Competency-Based Critical Care

Series Editors

John Knighton, MBBS, MRCP, FRCA Paul Sadler, MBChB, FRCAConsultant ConsultantIntensive Care Medicine & Anaesthesia Critical Care Medicine & AnaesthesiaPortsmouth Hospitals NHS Trust Queen Alexandra HospitalPortsmouth PortsmouthUK UK

Founding Editor

John S.P. LumleyEmeritus Professor of Vascular SurgeryUniversity of LondonLondonUK

and

Honorary Consultant SurgeonGreat Ormond Street Hospital for Children NHS Trust (GOSH)LondonUK

Other titles in this series

Renal Failure and Replacement Therapiesedited by Sara Blakeley

John P. Adams Dominic Bell Justin McKinlay (eds.)

Neurocritical Care

A Guide to Practical Management

EditorsJohn P. AdamsThe General Infirmary at LeedsGreat George StreetLeeds LS1 3EXUnited KingdomJohn.Adams@leedsth.nhs.uk

Justin McKinlayThe General Infirmary at LeedsGreat George StreetLeeds LS1 3EXUnited Kingdomjustin.mckinlay@leedsth.nhs.uk

Dominic BellThe General Infirmary at LeedsGreat George StreetLeeds LS1 3EXUnited Kingdomdominic.bell@leedsth.nhs.uk

ISSN 1864-9998 e-ISSN 1865-3383ISBN 978-1-84882-069-2 e-ISBN 978-1-84882-070-8DOI 10.1007/978-1-84882-070-8Springer London Dordrecht Heidelberg New York

British Library Cataloguing in Publication DataA catalogue record for this book is available from the British Library

Library of Congress Control Number: 2009931330

Springer-Verlag London Limited 2010Apart from any fair dealing for the purposes of research or private study, or criticism or review, as permitted under the Copyright, Designs and Patents Act 1988, this publication may only be reproduced, stored or transmitted, in any form or by any means, with the prior permission in writing of the publishers, or in the case of reprographic reproduction in accordance with the terms of licenses issued by the Copyright Licensing Agency. Enquiries concerning reproduction outside those terms should be sent to the publishers.The use of registered names, trademarks, etc., in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant laws and regulations and therefore free for general use.The publisher makes no representation, express or implied, with regard to the accuracy of the information contained in this book and cannot accept any legal responsibility or liability for any errors or omissions that may be made.

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

John Adams dedicates this book to his wife Kate to compensate for neglect of his responsibilities as husband and father. The families of his fellow editors did not specifically notice or comment and for this we are grateful.

Brain injury is a worldwide leading cause of mortality and morbidity and requires early and appropriate management to minimize these adverse sequelae. Despite such needs, access to specialist centers is limited, forcing both immediate and secondary care of these patients onto generalist staff. These responsibilities are made more problematical by differences in patient management between and even within specialist centers, due in part to an insufficient evidence-base for many interventions directed at brain injury.

This book is borne out of the above observations and is targeted at emer-gency and acute medicine, anesthetic and general intensive care staff caring for brain injury of diverse etiology, or surgical teams responsible for the inpatient care of minor to moderate head trauma.

Although explaining the various facets of specialist care, the book is not intended to compete with texts directed at neurosciences staff, but aims to advise on optimal care in general hospitals, including criteria for transfer, by a combination of narrative on pathophysiology, principles of care, templates for documentation, and highly specific algorithms for particular problems. It is intended that the content and structure can form the basis of guidelines and protocols that reflect the needs of individual units and that can be constantly refined. Our ultimate goal is to promote informed, consistent, auditable, multidisciplinary care for this cohort of patients and we hope that this text contributes to that process.

Preface

vii

We are indebted to our fellow authors who have not only made this book possible, but have approached the task with enthusiasm. All understand and endorse the importance of clear, comprehensive, evidence-based, and con-sistent advice in the support of colleagues caring for these patients outside the regional center.

We are also grateful for the observations of colleagues responsible for the eventual rehabilitation of these patients, mainly that even minor reductions in neurological deficit by early and appropriate care, can have a significant impact on quality of life, with proportional benefit not only for the patient, but family, health and social care institutions, and society. These observations justify the book and warrant implementation of the contained principles.

Finally, we thank Melissa Morton in the UK and Robin Lyon in New York for all their help and support in bringing this book to publication.

Acknowledgments

ix

Chapter 1 Brain Injury and Dysfunction: The Critical Role of Primary Management .............................................. 1M.D. Dominic Bell

Chapter 2 Monitoring the Injured Brain .............................................. 9Simon Davies and Andrew Lindley

Chapter 3 The Secondary Management of Traumatic Brain Injury ........................................................................... 19Dominic Bell and John P. Adams

Chapter 4 Critical Care Management of Subarachnoid Hemorrhage ........................................................................... 33Audrey C. Quinn and Simon P. Holbrook

Chapter 5 Central Nervous System Infections ..................................... 43Abigail Walker and Miles Denton

Chapter 6 Cervical Spine Injuries ......................................................... 51John P. Adams, Jake Timothy, and Justin McKinlay

Chapter 7 Recent Advances in the Management of Acute Ischemic Stroke ..................................................................... 61Ahamad Hassan

Chapter 8 Seizures on the Adult Intensive Care Unit .......................... 69Morgan Feely and Nicola Cooper

Chapter 9 Non-Neurological Complications of Brain Injury ............. 77John P. Adams

Chapter 10 Acute Weakness in Intensive Care ....................................... 89Louise Barnes and Michael Vucevic

Chapter 11 Coma, Confusion, and Agitation in Intensive Care ............ 97Matthew Clark and Justin McKinlay

Contents

xi

Contents

Chapter 12 Death and Donation in Critical Care: The Diagnosis of Brainstem Death ...................................... 105Paul G. Murphy

Chapter 13 Death and Donation in Critical Care: Management of Deceased Organ Donation ........................ 113Paul G. Murphy

Chapter 14 Imaging the Brain-Injured Patient ...................................... 121Tony Goddard and Kshitij Mankad

Chapter 15 Ethical Dilemmas Within Intensive Care ............................ 137M.D. Dominic Bell

Appendices .................................................................................................... 145

Index .............................................................................................................. 173

xii

John P. AdamsLeeds General InfirmaryLeeds Teaching Hospitals NHS TrustLeedsWest Yorkshire LS1 3EXUK

Louise BarnesHull Royal InfirmaryHull and East Yorkshire Hospitals NHS TrustHull HU3 2JZUK

Dominic BellLeeds General InfirmaryLeeds Teaching Hospitals NHS TrustLeedsWest Yorkshire LS1 3EXUK

Matthew ClarkDepartment of Anesthetics and Intensive CareLeeds General InfirmaryLeeds Teaching Hospitals NHS TrustLeedsWest Yorkshire LS1 3EXUK

Nicola CooperLeeds Teaching Hospitals NHS TrustLeeds General InfirmaryLeedsWest Yorkshire LS1 3EXUK

Simon DaviesDepartment of AnaestheticsYork Hospital NHS TrustYork HospitalYorkNorth Yorkshire YO31 8HEUK

Miles DentonLeeds General InfirmaryLeeds Teaching Hospitals NHS TrustLeedsWest Yorkshire LS1 3EXUK

Morgan FeelyDepartment of NeurologyLeeds General InfirmaryLeeds Teaching Hospitals NHS TrustLeeds General InfirmaryLeedsWest Yorkshire LS1 3EXUK

Tony GoddardDepartment of NeuroradiologyLeeds General InfirmaryLeeds Teaching Hospitals NHS TrustLeedsWest Yorkshire LS1 3EXUK

Contributors

xiii

Contributors

Ahamad HassanDepartment of NeurologyLeeds General InfirmaryLeeds Teaching Hospitals NHS TrustLeedsWest Yorkshire LS1 3EXUK

Simon HolbrookAcademic Unit of AnesthesiaSt. Jamess University HospitalLeedsWest Yorkshire LS9 7TFUK

Andrew LindleyLeeds Teaching Hospitals NHS TrustLeeds General InfirmaryLeedsWest Yorkshire LS1 3EXUK

Kshitij MankadDepartment of NeuroradiologyLeeds General InfirmaryLeeds Teaching Hospitals NHS TrustLeedsWest Yorkshire LS1 3EXUK

Justin McKinlayDepartment of Anaesthetics and Neurocritical CareLeeds General InfirmaryLeeds Teaching Hospitals NHS TrustLeedsWest Yorkshire LS1 3EXUK

Paul G. MurphyDepartment of AnaesthesiaLeeds General InfirmaryLeeds Teaching Hospitals NHS TrustLeedsWest Yorkshire LS1 3EXUK

Audrey C. QuinnLeeds General InfirmaryLeeds Teaching Hospitals NHS TrustLeedsWest Yorkshire LS1 3EXUK

Jake TimothyDepartment of NeurosurgeryLeeds General InfirmaryLeeds Teaching Hospitals NHS TrustLeedsWest Yorkshire LS1 3EXUK

Michael VucevicDepartment of AnestheticsLeeds General InfirmaryLeeds Teaching Hospitals NHS TrustLeedsWest Yorkshire LS1 3EXUK

Abigail WalkerDepartment of AnesthesiaChristchurch HospitalChristchurchCanterburyNZ

xiv

A/B/C Airway, breathing, circulationALI Acute lung injuryAPTT Activated partial thromboplastin timeBAL Bronchoalveolar lavageBIS Bispectral indexBP Blood pressureCMV CytomegalovirusCNS Central nervous systemCOAG Coagulation screenCPP Cerebral perfusion pressure (MAP-ICP)CRP C-reactive proteinCSF Cerebrospinal fluidCT Computed tomographyCVP Central venous pressureCXR Chest X-rayECG ElectrocardiogramEEG ElectroencephalogramESR Erythrocyte sedimentation rateEtCO2 End-tidal carbondioxide concentrationFBC Full blood countFiO2 Fraction of inspired oxygenGCS Glasgow coma scaleGluc GlucoseHAS Human albumin solutionHb HemoglobinHIV Human immunodeficiency virusHR Heart rateHSE Herpes simplex encephalitisIABP Invasive arterial blood pressureICP Intracranial pressureICU Intensive care unitINR International normalized ratioIV IntravenousLFTs Liver function testsLP Lumbar punctureMAP Mean arterial pressure

Glossary of Terms and Abbreviations

xv

Glossary of Terms and Abbreviations

MI Myocardial infarctionMRI Magnetic resonance imagingMRSA Methicillin-resistant Staphylococcus aureusNaCl Sodium chlorideNEAD Non-epileptic Attack DisorderNGT Nasogastric tubeNICE National Institute for health and Clinical ExcellenceNJT Nasojejunal tubeNPE Neurogenic Pulmonary EdemaNSAID Non-steroidal anti-inflammatory drugODM Oesophageal doppler monitorOGT Orogastric tubePaCO2 Partial pressure of carbondioxide (arterial blood)PaO2 Partial pressure of oxygen (arterial blood)PCR Polymerase chain reactionPCWP Pulmonary capillary wedge pressurePE Pulmonary embolismPEEP Positive end-expiratory pressurePbtO2 Partial pressure of brain tissue oxygenPPI Proton pump inhibitorPVS Persistent vegetative stateSaO2 Arterial oxygen saturationSpp SpeciesSjvO2 Jugular venous oxygen saturationTB TuberculosisU&Es Urea and electrolytesUK United KingdomVt Tidal volumeVTE Venous thromboembolismWCC White cell countWFNS World Federation of Neurosurgical Socities

xvi

Key Points

1. In traumatic brain injury, maintain mean arte-rial (MAP) blood pressure >80 mmHg.

2. Avoid hypoxia at all costs; keep PaO2 >13 kPa, using PEEP if necessary.

3. Keep PaCO2 4.55.0 kPa; hyperventilate only if there are signs of impending brainstem hernia-tion.

4. Keep the neck in neutral position; always con-sider the possibility of cervical spine injury.

5. Maintain 15 head up position (as long as MAP adequate).

6. Do not give mannitol if patient is hypotensive. Speak to a Regional Neurosurgical Center be-fore giving additional doses.

Introduction

The human brain, in structure and function, rep-resents the pinnacle of biological evolution. Even the most rudimentary non-volitional role of matching ventilation to demand or maintaining homeostasis is phenomenally complex for an organism vulnerable to disease or dysfunction of the component tissues and organs, and more par-ticularly when exposed to mechanical, chemical, and thermal hazard as every environmental extreme is challenged. The coordination of physi-cal movement, played out at the highest level in sport and the performing arts, rightly warrants recognition as a marker of complex neuronal

activity, but conventionally, as a form of intelli-gence, bows to the cognitive capacity of the human brain. Numerical and literary skills, communica-tion, memory, and knowledge are entry-level cog-nitive skills, with mans advances through understanding of both science and nature repre-senting a higher plane. Reasoning and judgment, coupled with awareness of the needs of others and social skills arguably constitute the highest form of human intelligence. Interlinked with this func-tion are those characteristics of personality and emotional status which generate individual uniqueness. These may be reflected in our achieve-ments, as in career choice, or functional and artis-tic creativity, or our behavior relating to those achievements, as in innovation, ambition, and leadership. These higher functions also have an emotional dimension covering conscience, charity and self-sacrifice, enthusiasm, and the ability to love, rejoice and grieve.

This refinement and complexity of normal cerebral function is, however, associated with certain inherent vulnerabilities carrying signifi-cant implications for the management of either primary or secondary brain pathology or dysfunc-tion. Tissues such as bone are able to regain normal architecture after injury, complex organs such as the liver and kidney are able to regenerate with resto-ration of original levels of function, and heart, lung, and pancreas are able to withstand devascularization and subsequent transplantation. The specializa-tion of cellular structure and function within the central nervous system, however, appears to exclude a capacity for repair and renewal after anything other than the most trivial insult. Brain

1Brain Injury and Dysfunction: The Critical Role of Primary ManagementM.D. Dominic Bell

1

M.D.D. Bell

tissue has a high requirement for oxygen and energy substrates to maintain both structure and function, leaving little reserve in the face of impaired delivery. Even with normal arterial oxygen content, circulatory arrest will result in loss of consciousness within 15 seconds, and given the high oxygen requirements simply to maintain cel-lular integrity, more than 5 minutes of circulatory arrest at normothermia will result in neuronal death and a significant multifaceted neurological deficit. These aspects demonstrate the exquisite vulnerability of the brain to the so-called secon-dary cerebral insults, with cellular hypoxia being the commonest final pathway.

There is a gradient of sensitivity of the different neural tissues to a global insult such as hypoxia, whereby the loss of higher function precedes loss of motor activity, with ventilatory effort main-tained until immediately prior to death. This pattern parallels the picture of recovery from such an insult, the extreme end of the spectrum being the persistent vegetative state, where the patient is self-ventilating, but has no awareness of environ-ment or self. This demonstrates that survival alone cannot be considered a satisfactory outcome from brain injury, and that all effort must be directed toward preventing, where possible, even the most subtle changes to personality and cognitive func-tion at the other end of the spectrum, that would require the skills of a clinical psychologist to objectively quantify. Failure to address these aspects results not only in significant disability for patient and family, but phenomenal burden and cost to society.

This edition of the series, devoted to neurocriti-cal care and the prevention or minimization of such avoidable neurological deficit, examines the theory and evidence-base behind the various man-agement strategies expected of a regional unit. The secondary aim is to define and promote principles of care that can be deployed by any discipline, at any level of seniority, at any location, at any time, for any patient, with any pathology, and at any stage. Such principles, both clinical and procedural, are essential, given that most neuropathology arises outside the setting of a specialist center, and many patients will not access that center, either because neurosurgical intervention is not required, other injuries require immediate management, or because of limited bed availability.

Given the vulnerability of the brain as outlined earlier, it is unacceptable if the patient accrues additional avoidable morbidity in these circum-stances, or indeed while awaiting or during trans-

fer to the regional unit, through ignorance. Clinical experience also highlights how patient care can be compromised due to a lack of clarity and consis-tency in the referral process and acceptance by the regional unit, resulting in a hiatus in care with neither party taking full responsibility for these aspects. Such a scenario is arguably more unac-ceptable than ignorance, and demands explicit policy from the center and audit of process to monitor compliance.

Role of the Regional Neurosurgical Center (RNC)

Fundamental to optimal patient management and any relationship with the regional center is an understanding of the specific services provided there. The greatest demand will be for care of trau-matic brain injury, followed by subarachnoid hemorrhage, but the centers also have an emerg-ing involvement in conventional strokes. Throm-bolysis or interventional radiology for an ischemic infarct are being increasingly adopted as appro-priate emergency care, mirroring the approach taken to occlusive coronary events. The implica-tions of managing these patients as medical emer-gencies cannot be overestimated, but the care and cost implications of the current conservative eval-uative approach to strokes are significant, regard-less of the impact on the patient and family.

The role of the regional center for this range of pathology can be summarized as intervention, neuro-specific monitoring, and advice for referring units. Given that vascular pathology is addressed in subsequent chapters, the role is only outlined in greater depth, as below, for traumatic brain injury:

1. To expedite removal of a significant intracra-nial hematoma

2. To monitor for the potential expansion of a less significant hematoma

3. To provide specialized monitoring (e.g., intrac-ranial pressure, jugular venous oximetry) to direct the neuro-intensive care of the diffuse axonal injury

4. To undertake radical surgical maneuvers for refractory intracranial hypertension, e.g., de-compressive craniectomy or lobectomy for ex-tensive contusion

Although it could be argued that a patient should be transferred to a specialized unit for

2

1. Brain Injury and Dysfunction: The Critical Role of Primary Management

imaging and assessment of the patient to make the above distinctions, CT scanning in the referring hospital has reduced the necessity for this and digital image transfer should improve the quality of discussion and decision-making. Furthermore, it is clearly not in the interest of the majority of patients to be transferred for the sole purpose of diagnosis.

Indications for Patient Transfer

Group 1: Transfer delayed only for correction of secondary cerebral insults or for life-saving surgery (e.g., expanding extradural hematoma with localizing signs).

Group 2: Requires urgent transfer following optimization and life and limb saving surgery (e.g., subdural hematoma with no mass effect).

Group 3(a): Patients should only be transferred after absolute stabilization given that the overall principles of care are to avoid secondary cere-bral insults, rather than to offer neuro-specific therapies (e.g., contusional injury with no mass effect).

Group 3(b): Some non-neurosurgical intensive care units (ICU) monitor ICP in cases of diffuse axonal injury; transfer may become necessary if the ICP subsequently becomes difficult to control.

Organizing the Response

Groups 1-3(a) above demonstrate the importance of the primary decision-making which involves diagnostic skills, confident liaison with the regional center, and an appropriate level of care in the event of retention of the patient. This respon-sibility usually falls to the attending anesthetist or intensive care specialist following initial stabiliza-tion in the emergency department. This individual has a pivotal role in coordinating this process and therefore assumes both clinical and logistical responsibilities (see Table 1.1).

Avoidance of Secondary Cerebral Insults

No treatment strategy can reverse neuronal death caused by the primary brain injury, but much can

be done to avoid preventable secondary neuronal death and subsequent deficit. These secondary insults share a final common pathway that takes areas of the brain compromised by the primary injury, or indeed the whole brain, toward irrevers-ible ischemia (see Fig. 1.1).

Secondary cerebral insults can be triggered by intracranial or systemic factors, which either reduce cerebral oxygen delivery or increase cere-bral oxygen consumption (Table 1.2). In addition, an increase in the volume of brain, blood, or CSF, or an expanding space occupying lesion (e.g., hematoma) may increase the pressure within the rigid skull and trigger global ischemia. Focal damage may be caused by local compression or shearing forces.

Cerebral Oxygen Delivery

Cerebral oxygen delivery depends upon:

(a) An adequate circulating volume at a perfusion pressure above the lower level of cerebral autoregulation.

(b) An adequate amount of oxygenated hemo-globin that dissociates appropriately at tissue level.

Cerebral Oxygen Consumption

To avoid excessive cerebral oxygen consumption in the context of compromised cerebral oxygen

Table 1.1. Roles of the attending specialist during the primary management of patients with traumatic brain injury

1. Primary resuscitation 2. Neurological assessment 3. Deciding on the need for intubation, sedation and ventilatory

support 4. Management of problems such as convulsions 5. Interpretation of CT scans adequate for prioritization of

treatment options 6. Prioritizing and expediting essential general surgical and

orthopedic interventions 7. Deciding on transfer or retention after such interventions 8. Maintaining neurological observations 9. Avoiding secondary cerebral insults or expansion of any

intracranial pathology10. Organizing further CT scans in the event of retaining a patient11. Maintaining dialog with the neurosurgeons and the

neurosurgical intensive care12. Deciding, in the face of massive injury, that no overall benefit

from transfer exists

3

M.D.D. Bell

delivery, it is essential to recognize and actively treat any seizure activity and to provide adequate analgesia and sedation, once a patient is intubated and ventilated. Pyrexia should be treated with active cooling measures once the patient is stabi-lized on the ICU. Hyperglycemia, which is believed to increase cerebral oxygen consumption, should be targeted during all epochs of care.

Expansion of Intracranial Contents

(a) Space-Occupying Lesions, for example, he-matomata or contusions

The key priority is to determine whether urgent neurosurgery is required. General supportive care includes avoidance of aspects that allow a hematoma to expand through loss or dilution of platelets or coagulation factors. Hypothermia, hypocalcemia, and administration of large volumes of colloid solutions should be avoided.

These aspects assume greatest significance in the context of a subdural or intracranial hematoma, where such attention may avoid the need for surgical intervention.

(b) Brain Edema Four Mechanisms:

1. Hydrostatic edema: occurs when arterial pressure exceeds the upper limit of auto-regulation or when there is venous congestion (head-down position, pressure on the jugular veins, high intrathoracic pressure).

2. Osmotic edema: non-ionic crystalloid solutions such as dextrose become, in effect, free water once the sugar component is metabolized.

3. Oncotic edema: due to low plasma proteins; can become important when the bloodbrain barrier (BBB) is damaged.

4. Inflammatory edema: the inflammatory response to insults such as trauma or hy-poxia can lead to increased capillary per-meability and disruption of the BBB. It is critically important to avoid preventable insults such as osmotic edema when this has arisen.

The management of cerebral edema and raised intracranial pressure traditionally involves admin-istration of mannitol. This can only be effective if the BBB is intact, there is mass rapid movement of water from the tissues into the circulating com-partment, and finally rapid excretion via the kidneys of both mannitol and water. The main role of mannitol is to temporarily reduce the amount of brain water and thereby reduce overall intracranial pressures and relieve pressure on vital structures such as the brainstem. This buys time before definitive neurosurgical intervention. By reducing the size of normal brain, abnormal areas including hematomata can expand, generating a greater shearing effect. If mannitol is used indis-criminately with a deranged BBB, the molecule can diffuse across and ultimately contribute to the development of osmotic edema. This is more likely to occur with hypotension and poor renal per-fusion such that the mannitol is not excreted.

Increase in Cerebral Blood Volume

1. Arterial: PaCO2 is the commonest avoidable cause of cerebral arterial vasodilatation.

2. Venous: discussed earlier, for example, neck po-sitioning, endotracheal tube ties.

MECHANISM OF ISCHEMIA

OEDEMA

PRESSURE

ISCHAEMIA

ANAEROBICMETABOLISM

ACID PRODUCTION

OSMOTIC PRESSURE

MEMBRANE DYSFUNCTION

OXYGEN REQUIREMENTS

EXCITATORY NEURO-TRANSMITTERS

SUPEROXIDES HYDROPEROXY RADICALS

CLOSEDBOX

SECONDARY INSULT S

CALCIUM INFLUX CALCIUM MEDIATED CELL DEATH

APOPTOSIS

Figure 1.1. Mechanism of ischemia in brain injury.

Table 1.2. Intracranial and systemic causes of secondary brain injury

Intracranial Systemic

Expanding contusion/hematoma HypotensionCerebral edema HypoxiaVascular injury/carotid dissection Hypo or HypercapniaSeizures PyrexiaHydrocephalus CoagulopathyVasospasm Hypo or

hyperglycemiaPneumocephalus AnemiaIntracranial infection Sepsis

4

1. Brain Injury and Dysfunction: The Critical Role of Primary Management

Cerebrospinal Fluid

The ventricular system and contained CSF are usually capable of reducing in size to accommodate brain edema without causing a rise in intracranial pressure. Pathologies such as subarachnoid hemorrhage and bacterial meningitis can cause obstructive hydrocephalus. This requires insertion of a ventricular drain.

Overall Management Strategy

Optimal patient care derives from an understanding of the common pathologies that compromise brain structure or function, and of the principles under-pinning appropriate treatment options. The key goal of this edition is to demystify this area of activ-ity and thereby empower clinicians caring for these patients, particularly within the primary receiving hospital, since it is in this setting that there is the greatest opportunity for patient harm through act, omission, or delay in accessing the regional center. The clinical aspects of care, both neuro-specific and general, need to be formalized through protocols to ensure consistency, regardless of grade or discipline of attendant staff. It is vital that the logistical aspects of care be similarly formalized, namely documenta-tion, particularly observation charts, investigations, involvement of other disciplines, communication, and any referral process to the regional neurosurgi-cal center. Only with such a structure will the right things be done on the right patient, in the right order, and at the right time. The challenge for clinicians working within a regional unit is to recognize the fundamental importance of achieving these goals in the referring hospital, and to actively promote and support such a system. The challenge for those working in the referring hospital is to ensure that this responsibility of the regional unit is discharged.

Such goals and the system directed at these are defined as care bundles: strategies to not only optimize care based on the strongest available evi-dence, but also to facilitate audit of process.

Readers are referred to the appendices for exam-ples of how the principles are translated into explicit recommendations for care within the authors region, with responsibility for dissemination and imple-mentation resting with the local critical care network1 There is, however, still much to be done to eradicate inconsistencies of care through ignorance and limited formalization of process, as much as limited availability in the regional centers. It is hoped that those readers who recognize the magnitude of the problem will be stimulated by this edition to confi-

dently address those issues, which are so critical for patient care and professional satisfaction.

Principles of Management of Brain Injury

The primary clinical management of any patient with a brain injury, regardless of the diagnosis or severity, consists of routine resuscitation maneuvers and diag-nosing the nature and severity of both CNS and non-CNS pathology. Consideration should always be given to the possibility of a lesion for which there is a specific surgical or medical intervention, or interim supportive measures that can prevent that lesion gen-erating morbidity or mortality. In the event of there being more than one pathology, clinical judgment has to determine the priorities of treatment.

Running parallel to that clinical process is a logistical process, which incorporates aspects such as teamwork, leadership, communication, prioriti-zation, documentation, and timekeeping.

The Clinical Process

1. Resuscitation: as per ALS/ATLS guidelines.2. Diagnosis: CNS pathology/non-CNS injury/

co-morbidity.

Indications for a CT brain scan after head injury are outlined in Table 1.3 (see NICE Guideline 2007).

(a) CNS pathology: diagnosis, CT findings, severity (GCS, pupils, focal neurology, seizures), trends, confounding variables (e.g., drugs, alcohol, hypotension, hypothermia).

(b) Non-CNS pathology: remember the possi-bility of spinal injury.

3. Consideration of need for neurosurgical referral: Use standardized form for transfer of information (see example, Appendix).

4. Neuro-specific observations/monitoring: Use a standardized chart.

5. Neuro-specific treatment: for example, mannitol (see Table 1.4), hypertonic saline (HSL), anticonvulsants.

6. Define priorities for treatment:

(a) Urgent transfer (b) Life or limb-saving surgery (c) General support and stabilization1 http://www.wyccn.org.uk/CareBund.htm

5

M.D.D. Bell

7. Avoidance of secondary insults: see targets outlined here.

8. Regular re-evaluation of all the aforementioned components.

The Logistical Process

1. Involve all relevant specialties2. Determine team leadership3. Establish documentation of observations4. Ensure explicit communication:

(a) Internally within the team (b) With key support specialties; radiology,

transfusion, pharmacy, etc. (c) With the regional neurosurgical center

5. Determine satisfactory timescale for:

(a) Diagnostic procedures (b) Care/interventions (c) Communication with neurosurgical center (d) Transfer

(e) Re-evaluation of all aspects of care6. Ensure documentation (using standardized

templates where available) of:

(a) Observations (b) All above clinical undertakings (c) Criteria for transfer (d) Results of discussion with regional center

7. Ensuring all appropriate support for any trans-fer is available (functioning equipment, trained personnel, means of communication).

8. Define criteria for stabilization prior to transfer.

Avoidance of Secondary Cerebral Insults

1. Maintaining cerebral oxygen delivery

(a) Adequate circulating volume: Aim for capil-lary refill time PEEP+5 with crystalloids (0.9%NaCl) up to 2 L followed by a colloid (e.g., voluven, gelofusine). Give blood and clotting factors to maintain Hb ~10 g/dL or hematocrit 30, INR 100,000.

(b) Adequate oxygenation: Maintain PaO2 >13 kPa with supplemental oxygen and PEEP if necessary. Intubate and ventilate for GCS 80 mmHg or within 15% of normal values if normally hypertensive. After volume resuscitation, vasopressors or inotropes may be required to maintain an adequate blood pressure, the choice depending upon the cardiovascular profile (see Appendix). Advanced monitoring (e.g., esophageal doppler, pulmonary artery catheter) may be required to guide this process, espe-cially if there is uncertainty about volume status.

2. Controlling cerebral oxygen consumption

(a) Control seizure activity: Seizure activity is usually treated with a benzodiazepine (e.g.,

Table 1.4. Indications for Mannitol

Unilateral pupillary dilatation, or unilateral progressing to bilateral dilatation (primary bilateral dilatation may represent fitting, drug intoxication or overdose, or overwhelming brain injury).

Dose: 0.5 g/kg (approximately 200 mL of 20% solution) over 1015 min. Can be repeated at 12 hourly intervals to maximum serum osmolarity of 320 mosmol/L or Na+ of 160 mmol/L. Speak to the Regional Neurosurgical Center prior to giving additional doses.

Alternative: Hypertonic saline (HSL) is being increasingly used for the same purpose with good effect. We use 30 mL of 20% HSL over 20 min via a CVC, with a similar serum [Na+] cut-off of 160 mmol/L.

Precautions: Mannitol should not be given to patients who are hypotensive or have evidence of inadequate renal perfusion. All patients require bladder catheterization.

Table 1.3. Indications for a CT brain scan after head injury

Depressedconsciouslevel Focalneurologicaldeficit Suspectedopenorbasalskullfracture Age>65,withlossofconsciousnessoramnesia GCS15withnofracture,butotherconcerningfeatures(severe

persistent headache, vomiting, seizure, altered behavior) Unabletoassessproperly(e.g.,alcohol,drugs) Priortoanesthesiafortreatmentofotherinjuries

http://www.nice.org.uk/nicemedia/pdf/CG56NICE Guideline.pdf

6

1. Brain Injury and Dysfunction: The Critical Role of Primary Management

lorazepam 24 mg IV bolus) in the first instance, followed by a longer-acting agent (e.g., phenytoin 15 mg/kg over 20 min). See Chap. 8 for detailed description, and the Appendix for the status epilepticus algorithm.

(b) Ensure adequate analgesia and sedation if intubated: Use fentanyl or alfentanil by infusion with propofol (midazolam can be used if there is cardiovascular instability). Maintain paralysis with infusion of muscle relaxant (e.g., cisatracurium or vecuronium) and monitor with a nerve stimulator. All head-injured patients require bladder catheterization.

3. Avoiding increases in intracranial pressure

(a) Avoid expansion of intracranial hematoma/contusion: Maintain normal clotting and platelet counts. Monitor calcium in face of massive transfusion. Consider Factor VIIa if intracranial hematoma or contusion in the face of nonsurgical major hemorrhage despite administration of platelets and clotting factors (see Chap. 3 for detailed description).

(b) Avoid brain edema: Use 0.9% NaCl, avoid dextrose.

(c) Avoid hyperemia: Maintain PaCO2 4.55.0 kPa. (d) Avoid venous congestion: 15 head up tilt.

Avoid external compression and high intrathoracic pressures.

For full details of the current NASGBI guide-lines for transfer of brain injured patients, visit www.nasgbi.org.uk.

Further Reading

Clayton TJ, Nelson RJ, Manara AR (2004) Reduction in mortality from severe head injury following intro-duction of a protocol for intensive care management. Br J Anaesth 93(6):761762

Modernisation Agency/Department of Health (2004) The Neurosciences Critical Care Report. London www.dh.gov.uk/publications

NICE (2007) Head Injury: Triage, assessment, investiga-tions and early management of head injury in infants, children and adults. London http://www.nice.org.uk/nicemedia/pdf/CG56NICEGuideline.pdf

The Neuro Anaesthesia Society of Great Britain and Ireland and The Association of Anaesthetists of Great Britain and Ireland (2006) Recommendations for the Safe Transfer of Patients with Brain Injury. London www.nasgbi.org.uk

7

Key Points

1. Repeated clinical assessment through the Glas-gow Coma Scale (GCS) is the cornerstone of neurological evaluation.

2. Ventilated head-injured patients with intracra-nial pathology on CT require ICP monitoring.

3. Invasive or noninvasive neuro-specific moni-toring requires careful interpretation when as-sisting goal-directed therapies.

4. Multimodal monitoring using a combination of techniques can overcome some of the limita-tions of individual methods.

Neuro-Specific Monitoring

Accurate neurological assessment is fundamental for the management of patients with intracranial pathology. This consists of repeated clinical exam-ination (particularly GCS and pupillary response) and the use of specific monitoring techniques, including serial CT scans of the brain. This chapter provides an overview of the more common moni-toring modalities found within the neuro-critical care environment.

In general terms, a combination of assessments is more likely to detect change than one specific modality. Real-time continuous monitoring (e.g. ICP) will provide more timely warning about adverse events (e.g., an expanding hematoma) as compared to static assessments such as sedation holds or serial CT brain scans.

Clinical Assessment

The Glasgow Coma Scale

The Glasgow Coma Scale (GCS) provides a stand-ardized and internationally recognized method for evaluating a patients CNS function by record-ing their best response to verbal and physical stimuli. A drop of two or more GCS points (or one or more motor points) should prompt urgent re-evaluation and a repeat CT scan. The GCS is described in detail in Chap. 10.

NB. Eye opening is not synonymous with awareness, and can be seen in both coma and Per-sistent vegetative state(PVS). The important detail is that the patients either open their eyes to command or fixes or follows a visual stimulus.

Pupillary Response

Changes in pupil size and reaction may provide useful additional information:

Sudden unilateral fixed pupil: Compression of the third nerve, e.g., ipsilateral uncal her niation or posterior communicating artery aneurysmUnilateral miosis: Horners syndrome (consider vascular injury)Bilateral miosis: Narcotics, pontine hemor- rhageBilateral fixed, dilated pupils: Brainstem death, massive overdose (e.g. tricyclic antidepressants).

2Monitoring the Injured BrainSimon Davies and Andrew Lindley

9

S. Davies and A. Lindley

In the non-specialist center, neurological assess-ment of the ventilated patient consists of serial CT brain scans, pupillary response, and assessment of GCS during sedation holds. A reduction in seda-tion level will usually be at the suggestion of the Regional Neurosurgical Center (RNC) and its timing will depend upon a number of factors. Responses such as unilateral pupillary dilatation, extensor posturing, seizures, or severe hyper-tension should prompt rapid re-sedation, repeat CT scan, and contact with the RNC. In the patient with multiple injuries, consideration must be given to their analgesic requirements prior to any decrease in sedation levels.

Invasive Monitoring

Intracranial Pressure Monitoring

Cerebral perfusion pressure (CPP) reflects the pressure gradient that drives cerebral blood flow (CBF), and hence cerebral oxygen delivery. Meas-urement of intracranial pressure (ICP) allows estimation of CPP.

CPP = Mean Arterial Pressure ICP

Sufficient CPP is needed to allow CBF to meet the metabolic requirements of the brain. An inade-quate CPP may result in the failure of autore-gulation of flow to meet metabolic needs whilst an artificially induced high CPP may result in hyperemia and vasogenic edema, thereby wors-ening ICP. The CPP needs to be assessed for each individual and other monito ring modalities (e.g., jugular venous oximetry, brain tissue oxygen-ation) may be required to assess its adequacy.

Despite its almost universal acceptance, there are no properly controlled trials demonstrating improved outcome from either ICP- or CPP-targeted therapy. However, in the early 1990s Marmarou et al. showed that patients with ICP values consi-stently greater than 20 mmHg suffered worse outcomes than matched controls, and poorer outcomes have been described in patients whose CPP dropped below 60 mmHg (Juul 2000; Young et al. 2003). As such, ICP- and CPP-targeted therapy have now become an accepted standard of care in head injury management.

The 2007 Brain Trauma Foundation Guidelines (Brain Trauma Foundation 2007) recommend

treating ICP values above 20 mmHg and to target CPP in the range of 5070 mmHg. Patients with intact pressure autoregulation will tolerate higher CPP values. Aggressive attempts to maintain CPP >70 mmHg should be avoided because of the risk of ARDS.

Measuring ICP

Intraventricular devices consist of a drain inserted into the lateral ventricle via a burr hole, and connected to a pressure transducer, manometer, or fiber optic catheter. This remains the gold standard but is associated with a higher incidence of infection and greater potential for brain injury during placement. It has the added benefit of allowing CSF drainage. Historically, saline could be injected to assess brain compliance.

Extraventricular systems are placed in paren-chymal tissue, the subarachnoid space, or in the epidural space via a burr hole. This can be inserted at the bedside in the ICU. These systems are tipped with a transducer requiring calibra-tion, and are subject to drift (particularly after long-term placement). Examples of extraven-ticular systems are the Codman and Camino devices. These devices have a tendency to underestimate ICP.

In general, both types of device are left in situ for as short a time as possible to minimize the risk of introducing infection. Prophylactic antibiotics are not generally used.

Indications for ICP monitoring

More specific indications:

Traumatic brain injury, in particular: Severe head injury (GCS 50 mmHgUsual treatment threshold is 20 mmHg

Head injury + ventilator + abnormal CT brain scan = ICP monitor

10

2. Monitoring the Injured Brain

Focal pathology on CT brain scan Head injury and age >40 Normal CT brain scan but systolic blood pressure persistently

S. Davies and A. Lindley

C-waves: Small oscillations in ICP that reflect changes in systemic arterial pressure.

With cerebral autoregulation intact, a rise in MAP produces vasoconstriction and a fall in ICP. However, when autoregulation fails, the circula-tion becomes pressure passive and changes in MAP are reflected in changes in the ICP. Continu-ous analysis of MAP and ICP allows a correlation coefficient called the pressure reactivity index to be derived (PRx). Positive values indicate dis-turbed cerebral vascular reactivity, whilst negative values indicate that reactivity remains intact (Gupta 2002).

Despite the fact that trial results have not always been compelling, most clinicians regard the ICP monitor as an essential tool that allows estimation of CPP (Czosnyka and Pickard 2004; Czosnyka et al. 1996), gives early warning of developing pathology, allows the response to therapy to be objectively measured, and also has value as a prognostic indicator (Joseph 2005).

Jugular Venous Oximetry (SjvO2)

SjvO2 is an indicator of global oxygen extraction of the brain. Jugular venous desaturation sug-gests an increase in cerebral oxygen extraction which indirectly implies that there has been a decrease in cerebral oxygen delivery, and hence perfusion.

The internal jugular vein drains the majority of blood from the brain, and in most patients the right lateral sinus is larger. Despite the fact that flow is different on the two sides, oxygen satura-tions are normally very similar. This also appears to be the case in diffuse brain injury, whilst in focal injuries there tends to be a greater difference in the saturations of the two veins.

Jugular venous saturations can be measured using the principle of infrared refractometry via a specially designed catheter (Gopinath et al. 1994).

SjvO2 values

5575% normal >75% luxury perfusion

2. Monitoring the Injured Brain

blood pressure may risk cerebral ischemia, especially in those patients with preoperative hypertension. SjvO2 monitoring allows the anesthetist to assess the degree to which blood pressure can be safely lowered during the oper-ative period. Similarly, a low PaCO2 will cause SjvO2 desaturation.

Problems with SjvO2 Monitoring

The major criticism of SjvO2 is that it is a measure of global oxygen delivery and does not reflect metabolic inadequacies in focal areas of injury, and hence may miss regional areas of ischemia. Inaccuracies can occur with catheter misplace-ment, contamination with extra cerebral blood, when the catheter abuts the vessel wall, or if thrombosis occurs around the catheter tip. Contraindications and complications are similar to those of an IJV central line.

Interpretation of Changes in SjvO2If cerebral oxygen delivery is impaired, oxygen extraction increases and SjvO2 decreases. If autoregulation is intact, CBF increases to meet metabolic demand and SjvO2 is restored. However, in the injured brain autoregulation is often impaired and cerebral ischemia ensues.

SjvO2: This implies inadequate cerebral oxygen delivery that may be due to decreased oxygen delivery (systemic hypoxia, anemia), decreased CBF (hypotension, raised ICP, excessive hypoc-apnia or vasospasm), or increased cerebral oxygen consumption (seizures, hyperthermia, pain)

SjvO2: This is somewhat more difficult to inter-pret, and may represent either hyperemia (e.g., when the autoregulation mechanisms are lost) or reduced oxygen consumption (e.g., hypothermia, deep sedation, or cerebral infarction).

Lactate Oxygen Index: During cerebral hypop-erfusion the brain can become a net producer of lactate, with the jugular venous lactate rising above arterial values. The lactate oxygen index is discussed in more detail in Chap. 3.

Brain Tissue Oximetry

Interest in measuring brain tissue oxygenation via implantable sensors has grown in recent years.

The Licox sensor is an implantable polarographic electrode that measures tissue oxygen tensions. It is inserted through a compatible bolt and ideally should be placed into the penumbral area of the injury. Oxygen diffuses from the tissue through the catheter into an electrolyte chamber where an electrical current is generated. Brain tissue oxygen tension is normally lower than arterial oxygen tension (1550 mmHg), whilst tissue CO2 is normally higher (range 4070 mmHg). The sensors are useful in monitoring local changes and trends in tissue oxygenation that might be missed by SjvO2 measurements.

At present it is primarily used in severe head injury and poor-grade subarachnoid hemor-rhage, and in conjunction with other monitoring moda lities. The technique allows a continuous method of monitoring of regional tissue oxygen-ation and in particular, monitoring areas of high ischemic risk, and is a promising and reliable clinical tool.

Noninvasive Monitoring

Transcranial Doppler Ultrasound

Transcranial Doppler is a noninvasive technique that calculates blood flow velocity in the cerebral vasculature. An ultrasound beam is reflected back by the moving blood stream at a different frequency than it was transmitted (Doppler shift), and from the Doppler equation the velocity of blood flow (FV) can be calculated. Changes in FV correlate well with changes in CBF, as long as the orienta-tion of the transducer and the vessel diameter remain constant. It is used clinically to diagnose vasospasm, to test cerebral autoregulation, and to detect emboli during cardiac surgery and carotid endarterectomy (Moppett and Mahajan 2004).

Normal Values

From the FV waveform systolic, diastolic, and mean velocities can be calculated. The mean FV in the middle cerebral artery (MCA) is usually 3590 cm/s and correlates well with CBF. The FV can be influenced by age, being lowest at birth (24 cm/s), highest at age 46 years (100 cm/s), and then declining until the seventh decade of life (40 cm/s). FV is also 35% higher in females and is also increased in hemodilutional states.

13

S. Davies and A. Lindley

Technique for Insonating the Middle Cerebral Artery (MCA)

The M1 branch of the MCA is the commonest vessel to be insonated, and is visualized through a transtemporal window (Fig. 2.4) with a 2 MHz pulsed Doppler signal. The anterior and posterior cerebral arteries can also be accessed through this window, whilst a transorbital approach allows access to the carotid siphon and the suboccipital route to the basilar and vertebral arteries.

Analysis of Doppler waveform

Analysis of the Doppler waveform gives rise to useful derived variables as well as blood velocity information.

Pulsatility Index (PI): FV sys FVdias/FVmean (normal value: 0.61.1)This reflects distal cerebrovascular resistance and correlates with CPP.Change in CBF with arterial CO 2 tension (cerebral vascular reactivity).

Uses of TCD in Intensive Care

Head Injury

Three distinct phases have been shown in severe head injury with regard to CBF and MCA FV.

Phase 1 occurs on the day of injury and has a normal CBF, normal MCA FV, and normal AVDO2. Phase 2 occurring 12 days post-injury, a hyper- emic state is encountered with an increased CBF, MCA FV and decreased AVDO2.

The final phase seen at days 415 is the vasos- pastic phase and is associated with a signifi-cantly decreased CBF and increased MCA FV. The use of TCD allows interpretation of the dynamic physiological changes seen in severe head injury, and in combination with other modalities allows perfusion and oxygenation to be optimized for the individual patient.The highest MCA FV recorded at any stage

had been shown to be an independent predictor of outcome from head injury, and the loss of autoregulation (calculated by regression of CPP on MCA FV) has also been shown to be a predictor of poor outcome from head injury.

Subarachnoid Hemorrhage

Vasospasm occurs in approximately 50% of people with subarachnoid hemorrhage between 217 days post-event, and is associated with significant morbidity and mortality. TCD may be used to detect vasospasm by the increase in MCA FV associated with vessel narrowing. Spasm is also assumed to be occurring when blood velocity is >120 cm/s (see Fig. 2.5a, b). High MCA FV is asso-ciated with worse-grade SAH, larger blood loads on CT (assessed by Fischer Grade) and hence worse outcome (Steiger et al. 1994).

Electroencephalography

An electroencephalogram (EEG) is obtained using the standardized system of electrode placement. Practically, this is not often readily available and requires expert interpretation. The EEG is affected by anesthetic agents and physiological abnormalities such as hypoxia, hypoperfusion, and hypercarbia.

A number of methods have been developed to simplify and summarize the EEG data.

Cerebral Function Monitor (CFM): This is a modified device from a conventional EEG. It uses a single biparietal or bitemporal lead, and is pro-cessed to give an overall representation of average cortical activity.

Cerebral function analyzing monitor: Developed from the CFM but displays information about both amplitude and frequency separately.

Bispectral Analysis: This modification of the EEG analyzes the phase and power between any two EEG frequencies. The bispectral index (BIS) is a dimensionless number statistically derived

Figure 2.4. Insonation of the middle cerebral artery through a trans-temporal window.

14

2. Monitoring the Injured Brain

from these phased and power frequencies and ranges from 0 to 100 (100-awake, 60-unconscious, 0-isoelectric EEG). This technology was derived with normal subjects and is not readily transferable to the injured brain, but may have a use in guiding sedation and analgesia.

Spectral Edge Frequency:

Compressed Spectral Array: Raw EEG data is processed into a number of sine waves (Fourier analysis). Power spectral Analysis then investi-gates the relationship between power and fre-quency of the sine waves over a short time period (Epoch). Compressed spectral array is obtained

by superimposing linear plots of successive epochs to produce a three-dimensional hill and valley plot (Fig. 2.6). The spectral edge frequency looks at the frequency below which a determined power of the total power spectrum occurs. SEF90 indicates a spectral edge frequency of 90% and is the frequency below which 90% of activity is occurring.

Application of the EEG in the ICU

Seizure management: Confirms the diagnosis of seizures and identifies a focal or lateralized source of activity. It also helps to distinguish between involuntary movements, posturing,

Figure 2.5. (a) and (b) TCD examination of a patient following a subarachnoid hemorrhage and endovascular coiling of an anterior communicating artery aneurysm. The patients GCS had dropped and they had developed a right-sided hemiparesis. The velocities on the right were normal whereas those on the left were high and indicative of vasospasm.

Figure 2.6. Hill and valley plot of the Compressed Spectral Array.

15

S. Davies and A. Lindley

and eye signs that are common in the intensive care, and true seizure activity.Nonconvulsive status epilepticus: This repre- sents a state that lasts more than 30 min with clinical evidence in alteration in mental state from normal, and seizure activity on the EEG. Between 4 and 20% of patients with status epi-lepticus have nonconvulsive episodes.Metabolic suppression: Burst suppression (isoelectric EEG) is a definable end point when pharmacological reduction of the cerebral metabolic rate of the injured brain is required for either neuroprotection or intractable intrac-ranial hypertension.Ensuring adequate sedation in patients who require prolonged neuromuscular paralysis.Prognosis: The EEG can be of prognostic value following brain injury, with absence of spon-taneous variability being associated with poor outcome.

Near Infrared Spectroscopy

While the criticism of jugular venous oximetry is that it is representative of global oxygen delivery, near infrared spectroscopy (NIRS) is a noninvasive tech-nique that measures regional cerebral oxygenation.

Light in the near infrared wavelength (7001,000 nm) can pass through bone, skin, and other tissues with minimal absorption, but is partly scat-tered and partly absorbed by brain tissue. The amount of light absorbed is proportional to the con-centration of chromophobes (iron in hemoglobin, and copper in cytochromes), and measurement of absorption at a number of wavelengths provides an estimate of oxygenation (Owen-Reece et al. 1999).

The probes illuminate a volume of about 810 mL of tissue and are ideally suited for use in neonates because of their thin skull, but have been used with success in adults.

Advantages of this technique are that it is non-invasive, and provides a regional indicator of cere-bral oxygenation. Its major limitation is its inability to distinguish between intra- and extra-cranial changes in blood flow.

Multimodal Monitoring

In any type of brain injury, the available monitoring modalities are prone to artifact and misinterpretation.

By utilizing more than one monitoring technique, the observer is more likely to determine whether a genuine change in cerebral physiology has occurred and what the most appropriate interven-tion should be. For instance, in traumatic brain injured patients we routinely monitor ICP, proc-essed EEG, SjvO2 and brain-tissue oxygen tension (PbtO2), allowing us to observe both local and regional changes in cerebral hemodynamics. General rules cannot always be applied to indi-vidual patients, and multimodal monitoring can allow more informed decision making such as determining CPP thresholds or the ability of the cerebral vasculature to autoregulate (Matta et al. 2000).

Conclusions

A wide range of monitoring techniques is available, each with their different strengths and limitations. Multimodal monitoring using a combination of techniques can overcome some of the limitations of the individual methods discussed. The choice of monitoring is often guided by clinical familiarity and local policy.

References

Brain Trauma Foundation Guidelines 2007. www.brain-trauma.org

Czosnyka M, Pickard JD (2004) Monitoring and inter-pretation of intracranial pressure. J Neurol Neurosurg Psychiatry 75:813821

Czosnyka M, Guazzo E, Whitehouse H, Smielewski P, Czosnyka Z, Kirkpatrick P et al (1996) Significance of intracranial pressure waveform analysis after head injury. Acta Neurochir (Wien) 138(5):531542

Gopinath SP, Robertson CS, Contant CF et al (1994) Jugular venous desaturation and outcome after head injury. J Neurol Neurosurg Psychiatry 57:717723

Gupta AK (2002) Monitoring the injured brain in the intensive care unit. J Postgrad Med 48(3):218225

Joseph M (2005) Intracranial pressure monitoring: vital information ignored. Indian J Crit Care Med 9(1): 3541

Juul N, Morris GF, Marshall SB, Marshall LF (2000) Intracranial hypertension and cerebral perfusion pressure: influence on neurological deterioration and outcome in severe head injury. the executive committee of the international selfotel trial. J Neuro-surg 92:16

16

2. Monitoring the Injured Brain

Marmarou A, Anderson RL, Ward JD et al. (1991) Impact of ICP instability and hypotension on outcome in patients with severe head trauma. J Neurosurg 75: S59S66.

Matta B, Menon D, Turner J (2000) Multimodal monitoring in neurointensive care. Textbook of Neuroanaesthesia and Critical Care. Greenwich Medical Media, Cambridge

Moppett IK, Mahajan RP (2004) Transcranial Doppler ultrasonography in anaesthesia and intensive care. Br J Anaesth 93:710724

Owen-Reece H, Smith M, Elwell CE, Goldstone JC (1999) Near infrared spectroscopy. Br J Anaesth 82:41826

Steiger HJ, Aaslid R, Stooss R, Seiler RW (1994) Transcra-nial Doppler monitoring in head injury: relations between type of injury, flow velocities, vasoreactivity, and outcome. Neurosurgery 34:7985

Young JS, Blow O, Turrentine F, Claridge JA, Schulman A (2003) Is there an upper limit of intracranial pressure in patients with severe head injury if cerebral perfusion pressure is maintained? Neurosurg Focus 15(6):E2

Anaesthesia UK. www.frca.co.uk

17

Key Points

1. The management of traumatic brain injury (TBI) has increasingly become more tailored to the individual patient; measuring adequacy of cerebral oxygenation may allow lower cerebral perfusion pressures to be targeted and more ra-tional adjustments of PaCO2 levels.

2. Patients with TBI who are hypothermic at pres-entation should not be rapidly rewarmed.

3. Hypertonic saline can be a useful alternative to mannitol in the management of intracranial hypertension.

4. Steroids are not currently recommended in the management of TBI.

5. Recombinant Factor VIIa may be useful in cases where correction of acidosis and hypothermia and administration of appropriate blood prod-ucts has failed to control continued nonsurgi-cal bleeding.

6. Decompressive craniectomy is a useful thera-peutic maneuver in selected cases of refractory intracranial hypertension.

The Secondary Management of Traumatic Brain Injury

The management of Traumatic Brain Injury (TBI) is challenging, right from the point of injury through to rehabilitation. Although this chapter sets out the evidence-base behind certain treat-

ment strategies, it will be clear to the reader that there is still no consensus position on many aspects of care. Traumatic Brain Injury constitutes the key cause of death in trauma, with trauma itself the principal cause of death and disability up to the age of 50. Given both the magnitude of the problem and the significant negative impact, there is an urgent need to promote both seamless and consistent care.

Pathogenesis of Brain Injury

Impact to the cranium may be wholly absorbed by fragmentation of the skull with no direct brain injury. Fractures in the temporo-parietal region may be associated with tears to the middle menin-geal artery and a resultant extradural hematoma, which if identified and evacuated quickly, is not usually associated with any significant longer-term implications. The underlying brain is, however, vulnerable to injury even without penetration of the skull. Internal movement results in compression, stretch, and shearing of neurons and supporting tissue, causing direct neuronal damage, hemorrhage, or contusion.

The MonroKellie doctrine describes the prin-ciple whereby skull contents of brain, blood, and CSF are normally in equipoise with a pressure of

D. Bell and J.P. Adams

thereby aggravating edema. Cellular dysfunction is associated with shifts in ionic concentrations, potassium leaving to be taken up by the glial cells, resulting in cytotoxic edema and astrocyte swell-ing. The concurrent influx of calcium and sodium into the neurons promotes release of excitatoxic neurotransmitters (e.g., glutamate), which further exacerbate calcium influx, and eventually results in irreversible change (see Chap. 1, Fig. 1.1). Destru ctive enzyme systems such as lipases and proteases are activated, aggravating cellular destruction with cell death ultimately triggered by the release of mitochondrial apoptotic proteins.

There appears to be a higher order of inflam-matory response than that seen after injury to other tissues or organs, compounded by the pres-ence of the rigid skull. This explains why certain patients who initially appear to have a relatively trivial injury progress to intractable intracranial hypertension (ICH) despite all appropriate care, culminating in death or profound deficit. Even those patients who eventually make a reasonable recovery may demonstrate a progressive rise in intracranial pressure (ICP) beyond a week after injury, in contrast to an inflammatory response after peripheral trauma, which typically peaks at 2448 h, and contrary to the perceived wisdom that high ICP will begin to abate after 48 h.

The absence of any treatment strategy that mod-ifies this complex and progressive inflam matory response illustrates the limitations of a neurosur-gical center to continuing the principle of avoiding secondary cerebral insults (Chap. 1, Table 1.2).

The Regional Neurosurgical Center (RNC) carries these principles further by taking active measures to increase oxygen delivery and reduce oxygen consumption and by controlling the volume of brain, blood, or CSF. This process requires spe-cialized monitoring to ensure that the strategies are effective (i.e., ICP is reduced) and that addi-tional cerebral insults are avoided (e.g., hyperven-tilation compromising cerebral oxygen delivery). Timely neurosurgical input allows rapid removal of any significant intracranial hematoma, monitor-ing for the potential expansion of a less significant hematoma and radical surgical maneuvers for refractory ICH (e.g., decompressive craniectomy).

Although flow diagrams are provided in the text and appendices, this chapter is not directed toward an empirical and prescriptive approach to care, but to an analysis of the various treatment options,

such that any practitioner can exercise profes-sional judgment when faced with different pathol-ogy, at a different time after injury, with a different clinical presentation.

Apart from obvious examples such as immedi-ate evacuation of an extradural hematoma, there are very few scenarios where there is a universally accepted unequivocal treatment strategy. With a subdural hematoma, bleeding arises from bridg-ing veins and the surface of the brain itself, and there is no single identifiable source to target. Any mass effect is as likely to be attributable to swell-ing of the underlying brain from the associated injury, as much as the hematoma. Further brain swelling is likely to take up any space created by removal of a hematoma. Embarking on surgery for removal of an intracerebral hematoma or con-tusion is even more problematical because of dif-ficulties in identifying, accessing, and controlling bleeding points, with the distinct possibility of collateral damage to surrounding vulnerable neural tissue. The decision in these scenarios is therefore based not just on evidence of a hema-toma radiologically, but also on location, size, pro-gression, associated intra and extra-cranial injury, co-morbidity, impact on neurological function, and the level of ICP.

The medical management of ICH is also vexed, with a wish to reduce cerebral oxygen demand through sedation making it impossible to under-take a functional neurological assessment. Further-more, sedative strategies have a negative impact on cardiovascular, respiratory, gastrointestinal, and immune status, all of which may at times generate significant secondary insults. Maneuvers under-taken to improve cerebral oxygen delivery such as increasing cerebral perfusion pressure (CPP) may constitute a secondary cerebral insult by other mechanisms. Strategies directed at control of a rising ICP such as hyperventilation or administra-tion of mannitol may similarly constitute second-ary insults. This chapter therefore aims to explore the benefits, hazards, and the weight of evidence to support the use of these various interventions.

Control of Cerebral Oxygen Demand

Seizure Control

Chapter 8 has a detailed description of the recog-nition and management of seizure activity; an

20

3. The Secondary Management of Traumatic Brain Injury

algorithm for the treatment of status epilepticus is also included in the appendices.

Sedation

Sedation not only reduces cerebral oxygen demand, but also contributes to the reduction of other secondary insults by facilitating airway control, optimizing ventilatory support, and reducing global oxygen demands. The negative aspects of sedation, however, affect most systems of the body and may ultimately contribute to both morbidity and mortality. The combination of sedation and positive pressure ventilation usually leads to a requirement for inotropic and/or vaso-pressor support, and not infrequently, one wit-nesses a rising requirement for these despite excluding failure of the pituitaryadrenal axis. This in turn is often associated with signs of coro-nary ischemia, particularly in previously fit young males.(Cremer et al. 2001) Adverse effect on gas-trointestinal function may be instrumental in the additional complication of VAP (ventilator-associ-ated pneumonia), as well as compromising nutri-tional status. Immune impairment may contribute to the development of infection, generating a common scenario whereby sedation has to con-tinue to manage gas-exchange problems caused by the acquired pneumonia. Sedation and immo-bility also predispose to thrombotic complications and skincare problems. One of the greatest diffi-culties with sedation, however, is the inability to make a functional neurological assessment, and clearly the longer sedation is continued, the greater the subsequent period in which the clinical picture may be compromised by drug accumulation or withdrawal phenomena, the latter possibly requir-ing the use of further sedative regimens.

Thiopentone creates singular difficulties in this regard, with an exceptionally long elimination half-life, no antagonist, and the ability after higher-level administration to mimic the signs of brainstem death with irregular, dilated, and unreactive pupils.Sedation should be provided:

1. When the patients level of consciousness is ob-tunded (GCS 8 or less) such that they cannot maintain or protect the airway or adequately self-ventilate and oxygenate

2. When intubation and ventilation is required to address other aspects of injury or disturbance of respiratory function, or

3. When ICP remains high, despite avoidance of all other cerebral insults and in the absence of any functional neurological activity.

Such primary sedation should ideally be noncu-mulative in the interests of early clinical assessment, with propofol/alfentanil a reasonable combination, but midazolam an acceptable addition or alternative to propofol, if high dosage causes cardiovascular problems or lipid accumulation. Remifentanil is gaining popularity as a single agent or in combi-nation with a sedative.

When using secondary sedation strategies for refractory ICH, it is essential to have a measurable endpoint (Winer et al. 1991). This includes the use of processed EEG and the achievement of burst suppression using the minimum amount of seda-tion necessary. In addition, the reduction in elec-trical activity should be accompanied by a fall in ICP or an increase in jugular venous oxygen saturation (SjvO2). The usual dose of thiopentone required to achieve the electrical end-point of burst suppression is a loading dose of 510 mg/kg, with a subsequent infusion rate of 510 mg/kg/h. If only used where high ICP is not responsive to all other strategies and following these principles, the hazards of barbiturate coma (pneumonia, sepsis syndrome, and hepatic dysfunction(Schwab et al. 1997) can be offset, not only against the control of ICH but also potential longer-term recovery ben-efits (Lee et al. 1994; The Brain Trauma Foundation. The American Association of Neurological Sur-geons 2000a; Dereeper et al. 2002). Recent local experience with brain-tissue oxygen measurement, however, has raised some concerns about the use of thiopentone. Despite seeing a fall in ICP and maintenance of a satisfactory SjvO2, thiopentone can lead to a reduction in brain-tissue oxygen levels (presumably risking a further ischemic insult), and therefore its use should probably be reserved for specialist centers.

Additional agents such as lidocaine (1 mg/kg 46 hourly) or ketamine may have a role in modi-fying surges in ICP in response to interventions such as suctioning, but again their use should probably be discussed with the RNC.

Hypothermia/Temperature Control

Barbiturate coma has the capacity to reduce brain oxygen and energy consumption by between 5060%. Induced hypothermia can reduce this further by

21

D. Bell and J.P. Adams

slowing cellular constitutive process. Control of pyrexia is universally accepted with symptomatic therapy (paracetamol, nonsteroidal antiinflamma-tory drugs, surface cooling) and by treating sources of infection, but the step beyond this to induction of hypothermia does not constitute routine practice.

Clinical studies have given conflicting results, (Marion et al. 1997; Clifton et al. 2001) but a more recent meta-analysis concluded that patients with high ICP refractory to all other maneuvers may benefit (Henderson et al. 2003). Given the positive conclusions of studies evaluating neurological outcome after induced hypothermia following cardiac arrest (The Hypothermia after Cardiac Arrest Study Group 2002), it is likely that this strategy will continue to be evaluated.

However, even mildmoderate (3335C) hypo-thermia can produce cardiovascular instability, disturbance of coagulation, and immune impairment. In addition, cerebral oxygen delivery may be com-promised through the reduction of cardiac output, cerebral vasoconstriction, increased plasma viscos-ity, and a shift in the oxygen dissociation curve to the left. It is our current practice to control pyrexia and promote normothermia rather than to induce hypothermia, unless the ICP is persistently elevated and unresponsive to other therapeutic maneuvers. Patients with traumatic brain injury who are hypo-thermic on admission should not be rapidly re-warmed as this may be associated with a poorer outcome.

Optimization of Cerebral Oxygen Delivery

An adequate circulating volume, with an adequate level of functional hemoglobin, with no cerebral vasoconstriction, and no factors compromising blood rheology or oxygen dissociation are key to ensuring cellular oxygen delivery. The brain requires a critical perfusion pressure above the lower limit of autoregulation. When ICP exceeds

20 mmHg flow through the microcirculation will be compromised, with an associated higher mortality (Johnston et al. 1970). However, it is simplistic to believe that increasing the CPP will compensate for this. If the blood brain barrier is disrupted and microvascular flow is impaired, an increase in MAP may simply aggravate brain swell-ing through the formation of hydrostatic edema. There is no evidence that increasing CPP improves the perfusion of pericontusional ischemic tissue (Steiner et al. 2003). A polarized debate continues on the correct approach in these circumstances, with proponents of the Lund philosophy (Grande 2004) (see Fig. 3.1) targeting the causes of high ICP, rather than relentlessly pursuing a fixed differen-tial pressure. Despite endorsement by national and international bodies, (The Brain Trauma Foundation. The American Association of Neurological Surgeons 2000b; Maas et al. 1997), there are no definitive controlled trials to conclusively prove that ICP-guided therapy is efficacious. Recommendations are based on the association between secondary insults and poor outcome (Jones et al. 1994), but the evidence base to take this one step beyond avoiding such insults to actively manipulating these variables is not fully established, with conflicting results in the literature. Although recent studies have demon-strated reduced mortality (Clayton et al. 2004) and improved outcome from protocolized ICP /CPP-directed care (Fakhry et al. 2004), other retro-spective cohort studies from hospitals with different strategies for head-injured patients demonstrated no evidence of benefit, as determined by the extended Glasgow Outcome Scale (Cremer et al. 2005). The CPP-directed strategy was however noted to be associated with prolonged mechanical venti-lation and increased levels of therapy intensity.

Our approach is to maintain a CPP of > 60 mmHg, with the target CPP being referenced to achie ving a SjvO2 value of greater than 60% and a lactate

The Lund Approach to severe traumatic brain injury

Pharmacological principles:

Reduction of capillary hydrostatic pressure with a metoprolol and clonidine

Reduction of cerebral blood volume with thiopental and dihydroergotamine

Reduction of stress response with opioids, benzodiazepines and thiopental

Maintaining normal colloid oncotic pressure with albumin, blood and plasma transfusions

Figure 3.1. The pharmacological principles of the Lund Approach to traumatic brain injury.

22

3. The Secondary Management of Traumatic Brain Injury

oxygen index of 10 gm/dl, incr e asing inspired oxygen concentration, and norma lizing CO2 can be associated with improved tissue oxygenation and potentially more favorable outcomes (Stiefel et al. 2005), whilst avoiding the hazards of pursuing CPP in the face of a relentless rise in ICP.

Control of ICP

Elevation of ICP after traumatic brain injury can be attributed to an expansion of the primary com-ponents (brain, blood, and CSF) or any focal pathology (hematoma or contusion), or a combi-nation of these factors.

Management is directed toward treatment or control of these contributing factors (see Fig. 3.3), with occasional symptomatic relief in the form of decompressive craniectomy. The warning signs and subsequent management of acute brain her-niation are outlined in Table 3.1

Management of Hematoma/Contusion

With the exception of an extradural hematoma, evacuation of a mass lesion is a significant under-taking that may in fact aggravate brain injury. There is no guarantee that either a hematoma or further brain swelling will not fill any space created, and surrounding ischemic tissue may suffer further collateral damage. Limiting progression involves scrupulous optimization of coagulation (platelets > 100,000 and INR and APTTR

D. Bell and J.P. Adams

Case History

Problem: An alcoholic patient with a recurrent extradural hematoma had ongoing nonsurgical bleeding from associated lower-limb injuries, despite provision of all blood products. The neurosurgeons were reluctant to operate without optimization of coagulation status.Treatment: Single dose of rFVIIa 120 mcg/kg over 3 min (Vialet et al. 2003) following administration of platelets, FFP, and cryoprecipitate and bicarbonate to adjust pH to 7.25.Outcome: Uneventful surgery, blood loss from other injuries minimized.Comment: No laboratory test for this treatment, the endpoint being clinical control of bleeding. Further dose to be given if no positive response within 15 min.

CIRCULATION

MAP > 80mmHg CPP > 60, if ICP measured*

Hb ~10g/dl Maintain adequate circul ating

volume (e .g. CVP, PC WP, ODM) If hypotensive, check for bleeding

Consider need for inotropes or vasopressors*

SEDATION

Propofol 1-6mg/kg/hrAlfentanil 1- 4mg/hr

Midazolam if unstabl eConsider paralysis

EEG for Thiopentone coma**

GENERAL MEAS URES

15 Head up tilt, neck neutral Check ETT ties, hard collar

OGT/ NGT Early enteral feeding

Metoclopramide if not absorbing H2 blocker / PPI

Insulin: maintain glucose 4-8mmo l/ l VTE prophylaxis*

VENTILATION

FiO2 1.0 until ABG V t 7-10ml/k g

Freq. 10-12/min PEEP 2.5-5cm H2O

PaO2>13kPa PaC O2 4.5-5kPa*

( until S jvO2 available)

MONITORING

ECG, SaO 2, EtCO 2IABP

CVP / PC WP / ODM Temperature ICP & CPP S jvO2, Pbt O2

EEG

INVESTIGATIONS

FBC, Clotting U&Es, LTFs

Glucose, Ca 2+

Arterial blood gases Group & Save

FLUIDS *

Daily Fluid Balance 0.9% NaCl maintenance ( unless

grossly hypernatremic) 6% starch ( up to 1.5l /day)

Blood, Hb ~ 10 g/dl Clotting products

(INR, APTT100)

PYREXIA

Culture blood, sputum, urine CRP

CXR, consider BAL Paracetamol 1g qd s

+/- NSAIDs** Active Cooling

Consider line change Antibiotics*

* see accompanying discussion in text ** discuss with RNC

Figure 3.3. Initial stabilization of the brain injured patient on ICU.

Manipulation of CSF

It is very rare for a rise in the volume of CSF to contribute to ICH after trauma, even when this is associated with traumatic subarachnoid hemor-rhage and a theoretical potential for aqueduct occlusion. The usual radiological picture is oblit-eration of the ventricular system, and in such circumstances there is nothing to be gained by manipulating CSF production medically, or by attempts at drainage surgically. The potential ben-efits of using loop diuretics, as discussed later in

24

3. The Secondary Management of Traumatic Brain Injury

the management of brain edema, do however include reduced CSF production, and in certain cases this may be of marginal benefit.

Manipulation of Intracranial Blood Volume

As an increase in blood volume may contribute to ICH, secondary cerebral insults such as hypercap-nia and venous obstruction must be avoided. Hyperventilation can reduce intra-cranial pressure by inducing cerebral vasoconstriction, but may worsen cerebral ischemia. It should only be used as a temporary measure to prevent imminent brainstem herniation unless SjvO2 and the lactateoxygen index are being measured. In a similar way, hyperoxia will also lead to cerebral vasoconstric-tion and a reduction in cerebral blood volume and as such, may also be a useful temporary holding measure when ICP is very high.

Manipulation of Brain Swelling

Osmotherapy

Prevention includes avoidance of hypotonic solu-tions such as 5% dextrose and by maintaining nor-moglycemia. With intact autoregulation, mannitol

increases cerebral blood flow by expanding the cir-culating volume and by improving rheology. This results in a reflex cerebral vasoconstriction and a rapid fall in ICP. A subsequent reduction in brain water then occurs because of the osmotic differen-tial. However, in contusional or diffuse axonal injury, the BBB is frequently disrupted. This may result in mannitol redistributing in the brain interstitium and contributing to the edema, rather than improv-ing it (Kaufmann and Cardoso 1992). Mannitol also increases serum sodium and osmolality, the latter becoming nephrotoxic at >320mOsm/L. The rise in serum sodium is also mirrored in the brain intersti-tium, this again generating osmotic edema. Man-nitol remains, therefore, most effective in shrinking relatively normal brain as a temporizing measure prior to definitive surgical relief of a mass lesion. It should be used as intermittent rather than continu-ous therapy and there is some evidence that it is more efficacious if used at higher dose (1.4 g/kg rather than conventional 0.5 g/kg) (Cruz et al. 2004). It is our practice to restrict mannitol to a dose of 0.51 g/kg every 6 h and to stop if serum osmolality exceeds 320 or serum sodium >160 mmol/l. We rarely use it beyond the first 2436 h after injury, after which our preference is to switch to hypertonic saline (HSL). This is claimed to be more effective than mannitol in reducing ICH, without compro-mising the hemodynamic status of the patient (Vialet et al. 2003; Munar et al. 2000). Sodium chlo-ride is completely excluded from the intact bloodbrain barrier (reflection coefficient = 1.0), and is theoretically a better osmotic agent than mannitol (reflection coefficient 0.9). Additional benefits may include antagonism of excitatory neurotransmitters. We use a regime of 30 ml of 20% HSL over 20 min through a central venous catheter. Repeated doses can be given after approximately 6 h as long as plasma [Na+] has not exceeded 160 mmol/l.

Loop Diuretics

Furosemide is effective in reducing brain water and is synergistic when used with mannitol. It reduces CSF production and increases sodium and water transfer through the arachnoid granu-lations. It also eliminates sodium and water through the kidneys, thereby avoiding the higher sodium levels seen with recurrent mannitol administration. Unlike mannitol, it does not con-tribute to brain edema. Our policy is to commence

Table 3.1. Recognition and management of acute brain herniation

Warning signs:Reduction in conscious levelUnilateral third Nerve palsyLateralising motor signs e.g., hemiparesis, extensor posturingHypertension, bradycardia or respiratory irregularity (Cushings Triad)

Management Rapidintubation and ventilation (great care needed to avoid

exaggerated pressor response to laryngoscopy and intubation experienced anaesthetist essential. Invasive blood pressure monitoring ideal, but do not delay establishing ventilation)

Hyperventilate to PaCO2 3.54.0 kPa as a temporary measure Mannitol 20% 0.5 g/kg over 10 min Sedation to reduce cerebral metabolic rate (e.g., propofol,

thiopentone) supplemented with opioid analgesic (e.g., fentanyl, alfentanil)

Head up position and good neck position to encourage venous drainage

MaintainadequateMAP (ideally 90100 mmHg) with pressor. Do not treat hypertension (may reduce cerebral perfusion)

100%O2 (hyperoxia) may reduce cerebral blood volume and ICP (especially in younger patients) and can be utilised whilst more definitive treatment is sought

ACT scanwillberequiredassoonasthepatientisstableenoughto be moved to exclude surgical lesions such as hydrocephalus or a haematoma

ContactRNC for further advice

25

D. Bell and J.P. Adams

an infusion at 0.3mg/kg/day and adjust to achieve neutral water balance over a 24-h period.

In instances where rapid control of ICH is required (i.e., herniation syndromes) and mannitol has not caused a significant diuresis, furosemide 0.250.5 mg/kg can be administered.

Steroids

The efficacy of steroids on modification of trau-matic edema or outcome has not been validated (Alderson and Roberts 1997). A recent large multi-center, prospective, randomized, placebo-controlled trial (CRASH study) demonstrated no reduction in mortality (Roberts et al. 2004),and their use is not currently recommended.

Management of ICH by Craniectomy

Following encouraging trial results (Guerra et al. 1999; Albanese et al. 2003), surgical intervention for ICH is currently being reevaluated. It should be considered in cases of intractable ICH unrespon-sive to all medical maneuvers, or when medical maneuvers are generating such significant side effects (most commonly myocardial ischemia) that morbidity or mortality are likely to arise from complications of treatment. The magnitude of this intervention should not be underestimated, with the possibility of uncontrolled bleeding and further brain injury. As our experience grows, it may be possible to identify those patients who will benefit most from early decompression. Although a randomized ICP rescue trial of thiopentone versus craniectomy is currently being carried out in the United Kingdom, (www.rescueicp.com), it is the authors current practice to consider the procedure for diffuse axonal injury in young patients with escalating vasopressor requirements and a LOI approaching 0.08 despite optimization.

Other Aspects of the Management of Traumatic Brain Injury

The general principles of intensive care such as early enteral nutritional support apply equally to the patient with traumatic brain injury (Fig. 3.3). However, certain aspects of care such as thrombo-prophylaxis, antibiotic therapy, and optimal timing of surgical intervention for other injuries are more contentious.

Neurosurgical patients are at a significant risk from thromboembolic complications. Each case has to be judged on individual merits depending upon CT findings, coagulation status, and associated injuries. Low molecular weight heparin (LMWH) is usually withheld until at least 2448 h after injury and possibly longer if surgical intervention is required, or if there is any radiological evidence of extension of focal pathology. Patients with trau-matic brain injury have at least a moderate risk of venous thromboembolism. LMWH for example, tinzaparin 35004500iu s.c. daily is now routinely started 2448 h after admission (obese patients may require larger doses based on body weight). Exceptions to this would include coagulopathy, low platelet count, hemorrhagic contusions, or imminent surgical intervention, but not the pres-ence of an ICP catheter as such. With any contrain-dication to LWMH, mechanical compression devices should be s