availability of jackboxes to support a system with additional electrodes. Most recordings also include a single channel ECG. Some recordings may also include channels for airflow and respiratory monitoring or EMG. An isolated ground electrode should always be placed and never connected to the EEG equipment or earth ground. Montages are a method of organizing electrical activity for lateralization and localization. EEG waveforms are a representation of the difference in electrical potentials between two electrodes in a pair and a montage helps to organize those waveforms for visual analysis. Initial recordings are made from a referential montage. The reference electrode should be an additional electrode beyond those placed for the 10–20 or 10–10 system. Common placement of the reference electrode is between Cz and Pz. Alternatives include Cz or an average reference constructed from all cerebral electrodes. Montages can be reformatted using digital systems. Montages can be modified to accommodate limitations in electrode placement due to skull defects or other intracranial monitoring equipment. Modifications should be symmetric if possible and clearly labeled by the EEG technician. Although the number of possible montages is limited primarily by the number of electrodes and electrode positions (21 possible montages in the 10–20 system), standardization is useful to provide a common “language” and method of information exchange among technicians and electroencephalographers. Unique montages are used for neonatal EEG and studies requested for determination of electrocerebral inactivity.

16 ●

●

●

Part I

Technical Issues

Standard guidelines for nomenclature of each electrode placement system are available and should be routinely used (8). There are variations in nomenclature, but standard terminology includes electrodes at positions Fp1/2, F3/4, C3/4, P3/4, F7/8, T3/4, T5/6, Fz/Cz/Pz, A1/2 with a ground/reference pair. Some centers use reduced montages and coverage for seizure detection in the ICU setting (Figure 2.3). Most reduced coverage montages include 10, 11, or 13 electrodes. In the ICU setting, where seizure detection is more important than precise localization and mapping, reduced montages may be sufficient. Several studies have examined the use and sensitivity of reduced coverage montages. Kolls et al. (5) evaluated the use of a “hairline EEG” reduced montage to determine the sensitivity of reduced coverage in the screening of nonconvulsive status epilepticus (NCSE) (9). EEG data was reformatted with electrodes Fp1/2, F7/8, T3/4, and T5/6 in three 6-channel montages and compared with original EEG interpretation. Sensitivity of seizure detection was 72% and authors did not recommend this format to screen for NCSE. Karakis et al. compared the sensitivity of a 7-electrode montage (Fp1/2, T3/4, O1/2, Cz) with a hairline montage described by Kolls et al. with 38 reformatted records reviewed by blinded attending physicians and residents (9). Reduced Electrode Montage-RM (9 Electrodes)

Full Electrode Montage-FM (19–20 electrodes)

FP1

F7

T7

FP1

FP2

F8 F3

FZ

C3

P7

FP2

F4

P3 O1

PZ OZ

T8

C4

CZ

P4

T7

C3

CZ

C4

T8

P8

O2

Spatial Zones 1. Left fronto-temporal zone (LFTZ) 2. Left parieto-occiptal zone (LPOZ) 3. Central zone (Cz) 4. Right fronto-occiptal zone (RFOZ) 5. Right parieto-occiptal zone (RPOZ)

O1

: FP1, F7, F3, T7 : P3, P7, O1 : C3, CZ, C4, FZ : FP2, F8, F4, T8, PZ : P4, P8, O2

O2

: FP1, T7 : O1 : C3, CZ, C4 : FP2, T8 : O2

FIGURE 2.3 Comparison of full vs. reduced electrode montages. FM and RM using the 10–20 system, and spatial zones of the scalp surface. FM, full electrode montage; RM, reduced electrode montage. Source: Reproduced with permission from Tekgul H, Bourgeois BF, Gauvreau K, Bergin AM. Electroencephalography in neonatal seizures: Comparison of a reduced and a full 10/20 montage. Pediatr Neurol. 2005;32(3):155–161.

Chapter 2 Electrodes and Montages

17

The average sensitivity for seizure detection was 92.5%, with specificity of 93.5% for the 7-electrode montage. The average sensitivity and specificity for the hairline montage was 85% and 97%. A recent study concluded that a 10-electrode reduced montage did not affect EEG interpretation or clinical prognosis in a group of 142 patients with postanoxic encephalopathy (10). EEG classification and prognosis were compared in the reduced 10-channel montage to the full 21-electrode montage. Interobserver agreement was good between reviewers who evaluated EEG using each montage.

III.

FURTHER CONSIDERATIONS/REMAINING QUESTIONS

A. Choice of electrodes ●

●

●

The cost of EEG electrodes, whether noninvasive or invasive, continues to decrease and may offer the opportunity for a greater number of centers to offer cEEG monitoring. Increasing options for disposable electrodes may lead to the need to balance issues of patient safety and technician time with environmental concerns about disposable materials. The growth of imaging-compatible electrodes may improve the ability for patients to get urgent scans in the ICU and may improve technician efficiency.

B. Options for montages and additional electrodes ●

Although there have not been many changes in recent years with regard to the standardization of montages for cEEG, options are increasing for the addition and use of multimodality monitoring in the ICU and the correlation of these physiologic parameters with EEG signals.

References 1. Sinha SR, Sullivan L, Sabau D, et al. American Clinical Neurophysiology Society Guideline 1: minimum technical requirements for performing clinical electroencephalography. J Clin Neurophysiol. 2016;33(4):303–307. doi:10.1097/WNP.0000000000000308 2. Herman ST, Abend NS, Bleck TP, et al. Consensus statement on continuous EEG in critically ill adults and children, Part II: personnel, technical specifications and clinical practice. J Clin Neurophysiol. 2015;32(2):96–108. doi:10.1097/WNP.0000000000000165 3. Acharya JN, Hani A, Cheek J, et al. American Clinical Neurophysiology Society Guideline 2: guidelines for standard electrode position nomenclature. J Clin Neurophysiol. 2016;33(4):308– 311. doi:10.1097/WNP.0000000000000316 4. Kuratani J, Pearl PL, Sullivan L, et al. American Clinical Neurophysiology Society Guideline 5: minimum technical standards for pediatric EEG. J Clin Neurophysiol. 2016;33(4):320–323. doi:10.1097/WNP.0000000000000321 5. Kolls B J, Husain A. M. Assessment of hairline EEG as a screening tool for nonconvulsive status epilepticus: response to Bubrick et al. Epilepsia. 2007;48(12):2375. 6. Vulliemoz S, Perrig S, Pellise D, et al. Imaging compatible electrodes for continuous electroencephalogram monitoring in the intensive care unit. J Clin Neurophysiol. 2009;26(4):236–243. 7. Herman ST, Abend NS, Bleck TP, et al. Consensus statement on continuous EEG in critically ill adults and children, Part II: personnel, technical specifications and clinical practice. J Clin Neurophysiol. 2015;32(2):96–108.

18

Part I

Technical Issues

8. American Clinical Neurophysiology Society Guideline 3: proposal for standard montages to be used in clinical EEG. J Clin Neurophysiol. 2016;33(4):312–316. doi:10.1097/ WNP.0000000000000317 9. Karakis I, Montouris GD, Otis JA, et al. A quick and reliable EEG montage for the detection of seizures in the critical care setting. J Clin Neurophysiol. 2010;27(2):100–105. doi:10.1097/ WNP.0b013e3181d649e4 10. Tjepkema-Cloostermans MC, Hofmeijer J, Hom HW, et al. Predicting outcome in postanoxic coma: are ten EEG electrodes enough? J Clin Neurophysiol. 2017;34(3):207–212.

Additional Reading Hirsch LJ, Brenner RP, eds. Atlas of EEG in Critical Care. Chichester, UK: John Wiley & Sons; 2010.

3 Networking, Remote Monitoring, and Data Storage Joshua Andrew Ehrenberg

IN THIS CHAPTER

Review of different types of network configurations Comparison of remote EEG monitoring and review systems Options for storage of digital EEG and video ● ● ●

KEY POINTS

There are various options for network configurations including stand-alone EEG networks, virtual local area networks (VLANs), and facility-integrated networks. Major considerations when choosing a method of remote access to EEG data include speed and ease of access, security, cost, resources available for maintenance, and number of concurrent users. Remote monitoring options include desktop mirroring, terminal services, and virtualized review systems. There are various archiving and data storage strategies that can be utilized to store the vast amount of data that is acquired with continuous EEG and video recording. The best method of data storage depends on what resources are available (time and equipment) as well as how much of the data is considered necessary to keep for prolonged periods. ●

●

●

I.

BACKGROUND

A. Growth of networks in health care settings and clinical neurophysiology ●

Computer networks and remote monitoring have been increasing in use in clinical neurophysiology.

19

20

●

●

Part I

Technical Issues

Computer networks in the health care setting have become essential. The majority of the growth has been driven by the need to view radiologic images, examine lab results, and create system-wide electronic medical records. Over the past few decades, local networks in epilepsy monitoring units (EMUs) and remote monitoring in intraoperative monitoring (IOM) have become commonplace. EMU implementations In the traditional EMU setting, networks are usually physically separate from the hospital or facility network, and have no shared connections. The EMU network usually connects to EEG recording units, data storage servers, and reading stations. This physically separate network configuration allows for collection of large amounts of video and EEG data as well as fast data transmission without impacting the hospital network. The disadvantage to this configuration is the inability to view data remotely, even from physicians’ offices within the hospital, unless directly connected to the EMU local network. Remote monitoring has been utilized in IOM for many years. IOM monitoring is typically arranged in a point-to-point configuration where the reviewer is connected directly to a single data acquisition unit and able to view data in real time. Remote monitoring in IOM was originally accomplished through phone line data connection, but is currently through a facility network where viewing takes place from a separate location, either a physician’s office or sometimes as distant as other states. This point-to-point configuration is not as effective for prolonged EEG monitoring where there are typically multiple patients, multiple EEG reviewers, and the need to review previously recorded data.

B. Network configurations ●

There are many network configurations and remote monitoring options that can be utilized in ICU continuous EEG (cEEG) monitoring. As with any aspect of clinical neurophysiology, there is no perfect design any more than there is a perfect montage, so applying knowledge of strengths and weaknesses as well as available resources is vital to determining the optimum design.

C. Data storage ●

EEG data storage has undergone many changes over the years. In the past, data storage has ranged from collections of piles of paper in large storage rooms to stacks of video cassette tapes. One of the major benefits of digital EEG is the ease of storing large amounts of data on fairly small media. This began as optical disks, then archiving to CD and later to DVD. More recently, storage to large external USB hard drives and “centralized” network storage has become popular. Recently, “cloud” storage has become prevalent which refers to a network, Internet-based service for hosting data storage across multiple sites. Advantages include ease of use, scalability to needs (“pay as you use”), and access from multiple devices. However, patient privacy and data security must be thoroughly evaluated. Data breaches, “ransomware,” and data integrity

Chapter 3 Networking, Remote Monitoring, and Data Storage

21

(speed of use for real time, and redundancy of raw data transfers) are serious concerns that must be weighed. Regardless of which data storage media is used, it is vital to ensure appropriate safeguards are in place to protect patient privacy (e.g., encryption), but these should be weighed against increased complexity and additional steps (1). Data file sizes Data file sizes can range from 1 or 2 GB (without video) to 20 or 30 GB (with video) per patient per day. “Clipped” segments are much smaller, with only small time spans of EEG and video data stored, usually for significant events such as seizures or representative samples of background patterns. One of the problems with ICU cEEG data storage is that activity not seen as significant now might later be decided to be important. However, storing the entire 20- to 30-GB file of patient data each day is cost prohibitive for most institutions. ●

II.

BASICS

A. Network configurations Networks are interconnections between computers and computerized devices, including EEG acquisition units, reading stations, data servers, desktop computers, switches, and routers. Network configurations include stand-alone, virtual local area network (VLAN) and full integration into a facility network. Each has individual strengths and weaknesses (Table 3.1). Stand-alone networks (Figure 3.1) Stand-alone networks were the most common implementation in EMUs a decade ago. It is a configuration where all of the involved computer devices are connected to each other but not connected to any external network. Early computer EEG systems typically had internal hard drives that could store no more than a few days of continuous data, so recording units were connected to large data storage servers that were quite costly and had to be mounted in a large network rack that stood 4 to 5 feet in height.

●

●

TABLE 3.1

Comparison of Different Network Configurations

NETWORK CONFIGURATION

SPEED

TECHNICAL IT STAFF

COST

Stand alone

High

Low

VLAN

Moderate

High

Facility network

Fluctuates Moderate

REMOTE CAPABILITIES

MULTIPLE FACILITIES

Moderate Low

Limited

Limited

High

High

Yes

Yes

Low

Moderate

Yes

Yes

IT, information technology; VLAN, virtual local area network.

COMPLEXITY

22

Overall network or Internet

Network card in back of computer preset MAC address: aa:23:45:67:89:ab Computer system settings (using example names and numbers) Computer Name: EEGreader1 Fully qualified domain name: EEGreader1.localEEG.edu IP address: 192.168.1.6 Subnet mask: 255.255.0.0 Gateway: 192.168.1.1

Additional network card in back of computer preset MAC address: 01:23:45:67:89:ab Computer system settings (using example names and numbers) Computer Name: EEGreader1 Fully qualified domain name: EEGreader1.hospital123.edu IP address: 100.35.219.6 Subnet mask: 255.255.0.0 Gateway: 100:35:219.1 Router, switch or hub in EEG area

Data file server with system settings Computer name: EEGfileserver Fully qualified domain name: EEGfileserver.localEEG.edu IP address: 192.168.1.59 Subnet mask: 255.255.0.0 Gateway: 192.168.1.1

Stand-alone EEG network

Router, usually in a network “closet” nearby

Higher level router, usually in a central IT network location Overalll network or Internet

Network card in back of computer preset MAC address: 01:23:45:67:89:ab Computer system settings (using example names and numbers) Computer Name: EEGreader1 Fully qualified domain name: EEGreader1.hospital123.edu IP address: 100.35.219.6 Subnet mask: 255.255.0.0 Gateway: 100:35:219.1

Router, possibly in same rack

Integrated facility EEG network

FIGURE 3.1 A graphical representation of stand-alone versus integrated facility EEG networks.

Data file server with system settings Computer Name: EEGfileserver Fully qualified domain name: EEGfileserver.hospital123.edu IP address: 100.35.220.59 Subnet mask: 255.255.0.0 Gateway: 100.35.22

Part I Technical Issues

Router, usually in a network “closet” nearby

Chapter 3 Networking, Remote Monitoring, and Data Storage

23

Benefits of a stand-alone network Maximum data speed as bandwidth is not shared with the facility network. Managed within an EEG department without the need for hospital information technology (IT) resources. Ability to be tailored to the specific needs of the facility. Security of the EEG data from external access. Drawbacks of a stand-alone network No remote monitoring capability. Lack of flexibility as the only recording locations are predefined and separately wired patient rooms. Hospital IT support staff is not usually available to identify and resolve technical issues. Virtual local area network VLANs are applicable to computers connected within an internal network. This is usually with computers and devices that are physically close to each other, within the same building, or on the same campus. Wide area networks (WANs) are applicable to computers that can connect over large distances, such as the Internet, or between different physical campuses of an institution. VLANs are “in house” networks or hospital intranets, whereas WANs utilize web-based portals over the Internet for access into the hospital’s VLAN. “Cloud” storage is an example of using a WAN. A computer’s network cable plugs into the wall and is usually connected on the other end (at some point) into a port on a router. This is somewhat analogous to an old-fashioned telephone switchboard. A router is a fairly simple computer that takes packets of information from computers and sends them, or routes them, to where they need to go, either to a system connected to the same router or to a system via a connected router. Advantages Department cost savings by utilizing existing hospital IT support for network management and troubleshooting. More flexible than a stand-alone network. Recording in additional locations does not require the physical installation of separate network cables. Capable of remote monitoring functions. Disadvantages Setting up and managing this type of configuration is relatively complex and requires significant IT resources. Facility network integration (Figure 3.1) Most facilities have existing networks connecting to various medical resources, imaging and electronic medical records, as well as email and the Internet. This is referred to as being on the facility “backbone.” This type of configuration allows for maximum flexibility recording from various locations and remote monitoring, but has some inherent drawbacks, such as fluctuations in speed due to shared bandwidth and increased IT security requirements. Facilities will usually have various rules for systems on their “backbone” to protect the integrity of the network as a whole, including antivirus software, automatic updates for the operating system, and user rights limitations.

24

Part I

Technical Issues

Other IT regulations include restricted program and system access for users.

B. Network configuration summary When setting up an ICU EEG monitoring program, it is important to consider network configuration options early in the planning process with consideration of both immediate needs as well as future growth. Sometimes, a proportionally higher initial cost will ensure that the IT infrastructure is in place to support use in future years.

●

●

C. Network speed ●

●

●

Modern facility network speeds range from 100 Mb/sec up to 1 Gb/sec, with external access at least 4 to 10 Mb/sec, even over wireless connections and the Internet. At speeds slower than 10 Mb/sec, only EEG can be seen and not in real time. At speeds of 100 Mb/sec, live EEG can be seen as well as recorded video. At speeds of 1 Gb/sec and faster, true real-time viewing of EEG and digital video is possible (Table 3.2). However, in network configurations that are directly on the facility network, these maximum speeds are shared by all the traffic that is concurrently being transmitted, so less-than-maximum speeds are typically seen.

D. Remote monitoring ●

The definition of remote monitoring in the context of ICU cEEG is the ability to see neurophysiologic data (both real time as well as previously recorded data) from a physically remote location.

TABLE 3.2 Speed Comparisons and the Realistic Expectations for Resulting Performance and Common Scenarios NETWORK SPEED

CONNECTION TYPE

REVIEW CAPABILITIES

TYPICAL SCENARIO

28.8 kb/sec or 56 kb/sec

Modems

Recorded EEG only

Direct connections from review station to acquisition

4 to 10 Mb/sec

Wireless connections across the Internet, CAT3 ethernet

Recorded EEG only

EEG review from home

100 Mb/sec

CAT5 ethernet, internal facility networks

Real-time or recorded EEG, recorded video

Facilities with network wiring installed more than 5 years ago

1 Gb/sec or faster

CAT5e, CAT6, newer internal facility networks

Real-time EEG and digital video

Modern facility networks

Chapter 3 Networking, Remote Monitoring, and Data Storage

Acquisition unit

25

Desktop mirroring

Terminal services

Citrix native

●

●

FIGURE 3.2 Remote monitoring options.

There are three major software options for implementation of remote monitoring: desktop mirroring, terminal services, and virtualized native Citrix applications (Figure 3.2). Desktop mirroring refers to the client computer simply viewing and interacting with the desktop running on the host computer. Terminal services consist of the client computer logging into a desktop that is running on a remote server. The remote server is also running other desktops. Virtualized Citrix native client (or other virtual machine/appliance software) is where the client computer connects to a dynamic computer created by the server. Software that is designed specifically to run in this type of environment and security infrastructure is called a native application. When assessing remote monitoring needs, there are six key aspects to consider: speed of use, ease of use, security, resource requirements, number of users, and functionality (Table 3.3). Speed of use Data files are usually physically located on either an acquisition station or a data server. Data files consist of not only the EEG but also of the digital video and other associated files. Remote monitoring requires the ability to run EEG reading software and also access to the EEG data files. If quantitative EEG (QEEG) software is also used, network access to these programs and files is also needed (except in desktop mirroring). Speed of use includes how fast the program opens (or programs in the case of using additional third-party software), how quickly individual EEG page displays can be reviewed, and how long it takes for one file to close and the next to open.

26

Part I

TABLE 3.3

Technical Issues

Comparison of the Key Aspects of Remote Monitoring Options DESKTOP MIRRORING

TERMINAL SERVICES

CITRIX NATIVE

Speed of use

Slow

Fast

Moderate to good

Ease of use

Poor

Moderate

Good

Security

Poor

Moderate

Good

Resource requirements

Low

Moderate

High

Number of users

1

1 to 3

Greater than 3

Functionality

Good

Moderate

Low to moderate

Example of how speed of use impacts EEG review time. Assume that there are 48 hours worth of video EEG data, located in four separate 12-hour files with QEEG data in 2-hour segments, and there are four patients to be reviewed. All QEEG segments are reviewed in addition to three 1-minute samples of raw EEG for each 2-hour QEEG segment. Fastest and slowest times for speed are based on anecdotal but real-world experiences and do not take into account additional time for the EEG reviewer to stop and analyze the EEG recording: Speed for program opening ranged from the fastest scenario at 10 seconds to the slowest scenario of 90 seconds. (The fastest speed of access was with an EEG review program over a terminal services environment with native trending, while the slowest program opening speed was with a Citrix native EEG review program, with third-party trending.) Page speed ranged from relatively instant for raw EEG page speed (up to 50 x) and QEEG segment speeds of less than 1 second per 2-hour segment to the slowest page speed of 2 pages/second and up to 1 minute per QEEG segment. Time to close one file and open the next ranged from 20 seconds in the fastest scenario to over 4 minutes in the slowest. Using the previous data, the fastest and slowest time to review these patients would be: FASTEST SCENARIO

SLOWEST SCENARIO

10 s (program opening)

90 s (program opening)

+5 s (to review 1 min of EEG)

+30 s (to review 1 min of EEG) (continued )

Chapter 3 Networking, Remote Monitoring, and Data Storage

27

(continued )

●

●

●

FASTEST SCENARIO

SLOWEST SCENARIO

x3 (1 min epochs reviewed per 2-hr QEEG

x3 (1 min epochs reviewed per 2-hr QEEG

section)

section)

x24 (2-hr QEEG sections)

x24 (2-hr QEEG sections)

+ 1 s (to “page” to next QEEG section)

+60 s (to “page” to next QEEG section)

x24 (2-hr sections)

x24 (2-hr sections)

+20 s (to close one file and open the next)

+240 s (to close one file and open the next)

x4 (12-hr files)

x4 (12-hr files)

x4 (patients)

x4 (patients)

Approx 30 min to review (31.27 min)

Approx 5.16 hr to review (310 min)

Ease of use Ease of use refers to how easy it is to open/access the program and open data files. This includes the number of times the mouse has to be clicked, and the number of times a user name and password have to be entered, and how many network addresses have to be remembered. The more integrated into the overall facility network, the easier remote monitoring program access should be. In the easiest real-world anecdotal example, a Citrix native application in an integrated networking system allows one entry of a user name and password, opening the application and selecting a patient, and double clicking a file to open it. In the most cumbersome example, a virtual private network (VPN) is used with remote desktop to a reading station or server, requiring sign-on to the VPN, running remote desktop connection, typing in an Internet protocol (IP) address, then signing in again through remote desktop connection, running the EEG application and selecting a patient, and double clicking on the file. Security Security in remote monitoring encompasses two separate concepts: (a) protecting the privacy of confidential patient data, and (b) controlling the ability of users to modify or delete data files. Facility networks are usually protected from access by anonymous outside users. Data traffic from outside the facility network that has not been authenticated is blocked. Resource requirements Vital resources include hardware, software, and technical staff to both implement and maintain the network. The easiest remote monitoring solution to implement is a desktop mirroring configuration. This usually requires no additional hardware, fairly inexpensive software, and minimal technical staff involvement.

28

●

●

Part I

Technical Issues

Terminal services remote monitoring requires a higher end application server, but software is included in most operating systems. Multiple concurrent users require additional terminal services access licenses. A moderate level of technical staff knowledge is needed but may be shared with hospital IT resources. For Citrix native EEG application implementation, at least one Citrix presentation server is needed, as well as EEG Citrix native application licenses which are usually more expensive than regular EEG application licenses. In addition, most Citrix native application implementations will require additional servers and software for redundancy in the event of system failure. Configuration and maintenance of Citrix implementations require a high-level IT technical staff. While the cost of Citrix has decreased over recent years, the increasing need for vendor-provided licensing, usually per user, has kept the overall cost of this solution relatively unchanged. Number of users In most large ICU EEG programs, there will likely be many concurrent users (epileptologists, fellows, residents, EEG techs), even more as cEEG services span multiple facilities. To a large degree, the number of concurrent users dictates the type of remote monitoring implementation that is warranted. Remote review with desktop mirroring is the slowest option and video review is not realistic. Most software desktop mirroring programs only allow one concurrent user. Therefore, desktop mirroring is only feasible with one remote user, where slow review speed is acceptable, and only one or two patients are reviewed at one time. Terminal services utilize a review station or application to which the user logs on. Each user receives a separate desktop presentation, and each desktop presentation shares the total system resources. The main drawback to a terminal services solution is that with more than three users, the shared resources are spread thin and many vendors do not support the configuration. Terminal services applications work best when fiscal and technical resources are available and only one to three concurrent users are expected. Citrix native applications utilize a Citrix presentation server to create “virtual” computers for each user. Programmatic speed is scaled to the speed of the network connection, which can prevent fluctuations in speed when reviewing but also limit maximal review speed. Citrix native remote review is best for more than three concurrent users, or when viewing EEG across multiple facilities. Functionality Functionality refers to what the remote user can accomplish including the ability to see live and recorded video and QEEG as well as perform manipulation of the recording unit or camera. For full functionality, a desktop mirroring scenario is best. Functionality using terminal services is comparable to an in-house EEG reading station. Citrix native implementations have the most limitations (largely dependent on existing hospital network architecture), but also have the lowest risk of users disrupting the acquisition unit.

E. Data storage ●

A strategy for data management is essential and involves understanding the size and types of files being acquired.

Chapter 3 Networking, Remote Monitoring, and Data Storage ●

●

●

●

●

●

29

It is important to realize that a single EEG recording consists of not just one file, but multiple linked files. There is usually one file for the EEG itself, in a vendor-proprietary format, which contains a numeric entry for each electrode for each time sample. This means that for a standard 16-channel montage, with a 500-Hz sample speed, there is an entry every 500th of a second for each electrode. There is also a separate set of digital video files as well as files that include synchronization information, annotations, and QEEG data. cEEG file sizes for standard 21-channel montage, at a sample speed of 500 Hz, per 12 hours range from 2 to 6 GB. The number of electrodes and the sample speed have a significant impact on this size, with a sample speed of 1,000 Hz doubling the file size. Digital video in an uncompressed format can run over 100 GB of data per hour, but with digital compression can reduce file sizes to 5 to 20 GB per 12 hours. MPEG4 format compresses video size by only recording pixels that change from the previous frame. So, in a video where a patient is laying completely still and there is no other movement, the file size will be relatively small. A video where there is a lot of constant movement will be much larger. NOTE: Flashing light, either from a monitor or the television, will create constant change in the brightness in the room, so an otherwise motionless scene will create a file as large as video with continuous motion. Complete video and EEG files should be stored for a period of time, but at some point will need to be archived for long-term storage. Approximately 150 to 200 patient days of data can be kept in entirety on a typical 1 to 2 TB data server. Most centers keep 30 or 60 days of patient data on local storage before moving to archive, unless the patient is still undergoing monitoring. Data storage strategies: How much data to keep? The first data storage strategy is similar to data storage in most typical EMUs. Each patient file is “pruned” to only include certain EEG and video segments of interest; either seizures, other significant events, or background samples. All other data are deleted. This leads to the smallest file size for archiving, but is the most limited in terms of later review. The second strategy is keeping all EEG and video data. ICU cEEG is a new enough technology that there is no consensus on what aspects of the data might prove important later. However, keeping all EEG and video data on every patient is almost logistically impossible with current storage limitations. The third data strategy is to keep all of the EEG data, but only video for significant events. The EEG is relatively small compared to the size of the digital video, and video data is less likely to be needed for future analysis. Data storage media (Table 3.4) There are many options for data storage with a wide range of advantages and disadvantages. External USB hard drives are more expensive than CD/DVD in terms of cost per megabyte/gigabyte. However, when considering EEG tech time for archiving, external USB hard drives are more cost effective. However, with both CD/DVD and external USB hard drives, the risk of file corruption should be considered as well as the possibility of physical loss of the media. Network-based solutions, either internally or over the Internet, have dropped in price over recent years. Owing to its ease of use, large and scalable storage capabilities, and data security (particularly from data breaches), this is the most commonly used data storage solution. “Cloud” storage would be an example of this design.

30

Part I

TABLE 3.4

Technical Issues

Comparison of Current Data Storage Options and Capabilities

MEDIA TYPE

SIZE

NUMBER OF PATIENT DAYS

CD

700 MB

Less than 1

Too small to be very useful for cEEG

DVD

4.7 GB

1 to 5 depending on how “pruned”

Must include the cost of the tech time involved in pruning (about 1 hr each), no backup of data

External USB hard drive

250 GB–2 TB

20 to 200

No backup of data in the event of drive failure

Network largescale storage

Relatively unlimited

Relatively unlimited

Cost has drastically decreased in recent years and has become standard at large centers

COMMENTS

Data format Most systems record EEG into proprietary formats. To view the data on another system it has to either be bundled with a program to view the data, or if using a different program on another computer, it must be converted into a format that can be imported by the other EEG program. There are three “universal” formats: EDF, ASCII, and TS1. EDF stands for European Data Format and can be imported by most systems. There are related formats, such as BDF and EDF +. ASCII is the raw text format of the file which is usually very large as no compression of the data is used. The American Clinical Neurophysiology Society has published more in-depth guidance on the format and interusability of EEG data (2). ●

III.

FURTHER CONSIDERATIONS/REMAINING QUESTIONS

A. Integration with electronic medical records ●

●

HL7 is a standard format for communication between medical systems that can allow orders placed in electronic medical records to populate EEG systems, and then data from the EEG to populate back to the medical record. Though it is relatively simple to implement, the hardware needed and the configuration costs are relatively high. Many hospital IT departments might be accepting of sending data to the EEG systems, but very reluctant to allow data back into the electronic medical record.

B. Rapidly changing technology ●

Computers, networks, and storage change at a very fast rate. According to Moore’s law, computer technology capabilities and speeds double every 2 years. This theory, developed in 1965, has held true through 2016 and will likely remain so through 2025 or even later.

Chapter 3 Networking, Remote Monitoring, and Data Storage

31

The current “cutting edge” of wireless networking with gigabit speeds will seem slow by tomorrow’s standards. System capabilities will also increase as there is more and more integration of the EEG network and systems into the main networks at facilities. Careful watch must be kept to balance ever-increasing costs associated with staying on the “cutting edge” with practical delivery of patient care.

ACKNOWLEDGMENT

My thanks to Sheri Richardson and Dave Huston for their contributions of computer expertise.

References 1. Maus D, Epstein CM, Herman ST. Digital EEG. In: Schomer DL, Lopes da Silva FH, ed. Neidermeyer’s Electroencephalography. Philadelphia, PA: Lippincott Williams and Walters Kluwer Business; 2011:119–142. 2. Halford JJ, Sabau D, Drislane FW, et al. American Clinical Neurophysiology Society Guideline 4: recording clinical EEG on digital media. J Clin Neurophysiol. 2016;33(4):317–319.

Additional Reading Herman ST, Abend NS, Bleck TP, et al. Consensus statement on continuous EEG in critically ill adults and children, Part II: personnel, technical specifications and clinical practice. J Clin Neurophysiol. 2015;32(2):96–108.

4 Staffing an ICU EEG Monitoring Unit Abdulrahman Alwaki, Joshua Andrew Ehrenberg, and Andres Rodriguez-Ruiz

IN THIS CHAPTER

American Society of Electrodiagnostic Technologists (ASET) and American Clinical Neurophysiology Society (ACNS) recommendations for technical staffing of an ICU EEG monitoring program Discussion of the roles and responsibilities of various technical and administrative staff needed to develop and maintain continuous EEG monitoring Examples of different staffing models that can be customized to fit the needs of institutions of various sizes and diverse patient populations ●

●

●

KEY POINTS

The field of ICU EEG monitoring has evolved considerably over the years, incorporating numerous technological advances. Experienced neurodiagnostic staff continue to be one of the most valuable yet limited resources needed for implementation of a successful continuous EEG (cEEG) program. Large cEEG programs require additional administrative staff, including information technology support, technical and medical directors, and EEG educators. Neurophysiologists who interpret cEEG recordings should have formal training, with emphasis in EEG recording in critically ill patients. There are a variety of staffing models that can be employed ranging from continuous monitoring by registered neurodiagnostic technologists to limited staffing options depending on specific institutional needs. ●

●

●

●

●

I. ●

32

BACKGROUND

The use of ICU EEG monitoring has risen considerably over the last decade at a rate of about 33% per year (1).

Chapter 4

Staffing an ICU EEG Monitoring Unit

33

In addition to technological advances, the need for more staff resources has spurred additional growth in the field. However, methods of EEG review and interpretation remain highly variable. A survey of EEG use in the ICU setting reported that the majority of physicians review each record two or more times per day, but most centers do not provide continuous EEG (cEEG) review by either a physician or neurodiagnostic technologist (2). Programs developing cEEG monitoring need to consider how to manage afterhours cEEG requests, address nursing concerns, communicate critical EEG results, and respond to urgent imaging needs. As ICU EEG monitoring has become a critical element of neurological care in this patient population, the development of different staffing models has become necessary. ●

●

●

●

II.

BASICS OF STAFFING ROLES AND RESPONSIBILITIES

A published consensus statement from the American Clinical Neurophysiology Society (ACNS) offers key recommendations regarding personnel, technical specifications, and clinical practice (3). The American Society of Electroneurodiagnostic Technologists (ASET) has also developed a guide for the expansion and implementation of institutional policies and procedures (4). These recommendations highlight qualifications and responsibilities of various levels of EEG personnel, standards for cEEG acquisition equipment, and suggestions for operating protocols. ●

●

●

A. Neurodiagnostic technologists The ACNS and the ASET outline minimal requirements for both basic and advanced levels within each neurodiagnostic technologist subspecialty (Table 4.1). Neurodiagnostic technologists perform hands-on initiation and maintenance of EEG recordings, and are expected to have general EEG technical knowledge. Required skills include the application of electrodes; efficient operation and placement of recording equipment; verification of network connectivity; identification of artifacts, including electrode malfunction; and the documentation of pertinent patient history and daily clinical changes. Electrode application sites and lead wires should be maintained daily, with a focus on preventing skin breakdown for which critically ill patients are at a high risk. It is critical for the technologist to perform a daily assessment for reactivity on each patient so any changes on their cEEG can be detected.

●

●

●

●

●

B. Neurodiagnostic specialists ●

●

●

It is important to emphasize that the role of the neurodiagnostic specialist is not to replace the ICU EEG-trained neurophysiologist, but rather to support and work under his or her direct supervision for the purpose of improving the efficiency and productivity of the department. Neurodiagnostic specialists require clinical knowledge and responsibility in addition to what basic technologists can be expected to provide (Table 4.1). The expectations for this role go beyond simple EEG pattern recognition, and include identification of changes in background activity, effects of medications on EEG patterns, and familiarity with the overall clinical status of the patient.

34

TABLE 4.1

Staffing Roles and Responsibilities ACNS JOB TITLE (4)

MINIMUM EDUCATION RECOMMENDATIONS

JOB DESCRIPTION

REGISTRATION

Neurodiagnostic technologist I (NDT I)

Neurodiagnostic technologist I (NDT I)

Associate degree or enrolled in neurodiagnostic program

Electrode application and maintenance

No registration

Neurodiagnostic technologist I (NDT I)

Neurodiagnostic technologist II (NDT II)

Associate degree or enrolled in neurodiagnostic program

Performs EEG under technical supervision

Eligible for registration in EEG by ABRET (R. EEG T.)

6 months NDT experience Neurodiagnostic technologist II (NDT II)

Neurodiagnostic technologist III (NDT III)

Associate degree or appropriate clinical experience

Perform EEG independently

Registration in EEG by ABRET (R. EEG T.)

ICU/cEEG specialist I

Neurodiagnostic specialist I (NDS I)

3 years of NDT experience, with 1–2 years in ICU cEEG

NDT III responsibilities: Identification of ictal and interictal patterns

Meets ASET National Competency Skill Standards for ICU EEG

Expertise in QEEG

ACNS: Certification in Long-Term Monitoring by ABRET (CLTM)

Notification of findings and descriptive analysis (continued )

Part I Technical Issues

ASET JOB TITLE (4)

TABLE 4.1

Staffing Roles and Responsibilities (continued ) MINIMUM EDUCATION RECOMMENDATIONS

JOB DESCRIPTION

REGISTRATION

ICU/cEEG specialist II with management duties

Neurodiagnostic specialist II (NDS II)

ASET: 2 years of ICU EEG experience

Development of technical policies and procedures

Certification in Long-Term Monitoring by ABRET (CLTM)

ACNS: 3 years of ICU EEG experience after CLTM

Supervision and training of NDT, nurses, and other ICU staff

ABRET, American Board of Registration of EEG Technologists, Neurodiagnostic Credentialing and Accreditation; ACNS, American Clinical Neurophysiology Society; ASET, American Society of Electroneurodiagnostic Technology, The Neurodiagnostic Society; cEEG, continuous EEG; QEEG, quantitative EEG. Source: From Ehrenberg JA, Rodriguez A, LaRoche SM. Staffing Considerations for ICU EEG Monitoring. 2017. With permission of Springer Nature.

Staffing an ICU EEG Monitoring Unit

ACNS JOB TITLE (4)

Chapter 4

ASET JOB TITLE (4)

35

36

Part I

Technical Issues

At a busy comprehensive epilepsy center, an average of four to six patients may be assigned to each neurodiagnostic specialist to ensure efficiency and quality of care. At least 3 years of EEG experience, including exposure to cEEG monitoring, and Certification in Long-Term Monitoring (CLTM) is recommended for a specialist position. Competency in the use of ACNS Standard Critical Care EEG (CC EEG) terminology is also crucial (5). ●

●

C. Non-EEG procedural staff ●

●

●

●

●

Non-EEG procedural staff include those who typically have their primary function in some other clinical role, such as nursing care, but who have received limited, targeted training to function as ancillary staff. Some care models include nursing staff to initiate EEG recordings after normal hours of operation which may include use of electrode templates or EEG caps to guide electrode placement or use of limited montages such as a “hair-line” recording (see Chapter 2) (6). Increased use of QEEG measures has made bedside monitoring for identification of significant EEG changes, including seizures, feasible for non-EEG trained staff. Studies have demonstrated that with minimal training, critical care personnel (i.e., attending physicians, fellows, and critical care nurses) are able to detect seizures using QEEG (7,8,9). In one study, little difference was observed in the sensitivity of seizure detection between neurophysiology fellows, critical care nurses, and attending physicians (9).

D. Neurophysiology administrative staff ●

●

●

●

Administrative staff support is crucial. Key administrative tasks include procedure billing, productivity tracking, patient appointments, and staff scheduling. While these team members usually support the entire EEG and epilepsy monitoring program, it is important to consider the additional workload presented by the development and growth of ICU EEG monitoring service. The neurophysiology technical director offers neurodiagnostic expertise pertaining to the technical aspects of developing and maintaining the cEEG program. Duties include input into selection of EEG equipment, oversight of service contracts, and ensuring technical standards and lab accreditations are in place. The technical director often bridges the gap between the priorities of the clinical team and objectives of hospital administration, which may not always be congruent. The neurophysiology medical director is typically a physician with a leadership role within the ICU EEG team and ensures that facility policies and procedures are in compliance with current medical standards. While managing day-to-day operations and logistics, the neurophysiology director’s overarching goal is keeping patient care a top priority while ensuring program growth and development. Information technology support staff members have become essential for the operation of an ICU EEG program. Information technology (IT) support works closely with the technical and medical directors to ensure security of computer data, as well as address network management, data storage, and remote monitoring.

E. Neurophysiologist/electroencephalographer ●

A team of neurophysiologists who have completed fellowship training in clinical neurophysiology with an emphasis on cEEG monitoring is recommended in order to provide expert interpretation of the long-term EEG recordings in critically ill patients.

Chapter 4

Staffing an ICU EEG Monitoring Unit

37

Physicians interpreting ICU EEG recordings are also expected to maintain certification by the American Board of Clinical Neurophysiology (ABCN) and/or the American Board of Psychiatry and Neurology (ABPN) with subspecialty in clinical neurophysiology or epilepsy (1). Training should include recognition of seizures, status epilepticus, ischemia, and other EEG findings seen in the setting of acute brain injury, including periodic and rhythmic patterns (1). Competency in the analysis and utility of QEEG is recommended, including knowledge of limitations of QEEG. In comprehensive centers, the attending neurophysiologist often works very closely in conjunction with fellows, technical staff, and other team members to make frequent interpretation more manageable. During after-hours and weekend shifts, physician staffing is often limited, so assistance from advanced neurodiagnostic technical staff becomes essential.

●

●

●

●

●

F.

Staffing models (Table 4.2)

●

●

Continuous monitoring staffing model (Table 4.2) From an ideal patient care perspective, continuous and real-time EEG interpretation would be provided 24 hours a day for each patient. This model requires a large investment of resources and is rarely employed. Typically, only comprehensive medical centers are able to provide this level of service where the burden of around-the-clock staffing can be shared. Resources at this level of service typically include in-house technologists and a team of neurodiagnostic specialists who monitor live EEG recordings in real time from a centralized workstation either within the hospital (typically in the ICU or epilepsy monitoring unit [EMU]) or from a remote location. Neurophysiologists (fellows or attendings) must be available 24/7 for interpretation of EEG findings determined to be of concern, and be able to convey these findings quickly to the clinical team responsible for patient care. A communication protocol is recommended to expedite the flow of information from the neurodiagnostic specialist to neurophysiologist to clinical team. Some comprehensive epilepsy centers offer a variation of this model, providing continuous, real-time monitoring during the day and evening, with only intermittent review throughout the night. Hybrid staffing model (Table 4.2) This model includes the use of both EEG-specific and non–EEG-specific staff. The EEG-specific staff normally provides coverage during daytime shifts while after-hours and weekend shifts are covered by either on-call EEG technologists or non-EEG trained staff. Utilizing on-call EEG technologists can be a cost savings (compared to 24/7 staffing) if overnight callbacks are infrequent. However, if EEG staff is routinely called in for urgent cEEG hookups, it is more cost effective to transition a continuous staffing model which also minimizes staff burnout. Non–EEG-trained staff who are expected to initiate EEG studies are required to possess basic skills of electrode placement (often employing limited montages or EEG templates or caps), EEG initiation, and troubleshooting. The employment of the hybrid model has been formally assessed and found to significantly increase the availability and speed of EEG initiation, with minimal impact on the short-term quality of EEG recording.

38

TABLE 4.2

Staffing Models HOSPITAL TYPE

CEEG VOLUME

STAFF REQUIRED

NUMBER OF ACQUISITION UNITS

IT REQUIREMENTS

Continuous

Large academic center

>6 per day

5+ NDT I/II/III

8+

Cloud-based or large external server

Large health care system

3+ NDS I/II

Dedicated IT support

3+ clinical neurophysiologists Neurointensivist or neurohospitalist support Hybrid

Small academic center

3 to 6 per day

3 to 4 NDT I/II/III

6 to 7

Medium-size local server

1 to 2 NDS I/II Midsize hospital

Shared IT support 1 to 3 clinical neurophysiologists General neurologist or neurohospitalist support

Limited

Small community hospital

≤2 per day

1 to 2 NDT I/II/III

2 to 3

1–2 clinical neurophysiologists May elect for externalizing EEG services Onsite or on-call general neurologists cEEG, continuous EEG; NDS, neurodiagnostic specialist; NDT, neurodiagnostic technologist. Source: From Ehrenberg JA, Rodriguez A, LaRoche SM. Staffing Considerations for ICU EEG Monitoring. 2017. With permission of Springer Nature.

Disk storage or small size local server

Part I Technical Issues

MODEL TYPE

Chapter 4

Staffing an ICU EEG Monitoring Unit

39

●

Implementation of this model requires additional training and workload for ICU nursing staff. If prolonged EEG monitoring is anticipated, it would be anticipated that electrodes initially applied by non-trained personnel via reduced montage or templates would be revised using standardized methods as soon as trained EEG staff are available. Limited staffing model (Table 4.2) This model can be considered in smaller institutions where less-frequent cEEG is expected, fewer trained technologists are available, and institutional finances are constrained. In smaller centers, cEEG interpretation and routine maintenance can be offered during daytime shifts, with little to no coverage after-hours or during weekend shifts. Often, this model serves as the initial step toward later development of a more comprehensive monitoring service. Critical patients (i.e., nonconvulsive status epilepticus) who require more frequent EEG review should be considered for transfer to a comprehensive epilepsy center that can provide the necessary cEEG monitoring and interpretation. On-call versus in-house staffing models The use of on-call versus in-house technical staff for coverage of after-hours services is both a clinical and financial decision for a program. Institutions will often utilize an on-call model during the initial program development and later transition to full-time in-house staff as patient volume grows. The on-call staffing model allows for expanded night and weekend coverage without significantly increasing the number of employed technical staff. Extensive overtime labor can become expensive, and frequent overnight callbacks can contribute to staff burnout. In-house staff provide continuous coverage without premium on-call costs and allow for rapid initiation of EEG monitoring. Maintaining in-house technical staff 24 hours a day may not be cost effective if patient volume is not adequate to support the cost of labor. A cost–benefit analysis should be done to compare the wage compensation of on-call staffing versus in-house staffing to determine the ideal model for a given institution’s patient volume. ●

G. Externalizing services ●

●

●

●

Contract EEG services can be utilized depending on the needs of the individual institution. Contract staff typically fill one particular level of experience or may include all levels of continuous EEG needs (ranging from technical services, such as electrode placement and EEG monitoring, to clinical services of EEG interpretation). Facilities may employ contract services for night or weekend shifts when in-house staffing is insufficient or when patient volume rises above normal capacity of in-house resources. Despite potential benefits of such an arrangement, limitations do exist and should also be considered. Technical staff initiating EEG recordings must be familiar with the inpatient environment of the institution, which may require significant training as well as credentialing.

40

●

●

●

Part I

Technical Issues

Neurodiagnostic specialists as well as neurophysiologists who are contracted for remote EEG monitoring and interpretation can more easily adapt to a wide range of inpatient settings, but they must be familiar with each hospital’s method of communication and reporting. For efficient utilization of external services, a small group of contracted specialists could be assigned to a singular institution so that repetitive training is at a minimum. Overhead costs of contract services should be considered as most exceed that of hiring a full-time, employed staff. Additional factors to consider include travel time, logistics of remote access, and networking.

III.

FURTHER CONSIDERATIONS/REMAINING QUESTIONS

A. Benefits of real-time neurodiagnostic technologist monitoring model Although identification of seizures and other relevant EEG changes in real time would seem to be the optimal model for patients who are critically ill, data is needed to establish the benefit of real-time monitoring. ●

B. Impact of limited staffing models ●

●

●

Data is needed to determine the impact on nursing services when cEEG responsibilities are added to the complex care already provided by critical care nurses. Trending and automated event detection can be utilized by non-EEG personnel to help identify important EEG changes, but they are imperfect. False-positive events could lead to excessive calls to neurodiagnostic technologists and neurophysiologists, or unnecessary use of medications. False-negative events could lead to inappropriate or delayed treatment. Although limited staffing models may not allow for continuous monitoring, the argument could be made that “some EEG monitoring is better than no EEG monitoring,” particularly in resource-poor areas.

References 1. 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. 2. Gavvala J, et al. Continuous EEG monitoring: a survey of neurophysiologists and neurointensivists. Epilepsia. 2014;55(11):1864–1871. 3. Herman ST, et al. Consensus Statement on Continuous EEG in Critically Ill Adults and Children, Part II. J Clin Neurophysiol. 2015;32(2):96–108. 4. ASET The Neurodiagnostic Society. Neurodiagnostic Practice Levels [cited June 9, 2015]. http:// www.aset.org/files/public/Neurodiagnostic_Practice_Levels.pdf 5. Hirsch LJ, LaRoche SM, Gaspard N, et al. American clinical neurophysiology society’s standardized critical care EEG terminology. J Clin Neurophysiol. 2013;30(1):1–27. 6. Kolls BJ, Lai AH, Srinivas AA, Reid RR. Integration of EEG lead placement templates into traditional technologist-based staffing models reduces costs in continuous video-EEG monitoring service. J Clin Neurophysiol. 2014 Jun;31(3):187–193. 7. Dericioglu N, Yetim E, Bas DF, et al. Non-expert use of quantitative EEG displays for seizure identification in the adult neuro-intensive care unit. Epilepsy Res. 2015;109:48–56.

Chapter 4

Staffing an ICU EEG Monitoring Unit

41

8. Swisher CB, White CR, Mace BE, et al. Diagnostic accuracy of electrographic seizure detection by neurophysiologists and non-neurophysiologists in the adult ICU using a panel of quantitative EEG trends. J Clin Neurophysiol. 2015;32(4):324–330. 9. Topjian AA, Fry M, Jawad AF, et al. Detection of electrographic seizures by critical care providers using color density spectral array after cardiac arrest is feasible. Pediatr Crit Care Med. 2015;16(5):461–467.

PART II: INDICATIONS

5 Status Epilepticus Sebastian Pollandt and Thomas P. Bleck

IN THIS CHAPTER

Current definitions of status epilepticus Clinical semiology of status epilepticus in the critically ill population Utility of continuous EEG monitoring for status epilepticus ● ● ●

KEY POINTS

Continuous EEG (cEEG) can provide invaluable information and guide appropriate patient management in the setting of status epilepticus (SE). cEEG is indicated in the following clinical situations: Following generalized convulsive status epilepticus (GCSE) in patients with persistent encephalopathy to assess for ongoing nonconvulsive status epilepticus (NCSE). For monitoring the response of SE to treatment, especially when using continuous intravenous anesthetic medications. To assess for NCSE in patients with unexplained coma or altered mental status. For determination of whether repetitive, involuntary movements represent SE versus nonepileptic events. ●

●

I. ●

BACKGROUND

The incidence of status epilepticus (SE) is estimated to be up to 61 episodes per 100,000 per year, with an overall mortality of 20% (range 1.9%–40%) (1). In the past two decades, there have been considerable advances in the understanding of its pathophysiology, causes, clinical features, EEG correlate prognosis, and treatment.

A. Definitions of SE ●

42

In 2015, the International League Against Epilepsy (ILAE) Task Force on Classification of Status Epilepticus proposed a conceptual definition that encompasses all types of SE (2):

Chapter 5

43

Status Epilepticus

SE is a condition resulting either from the failure of the mechanisms responsible for seizure termination or from the initiation of mechanisms which lead to abnormally prolonged seizures (after time point t1). It is a condition that can have long-term consequences (after time point t2), including neuronal death, neuronal injury, and alteration of neuronal networks, depending on the type and duration of seizures (Figure 5.1) (3).

●

The term t1 refers to the time point when a seizure should be considered prolonged and unlikely to terminate without intervention. The term t2 refers to the time point beyond which there is a risk of long-term consequences. The respective t1 and t2 vary significantly for the numerous forms of SE and are largely based on animal data and limited clinical evidence. Experts suggest a t1 of 5 minutes for generalized convulsive status epilepticus (GCSE) and 10 minutes for focal SE with impaired consciousness. Long-term consequences (t2) can be expected after 30 minutes of GCSE and 60 minutes of focal SE with impaired consciousness. The task force also developed a new classification system built on four axes: (a) semiology, (b) etiology, (c) EEG correlates, and (d) age (2). Treatment should be started.

Status epilepticus should be controlled.

t2

t1 Stage I (A)

5 min

30 min Stage II

Stage I Generalized tonic-clonic status epilepticus

(B)

t1

t2

10 min

60 min

Focal status with impairment of consciousness

t1

10−15 min

(C)

Stage III

t2

unknown

Absence status epilepticus

FIGURE 5.1 ILAE definition of status epilepticus. Operational dimensions with t1 indicating the time that emergency treatment of SE should be started and t2 denoting the time at which long-term consequences may be expected. Time (t1) is when a seizure is likely to be prolonged leading to continuous seizure activity. Time (t2) is when a seizure may cause long-term consequences (including neuronal injury, neuronal death, alteration of neuronal networks, and functional deficits). For generalized tonic–clonic status, the stages have been added (Stage I 5–10 min; Stage II 10–30 min; Stage III > 30 min). ILAE, International League Against Epilepsy; SE, status epilepticus. Source: Reprinted with permission from Trinka E, Kälviäinen R. 25 years of advances in the definition, classification and treatment of status epilepticus. Seizure. 2017;44:65–73.

44

Part II

Indications

Semiology forms the cornerstone of this classification scheme. Clinical semiology is differentiated along two taxonomic criteria: motor activity and impairment of consciousness. SE with prominent motor symptoms, including convulsive status epilepticus (CSE). Overt CSE is readily diagnosed by clinical presentation and its diagnosis typically does not require continuous EEG (cEEG). SE without prominent motor symptoms, including nonconvulsive status epilepticus (NCSE). Each of the previous groups can be further divided on the basis of the degree of impairment of consciousness (Table 5.1). ●

●

TABLE 5.1

Axis 1: Classification of SE

(A)

WITH PROMINENT MOTOR SYMPTOMS

A.1

Convulsive SE (CSE; synonym: tonic–clonic SE)

A.1.a.

Generalized convulsive

A.1.b.

Focal onset evolving into bilateral convulsive SE

A.1.c.

Unknown whether focal or generalized

A.2

Myoclonic SE (prominent epileptic myoclonic jerks)

A.2.a.

With coma

A.2.b.

Without coma

A.3

Focal motor

A.3.a.

Repeated focal motor seizures (Jacksonian)

A.3.b.

EPC

A.3.c.

Adversive status

A.3.d.

Oculoclonic status

A.3.e.

Ictal paresis (i.e., focal inhibitory SE)

A.4

Tonic status

A.5

Hyperkinetic SE (continued )

Chapter 5

Status Epilepticus

TABLE 5.1

Axis 1: Classification of SE (continued )

(B)

WITHOUT PROMINENT MOTOR SYMPTOMS (I.E., NONCONVULSIVE SE [NCSE])

B.1

NCSE with coma (including so-called “subtle” SE)

B.2

NCSE without coma

B.2.a.

Generalized

B.2.a.a

Typical absence status

B.2.a.b

Atypical absence status

B.2.a.c

Myoclonic absence status

B.2.b.

Focal

B.2.b.a

Without impairment of consciousness (aura continua, with autonomic, sensory, visual, olfactory, gustatory, emotional/psychic/experiential, or auditory symptoms)

B.2.b.b

Aphasic status

B.2.b.c

With impaired consciousness

B.2.c

Unknown whether focal or generalized

B.2.c.a

Autonomic SE

45

EPC, epilepsia partialis continua; SE, status epilepticus. Source: Reprinted with permission from Trinka E, Cock H, Hesdorffer D, et al. A definition and classification of status epilepticus—report of the ILAE Task Force on Classification of Status Epilepticus. Epilepsia. 2015;56(10):1515–1523.

●

●

In the ICU, NCSE frequently presents as decreased level of consciousness including coma. It is common across a number of etiologies (Figure 5.2; also see Chapter 22). As other conditions commonly seen in the ICU can also present with decreased level of consciousness, cEEG is required to make a definitive diagnosis of NCSE. Diagnosing NCSE and distinguishing it from these other conditions is a primary indication for cEEG monitoring in the ICU. Because outcomes are strongly related to the duration of SE, prompt identification of SE is critical. The second axis is based on the etiology of SE, and is divided into two groups: (a) known or symptomatic, and (b) unknown or cryptogenic. The third axis of the classification comprises EEG correlates of SE including location, name, and morphology of the EEG pattern and effect of treatment. In convulsive SE, the clinical presentation is most often clear and EEG is often obscured

46

Part II

Indications

60%

Percent with NCS/NCSE

50%

40%

NCS (%) NCSE (%) 30%

20%

10%

(D

a eL fte or r G n enz CS = o, E 16 1 4 99 C 8) (C N la S i as n s f n en ect = , 2 io 35 0 n 04 ) (C la as IC n sen H = 10 , 20 2 07 ) (C la as TB se I n = n, 2 51 0 04 ) (C la A as M s n en S = 10 , 20 5 04 ) (C la S as A n sen H = 11 , 20 6 06 ) (C la as AI s S n en = ,2 56 0 04 ) (C re H pe IE n au = ,2 54 01 3) (P ol S la D n n dt, H = 76 201 7)

0%

FIGURE 5.2 Incidence of nonconvulsive seizures in patients undergoing cEEG. Incidence of NCS (blue bars) and NCSE (red bars) in patients undergoing cEEG monitoring. AIS, acute ischemic stroke; AMS, altered mental status; CNS, central nervous system; GCSE, generalized convulsive status epilepticus; cEEG, continuous EEG; HIE, hypoxic–ischemic encephalopathy; ICH, intracranial hemorrhage; NCS, nonconvulsive seizures; NCSE, nonconvulsive status epilepticus; SAH, subarachnoid hemorrhage; SDH, subdural hematoma; TBI, traumatic brain injury.

●

by artifact; thus, the EEG may be of little value. Whereas in NCSE (especially with impaired consciousness/comatose state), EEG review is often needed for an accurate diagnosis. Unified EEG terminology and criteria for the diagnosis of NCSE have also been published, termed Salzburg criteria (4), and have been validated in a subsequent study (5). EEG criteria for SE are discussed in detail in Chapter 22. The fourth axis of age recognizes that there are specific electroclinical diagnoses that commonly occur in various age groups; for example, febrile SE in infants and myoclonic SE of juvenile myoclonic epilepsy (JME) in adolescents and adults.

B. Clinical semiology ●

●

●

Seizures in the ICU population are often very subtle and only 8% to 32% of patients have a clear clinical correlate associated with seizures. Critically ill patients also frequently exhibit involuntary movements such as clonus, asterixis, postanesthetic shivering, tremor, posturing, or myoclonus that can mimic clinical seizure activity. If movements are repetitive and/or rhythmic, they may be assumed to represent epileptic seizures, leading to addition of unnecessary medications that increase potential for adverse effects and drug interactions.

Chapter 5

Status Epilepticus

47

EEG monitoring is the only method of distinguishing involuntary movements of ictal origin from other physiological or nonphysiological movements. Video analysis in addition to EEG is critical to establish that the movements in question have been captured and adequately assessed. In addition, video aids in the diagnosis of focal motor seizures that may not have an EEG correlate but are highly suggestive of epileptic seizures based on visual analysis of the semiology. A recent retrospective study analyzed records from 626 patients who underwent cEEG (154 for event characterization and 472 for altered mental status) (6). Seizures were captured in 48 patients (31.2%) undergoing cEEG monitoring for characterization of clinical events. This was not significantly different from the incidence of seizures in patients undergoing cEEG for altered mental status (N = 133, 28.2%). For patients who underwent cEEG monitoring to characterize motor events, the most common event type was limb myoclonus/tremor, seen in 53 patients (34.4%), followed by extremity weakness in 21 patients (13.6%), facial and periorbital twitching in 19 patients (12.3%), and eye deviation/eye movement abnormalities in 15 patients (9.7%). Patients undergoing cEEG monitoring for facial/periorbital twitching were significantly more likely to have electrographic seizures (78.9%, p