Publications of the University of Eastern Finland

Dissertations in Health Sciences

isbn 978-952-61-1006-6

Publications of the University of Eastern FinlandDissertations in Health Sciencesd

issertation

s | 149 | Ta

deu

sz Mu

sialo

wicz | E

EG

-based Monitorin

g during G

eneral A

naesth

esia and S

edation

Tadeusz MusialowiczEEG-based Monitoring

during General Anaesthesia and Sedation

Studies on Cardiac Surgery

and Status Epilepticus Patients

Tadeusz Musialowicz

EEG-based Monitoring during General Anaesthesia and SedationStudies on Cardiac Surgery and Status Epilepticus

Patients

The use of anaesthetic agents

during general anaesthesia

and sedation in the intensive

care unit induce changes in

patient’s electroencephalographic

(EEG) waveforms of the brain

cortex. The utility of EEG-based

neuromonitoring during general

anaesthesia for patients undergoing

cardiac surgery and postoperative

sedation and during the treatment

of status epilepticus in the intensive

care unit was evaluated in this

study. The dissertation provides new

information about the differences

between EEG-based monitoring

techniques and their feasibility to

assess hypnosis in cardiac surgery

and in intensive care settings.

TADEUSZ MUSIALOWICZ

EEG-based Monitoring during General

Anaesthesia and Sedation

Studies on Cardiac Surgery and Status Epilepticus Patients

To be presented by permission of the Faculty of Health Sciences, University of Eastern Finland for public examination in Kuopio University Mediteknia, Auditorium 1,

on Friday, January 18th 2013, at 12 noon

Publications of the University of Eastern Finland Dissertations in Health Sciences

Number 149

Department of Anaesthesiology and Department of Intensive Care Medicine, Kuopio University Hospital, Institute of Clinical Medicine, School of Medicine, Faculty of Health Sciences,

University of Eastern Finland Kuopio

2013

Kopijyvä OY Kuopio, 2013

Series Editors:

Professor Veli-Matti Kosma, M.D., Ph.D. Institute of Clinical Medicine, Pathology

Faculty of Health Sciences

Professor Hannele Turunen, Ph.D. Department of Nursing Science

Faculty of Health Sciences

Professor Olli Gröhn, Ph.D. A.I. Virtanen Institute for Molecular Sciences

Faculty of Health Sciences

Distributor: University of Eastern Finland

Kuopio Campus Library P.O.Box 1627

FI-70211 Kuopio, Finland http://www.uef.fi/kirjasto

ISBN (print):978-952-61-1006-6 ISBN (pdf): 978-952-61-1007-3

ISSN (print):1798-5706 ISSN: (pdf): 1798-5714

ISSN-L: 1798-5706

III

Author’s address: Department of Anaesthesiology and Department of Intensive Care Medicine Kuopio University Hospital

University of Eastern Finland KUOPIO FINLAND Supervisors: Docent Ilkka Parviainen, M.D., Ph.D. Department of Intensive Care Medicine Kuopio University Hospital University of Eastern Finland

KUOPIO FINLAND Professor Markku Hynynen, M.D., Ph.D. Department of Anaesthesiology Jorvi Hospital, Espoo Helsinki University Hospital University of Helsinki HELSINKI FINLAND

Reviewers: Professor Seppo Alahuhta, M.D., Ph.D. Department of Anaesthesiology Oulu University Hospital

University of Oulu OULU FINLAND

Docent Anne Vakkuri, M.D., Ph.D.

Department of Anaesthesiology Peijas Hospital, Vantaa Helsinki University Hospital University of Helsinki HELSINKI FINLAND

Opponent: Professor Arvi Yli-Hankala, M.D., Ph.D.

Department of Anaesthesiology Tampere University Hospital University of Tampere TAMPERE FINLAND

IV

V

Musialowicz, Tadeusz. EEG-based monitoring during general anaesthesia and sedation. Studies on cardiac surgery and status epilepticus patients. University of Eastern Finland, Faculty of Health Sciences, 2013. Publications of the University of Eastern Finland. Dissertation in Health Sciences number 149. 2013. 92 p. ISBN (print):978-952-61-1006-6 ISBN (pdf): 978-952-61-1007-3 ISSN (print):1798-5706 ISSN: (pdf): 1798-5714 ISSN-L: 1798-5706 ABSTRACT

The aim of this study was to evaluate and compare the utility of different EEG-based neuromonitoring methods during general anaesthesia for cardiac surgery and sedation in the ICU. In study I, midlatency auditory-evoked potentials (MLAEPs) were evaluated as a measure of the sedation level in cardiac surgery patients during postoperative sedation. The purpose of this study was to evaluate whether graded changes in sedation induced by pre- and postoperative sedative drugs can be detected by any of the components of MLAEPs. The latency of the Nb component increased significantly after premedication and during deep and moderate sedation, and remained increased the day after surgery. The Na-Pa amplitude was decreased during deep sedation compared with baseline and moderate sedation. In study II, MLAEPs were evaluated in cardiac surgery patients when anaesthesia was guided with the aid of the bispectral index (BIS). It was tested whether the parameters derived from EEG indicate similar states of anaesthesia with the two different anaesthesia techniques. At similar BIS levels, the latency of the Nb component of MLAEPs was significantly increased after intubation, and both the Nb and Pa component was significantly increased after sternotomy in the patients anaesthetised with isoflurane compared with propofol. Our results suggest that the parallel use of these two methods can show differences in the components of anaesthesia. In study III, the feasibility of the BIS index to assess the EEG burst suppression pattern during anaesthesia for refractory status epilepticus (RSE) in the ICU was evaluated. An excellent correlation was found between the BIS index and bursts per minute in the raw EEG. BIS monitoring represents a useful alternative for titration of an anaesthetic agent to reach sufficient suppression in the EEG. In study IV, state entropy (SE) values were compared with the BIS index in patients undergoing general anaesthesia for cardiac surgery. Response entropy (RE) as a predictor for surgical noxious stimulation was also evaluated. A poor agreement between SE and BIS index values was found, but RE was a good predictor of arousal after surgical stimulation. In conclusion, in EEG-based monitoring for the assessment of hypnosis during cardiac surgery, different indices show poor agreement (BIS vs. SE) and indicate different components of anaesthesia (MLAEP vs. BIS). In the ICU settings, EEG-based methods can be used to titrate various levels of sedation and to monitor the response to RSE treatment.

National Library of Medical Classification: QU 34, WG 169, WL 150, WL 385, WO 200, WO 275, WV 270

Medical Subject Headings: Anesthesia, General; Consciousness Monitors; Conscious Sedation; Deep Sedation; Electroencephalography; Entropy; Evoked Potentials, Auditory; Cardiac Surgical Procedures; Intensive Care Units; Status Epilepticus

VI

VII

Musialowicz, Tadeusz. EEG:hen pohjautuva yleisanestesian ja sedaation valvonta sydänleikkattulla ja status-epileptikus potilailla. Itä-Suomen yliopisto, Terveystieteiden tiedekunta, 2013. Publications of the University of Eastern Finland. Dissertation in Health Sciences number 149. 2013. 92 p. ISBN (print):978-952-61-1006-6 ISBN (pdf): 978-952-61-1007-3 ISSN (print):1798-5706 ISSN: (pdf): 1798-5714 ISSN-L: 1798-5706 TIIVISTELMÄ

Tämän väitöskirjatyön tavoitteena oli arvioida ja verrata eri EEG-perusteisten mittareiden käyttökelpoisuutta sydänanestesian ja tehohoidon sedaation aikana. Ensimmäisessä osatyössä tutkittiin keskilatenttisia kuuloherätevasteita postoperatiivisessa sedaation asteen arvioinnissa sepelvaltimo-ohitusleikkauspotilaalla. Tutkimuksen tarkoituksena oli todeta, pystyykö kuuloherätevasteiden eri kompononttien avulla arvioimaan sedaation eri vaiheita ja sedaatiivisten lääkkeiden vaikutuksia ennen leikkausta ja sen jälkeen. Kuuloherätevasteiden Nb-komponentin latenssi piteni esilääkitysten jälkeen, syvässä ja kohtalaisessa sedaatiossa ja leikkauksen jälkeen seuraavana päivänä. Na-Pa komponentin amplitudi oli pienempi syvässä sedaatiossa kuin leikkausta edeltävänä päivänä ja kohtalaisessa sedaatiossa. Toisessa osatyössä tutkittiin keskilatenttisia kuuloherätevasteita sepelvaltimo-ohitusleikkauspotilaalla, kun anestesian syvyyttä ohjattiin bispektraali (BIS) indeksin avulla käyttäen kahta eri anestesiamenetelmä. Tutkimuksen tarkoitus oli selvittää, kuvastavatko kaksi eri EEG:hen pohjatuvaa anestesian komponentteja monitoroivaa menetelmää anestesian tilaa samalla tavalla sydänleikkauksen eri vaiheessa mitattuna. Kun BIS-indeksi oli samalla tasolla molemmissa anestesiaryhmissä, Nb- ja Pa latenssit erosivat ryhmien välillä kipuärsytyksen jälkeen. Tulokset viittaavat siihen, että BIS mittaa anestesian hypnoottista komponenttia, kun taas herätepotentiaalit kuvastavat anestesian hypnoottisen ja analgeettisen komponentin välistä tasapainoa. Kolmannessa osatyössä tutkittiin BIS-monitorin soveltuvuutta tunnistaa purskevaimentuma status epileptikuksen hoidossa anestesian aikana. Sekä BIS indeksi että BIS-supressiosuhde korreloivat hyvin EEG:n purskevaimentuman kanssa. BIS- monitoroinnin avulla voidaan ohjata anestesia-aineiden annostelua ylläpitämään purskevaimentuma status epileptikuksen hoidossa. Neljännen osatyön tarkoituksena oli verrata BIS- ja entropiamonitorin indeksien yhtäpitävyyttä ja korrelaatiota sydänleikkauksen anestesian yhteydessä. Samalla tutkittiin, onko vaste-entropia (Response entropy, RE) parempi kirurgisen stimulaation indikaattori kuin BIS, otsalihasten elektromyografia (EMG) tai hemodynaamiset vasteet. Tase-entropia (State entropy, SE) ja RE korreloivat hyvin BIS:n kanssa verrattaessa koko anestesian aikaista keskiarvoa, mutta leikkauksen eri vaiheissa SE:n ja BIS:n välillä oli huomattavia eroja. RE oli hyvä nosiseption indikaattori kirurgisen stimulaation jälkeen. Johtopäätöksenä voidaan todeta, että eri EEG:hen perustuvat anestesian mittarit eroavat kyvyssä kuvastaa anestesian eri komponentteja. Tehohoidossa näitä sedaation mittareita voidaan käyttää sedaation asteen säätelyssä ja status epileptikuspotilaan anestesian seurannassa.

Yleinen Suomalainen asiasanasto: EEG; monitorointi; nukutus; sedaatio; tehohoito; epilepsia; sydänkirurgia

VIII

IX

I dedicate this thesis to my wife

Dorota and to my sons Antti and

Aleksander.

Salus aegroti suprema lex

X

XI

Acknowledgements

This study was carried out in the Department of Anaesthesiology and Intensive Care Medicine, in collaboration with Departments of Clinical Neurophysiology and Neurology, Kuopio University Hospital, Kuopio, Finland.

I would like to thank Professor Esko Ruokonen who first inspired me to start clinical research work with two glasses of “Jack Daniels” at the Helsinki airport in the beginning of the year 2000.

I wish to thank my principal supervisors Docent Ilkka Parviainen and Professor Markku Hynynen. I appreciate Ilkka for his clinical teaching in the ICU and scientific mentoring of this work. I thank Markku for his guidance and tuition during this work and for in-depth revision of the thesis.

I wish to thank my co-authors, Heidi Yppärilä-Wolters, Minna Niskanen, Stephan M. Jakob, Otto Pitkänen, Mikko Pöyhönen, Pekka Pölönen, Esa Mervaala, Reetta Kälviäinen, Pasi Lahtinen, Jouni Kurola and Ari Uusaro.

I am grateful to my official reviewers, Docent Anne Vakkuri and Professor Seppo Alahuhta for their expert advice and constructive criticism of the manuscript.

Sincere thanks belong to all my colleagues at the Departments of Anaesthesiology and Intensive Care Medicine for their collaboration during this work.

I wish to express my appreciation to the study nurses Petri Toroi, Heikki Ahonen and Timo Tuovinen, for their assistance in data acquisition.

Finally, I owe my loving thanks to my family, my wife Dorota for being the heart and soul in my life and sons Antti and Aleksander.

I have received financial support from Kuopio University Hospital, in the form of EVO grant.

Kuopio, January, 2013

Tadeusz Musialowicz

XII

XIII

List of the original publications

This dissertation is based on the following original publications:

I Musialowicz T, Hynynen M, Yppärilä H, Pölönen P, Ruokonen E, Jakob SM. Midlatency auditory-evoked potentials in the assessment of sedation in cardiac surgery patients. Journal of Cardiothoracic and Vascular Anesthesia 18:559-562, 2004.

II Musialowicz T, Niskanen M, Yppärilä-Wolters H, Pöyhönen M, Pitkänen O,

Hynynen M. Auditory evoked potentials in bispectral index-guided anaesthesia for cardiac surgery. European Journal of Anaesthesiology 24:571-579, 2007.

III Musialowicz T, Mervaala E, Kälviäinen R, Uusaro A, Ruokonen E, Parviainen I.

Can BIS monitoring be used to assess the depth of propofol anesthesia in the treatment of refractory status epilepticus? Epilepsia 51:1580-1586, 2010.

IV Musialowicz T, Lahtinen P, Pitkänen O, Kurola J, Parviainen I. Comparison of

spectral entropy and BIS VISTA™ monitor during general anesthesia for cardiac surgery. Journal of Clinical Monitoring and Computing 25:95-103, 2011.

The publications were reprinted with the permission of the copyright owners. In addition, previously unpublished data are also presented.

XIV

XV

Contents

1 INTRODUCTION 1

2 REVIEW OF THE LITERATURE 4 2.1 Mechanism of general anaesthesia 4

2.1.1 Physiological sleep versus anaesthesia 4 2.1.2 Central nervous system during anaesthesia 5

2.2 Levels of anaesthesia 7 2.3 Awareness during general anaesthesia 9

2.3.1 Definition of awareness 9 2.3.2 Incidence of awareness during general anaesthesia 10 2.3.3 Awareness during cardiac surgery 10 2.3.4 Risks factors of awareness 10

2.4 Assessment of anaesthesia depth 12 2.4.1 Clinical variables 12 2.4.2 EEG-based monitoring 13 2.4.3 Auditory evoked potentials 14

2.4.3.1 AEP anaesthesia monitor 16 2.4.4 Bispectral index 17

2.4.4.1 BIS evaluation studies 17 2.4.4.2 BIS and awareness during general anaesthesia 19 2.4.4.3 BIS and postoperative outcomes 21 2.4.4.4. Automatic anaesthesia delivery system 22

2.4.5 Spectral entropy 22 2.4.5.1 Entropy evaluation studies 23 2.4.5.2 Assessment of nociception during general anaesthesia 27

2.5. Assessment of the sedation level in the ICU 29 2.5.1 Clinical methods of sedation assessment 30 2.5.2 EEG-based monitoring of ICU sedation 32 2.5.3 Auditory evoked potentials and ICU sedation 32 2.5.4 Bispectral index and ICU sedation 33

2.5.4.1 BIS and sedation scales 33 2.5.5 Spectral entropy and ICU sedation 34

2.6. Anaesthesia of cardiac surgery 36 2.6.1 Hypothermia, cardiopulmonary bypass and EEG- based monitoring 37 2.6.2 Postoperative sedation after cardiac surgery 38

2.7. Refractory status epilepticus in the ICU 39 2.7.1. Anaesthesia and goal of the treatment 39

XVI

2.7.2. Monitoring sedation depth during status epilepticus treatment 40

3 AIMS OF THE STUDY 43

4 PATIENTS AND METHODS 45

4.1 Study patients 45 4.2 Anaesthesia, sedation protocol and cardiopulmonary bypass 46 4.3 Data collection 47 4.4 Statistical analysis 50

5 RESULTS 52

5.1 MLAEPs during ICU sedation 52 5.2 MLAEPs versus BIS during anaesthesia for cardiac surgery 53 5.3 BIS monitoring during the treatment of status epilepticus in ICU 54 5.4 BIS versus spectral entropy during anaesthesia for cardiac surgery 55

6 DISCUSSION 56

6.1 Assessment of sedation by MLAEPs after cardiac surgery 56 6.1.1 MLAEPs and anaesthesia 56 6.1.2 MLAEPs and Ramsay score 57 6.1.3 MLAEPs after premedication 57

6.2 MLAEPs in Bispectral index-guided anaesthesia during cardiac surgery 58 6.2.1 MLAEPs and BIS after noxious stimulation 58 6.2.2 TIVA versus volatile anaesthesia 59

6.3 Bispectral index monitoring during the treatment of status epilepticus 61 6.3.1 BIS and burst suppression in the EEG 61 6.3.2 BIS and epileptiform activity in the EEG 62

6.4 Comparison of BIS VISTA™ and entropy monitoring during cardiac surgery 63 6.4.1 Correlation and agreement 64 6.4.2 EMG activity, BIS and RE 65 6.4.3 RE and noxious stimulation 65 6.4.4 Entropy and muscle relaxation 66

6.5 Limitations of the study 67 6.6 General conclusion 68

7 CONCLUSIONS 69

8 REFERENCES 70

ORIGINAL PUBLICATIONS I-IV

XVII

Abbreviations

AAI A-line™ autoregressive index

AEPs Auditory evoked potentials

BAEPs Brainstem auditory evoked potentials

BIS Bispectral index scale

BS Burst suppression

CABG Coronary arteries bypass grafting

CBF Cerebral blood flow

CNS Central nervous system

CPB Cardiopulmonary bypass

EEG Electroencephalogram

EMG Electromyography

ETAC End-tidal anaesthetic-agent concentration

FTCA Fast track cardiac anaesthesia

GABA Gamma-aminobutric acid

GCS Glasgow coma scale

CLADS Closed-loop anaesthesia delivery system

GMR Glucose metabolic rate

HRV Heart rate variability

ICU Intensive care unit

MAC Minimum alveolar concentration

MLAEPs Midlatency auditory evoked potentials

NMBA Neuromuscular blocking agent

NMDA N-methyl-D- asparate

NMT neuromuscular transmission monitor

LOC Loss of consciousness

OAAS Observer’s assessment of the alertness and sedation

PET Positron emission tomography

PPGA Photoplethysmographic pulse wave amplitude

PTDS Posttraumatic distress syndrome

RASS Richmond agitation-sedation scale

RE Response entropy

RN Response index of nociception

RS Ramsay score

SAS Sedation-agitation scale

SE State entropy

SPI Surgical pleth index

SR Suppression ratio

XVIII

SSEPs Somatosensory evoked potentials

SSI Surgical stress index

TIVA Total intravenous anaesthesia

TOF Train-of-four

XIX

1 Introduction

General surgical anaesthesia is defined as a state of reversible unconsciousness with amnesia, analgesia, and immobilisation of the patient caused by intravenous and inhalation anaesthetic drugs (John and Prichep 2005, Apfelbaum et al. 2006). The depth of general anaesthesia was traditionally assessed by haemodynamic signs like tachycardia and hypertension, and autonomic signs like motor response, lacrimation, and changes in respiratory frequency, and pupil size (Guedel 1937, Thornton and Jones 1993), which do not correlate well with a patient’s level of consciousness (Moerman et al. 1993, Domino et al. 1999). Electroencephalogram (EEG) patterns are known to change with the patient’s depth of sedation and his or her overall anaesthesia level (Gibs et al. 1937, Martin et al. 1959). Different anaesthetics induce different changes in the EEG pattern (Yli-Hankala 1990a), and this disparity restricts the clinicians’ use of a raw EEG in their assessment of the depth of anaesthesia (Thornton 1991). Assessing the level of anaesthesia with only raw EEG is difficult, because visual interpretation of the EEG signal requires neurophysiological knowledge. Moreover, most clinicians are unfamiliar with the task (Barnard et al. 2007). Thus, the processed automatic EEG-based monitoring of hypnosis with use of algorithms to analyse the EEG signal has been under active development in the past decades.

Auditory evoked potentials (AEPs) reflect the response in the EEG to auditory stimulation, usually as a click through headphones. The auditory sensory mechanism is reported to be the last to disappear during the induction of anaesthesia and the first to recover during the return of consciousness (Jones and Konieczko 1986). Because the brainstem auditory evoked potentials (BAEPs) mainly remain unaffected by intravenous anaesthetic agents (Savoia et al. 1988), the relation between midlatency auditory evoked potentials (MLAEPs) and the depth of anaesthesia has been under active research (Thornton et al. 1983, 1984, 1985, 1989a, De Beer et al. 1996a). The decrease of the amplitudes and increase in the latencies as a linear manner with increasing concentration of volatile and intravenous anaesthetics has been reported (Thornton 1991).

The bispectral index scale (BIS) was the first commercially available anaesthesia depth monitor used in clinical practice as a measure of the pharmacodynamic anaesthetic effect on the central nervous system (CNS) (Sigl and Chamoun 1994, Glass et al. 1997, Johansen 2006). The algorithm used to generate the BIS is proprietary of Covidien and not freely available, but the basic principles have been described (Rampil 1998a, Bruhn et al. 2000). During moderate sedation and light anaesthetic states, the BIS index is calculated from the

thesia, it is calculated from the SyncFastSlow algorithm. Finally, it is calculated from the Burst Suppression Ratio/Quazi-algorithm during very deep anaesthesia levels (Rampil 1998a). The BIS monitor also analyses the suppression ratio (SR), which represents the percentage of the previous 63-second epoch of

BetaRatio algorithm. During surgical anaes

2

EEG that is isoelectric. Electromyography (EMG) of forehead muscle activity and signal quality analyses are also incorporated. The BIS monitor provides a BIS index between zero (indicating that the patient is under deep anaesthesia and isoelectric EEG) and 100, indicating that the patient is fully awake. The BIS index during surgical anaesthesia should be between 40 and 60 (Glass et al. 1997, Sebel et al. 1997). The BIS index monitor was first clinically validated with volatile anaesthetics and intravenous anaesthetics (Glass et al. 1997). The use of a BIS index monitor to assess the level of consciousness reduces the amount of anaesthetics used (Gan et al. 1997, Yli-Hankala et al. 1999), patient recovery time (Gan et al. 1997, White at al. 2004), patient discharge time from the hospital (Gan et al. 1997), and the incidence of awareness during general anaesthesia (Ekman et al. 2004, Myles et al. 2004).

The spectral entropy monitor is the most recent development in the field of monitoring hypnosis during general anaesthesia. The research work and validation of the spectral entropy technology was done in Finland. The spectral entropy algorithm is a mathematical calculation of the irregularity of an EEG signal (Viertiö-Oja et al. 2004, Vakkuri et al. 2004). In the new entropy monitor, two parameters are calculated: state entropy (SE) and response entropy (RE). The SE reflects the hypnotic level of anaesthesia and the patient’s brain cortical activity and is computed within an EEG range of 0.8 to 32 Hz. RE includes EEG and EMG components calculated within an EEG range of 0.8 to 47 HZ. RE also reflects forehead muscle activity. EMG activity indicates that the patient is responding to external stimuli (Vakkuri et al. 2004 and 2005). The BIS and spectral entropy monitors use a different algorithm for the EEG-analysis, but the information about the level of hypnosis should be near each other in levels above 60 and below 40 and almost equal between recommended range for surgical anaesthesia (40 – 60) (Vakkuri et al. 2005, Leffol-Masson et al. 2007). The clinical research comparing the monitors’ indices in the same patient during general anaesthesia is important for understanding the differences between monitors in daily clinical practice (Vakkuri et al. 2004, Vanluchene et al. 2004).

The use of this brain monitoring technology for measuring anaesthesia during cardiac surgery is challenging, because special characteristics, such as cardiopulmonary bypass (CPB) and hypothermia, generate changes in the EEG as well as in EEG-based monitoring (Bashein et al. 1992, Edmonds et al. 1996, Doi et al. 1997a, Mathew et al. 2001). Cardiac surgery also requires the use of CPB, in which blood pressure and heart rate as indicators of anaesthesia depth are absent (Kertai et al. 2012). The use of drugs for hypertension and β blockers and the intraoperative use of vasoactive drugs in this group of patients can mask the haemodynamic signs of inadequate anaesthesia and is potentially a confounding factor (Kertai et al. 2012). The objective monitoring of the depth of anaesthesia with EEG-based devices is vital, because every patient differs in requirements of sufficient dosage of anaesthetic agents. Young and healthy patients, who usually possess a relatively large therapeutic window, need more anaesthetic agents than older cardiac surgery patients who possess comorbidities.

3

The intensive care unit (ICU) environment, artificial ventilation, and invasive monitoring are very stressful for patients, and proper sedation is essential to reduce patient anxiety (Hansen–Flaschen et al. 1991, Nasraway et al. 2002a, Jacobi et al. 2002). The goal of ICU sedation is a tranquil, co-operative patient, who requires usually a light state of sedation (Shapiro et al. 1995). Too deep sedation is associated with cardio-respiratory depression, prolonged recovery-time, delayed extubation, and even a prolonged stay in the ICU with increased costs (Brook et al. 1999, Ostermann et al. 2000, Kress et al. 2000). In their daily practice, clinicians assess the level of sedation by using subjective scoring systems, such as the Ramsay Score (RS), Sedation-Agitation Scale (SAS), or Richmond Agitation-Sedation Scale (RASS), all of which classify the patient’s response to a command or painful stimulus (De Jonghe et al. 2000, Ostermann et al. 2000, Payen et al. 2007)). However, the predictive value of sedation scales is limited due to many confounding factors, such as motor deficit, altered states of consciousness, and inter-rater variation in interpretation. These limitations call for new tools using objective EEG-based monitoring of sedation levels. In spite of active clinical research in this field, there exists no validated EEG-based monitor for sedation assessment in ICU settings. On the other hand, in some specific situations during ICU treatment, like in the treatment of refractory status epilepticus (RSE), deep anaesthesia (to the point of instigating a burst suppression pattern in the EEG) is required to stop electrical activity in the brain (Meierkord et al. 2006). Continuous EEG monitoring during treatment of status epilepticus is essential to guide the depth of anaesthesia. Unfortunately, it is not always available.

The aim of this study was to evaluate the utility of EEG-based neuromonitoring during general anaesthesia for patients undergoing cardiac surgery, postoperative ICU sedation and during the treatment of status epilepticus in the ICU. It was evaluated by comparing the information from AEP to the information from BIS (study II) and by comparing the information from the spectral entropy to the BIS (study IV) during cardiac surgery with CPB. In the two other studies, the AEP was evaluated during short-duration postoperative sedation after cardiac surgery (study I), and the utility of the BIS during deep anaesthesia for the treatment of a refractory status epilepticus in the ICU (study III).

4

2 Review of the literature

2.1 MECHANISM OF GENERAL ANAESTHESIA

2.1.1 Physiological sleep versus anaesthesia

General anaesthesia is a reversible state of unconsciousness and insusceptibility to pain that a clinician produces within a patient by administrating inhalation or intravenous anaesthetic and analgesic agents to the patient (Campagna et al. 2003, Brown et al. 2010). Anaesthesia provides loss of consciousness, analgesia, and patient immobility. Balanced general anaesthesia consists of three components: hypnosis, analgesia, and muscle relaxation. During adequate general anaesthesia, all three components are in balance. Hypnosis causes lack of awareness and amnesia, and analgesia induces antinociception and autonomic stability. Both of them are relevant to the anaesthetic and analgesic drugs action on the CNS. In spite of active scientific research, the medical field possesses no exact knowledge regarding the action of general anaesthetics on the human brain (John and Prichep 2005). To acquire a better understanding of the state of anaesthesia, we should know the differences between normal physiological sleep, anaesthesia, and the comatose state caused by brain injury. Sleep is a state of decreased arousal that cycles at 90-minute intervals between two states: rapid-eye-movement (REM) sleep and non-REM sleep. The differences between physiological sleep and anaesthesia appear in Table 1 (Brown et al. 2010).

Table 1. Differences between normal sleep and general anaesthesia.

Sleep Anaesthesia

Endogenously generated Drug induced Onset and duration are disrupted by Onset and duration depend stress and environment factors on drug dose Rhythmic cycling between wake State varies with dose and sleep stages Arousal threshold (non-REM vs. REM) Arousal threshold suppressed Role in memory consolidation Amnesia Return to wakefulness in minutes Return to wakefulness in hours No “side effect” Semiconscious, nausea Metabolic rate low in non-REM and Decrease in metabolic rate increased in REM phase

5

2.1.2. Central nervous system during anaesthesia

During the preoperative visit, the anaesthesiologist usually discusses with patients about “being put to sleep” instead of coma, but in fact, general anaesthesia is a reversible drug-induced coma (Brown et al. 2010). During the deep coma state, patient responsiveness to pain stimulation disappears, just as it disappears during surgical anaesthesia. The EEG pattern during a deep coma is also similar to the EEG pattern that appears during deep anaesthesia with high-amplitude, low frequency activity (Young 2000). The anatomical structures that are directly involved in human consciousness include the peripheral sensors, ascending sensory pathway, thalamus, ascending reticular activating system, reticular nucleus, and functional cortical regions (Bonhomme and Hans 2004). There is no single “anaesthesia centre” in the brain. Anaesthesia causes global reduction in cerebral blood flow (CBF) and brain metabolism, and these causes arise from the hyperpolarisation of neuronal pathways to the cortex. During a positron emission tomography (PET) study by Kaisti et al. (2002), propofol and sevoflurane induced the global reduction in CBF. Both anaesthetic agents reduced relative CBF in the cuneus, precuneus, posterior limbic system, and the thalamus or midbrain; additionally, propofol reduced relative CBF in the parietal and frontal cortex. In another PET study by Schlüzen et al. (2010), sevoflurane caused a global whole-brain metabolic reduction of the glucose metabolic rate (GMR) in all regions of the human brain, with the most marked metabolic suppression in the lingual gyrus, thalamus, and occipital lobe. During the PET study by Jeong et al. (2006), propofol suppressed the GMR of the neocortex area more than did sevoflurane and sevoflurane suppressed the GMR of the paleocortex and telencephalon more than did propofol. Thus, the different anaesthetic agents (i.e., intravenous and volatile) offer different mechanisms of action on the brain structure.

In the unique study by Långsjö et al. (2012) the changes in the level of consciousness and regional CBF was visualising by PET scanning, using two different anaesthetic drugs (dexmedetomidine and propofol). During the emergency of consciousness the activation of a core network in the subcortical and limbic regions, localised in deep, phylogenetically old brain structures were noticed in PET scanning. The result of this study suggests that the neural network of conscious state involves not only the neocortex, but also deep parts of the brain. Consciousness is complex concept and has two main components: awareness of environment and of self (content of consciousness) and wakefulness (the level of consciousness). The figure 1 shows the simplified illustration of components of consciouseness and them relations to anaesthesia, coma and vegetative state (Laureys 2005).

6

Figure 1. Simplified illustration of the two major components of consciousness: the level of consciousness (i.e. wakefulness or arousal) and the content of consciousness (i.e. awareness or experience). Reprinted with permission from Laureys S. Trends Cogn Sci 2005;9:556-559. Copyright by Elsevier.

Anaesthetics also act in the different parts of brain structure producing amnesia, unconsciousness, and immobilisation. Anaesthetics produce unconsciousness by blocking the activity of the thalamocortical system (John and Prichep 2005). The amnestic effect of anaesthesia probably comes from anaesthesia’s blocking of the dorsalolateral prefrontal cortex (John and Prichep 2005). The anaesthetic dose required for patients’ loss of consciousness (LOC) and amnesia is not enough to achieve patients’ immobility during surgical stimulation (Chortkoff et al. 1995, Campagna et al. 2003). In everyday clinical practice, anaesthesiologists see patients in a deep level of hypnosis who nevertheless move their voluntary muscles in reaction to pain stimulation. The thalamus and cerebral cortex are more sensitive to anaesthetic agents than is the spinal cord, in which motor response is blocked (Campagna et al. 2003). As a result, a higher concentration of anaesthetic is required for immobilisation. The volatile anaesthetics also depress the spinal cord and assure better immobilisation than do intravenous anaesthetics (Campagna et al. 2003). In a high blood concentration, anaesthetics do not act specifically and instead depress a wide variety of neuronal receptors. However, at clinical concentrations, they are more specific to one receptor side. On the molecular level, the presence of postsynaptic inhibitory ligand-gated ion channels is a potential side effect of anaesthetic drug action. The inhibitory synaptic receptors, i.e. gamma-aminobutric acid subtype A (GABAA), are the site of action for propofol and volatile anaesthetics (John and Prichep 2005). Volatile anaesthetics also suppress the serotonin, acetylcholine, and glutamate receptors (Campagna et al. 2005). Nitrous oxide and ketamine act upon the CNS differently than do other anaesthetics and induce different changes in the EEG (Yli-Hankala 1990a, Yli-Hankala et al. 1993). Table 2 shows the CNS receptors involved during general anaesthesia.

7

Table 2. The CNS neurotransmitters and receptors affected during general anaesthesia (John

and Prichep 2005).

� gamma-aminobutric acid subtype A (GABAA)

� alfa-amino-3-hydroxy-5-methyl-4-isoxa-zole propionis acid

(AMPA) receptors

� glycene receptors

� serotonin type 2 (5-HT2a) receptors

� N-methyl-D-asparate (NMDA) receptors

� sodium channels

� potasium channels

� alfa2 adrenoreceptors

� acetylcholine receptors

� glutamate receptors

Unconsciousness and amnesia during general anaesthesia are brain cortical structure phenomena. Antinociception, patients’ immobility, and patients’ autonomic stability are subcortical components (Jameson and Sloan 2006), but emergency of consciousness is also associated with activation of subcortical structure of brain (Långsjö et al. 2012). In the unique study (Velly et al. 2007) of propofol and sevoflurane dynamic action on the brain structure the cortical and subcortical (deep brain electrodes), EEG was registered during general anaesthesia with propofol and sevoflurane. Cortical EEG was depressed dramatically during LOC, but not the subcortical EEG. Therefore, the subcortical EEG values were able to predict the patient’s movement in response to surgical stimulation.

2.2 LEVELS OF ANAESTHESIA

EEG measures the neuronal electrical activity of brain cortex cells from the scalp of skull. The EEG electrodes are used according the international 10-20 system. The EEG tracks typical frequency waves with amplitudes and latencies divided to five frequency bands (Gugino and Yli-Hankala 2004):

� Delta (δ) frequency range 1.5-3.5 Hz

� Theta (θ) 3.6-7.5 Hz

� Alpha (α) 7.6-12.5 Hz

� Beta (β) 12.6-25 Hz

� Gamma (γ) 25.1-50 Hz

The entry of sedative drugs and anaesthetics into the patient’s bloodstream changes the

8

EEG patterns that are dependent on drug concentration levels in the plasma.

The duration of anaesthesia can be divided into the three periods: induction, maintenance, and emergency (Brown et al. 2010). In each period, we can notice different clinical signs as well as EEG patterns. When a patient wakes with closed eyes before induction, the active EEG with alpha activity (10 Hz) dominates. After administration of a small amount of anaesthetic, the patient is sedated yet easily aroused. When we increase the dose of the drugs, the paradoxical excitation state occurs, which is characterised by euphoria, defensive movements, and an increase in the beta activity (12.5-25 Hz) of the EEG (Rampil 1998a, Brown et al. 2010). After more hypnotics are given, irregular respiratory patterns at the point of apnoea develop. Meanwhile, the patient no longer responds to oral commands, eye tracking stops, and the oculocephalic, eyelash, and corneal reflexes are lost. The blood pressure may have increased or decreased, and the heart rate has usually decreased. After the administration of muscle relaxants, tracheal intubation is performed.

The anaesthesia maintenance period is subdivided into three phases according to the changes in the EEG pattern. During the light state of anaesthesia, phase I, EEG beta activity (12.5-25 Hz) decreases and alpha (7.6-12.5 Hz) and delta (1.5-3.5 Hz) activity increases. In phase II, the intermediate state, beta activity decreases and alpha and delta activity increases especially in the anterior EEG leads. In phase III, the burst suppression pattern of EEG with intermittent alpha and beta activity is noticeable. In the deep state of anaesthesia, phase IV, the EEG is isoelectric and likewise in the coma state (Young 2000, Brown et al. 2010). A burst suppression pattern or isoelectric EEG may be purposely induced by anaesthetic drugs to stop seizures during RSE treatment. Mechanically controlled ventilation, opioids for analgesia, thermoregulatory support, and cardiovascular drugs are needed in those phases.

Emergency from anaesthesia is divided into three phases and is a passive process. Its sequence depends on the anaesthetics administered, the patients’ age, sex, comorbidities, and the duration of surgery. After cessation of anaesthetic drugs and reversal of the neuromuscular blockade, the first clinical sign of emergency phase I, is the return of irregular breathing. In the EEG, an increase in alpha and beta activity is typical for this period. In phase II, blood pressure and heart rate increases, and autonomic responsiveness returns, such as salivation (7th and 9th cranial nerve nuclei), tearing (7th cranial nerve nuclei), grimacing (5th and 7th cranial nerve nuclei), swallowing, gagging, and coughing (9th and 10th cranial nerve nuclei). In the EEG pattern, alpha and beta activity also increased in this phase. Return of muscle tone (spinal cord) allows possible extubation. In phase III, the patient opens his eyes and can respond to commands. Awake patterns now appear in the patient’s EEG (Brown et al. 2010).

9

2.3 AWARENESS DURING GENERAL ANAESTHESIA

2.3.1 Definition of awareness

When William Morton gave his first anaesthesia demonstration with ether in Boston 1846, his patient was aware during the procedure. The addition of neuromuscular blocking agents (NMBA) drugs to general anaesthesia had eliminated patient movement, as a sign of awareness. Awareness during surgery under general anaesthesia with total muscle relaxation is the most terrifying scenario for the patient as well as for the anaesthesiologist. During the American Society of Anaesthesiologists’ (ASA) task force on intraoperative awareness, the task force defined awareness as a “a patient becomes conscious during a procedure performed under general anaesthesia and subsequently has recall of these events” (Apfelbaum et al. 2006). This definition is limited only to explicit memory, which means that the patient remembers this event. It does not cover any implicit memories. The novel Michigan Awareness Classification Instrument (Mashour et al. 2010) recognised the five class of awareness during general anaesthesia: class 1, indicates isolated auditory perceptions; class 2, tactile perception (e.g., perception of surgical manipulation or endotracheal intubation); class 3, pain; class 4, paralysis (e.g., a feeling that one cannot move, speak, or breathe); class 5, paralysis and pain. Letter D with class number indicates that there is associated distress e.g., reports of fear, anxiety, suffocation, sense of doom, sense of impending death.

2.3.2 Incidence of awareness during general anaesthesia

Awareness under general anaesthesia has been under active research in the past decade. The incidence of awareness is estimated to be approximately 0.1-0.4% (Ranta et al. 1998, Sandin et al. 2000, Sebel 2004a). One study interviewed 11785 anaesthetised patients 1-3 days and 7-14 days after operation, but before leaving the post-anaesthesia care unit. The incidence of awareness was 0.18 % when NMBA drugs were used and 0.1% when they were not used (Sandin et al. 2000). These incidence figures of awareness are relatively low, but if one takes into consideration the fact that 2 million general anaesthesias are administrated per year in the USA, then one will discover that 26000 patients have recall of events occurring during surgery. In a prospective awareness study conducted on 11101 Chinese patients, the incidence of definitive recall of events occurring during general anaesthesia was much higher, i.e. 0.41%, than in western countries (Xu et al. 2009). The risk factors of awareness that Xu et al. delineated were high ASA status, patients who had previously received general anaesthesia, and total intravenous anaesthesia (TIVA) with propofol. On the other hand, a study measured the incidence of awareness in 1000 non-cardiac operations with TIVA and neuromuscular blocking agents, to examine the influence of propofol-based TIVA on incidence rates (Nordström et al. 1997). This study did not reveal any increased risk for awareness with TIVA in this group of patients.

10

2.3.3 Awareness during cardiac surgery

The incidence of awareness in cardiac procedures has been reported greater than that in general surgery. The comorbidity of cardiac patients and haemodynamic instability during the surgery and CPB bring about the tendency to use light anaesthesia (Serfontein 2010, Kertai et al. 2012). In numerous previous studies, the incidence of awareness during cardiac surgery has been reported to be as high as 1.5-4% (Ranta et al. 1996) or even rates up 24% have been reported (Serfontein 2010). Patients undergoing cardiac surgery are at risk of awareness (Phillips et al. 1993) with an incidence of 1.14% (700 patients), but in more recent studies, this finding was not repeated. In the study of Ranta et al. (2002) the incidence of awareness in 929 cardiac patients was similar to the incidence reported during general surgery (0.5%). After the introduction of fast track protocols for cardiac surgery, a low figure of 0.3% on the incidence of awareness was reported (Dowd 1998). An academic dissertation concerning awareness during general anaesthesia (Ranta 2002) reported that the incidence of undisputed recall during cardiac surgery was 0.3 %, and the incidence of all possible cases of awareness was 2.4%. Ranta concluded that the much higher risk of awareness during cardiac surgery reported previously has declined so that it resembles the level observed during general surgery. In addition, a recent study on the new fast track anaesthesia techniques for cardiac surgery noted that the incidence of perioperative recall in 417 patients was 0% (Groesdonk et al. 2010). In the BAG RECALL trial (Avidan et al. 2011) comparing BIS protocol and end–tidal anaesthetic concentration (ETAC) management for the prevention of awareness, 6041 patients with high risk of awareness were included. A total of 19 cases of definite or possible intraoperative awareness (0.66%) occurred in the BIS group, as compared with 8 (0.28%) in the ETAC group, but at least 3 of the patients who experienced awareness were undergoing CPB for cardiac surgery.

Also in the review of 271 cases of awareness during general anaesthesia, the cardiac surgery was a risk of awareness (Ghoneim et al. 2009). In the review article about awareness in cardiac anaesthesia (Serfontein 2010) author concluded that cardiac surgical patients are a high-risk group for experiencing intraoperative awareness. The periods of anaesthesia and surgery when the patient is at risk of awareness are the intubation, intense surgical stimulation (sternotomy), haemodynamic instability during operation, and re-warming phase of CPB.

2.3.4. Risks factors of awareness

The American Society of Anaesthesiologists (ASA) has published descriptions of the conditions that increase the risk of awareness during general anaesthesia and guidelines to prevent intraoperative recall (Apfelbaum et al. 2006,). They are presented in Tables 3 and 4.

11

Table 3. The conditions with risk of awareness during general anaesthesia (Apfelbaum et al. 2006).

Major risks

� Caesarean section under general anaesthesia � Long term use of benzodiazepines, opioids and anticonvulsant agents � History of awareness � Difficult intubation � Heavy alcohol intake � Acute trauma � High risk cardiac surgery (EF <40%) � Off-pump cardiac surgery � Trauma with hypovolemia � ASA physical status IV or V � Severe aortic stenosis � Pulmonary hypertension Minor risk

� COPD � Preoperative use of β-blockers � TIVA � Low exercise tolerance (no musculoskeletal dysfunction) � Smoking of two or more packs of cigarettes per day � Obesity (BMI > 30)

Table 4. Schema to reduce intraoperative awareness in high-risk patients (Apfelbaum et al. 2006).

� Preoperative evaluation and identification of patients at risk

� Inform patient about risk of awareness

� Preoperative checklist of anaesthetic equipment

� The use of benzodiazepine or scopolamine premedication

� The use of EEG-based monitors of hypnosis (BIS, entropy)

� Consider the use of a large dose of anaesthetics

� The use of balanced anaesthetic techniques

� The use of combined anaesthetic techniques (TIVA and volatile anaesthetics)

� Consider avoiding total muscle paralysis

In the review of 271 cases of awareness during general anaesthesia two main risk factors of awareness were the absence of a volatile agent or propofol during maintenance of anaesthesia (high-dose opioid anaesthesia) and history of previous awareness (Ghoneim et al. 2009). In this study the investigators found no differences in the frequency of use of NMBA drugs in the awareness sample compared with 19504 patients who did not have awareness. The malfunction of the anaesthesia drug delivery systems, such as an infusion pumps and vaporiser, were also risk factors of awareness in this study. If the anaesthesiologist suspects intraoperative consciousness or patient responsiveness, he/she should administer immediately a bolus of propofol (e.g., 0.5mg/kg) or midazolam (0.06

12

mg/kg) and increase the dose of the main anaesthetic. The anaesthesia delivery system (infusion pumps, vaporiser) should also be checked, because any malfunction in this system can lead to a disturbance in anaesthesia maintenance. In the postoperative phase, an interview should be conducted using the Brice questionnaire (Brice at al. 1970) after the operation, the next morning after the operation, and on the third postoperative day.

Table 5. Brice questionnaire.

Patients should be asked the following five questions:

� What was the last thing that you remembered happening before you went to sleep?

� What is the first thing you remember happening on waking?

� Did you dream or have any other experiences whilst you were asleep?

� What was the worst thing about your operation?

� What was the next worst?

The development of posttraumatic stress disorder (PTSD) as a psychological consequence of awareness should be remembered, and patient support and psychiatric consultation should be offered, if necessary (Ranta et al. 1998, Apfelbaum et al. 2006).

2.4 ASSESSMENT OF ANAESTHESIA DEPTH

The monitoring of anaesthesia depth is complex due to the different aspects of anaesthesia: unconsciousness, amnesia, antinociception, and immobility with muscle paralysis. The first three components of anaesthesia are related to the anaesthetics and sedation agent’s action on the CNS, whereas the neuromuscular blockade is a peripheral phenomenon. The depth of general anaesthesia was traditionally assessed by haemodynamic signs such as tachycardia, hypertension, and autonomic signs, such as motor response, lacrimation, and changes in respiratory frequency. Clinicians monitor neuromuscular blockade with a neuromuscular transmission (NMT) monitor. The monitoring of analgesia is complex and not strait full, the changes in the blood pressure and heart rate, Surgical Stress Index (SSI), Response entropy and skin conductance were used. In the next chapters the monitoring of the hypnotic component of anaesthesia is discussed.

2.4.1 Clinical variables After introduction of volatile anaesthetics, the depth of anaesthesia was assessed according to the responses of the patient’s cardiovascular system and to the patient’s autonomic signs, mainly the changes in respiratory frequency. The increase in perspiration, movement, tearing, sweating, and changes in pupil size can be the somatic signs of patient arousal. During the use of volatile diethyl ether, Guedel’s four-stage anaesthesia classification (which is based on vital signs) was used (Guedel 1937). Stage I lasts from the beginning of induction of general anaesthesia to loss of consciousness (stage of sedation or disorientation). Stage II begins from loss of consciousness and extends to the onset of

13

automatic breathing. Eyelash reflex disappears at stage II, but other reflexes remain intact. As a result, coughing, vomiting, and struggling may occur. Stage III of surgical anaesthesia is divided into three planes: eyelid reflex lost, paralysis of the intercostal muscle, and paralysis of the diaphragm. Stage IV occurs with the overdose of anaesthetics, and presents with respiratory arrest and vasomotor collapse.

During inadequate levels of anaesthesia and ongoing surgical stimulation, blood pressure and heart rate can raise significantly. However, the use of muscle relaxants mean that vital signs such as patient`s movement or respiratory frequency are unreliable for monitoring the level of anaesthesia. Moreover, the cardiovascular signs like blood pressure and heart rate can be influenced by antihypertensive drugs and β-blockers, so cardiovascular signs do not correlate well with the level of hypnosis during general anaesthesia. In the Moerman et al. (1993) study, twelve definitive cases of awareness were found, and the anaesthesia records were checked for changes in blood pressure and tachycardia. High blood pressure was found in five cases, but in three cases, there was no rise in blood pressure or in heart rate. The ASA Closed Claims project (Domino et al. 1999) demonstrated that blood pressure increased only in 15% in all cases of awareness (n=79); tachycardia appeared in 7% of the cases and patient movement occurred in one case (most patients received muscle relaxants).

Heart rate variability (HRV) provides a non-invasive marker of the autonomous nervous system’s function. It is well known that the status of the autonomous nervous system is modified by anaesthetic or sedative drugs, and that noxious stimuli cause sympathetic activation. General anaesthesia markedly depresses HRV, so a couple studies have suggested that HRV reflects the brainstem’s response to anaesthesia (Pomfrett 1999). Hence, HRV was also studied for monitoring the depth of anaesthesia, and a correlation between EEG and HRV during anaesthesia was reported (Yli-Hankala et al. 1990b, Jäntti and Yli-Hankala 1990, Pomfrett et al. 1994).

2.4.2 EEG-based monitoring

EEG measures the neuronal electrical activity from the cortex of brain. Gibbs et al. (1937) first noticed the EEG changes caused by anaesthetic drugs. Sedative and anaesthetic drugs cause changes in the EEG patterns that are dependent on the anaesthetic drug’s concentration in the plasma. During relaxation with the eyes closed, alpha activity is dominant in the EEG recording. In contrast, light sedation decreases the alpha power in the EEG and increases the beta power. As the anaesthesia deepens, the lowest frequency brain activity (delta and theta) increases in quantity while the alpha and beta activities drop in quantity. In deep anaesthesia cortical activity is altogether silent, and an isoelectric EEG with a burst suppression pattern dominates. The use of raw EEG to monitor depth of anaesthesia requires visual evaluation and knowledge of neurophysiology. Thus, clinical use of raw EEG is restricted to the monitoring of sleep, epileptic activity, and brain abnormalities like hypoxia or ischemia. Clinicians incorporate the computerisation of the EEG signal using two techniques of EEG analysis: time-domain and frequency-domain

14

(Levy et al. 1980, Freye and Levy 2005). The time- domain is used only in clinical practice to calculate the burst suppression ratio during the EEG-based deep anaesthesia monitoring. The frequency domain can be characterised either by voltage or by power versus frequency. The transformation from time-domain to frequency-domain involves digitalisation and Fast Fourier transformation. The power spectrum shows EEG data in two dimensions: amplitude versus frequency, but the spectrum is still complex. Spectral edge frequency, the median frequency, and the peak power frequency are used to describe EEG signals (Jameson and Sloan 2006).

The monitoring of the depth of anaesthesia should fulfil certain aspects: it should show changes in anaesthesia level cause by changes in the anaesthetic blood concentrations, be similar with the use of different agents, show changes with surgical stimulation and indicate awareness with a short delay (Thornton 1991). Since the use of anaesthetics and sedative drugs cause changes in the EEG pattern, processed EEG-based anaesthesia depth monitoring with automatic algorithm EEG analysis has been developed. Bispectral EEG analysis incorporated a more complex algorithm and included phase coupling among frequency bands. The BIS monitor was the first commercially available monitor of the hypnotic component of anaesthesia (Rampil 1998a, Johansen 2006). In Table 6, EEG-based monitors of hypnosis during anaesthesia are presented.

Table 6. EEG-based monitors of hypnotic component of anaesthesia.

� AEP Monitor™ (Danmeter) � BIS™ Monitor (Covidien) � Cerebral State Monitor™ (Danmeter) � Narcotrend™ Monitor (Schiller) � Patient State Monitor (Hospira) � SNAP™ Monitor (Stryker) � Entropy Module (GE Healthcare)

2.4.3 Auditory evoked potentials

Evoked potentials reflect the electrical activity in the EEG produced by sensory stimulus. Input stimuli are categorised according to their sources: visual evoked potentials, somatosensory evoked potentials, and auditory evoked potentials. The AEP in the EEG reflect responses to auditory stimulation. The passage of the auditory stimulus from the cochlea to the cortex produces waveforms consisting of 11 waves. The AEP have been under active research in the assessment of sedation and anaesthesia (Thornton et al. 1983, 1988, 1989a, De Beer et al. 1996a, 1996b). They can be divided in three different groups based on the delay following the auditory event: brainstem auditory evoked potentials (BAEPs), which are recordable at 1-10 ms after the auditory stimulus; midlatency auditory evoked potentials (MLAEPs), recordable 10-50 ms after stimulus, and late latency auditory evoked potentials (LLAEPs), recordable after 50 ms after stimulus (Thornton 1991 and

15

1993). To create a distinguishable response, a computer can extract the AEP from background EEG by averaging a large amount of responses, and it displays the results as waveforms. The MLAEP waveforms consist of a series of positive and negative waves (No, Po, Na, Pa, Nb) that are characterised by their latency (the time from stimulus to the occurrence of the wave) and their amplitudes, which both represent the passage of electrical activity along the auditory pathway from cochlea to cortex. They originate from the medial geniculate and primary auditory cortex. The typical peaks of AEPs are shown in Figure 2.

Figure 2. The typical peak responses in the EEG after auditory stimulation: brainstem auditory

evoked potentials (I, II, III, IV, V, VI ), midlatency auditory evoked potentials (No, Po, Na, Pa,

Nb), and late latency auditory evoked potentials (P1, N1, P2, N2). Reprinted with permission of

Danmeter, Denmark.

The BAEPs do not change with intravenous anaesthetic agents like propofol or thiopental, and they are not used to measure the depth of anaesthesia (Thornton 1991). The MLAEPs show the dose dependent changes within general anaesthetics similar with different drugs used. Therefore, they are suitable for measuring the pharmacodynamic effect of anaesthetics on the EEG. Decreases in amplitudes and increases in latencies of MLAEPs are detected during progressive stages of sedation and anaesthesia (Thornton 1991), and they have been evaluated with halothane and enflurane (Thornton et al. 1984), etomidate (Thornton et al. 1985), isoflurane (Heneghan et al. 1987), and propofol (Thornton et al. 1989b). Surgical stimulation increases the amplitudes of both Pa and Nb parameters, both of which indicate arousal (De Beer et al. 1996b), which can be blocked by a bolus dose of opioids (Shinner et al. 1999). In one study of MLAEPs as an indicator of awareness during anaesthesia, the Nb latency shorter than 44.5 ms was associated with risk of awareness (Thornton et al. 1989a). AEP responses, like other form of EEG-based monitoring, are affected by technical and subject–related factors. Technical factors include the placement of EEG electrodes, electrical noise in the operating room (amplifiers are used to eliminate electrical artefacts), type of auditory stimulus (usually click and binaural stimulation),

16

stimulus intensity, and presentation rate (threshold of 70 dB and 10 per second).

The patient-related factors affecting the AEP responses are:

� significant hearing deficit � neurological diseases that impair the auditory nerve tract, such as multiple � sclerosis, tumours of the CNS, and schizophrenia. � age and sex � hypothermia

During deep anaesthesia, characterised by burst suppression patterns in the EEG, the MLAEP measurements lose their accuracy because of the flat EEG and the MLAEP waveforms are totally suppressed (Thornton 1991).

2.4.3.1 AEP anaesthesia monitor

The first commercially available AEP monitor, called A-line™ (Danmeter, Odense, Denmark), was not able to recognise the burst suppression pattern during deep anaesthesia. The second commercial anaesthesia depth monitor, which uses MLAEP, is called the AEP Monitor/2 (Danmeter, Odense, Denmark), and this monitor uses a composite AEP/EEG A-line autoregressive index (AAI). The AAI index measures the depth of anaesthesia by using two different sources. First, the index actively measures the patient’s AEP responses (latencies and amplitudes) to an auditory click stimulus delivered by headphones. Then, it passively measures the EEG β-ratio. During deep anaesthesia, the monitor collects information from the burst suppression algorithm and BS is calculated as a percentage of a 30-second window (Nishiyama and Hanaoka 2004, Bonhomme et al. 2006a). When patients are awake, the AAI values are 50-99, LOC is 40, and surgical anaesthesia 15-30. The AAI index was validated with the Observer’s Assessment of Alertness and Sedation Score (OAAS) during the propofol and midazolam anaesthesia (Ge et al. 2002). In the study of Struys et al. (2002) the accuracy of the AAI and BIS to predict effect-site concentration of propofol and sedation level was compared. Both AAI and BIS predicted propofol effect-site concentration, revealed information on the level of sedation and loss of consciousness but did not predict response to noxious stimulus. The next study (Struys et al. 2003) was conducted to compare the performance and accuracy of BIS and AAI to predict the propofol effect-site concentration, to measure the loss of responses to different stimulation defined as loss of response to verbal command, eyelash reflex, and noxious stimulus during stepwise increased levels of propofol infusion with and without remifentanil. The overall prediction probability to measure the hypnotic component of anaesthesia remained accurate for BIS and AAI, and their ability to detect loss of consciousness defined by OAAS and loss of response to eyelash reflex remained also accurate.

In the AEP monitor clinical utility studies by Recart et al. (2003a and 2003b) the use of an AEP monitor for titrating the volatile anaesthetic led to a significant reduction in the anaesthetic requirement compared to standard clinical practice. In Recart’s study, the use of the AEP monitor resulted in a shorter post-anaesthesia care unit stay and improved the

17

quality of recovery from the patient's perspective. In a study that compared BIS and AAI (Nishiyama et al. 2004), the AAI index increased significantly in response to nociceptive stimulation (intubation, skin incision), whereas the BIS index remained constant. One explanation is that the BIS and AEP measure different aspects of brain activity. AEP is thought to reflect the transfer of auditory stimulation to the auditory cortex and to measure anaesthesia depth, not simply the hypnotic component; in contrast, BIS reflects brain cortical activity and seems to be better at tracking the entire gradual progression of the sedative and anaesthetic state. In a study comparing the performance of BIS and AAI index in predicting consciousness, explicit and implicit memory during moderate sedation with propofol, both correlated well with loss of consciousness defined by OAAS (identical P(k) of 0.87), but did not allow prediction of postoperative explicit or implicit recall (Hadzidiakos et al. 2006). Disadvantages of the AEP anaesthesia depth monitoring include problematic monitoring of patients with hearing impairments, which can influence the AEP. The inter-personal variability of the hearing threshold means that a constant click stimulus produces different responses in different individuals. To my knowledge, the commercial AEP monitor by Danmeter is no longer available.

2.4.4 Bispectral index

The bispectral index (Covidien, MA, USA) was the first commercially available monitor of the hypnotic effect of anaesthesia. It monitors the pharmacodynamic effect of anaesthetic agents on the brain (Rampil 1998a, Johansen 2006, Punjasawadwong et al. 2010). By December 2012, the BIS bibliography included a huge amount of scientific articles, numbering over 1800 (BISeducation.com). The BIS technology was developed over a course of 10 years in the USA with the contribution 30 institutions. The United States Food and Drug Administration accepted the BIS monitor as a measure of hypnotic level of anaesthesia in 1996. The BIS is an EEG-based monitor that continuously monitors a patient’s level of hypnosis during general anaesthesia or sedation. The BIS technology combines multiple EEG signal analyses: bispectral, power spectral, and time domain analysis (Sigl and Chamoun 1994, Rampil 1998a, Bruhn et al. 2006). The BIS algorithm analyses the EEG and compares this analysis with a previous database of 5000 patients who received the same anaesthetic agents as those used during the analysis (Rampil 1998a). The BIS monitor uses a BIS index between zero and 100. Surgical anaesthesia is between 40 and 60 (Rampil 1998a).

2.4.4.1 BIS evaluation studies

Most of the intravenous anaesthetics (propofol, tiopenthal) and volatile anaesthetics (isoflurane, sevoflurane, and desflurane) cause gradual changes in the BIS index with increased drug concentration. During a multicentre BIS evaluation study (Glass et al. 1997) with propofol, isoflurane, and midazolam, the researchers discovered a good correlation between BIS and OAAS score with a very high predictive performance for correctly indicating probability of LOC. Another multicentre study, which monitored the anaesthetic effects on BIS, demonstrates that dosing anaesthetic drugs to lower BIS values achieves a

18

lower probability of movement in response to surgical stimulation (Sebel et al. 1997). In the BIS utility study (Gan et al. 1997) 302 patients were randomised to a standard practice group and a BIS group. BIS monitoring allowed a decrease in the propofol use, faster emergency from anaesthesia and reduced recovery time. BIS monitoring improved recovery from anaesthesia and reduced the amount of propofol (29%) and isoflurane (40%) used (Yli-Hankala et al. 1999) as well as the amount of sevoflurane (Aime et al. 2006). BIS monitoring reduced the incidence of the postoperative nausea and vomiting (Nelskylä et al. 2001). In a meta-analysis of BIS trials concerning ambulatory anaesthesia in 1390 patients across 11 randomised trials, the BIS-guided anaesthesia reduced anaesthetics consumption by 19%, reduced occurrence of nausea/vomiting from 38% of all patients to 32%, and reduced recovery time; however, the average recovery time was only 4 minutes shorter in the BIS group. There was no time difference between groups regarding discharge from the ambulatory surgery unit, and the total cost of patient stay was higher in the BIS group (Liu 2004). On the other hand, in the recent Cochrane review, including 31 randomised studies, there was no overall reduction in dose of propofol and volatile anaesthetics when using BIS monitoring (Punjasawadwong et al. 2010). In the large group of 1100 patients participating in the B-Unaware trial, the relationships during maintenance of anaesthesia between end tidal anaesthetic concentration (ETAC), BIS index and patients characteristic were examined (Whitlock et al. 2011). BIS frequently correlated poorly with ETAC and was often insensitive to changes in inhalation anaesthetic concentration. The use of nitrous oxide had no impact on the BIS values (Rampil et al. 1998b). Also, ketamine did not change BIS index or cause a paradoxical rise in BIS values (Sakai et al. 1999).

The EEG signal can be contaminated by artefacts from the environment of the operating room, where there are other electrical devices. During BIS monitoring for assessment of hypnosis, various conditions that could interfere with the BIS index and indicate an incorrect hypnotic state (Dahaba 2005) are presented in Table 7.

19

Table 7. Conditions causing disturbance in the BIS index values.

Effect BIS change

Paradoxical BIS changes with anaesthetics N2 O termination BIS ↓ Ketamine BIS ↑ Isoflurane BIS ↑ Halotane BIS ↑ Electric devices Pacemaker BIS ↑ Diathermy Disturbed Effect of different clinical conditions Hypoglycemia BIS ↓ Cardiac arrest BIS ↓ Hypovolemia BIS ↓ Cerebral ischemia BIS ↓ Hypothermia BIS ↓ Abnormal EEG patterns Alzheimer dementia BIS ↓ Brain injury BIS ↓ Brain death BIS 0

2.4.4.2 BIS and awareness during general anaesthesia

The anaesthesiologist has probably a new method to decrease the risk of awareness and intraoperative recall, which relies on processed EEG monitoring. When keeping the BIS under 60 during general anaesthesia, the possibility of patient’s recall of intraoperative events is unlikely (Punjasawadwong et al. 2010). The first clinical trial demonstrating that BIS monitoring reduces intraoperative awareness was the Ekman study (Ekman et al. 2004). They compared 4945 patients under general anaesthesia with intubation, muscle relaxation, and BIS monitoring with 7826 control patients from a study carried out 4 years previously (the latter were under the same conditions except that BIS was not used). A 77% reduction in the incidence of awareness in the BIS-monitored group was reported. The incidence of awareness in the BIS group was 0.04% compared with 0.18% in the previous study. In this study, higher BIS values (> 60) were found in the patients with awareness during general anaesthesia. It is possible that there was a change in anaesthetic practice in the institutions between the previous study and current investigation with BIS monitor. The study was also underpowered. If incidence of awareness is 0.01% and BIS produces 50 % reduction in incidence, almost 41 000 patients should be included in a prospective randomised trial (Sebel 2004b).

In the B-Aware trial (Myles et al. 2004), 2463 patients who were at high risk of awareness during anaesthesia were randomly allocated either to the BIS group or to the routine

20

practice group. BIS-guided anaesthesia reduced the risk of awareness by 82%. With a cost of routine BIS monitoring at US 16 dollars per use with a number needed to treat of 138, the cost of preventing one case of awareness in high-risk patients was about 2200 dollars in this study. If the patients in the Myles study were patients with a low risk of awareness during general anaesthesia, again almost 41 000 patients should have been included in the study. On the other hand, Avidan and colleagues (2008) did not confirm BIS monitoring’s ability to prevent awareness during general anaesthesia in patients with higher risk of awareness. Two thousand patients were assigned to BIS-guided anaesthesia (BIS 40-60) or to ETAC, the latter of which was administered within a range of minimum alveolar concentration (MAC) between 0.7-1.3. Two cases of definite awareness occurred in both groups (absolute difference 0%), and a reduction of volatile anaesthetics was not detected in the BIS group. The investigators concluded that these results do not support routine BIS monitoring as a part of standard anaesthesia practice. In this study patients with low risk of awareness (patients undergoing open-heart surgery) were included which will lead to a reduction in trial events. Also the study protocol with the BIS ranges was not well achieved (Orser 2008). In a large, multicentre, randomised study of 6014 patients at high risk of awareness who were randomly assigned to BIS-guided anaesthesia (BIS 40-60) or to ETAC-guided anaesthesia (MAC 0.7-1.3) (Avidan et al. 2011). Seven patients in the BIS group had definite intraoperative awareness whereas 2 patients had in the ETAC group. Possible awareness in the BIS group was ascribed to 19 patients versus 8 in the ETAC group. Total definite and possible awareness was 0.47% in all patients. In conclusion, the BIS-guided anaesthesia was not superior to the ETAC–guided anaesthesia in terms of awareness.

The results and conclusions of Avidan’s study are controversial. First of all, in this study, 41 % of awareness occurred when ETAC and BIS were in the study protocol’s target ranges; therefore, intraoperative awareness was not preventable with either monitoring method. ETAC is a measure of volatile drug concentration, and does not measure the response of the brain to the anaesthetic dose. Volatile anaesthetics were used in the control group, but quite often, such as during cardiac surgery, anaesthesia had been maintained by TIVA. In such a case, one cannot use ETAC for anaesthesia depth control. Pure volatile anaesthesia can produce hypotension, especially in the sickest patients, and cause unnecessary cardiovascular and cerebral risk. Monitors are meant to supplement, not supplant, clinical decision making, and depth-of-anaesthesia monitors that reduce complex neurobiology to simple numbers are no exception. There is no study in the literature focusing in the use of BIS to reduce the awareness only in the cardiac surgery patients (Kertai et al. 2012), but three clinical trials (Myles et al. 2004, Avidan et al. 2008 and 2011) of the impact of BIS monitoring on awareness included a large number of cardiac surgery patients (27%, 49%, 36%). The post hoc analysis of incidence of awareness in this group of patients did not reveal a reduction of awareness with BIS monitoring (Kertai et al. 2012).

21

2.4.4.3 BIS and postoperative outcomes

In a couple of studies, the association between too deep of an anaesthesia state and postoperative, long-term outcomes was reported. In the study of Monk and colleagues (2005), association between deep cumulative anaesthesia and one-year mortality for non-cardiac surgery suggests that intraoperative anaesthetic management may affect outcomes over longer time periods than previously appreciated. In that study, one-year mortality was 5.5% in all patients (n = 1064) and 10.3% in patients 65 years or older (n = 243). Significant, independent predictors of mortality were patient comorbidity, intraoperative systolic hypotension, and cumulative deep hypnotic time (BIS < 45). The association between duration of low BIS values (BIS < 45) and intermediate-term mortality was also found in 460 cardiac surgery patients (Kertai et al. 2010). The association was not dependent on the duration of anaesthesia or the concentration of volatile anaesthetics used. In 4086 patients with malignant disease who were monitored by BIS during surgery, 2-year mortality was associated with low BIS values (BIS < 45) and was also a significant predictor of the 1 and 2-year mortality in this patient group (Lindholm et al. 2009). In this study when the initial multivariate regression was repeated using pre-existing malignancy status among the co-varieties in the model, the previously significant relation between 1, and 2-yr mortality and (BIS <45) did not reach statistical significance. The effect of BIS monitoring on long-term survival in the 2463 patients from the B-Aware trial was published by Leslie et al. (2010). The median follow up time was 4.1 years, and the primary end point of the study was survival. The second objectives were incidences of myocardial infarction and stroke. There was no difference between the BIS group and standard anaesthesia care in the risk of death. However, the hazard ratio for death was higher in the patients with deep anaesthesia (BIS < 40 for over 5 minutes) compared with other BIS-monitored patients. In addition, the odds ratios for myocardial infarction and stroke were higher in this patient group. Those studies (Monk et al. 2005, Kertai et al. 2010, Lindholm et al. 2009) suggested the potential impact of anaesthesia management on morbidity and mortality, but they are observational studies and relationship between the level of anaesthesia and patients outcomes is still an open question.

For the patient as well as the anaesthesiologist, the primary concern related to anaesthesia is lack of awareness during surgery. For this reason, a deep level of anaesthesia is commonly used. High doses of anaesthetic agents are often used to lower blood pressure during surgery. In light of those studies, anaesthetic management should find a more sophisticated approach with a specified BIS goal.

In conclusion, the results of various studies that measured BIS index to guide the level of anaesthesia and to prevent awareness are still contradictory (Punjasawadwong et al. 2010). The BIS index appears to be useful for patients who have a high risk of awareness (Myles et al. 2004), but a study which used ETAC guided anaesthesia with volatile anaesthetic did not demonstrate the index’s utility (Avidan et al. 2008 and 2011).

22

2.4.4.4. Automatic anaesthesia delivery system

In the future, general anaesthesia may be maintained by fully automatic anaesthesia delivery systems. Infusion pumps will recognise the patient’s own pharmacodynamic profile to achieve a constant anaesthetic concentration in the blood and brain (Agarwal et al. 2009). The anaesthesiologist will set the target concentration of anaesthetics in the beginning of the procedure and target the anaesthesia depth. The computer control closed-loop anaesthesia delivery system (CLADS) will change the drug concentration according the feedback from anaesthesia information system and processed EEG monitoring (be it BIS or entropy). The BIS is used in the automated anaesthesia systems with target control infusion of anaesthetics and closed-loop feedback with target BIS values. Struys’ study (2001) compared closed-loop anaesthesia delivery systems with BIS target of 50 with a standard practice group. Better control of BIS target and systolic blood pressure and faster recovery were found in the closed-loop anaesthesia than in the control group. In a randomised, multicentre study (Liu et al. 2006), 160 patients undergoing minor or major surgery were allocated either to the closed-loop or to manual target control propofol infusion. The BIS target during anaesthesia maintenance was 40-60. Adequate level of anaesthesia in BIS range 40-60 was achieved better in the closed-loop group (21%), and total propofol consumption and time to the tracheal extubation were lower in this group. In recent studies (Solanki et al. 2010, Hemmerlig et al. 2010), BIS controlled closed-loop propofol administration system performs better than propofol manual administration for hypnosis control during surgical anaesthesia and postoperative sedation. In Solanki’s study, 41 cardiac surgery patients were randomly allocated to postoperative propofol sedation using either closed-loop anaesthesia delivery system or manual titration of propofol. The target value of BIS sedation was 70. The achieved time of sedation with target BIS 70 ± 10 was significantly higher in the CLADS group compared with manual group (p=0.002). In Hemmerlig’s study, 40 patients undergoing major surgery were randomly allocated either to the CLADS group with a target BIS of 45 or to the group with a manual administration of propofol. The performance of CLADS group was better than the control group, keeping BIS nearer to 45. Control of anaesthesia (BIS 45) occurred significantly more often (p=0.001) in the CLADS group.

2.4.5 Spectral entropy

Entropy is a name from thermodynamics employed in the analysis of complex biological signals and EEG research. Steyn-Ross from New Zealand adopted the entropy thermodynamic concept to in-depth anaesthesia research (Steyn-Ross et al. 1999). The spectral entropy, which is a measure of the hypnotic level of anaesthesia was conceptualised in Finland in 1999 by Datex-Ohmeda. In the year 2000, the abstract by Viertiö-Oja and colleagues, entitled “New Method to Determine Depth of Anaesthesia from EEG Measurements,” was recognised by The Society of Technology in Anaesthesia as the best abstract in anaesthesia clinical technology research. The entropy estimates the

23

complexity and irregularity of the signal (Viertiö-Oja et al. 2004) and quantifies the amount of disorder in the EEG frequency space. If the signal has a high complexity, then its entropy is high (Fig. 3A). If the signal is regular and totally predictable, then the value of entropy is low (Fig. 3B). One can compute the signal of EEG entropy with the time and frequency domain and with the Fourier transformation to the time-frequency balanced spectral entropy. If the EEG signal includes a wide spectrum of frequencies, then the spectral entropy is high. Inversely, if the complexity of EEG signal is low and features only a few frequencies, then the entropy value is low.

A)

B)

Figure 3 A and B. In the figure A with the patient awake, the EEG complexity is high with high

entropy value. In the figure B the EEG signal during anaesthesia is regular with only few

frequencies and entropy value is low. Reprinted with the permission of General Electric,

Healthcare, Finland.

Two components of spectral entropy are calculated. The response entropy is computed from EEG and EMG frequency ranges of 0.8 to 47 Hz. An increase in the EMG is an indicator of patient arousal, for example, in response to nociception stimulation. The RE entropy’s shortest sampling window is 1.92 s, which allows the RE to make a fast response during the monitoring to EMG activation (32-47 Hz). The state entropy is computed from range 0.8-32 Hz, and its time window sampling lies between 15-60s. The RE-SE difference is an indicator of EMG activation. The burst suppression ratio is also included in the measurement of deep anaesthesia with 1-min windows and a burst suppression detection technique was described by Särkelä and colleagues (2002).

2.4.5.1. Entropy evaluation studies

Vakkuri et al. (2004) validated spectral entropy as an accurate assessment of anaesthesia depth. They ran spectral entropy on 70 patients who had received sevoflurane, propofol, or thiopental. RE and SE distinguished very well between conscious and unconscious states and showed high sensitivity and specificity in the detection of LOS. RE decreased earlier than BIS and SE during the patients’ emergence from anaesthesia. Moreover, Ellerkmann et al. (2004) found that SE and RE were useful EEG-based measures of the effect of sevoflurane and that they both decreased continuously with the increase in the

0 50 100 150 200 250 300 350 400 450 500-1

-0.5

0

0.5

1

0 50 100 150 200 250 300 350 400 450 500-1

-0.5

0

0.5

1

High

Entropy

Low

Entropy

24

concentration of sevoflurane. A multicentre study by Vakkuri and colleagues (2005) looked at 368 patients under general anaesthesia with propofol-alfentanil-nitrous oxide. The reduction in dose of propofol and faster emergence from anaesthesia were detected in the entropy group. In the entropy group, the values of the entropy were higher than the BIS index during the whole operation, with the gap being the widest in the last 15 minutes. The SE, RE, BIS, and sedation level were recorded for 20 patients during minor gynaecologic surgery (Schmidt et al. 2004). A record was made every 20s during stepwise increase of propofol until the patients lost response. The SE, RE, and BIS revealed similar information about the level of sedation measured by prediction probability and allowed the researchers to distinguish between the different steps of anaesthesia.

In the study by Ellerkmann et al. (2006), the dose-response relationship of SE and RE during propofol anaesthesia in comparison with the BIS was investigated. The prediction probability (P(k)) was calculated to evaluate the performance of SE, RE, and BIS to predict changing propofol effect-site concentrations. The calculations revealed a significant difference between SE and RE compared with BIS, with P(k) = 0.77 +/- 0.09, 0.76 +/- 0.10, and 0.84 +/- 0.06, respectively. In this study, BIS was better than SE and RE at predicting the propofol effect-concentration. Contradictory results were discovered during a propofol anaesthesia-based study (Ianuzzi et al. 2005) that compared the BIS and SE for predicting loss of verbal contact and LOC during steady-state conditions. SE had higher correlation with effect-site concentration than did BIS and may therefore be more useful than BIS in predicting both LVC and LOC. Rinaldi et al. (2007) evaluated BIS and SE in terms of their correlations with different end tidal concentrations of sevoflurane. During anaesthesia, the correlation with end tidal (mean 1.5%) sevoflurane was -0.75 for SE and -0.70 for BIS. Another study (Moller and Rampil 2008) examined volunteers to assess the correlation of entropy values to word recall and the correlation of motor response to verbal command during emergency from either propofol or sevoflurane. When clinicians gave patients a verbal command to move, entropy increased in a dose and drug-dependent fashion. Entropy parameters measured by prediction probability were reliable predictors of recall. In the study of Kaskinoro et al. (2011) spectral entropy and BIS values were tested to differentiate consciousness from unconsciousness during increasing doses of three different anaesthetic agents. Thirty healthy male volunteers aged 19-30 yr were recruited and divided into three 10-volunteer groups to receive either dexmedetomidine, propofol, or sevoflurane in escalating concentrations at 10 min intervals until LOC was reached. Because of wide inter-individual variability, BIS and entropy were not able to reliably differentiate consciousness from unconsciousness during and after stepwise increasing concentrations of these anaesthetics.

In a study that featured propofol and nitrous oxide during the induction of anaesthesia (Anderson and Jakobsson 2004), entropy showed a successive decrease with propofol; however, loss of consciousness with nitrous oxide was not associated with change in entropy indices. The poor performance of spectral entropy during S-ketamine anaesthesia was also detected (Maksimow et al. 2006), because S-ketamine induced more high

25

frequency EEG activity in the gamma band. Compared with standard practice, using spectral entropy or BIS during sevoflurane anaesthesia reduced sevoflurane consumption by 29% (Aime et al. 2006). The spectral entropy guided induction of general anaesthesia with propofol reduced the propofol requirements and allowed better cardiovascular stability in 72 elderly patients (Riad et al. 2007). In elderly patients (> 65 yrs), the BIS and spectral entropy indices were significantly affected by their age (Lysakowski et al. 2009). At LOC with propofol induction, all indices were significantly higher in the elderly than they were in young patients, at 70 (range 58-91) versus 58 (40-70), respectively. In a study that compared standard anaesthesia practice and entropy guided propofol-remifentanil anaesthesia (Gruenewald et al. 2007), the standard group of patients received more propofol and less remifentanil. Prediction probability for LOC was the best for BIS (0.96) and RE (0.93), but patients’ emergency from anaesthesia was best predicted by RE (0.94). A PET-based study by Maksimow et al. (2005) aimed to find a correlation between spectral entropy and regional CBF during general anaesthesia with propofol and sevoflurane. The reduction in cortical and global cerebral blood flow was associated with significant decreases in spectral entropy.

There is a conceptual difference in the BIS algorithm and spectral entropy concept, which is detailed in Table 8. The BIS algorithm uses the bispectrum to combine information of the β-ratio at light levels of anaesthesia with information of burst suppression at the deepest levels of anaesthesia. The entropy applies exactly the same mathematical formulation at all levels of anaesthesia, starting from the state in which the patient is completely awake to total suppression. Table 8. Conceptual differences in the spectral entropy and bispectral index algorithms used to the EEG signal analysis.

Spectral Entropy

BIS

Patient awake 90-100 Light sedation (60-90)

Entropy

BetaRatio

Surgical levels of anaesthesia (30-60)

Entropy

Bispectrum (SynchFastSlow)

Deep anaesthesia (0-30)

Entropy

Burst Suppression Ratio

26

The RE index is a new contribution to the measurement of anaesthesia depth with higher EMG frequency in its calculations. The increase in the patient’s forehead EMG can be a sign of arousal, that is, inadequate analgesia. The values of response and state entropy are equal during general anaesthesia when the EMG power (sum of the spectral power between 32 and 47 Hz) is zero. The difference between RE and SE (RE-SE) is an indicator of upper facial muscle activity reflecting changes, for example, in the arousal reactions that the patient displays in response to painful stimuli (Vakkuri et al. 2004). In studies regarding RE responsiveness to pain stimulation (Wheeler et al. 2005, Weil et al. 2008), RE increased after surgical stimulation and was a predictor of patients’ movement after stimulation, but there was no correlation between haemodynamic response and RE (Weil et al. 2008). Authors concluded that high values (RE > 55) before stimulation should be avoided in order to decrease the risk of motor response. Valjus et al. (2006) randomly assigned 51 patients to receive esmolol or remifentanil. All patients’ RE was compared with their SE responses to the pain stimulation. The hypothesis was that the RE values should be higher in the esmolol group after stimulation, but no significant differences were detected between SE and RE in either group. Authors concluded that RE seems not to be more sensitive than SE in guiding the use of opioids during general anaesthesia. In addition, other studies suggest that the level of neuromuscular blocking may confound the clinician’s interpretation of the RE values (Liu H et al. 2005, Kawaguchi et al. 2009). Liu H et al. (2005) studied the influence of the bolus of the NMBA drug on BIS and entropy values. Under steady propofol-remifentanil anaesthesia in lightly anesthetised patients, a bolus of the NMBA administration decreased bispectral index and response entropy, whereas the bolus did not decrease state entropy values. In the study of the influence of the level of muscle relaxation on the RE-SE values after intubation, the patients who received a higher dose of rocuronium displayed stronger suppression of RE-SE response (Kawaguchi et al. 2009). The authors concluded that the estimates of nociception using RE-SE should be interpreted carefully in different states of muscle paralysis during general anaesthesia.

The performance of the entropy module in estimating nociception during sevoflurane anaesthesia was studied in 40 female patients (Takamatsu et al. 2006). After surgical stimulation, the RE-SE difference increased significantly, and RE (29 vs. 38) and SE (24 vs. 37) before skin incision were significantly lower in patients who did not move. The authors concluded that noxious stimulation increased the RE-SE difference. However, an increase in the difference does not always indicate inadequate analgesia and should be interpreted carefully during anaesthesia. Forty patients undergoing laparoscopic gastric banding were randomly assigned to receive either fentanyl or dexmedetomidine infusion, with desflurane concentration adjusted to maintain RE at 45 (Feld et al. 2006). BIS, RE, and EMG were evaluated during surgical stimulation. Fifteen of the 40 patients showed activation of RE above 60 during surgery also in addition to increased EMG and BIS. In 31 patients with propofol-nitrous oxide and propofol-remifentanil anaesthesia without neuromuscular blockers, the responsiveness of RE-SE and EMG to the noxious stimulation was evaluated (Aho et al. 2009). RE, SE, and RE-SE increased during intubation, and elevated RE was

27

followed by increased SE values, implying that RE-SE difference is not a good indicator of nociception. Another study (Aho et al. 2011) recorded EEG, EMG, and entropy values simultaneously in 38 patients before and after skin incision. It also studied the effect of rocuronium on EMG arousal. EEG arousal was classified as increase in the β and δ pattern activity. Both β and EMG arousals increased SE and RE. A significant increase in RE-SE values was only seen in patients without rocuronium. Thus, neither the RE-SE difference nor RE alone has been proven as predictors of nociception and analgesia level during general anaesthesia in clinical studies. Seitsonen et al. (2005) examined the relationship between motor reactions and physiological EEG, ECG, and photoplethysmography variables during skin incision in sevoflurane anaesthesia. The authors suggested that a combination of information from different sources may be required for monitoring the adequacy of analgesia during anaesthesia. The spectral entropy of EEG has been used in the development of the response index of nociception (Rantanen et al. 2006). This multiparameter approach combined photoplethysmographic waveform, SE, RE, and HRV to detect noxious stimuli during surgery. Later on it was further developed into a surgical stress index (SSI), which uses only information from the phtoplehtysmography (Huiku et al. 2007).

2.4.5.2 Assessment of nociception during general anaesthesia

Monitoring of pain perception during general anaesthesia, nociception, is challenging. Continuous interaction between analgesic and anaesthetic drugs during general anaesthesia has made it difficult to monitor separately hypnosis and analgesia. Autonomic signs such as tachycardia, hypertension, sweating and lacrimation are not specific responses to nociception and in the past years the multimodal monitoring including information from haemodynamic parameters, EEG, and pulse plethysmography was developed to monitor analgesia. Other monitoring methods such as activation of frontal facial muscles (Hynynen et al. 1985), changes in the skin conductance (Storm et al. 2002), suppression of the photoplethysmographic pulse amplitude (Seitsonen et al. 2005, Korhonen and Yli-Hankala 2009) have also been proposed to assess nociception during general anaesthesia.

The surgical stress index, a numerical parameter of the level of surgical stress delivered from heart beat interval and photoplethysmographic pulse wave amplitude (PPGA) has been proposed as a simple numerical measure of the surgical stress by Huiku and colleagues (2007). SSI was validated during propofol-remifentanil general anaesthesia, and reacted to surgical stimulation and changes in analgesic drug concentration (Huiku et al. 2007). SSI was evaluated in thirty gynaecological patients receiving randomly esmolol and remifentanil before surgical stimulation (Ahonen et al. 2007). In the esmolol group SSI reacted to the skin incision and was higher after trocar insertion. Authors concluded that SSI seems to reflect the level of surgical stress and may be helpful to guide the use of opioids during general anaesthesia. Changes in a SSI in response to surgical stimulation during propofol-remifentanil anaesthesia was compared with spectral entropy parameters,

28

HR, and PPGA in forty ASA I patients (Struys et al. 2007). SSI was better as a measure of nociception-antinociception balance than SE, RE, HR, and PPGA. In the twenty six patients undergoing shoulder surgery with plexus block and general anaesthesia the performance of SSI was compared with patients without plexus block. SSI was lower in the patients with plexus block and it was better in detecting nociceptive stimuli than HR, blood pressure and response entropy (Wennervirta et al. 2008). In the pilot study by Chen and colleagues (2010), SSI-guided analgesia was compared with standard clinical practice in eighty patients under general anaesthesia for ear, nose and throat surgery. In the SSI-guided group less remifentanil was administered and also less haemodymanic instability and movement during surgery was noticed compared with the standard practice group. In the interesting study by Ilies et al (2010), SSI performance as a measure of nociception-antinociception balance was compared under spinal and general anaesthesia. In awake patients under spinal anaesthesia, SSI did not reflect the analgesia balance, probably due to influence of patient mental stress on sympathetic nervous system. The authors concluded that SSI is not only affected by the nociception stimulus, but also by the depth of anaesthesia or sedation.

Recently SSI was renamed as the Surgical Pleth Index™ (SPI). SPI index was evaluated in thirty three patients undergo intracranial neurosurgery (Bonhomme et al. 2011). Patients were randomly assigned to three different remifentanil concentration groups. SPI was compared with MAP and HR. An influence of intravascular volume status on SPI was also studied. The performance of the SPI during noxious stimulation due to a neurosurgical head holder was comparable with that of MAP and HR. Low intravascular volume status and chronic treatment of hypertension lowered PSI response to stimulation. In the study of Hans et al. (2012) during stable anaesthesia and surgery, SPI also changed in response to fluid therapy, through pulse wave amplitude modifications by intravascular volume status. Hence, there are confounding factors in the interpretation of SPI as a measure of the nociception-anti-nociception balance during general anaesthesia.

The more sophisticated multiparameter approach for the nociception measurement was developed by the same research group (Rantanen et al. 2006), which was working also with SSI. In the Rantanen study, it first was estimated which haemodynamic and EEG parameters reflected the nociceptive–anti-nociceptive balance. HRV, RE, RE-SE and photoplethysmography were the best predictors of the Clinical Signs-Stimulus-Antinociception score system. Those parameters were used to develop an algorithm for response index of nociception (RN) calculation. The RN index is scaled from 0 to 100. In sixty patients under general anaesthesia for gynaecological surgery, RN index was evaluated during noxious stimulation and the pre- and after event values were compared (Sarén-Koivuniemi et al. 2011). An increase in the RN index was associated with patient movement and the index also increased after intubation and skin incision.

29

2.5 ASSESSMENT OF THE SEDATION LEVEL IN THE ICU

Patients undergoing intensive care can experience high amounts of stress from invasive monitoring, artificial ventilation, and the ICU environment itself, so sedation is an essential part of ICU treatment. Seventy per cent of ICU patients suffer from anxiety and distress during the ICU stay (Fraser et al. 2000), and more than 80% of patients under mechanical ventilation may develop delirium in the postoperative phase of treatment (Ely et al. 2001). Twenty two per cent of ICU patients (8-51%) developed Posttraumatic Stress Disorder (Davydow et al. 2008).

Sedation is defined as diminished cognitive function (cortical activity) with intact respiratory and cardiovascular function (Brown et al. 2010). The goal of ICU sedation according to Society of Critical Care Medicine guidelines recommendations (Jacobi et al. 2002) is tranquil and painless patient with possibility of communication. One retrospective study evaluated the use of sedatives and NMBA drugs during mechanical ventilation of over 5000 patients across numerous ICU settings (Arroliga et al. 2005). Sixty-eight per cent of these patients received sedation during mechanical ventilation, and 13% received NMBA drugs during treatment. In a large, multicentre study of sedation practice on 1381 adult patients across 44 French ICUs, 57% were deeply sedated on day 2 and 41% on day 6 (Payen et al. 2007). The guidelines of ICU sedation (Shapiro et al. 1995, Jacobi et al. 2002) recommend the use of bolus sedatives or sedatives in regular intervals if they are required, but quite often, continuous infusion is used (Hansen-Flaschen et al. 1991). The most frequently used drugs in Europe are benzodiazepines (midazolam, diazepam, and lorazepam), haloperidol, and propofol for sedation, and morphine and fentanyl for analgesia (Soliman et al. 2001). Dexmedetomidine is a new, promising, highly selective alpha-2 agonist for ICU sedation, which offers good anxiolysis without respiratory depression (Ruokonen et al. 2009).

In critically ill patients, sedation is sometimes used for a long time and thus prolongs the patient’s mechanical ventilation and ICU stay. In a study by Kollef et al. (1998) which included 242 ICU patients requiring mechanical ventilation, 93 received continuous i.v. sedation while 149 (61.6%) patients received either a bolus administration of i.v. sedation (n = 64) or no i.v. sedation (n = 85). The duration of mechanical ventilation was significantly longer for patients receiving continuous i.v. sedation compared with patients in the no-sedation group. Brook et al. (1999) randomly assigned 321 mechanically ventilated patients to the sedation protocol group with Ramsay sedation score 3 or to the standard sedation group. In the protocol group, the median duration of mechanical ventilation was significantly shorter (55 vs. 119 hours), and ICU and hospital stays were also shorter than the stays of the standard sedation group. In the randomised, controlled trial by Kress and colleagues (2000), 128 adult patients who were receiving mechanical ventilation and continuous infusions of sedative drugs were randomised into two groups of sedation protocol. In the intervention group, the sedative infusions were interrupted daily until the patients were awake. In the control group, the infusions were interrupted only at the

30

discretion of the clinicians. The daily interruption in the sedation decreased significantly the duration of artificial ventilation (4.9 vs. 7.3 days) and ICU stay (6.4 vs. 9.9 days) compared with control group. After this study, the daily interrupted sedation protocol was implemented to clinical practice in the ICU setting. Since the sedation protocol’s implementation in the ICU practice, many studies have proven this sedation’s aid in weaning time and ICU length of stay (Jakob et al. 2007). In a large, randomised “Awakening and Breathing” controlled trial with 336 ICU patients (Girard et al. 2008), patients in the group with daily interruption of sedatives with spontaneous breathing trials spent more days breathing without assistance (14.7 days vs. 11.6 days), and were discharged from intensive care (9.1 days vs. 12.9 days) and from the hospital earlier (14.9 days vs. 19.2 days). The reduction of costs has also been reported with the use of protocol sedation management (Mascia et al. 2000).

According to those studies, appropriate ICU sedation plays an important role in the critical care practice, but the problem is that there is still no true gold-standard method for monitoring the depth of sedation. In daily practice, the level of sedation is assessed by using clinical scoring systems that classify the patient’s response to a command or to painful stimulus. However, the predictive value of scores is limited due to many confounding factors, such as motor deficit, altered states of consciousness, and inter-rater variation in interpretation. Since sedation is basically a depression of the functioning of the CNS, methods measuring CNS function are the ones most likely to be related to the depth of sedation. EEG-based monitoring has been tested during ICU sedation, but the results of these studies are ambivalent. The next chapters present the different methods of sedation assessment.

2.5.1 Clinical methods of sedation assessment The assessment of a patient’s level of sedation is a routine practice in patient treatment in the ICU. Clinical experience shows us that some ICU patients during treatment need very deep sedation to tolerate mechanical ventilation while others are comfortable without sedation or light sedation. Over-sedation causes the problems described in the previous chapter and is more difficult to assess than under-sedation. The use of clinical sedation scores allow clinicians to estimate the intensity of deep sedation so that they can change doses of sedation drugs. The ideal sedation scoring system should be simple to use, non-invasive, and sensitive to different sedation drugs. For more objective and consistent assessment, 25 different clinical scoring systems were developed that use patients’ clinical responses to verbal comments and pain stimulation (De Jonghe et al. 2000). The Glasgow Coma Scale (GCS) was the first one developed to assess the level of consciousness of trauma patients using ocular, verbal, and motor activity, but it cannot be used for intubated patients and thus is not useful for assessing sedation in ICU practice. The most popular sedation scales used in ICU sedation assessment are presented in Table 9.

31

Table 9. Sedation score scales.

Ramsay Sedation Score (RS) (Ramsay et al. 1974)

Sedation Agitation Scale (SAS) (Riker et al. 1999)

Vancouver Interactive and Calmness Scale (VICS) (de Lemos at al. 2000)

Richmond Agitation-Sedation Scale (RASS) (Sessler at al. 2002)

Adaptation to Intensive Care Environment (ATICE) instrument (De Jonghe at al. 2003)

Minnesota Sedation Assessment Tool (MSAT) (Weinert at al. 2004)

The Ramsay score was widely use in ICU to assess patient’s mental status and sedation. Six levels of sedation are formulated: 3 levels when the patient is awake, and 3 levels when the patient is sedated (Table 10).

Table 10. Ramsay sedation scale.

Patient description

1. Anxious and agitated 2. Awake and cooperative, tranquil 3. Respond to verbal commands 4. Brisk response to a light glabellar tap or loud auditory stimulus 5. Sluggish response 6. No response

The Richmond Agitation-Sedation Scale is a well validated score and mostly used in the sedation assessments of ICU settings. The RASS score ranges from +4 to -5 and relates to patients’ behaviours. The RASS is followed by an assessment of response to voice (score -1 to -3) and then an assessment of response to physical stimulation, such as shaking the shoulder. If the patient does not respond to the shake, then the clinician rubs the sternum (score -4 to -5). In the large, multicentre, prospective cohort study of 313 patients receiving mechanical ventilation, the RASS demonstrated excellent inter-rater reliability with the Ramsay score, and the RASS was superior to the GCS. RASS also correlated well with the attention screening examination (r = 0.78), successful extubation, and BIS scores (r = 0.63, Ely et al. 2003). The degrees of RASS are presented in Table 11.

32

Table 11. Richmond Agitation-Sedation Scale. Score Patient description

+4 Combative Overtly combative or violent, immediate danger to staff +3 Very agitated Frequent non purposeful movement or ventilator dys-synchrony +2 Agitated Anxious or apprehensive but movements not aggressive or vigorous 0 Alert and calm -1 Drowsy Not fully alert, with eye contact to voice -2 Light sedation Briefly (<10 seconds) awakens with eye contact to voice -3 Moderate sedation Any movement (but no eye contact) to voice -4 Deep sedation No response to voice, but any movement to physical stimulation -5 Unarousable No response to voice or physical stimulation

The problem with the RASS scale (and other sedation scales) is that the ratings come from subjective assessments of sedation according the patient response to voice or pain stimulation. When the patient receives neuromuscular blocking drugs, is having disturbances in consciousness, or is deeply sedated, the uses of these scales are limited. Thus, EEG-based monitors, which are used during anaesthesia, have been under active research in ICU sedation assessment (LeBlanc et al. 2006).

2.5.2 EEG-based monitoring of ICU sedation

The EEG is sensitive to sedation drugs used in the ICU. Thus the EEG-based monitoring of sedation, AEP, BIS, and entropy were under active research as objective tools for assessment of ICU sedation. Regrettably, the use of these monitors in ICU sedation has disappointed investigators, and there are no studies demonstrating that the use of EEG-based monitors in ICU sedation improves outcomes and quality of sedation (LeBlanc et al. 2006).

2.5.3 Auditory evoked potentials and ICU sedation

AEPs were the first EEG-based monitors tested in studies on ICU sedation. AEPs were recorded in paralysed patients with mechanical ventilation before, during, and after physiotherapy. The Nb latency responded to patient arousal during the procedures at constant levels of sedation (Sneyd et al. 1992). A neurophysiological study (Schulte-Tamburen et al. 1999) compared AEP with five different scoring systems. The Ramsay score correlated best with the Nb latency of AEP. Nevertheless, the Ramsay score’s correlation was poor with the light levels of sedation, so AEP can only be recommended for patients under deep levels of sedation. Lu et al. (2008a) compared the AAI index to the BIS index during ICU sedation in 38 patients after administration of muscle relaxant. The AAI index (34 vs. 15) and BIS (58 vs. 44) both decreased significantly after the patients received neuromuscular blocking drugs, demonstrating that the electromyographic activity of the forehead muscles influenced the AAI and BIS values. In 77 critically ill children, the AEP

33

index and BIS were compared with the Ramsay sedation scale without muscle paralysis (Lamas et al. 2008). They found a moderate-to-good correlation and agreement between the clinical scale and BIS as well as the AEP index. In a study of 90 patients during short sedation after major surgery, the auditory evoked potential (AAI index) and BIS monitors revealed an acceptable correlation with the Ramsay sedation scale, but both were significantly influenced by EMG activity (Lu et al. 2008b). The authors concluded that they can be used to measure sedation in the paralysed patient and can be used during deep sedation when sedation scores are not reliable. In the von Dossow et al. (2009) study evaluating the correlation and agreement between the BIS or AAI and a Ramsay score in the 40 patients after cardiac surgery, both BIS and AAI correlated well with RS overall and also at different levels of sedation. The long-latency auditory-evoked potentials as an electrophysiological monitoring of ICU sedation state were also tested. They can monitor sedation with midazolam, but not with dexmedetomidine (Haenggi et al. 2006).

2.5.4 Bispectral index and ICU sedation

The goal of ICU sedation is a light state of sedation level with tranquil and arousable patients. At this level of sedation, BIS cannot function properly because of EMG artefacts. Simmons et al. (1999) studied BIS and SAS during sedation for mechanical ventilation under neuromuscular blockade. They discovered that BIS performed well as a monitor of sedation, but it is important to note that their study’s patients were in deep sedation. In the Nasraway et al. (2002b) study, only a moderate correlation (0.5) between BIS and SAS was detected. In 45 patients who required muscle relaxation during ICU treatment under sedation guided by the SAS, Vivien et al. (2003) measured BIS before and after NMBA administration. They found a significant decrease in the BIS values (67 vs. 43) after the administration of NMBA. They concluded that high muscular activity in the non-paralysed patient is a confounding factor during BIS sedation monitoring and that BIS can overestimate the sedation level. The same results and conclusions were made in a study of BIS under midazolam and isoflurane ICU sedation (Sackey et al. 2007). BIS does not reliably predict sedation depth as measured by clinical evaluation in non-paralysed ICU patients and also a strong correlation between BIS (0.75) and EMG was found. In the recent study by Dahaba et al. (2012) BIS was measured before and after reversal of NMBA drugs with sugammadex or neostigmine. BIS increased only in patients with EMG activity before reversal of the NMBA. In the randomised study by Weatherburn et al. (2007) in mechanically ventilated patients with midazolam and morphine sedation, BIS monitoring did not reduce the amount of sedation used, the length of mechanical ventilation time, or the length of ICU stay.

2.5.4.1 BIS and sedation scales

In the study by Arbour et al. (2009), also only a moderate correlation (0.52) was detected between BIS and SAS scale in the 40 mechanically ventilated patients. However, Arbour and colleagues also did notice a strong correlation between BIS and EMG (0.75). The authors concluded that BIS values during ICU sedation should be interpreted with caution,

34

because EMG activity and other factors seem to confound BIS scores. Many reports have demonstrated large variability and overlap in BIS values at distinct depths of sedation and that BIS is not independent of drugs used (Iselin-Chaves et al. 1998, Mi et al. 1999, Lysakowski et al. 2001). When a single drug is used, the BIS may provide a reliable monitoring for sedation, while administration of opioids together with hypnotics may substantially change the predictive value of the BIS (De Deyne et al. 1998). One possible explanation for this may be that small doses of opioids produce minimal electrophysiological alterations that change the BIS values. In the article by LeBlanc et al. (2006), they reviewed BIS during ICU sedation in 19 studies that compared BIS with sedation scales. The interpretation of data was difficult and randomised, controlled trials demonstrating improved outcomes with BIS monitoring have not been reported. Authors recommended that further studies with BIS are needed to determine the optimal method of obtaining BIS data and evaluating the impact of BIS on relevant patient outcomes. Therefore, the role of BIS monitoring in the ICU has yet to be determined.

Any frontal EMG activity during BIS monitoring can disturb the BIS index values by producing artefacts in the EEG at frequencies over 30 Hz, but in the new BIS VISTATM monitor algorithm should eliminate artefacts from the EMG activity. In a recent study (Karamchandani et al. 2010), The ability of BIS to assess sedation level during propofol-fentanyl-based sedation during mechanical ventilation in the ICU and correlation with RASS was evaluated. The good correlation (0.56) between BIS and RASS was found with the median BIS value of 56 (42-89) and a RASS scale from 0 to -3. The BIS index of 70 demonstrated high sensitivity (85%) and specificity (80%) while differentiating adequate sedation from deep sedation. In the a prospective observational trial (Ogilvie et al. 2011) in 94 mechanically ventilated trauma patients sedated with propofol alone or in combination with midazolam, BIS, RASS, EMG, and HRV , as a test of autonomic function, were measured for 45 minutes during daily spontaneous awakening trials. BIS was not altered by age, heart rate, or HRV and was similar with propofol alone or propofol plus midazolam, but the presence of brain injury or the use of NMBA drugs influenced the BIS values. BIS was reliable and has advantages over RASS with sensitivity: 90% versus 77% (p = 0.034) and specificity: 90% versus 75% (p = 0.021) in patients without NMBA agents.

2.5.5 Spectral entropy and ICU sedation

Mahon et al. (2008) evaluated spectral entropy values during moderate propofol sedation in healthy adult patients. They assessed sedation level with the AAOS scale and multichannel EEG monitoring. A statistically significant decrease in SE and RE values was demonstrated in patient/time units in which clinical or EEG evidence of sedation was present, so they concluded that spectral entropy offers potential as a monitor of propofol-induced sedation. Hernandez-Gancedo et al. (2007) went deeper by evaluating the correlation between the Ramsay sedation scale, BIS, and entropy values during ICU sedation in the postoperative period. The median values of RS was 6 (deep sedation), and the mean values of SE 53 ± 27, RE 60 ± 30 and BIS 62 ± 24. SE (0.71) and BIS (0.78) correlated well with Ramsay deep

35

sedation level; therefore, the authors concluded that entropy is not superior to BIS for the assessment of sedation. In the study by Maksimow et al. (2007), the utilisation of an entropy monitor with the dexmedetomidine-induced sedation was evaluated. The Entropy decreased from 76 +/-8 to 43 +/-10 (p < 0.001) after loss of consciousness, and increased from 14 +/-4 to 63 +/-13 (p < 0.001) upon regaining of consciousness. Haenggi et al. (2008) used BIS and entropy during the sedation of 44 ICU patients guided by the Ramsay scale. They detected that unpleasant or painful stimulation procedures are associated with changes in BIS and entropy values. However, clinical indications for drug administration were not reflected by these electroencephalographical parameters, and barely by sedation level (Ramsay score) before drug administration or tracheal suction. They concluded that this finding precludes incorporation of entropy and BIS as target variables for sedation and analgesia protocols in critically ill patients.

Walsh and colleagues (2008) used the spectral entropy module and Ramsay sedation scale in 30 intubated and mechanically ventilated patients to assess the depth of sedation. During light sedation (Ramsay 3-4), wide variations were detected, which were caused by the frontal EMG. They concluded that entropy values do not discriminate light sedation levels well, and a major confounding factor is EMG activity. In the healthy volunteers (Haenggi et al. 2009a) sedated by midazolam/remifentanil or dexmedetomidine/remifentanil, a large inter-individual variation of BIS-index and entropy was found. Entropy values showed greater variability than did BIS index values, and the variability was greater during dexmedetomidine/remifentanil sedation than during midazolam/remifentanil sedation. Inter-individual variability of BIS and entropy values in sedated volunteers makes the determination of sedation levels by processed EEG variables impossible. The same research group studied the use of the BIS, entropy, and early auditory response potentials for the discrimination of sedation levels assessed by RASS during the propofol/remifentanil postoperative sedation in the ICU (Haenggi et al. 2009b). Their conclusion was the same as in their previous study; the high inter- and inter-individual variability of entropy and BIS precludes define a target range of values to predict the sedation level in critically ill patients by using these parameters.

The responsiveness index (RI) is a new promising sedation assessment method that use the patients’ frontal EMG responses (Walsh et al. 2011), is. In a study of 30 ICU patients, the new algorithm of frontal EMG, RI was compared with the Ramsay score and spectral entropy. RI correlated well with the Ramsay category, and discrimination between 'lighter' vs. 'deeper' (Ramsay 1-3 vs. 4-6) was good for general ICU patients (P(k)=0.80) and excellent for cardiac surgery patients (P(k)=0.96). In this study RI performance as a measure of sedation level was significantly better than entropy.

36

2.6 ANAESTHESIA FOR CARDIAC SURGERY

Cardiac surgery patients with impaired cardiac function and comorbidities are vulnerable to haemodynamic compromise during anaesthetic administration (Kertai et al. 2012). For over three decades, high doses of opioid anaesthesia were used for cardiac surgery to prevent haemodynamic instability and myocardial depression with ischemia (Lowenstein et al. 1969). Morphine, fentanyl or alfentanil (Hynynen et al. 1986a and 1986b) in large doses were predominantly used and resulted in prolonged ICU stay after cardiac surgery (Mangano et al. 1992). It was also believed that in the postoperative period, the high-dose opioid analgesia in the ICU protected patients from heart ischemic episodes. For example, in the Mangano trial (1992), the analgesic and sedative-hypnotic medications (morphine vs. sufentanil) were discontinued at 18 hours following myocardial revascularisation.

The introduction of fast track cardiac anaesthesia (FTCA) (Silbert et al. 1998) with the use of short-acting opioids with propofol (Ahonen et al. 2000) as well as volatile anaesthetics or midazolam maintenance, allows clinicians to reduce ICU and hospital stay and to reduce cost (Cheng et al. 1996a, 1996b and 1998). In a broad, retrospective, cohort study (Svircevic et al. 2009) of 7989 patients undergoing cardiac surgery, the FTCA was compared with a historical control group of high-dose opioid anaesthesia. The primary endpoint was the incidence of in-hospital mortality and incidence (as secondary outcomes) of myocardial infarction, stroke, and renal failure. There was no increased risk of adverse outcomes in either anaesthesia management group, and the FTCA was a safe method for the management of patients undergoing a wide variety of cardiac operations. The FTCA is also safe anaesthetic management method in terms of intraoperative awareness. In (1998) clinical trial by Dowd et al. in 617 patients undergoing cardiac surgery with FTCA, the incidence of awareness was only 0.3%. The authors concluded that this low incidence of awareness may be related to the use of a balanced anaesthetic technique involving the continuous administration of volatile (isoflurane) or intravenous (propofol) anaesthetic agents before, during, and after cardiopulmonary bypass. The new, short-acting opioid remifentanil has also been employed in cardiac surgery anaesthesia with no adverse complications (Möllhoff et al. 2001).

The incidence of major, neurological complications after cardiac surgery, such as stroke, is estimated between 0.9-5.4 % (Ahlgren and Aren 1998). Mild complications such as transient ischemic attack, neuropsychological impairment, patient confusion, and patient delirium all range from 28% to 79% (Roach et al. 1996, Wolman et al. 1999). They are probably related to the reversible brain oedema that CT scans can detect immediately after a cardiopulmonary bypass (Harris et al. 1993). From the beginning of cardiac surgery, the EEG monitoring was used during the operation to recognise and prevent brain ischemia, hypoxia, or other pathological situations (Stump 1999). In 61 patients undergoing cardiac bypass surgery (Toner et al. 1997), the changes in the EEG monitoring were only transient and did not

37

correlate with the neurological deficit occurring after the operation. In the study by Bashein and colleagues (1992), they discovered a weak association between intra-operative EEG deficit and early neurological outcome as well as no relation with late outcomes. Roach and colleagues (1999) studied the neuroprotective effect of the burst suppression pattern in the EEG during cardiac surgery and this pattern’s correlation with neurological deficit in the post-operative period. Two hundred and twenty patients undergoing cardiac valve replacement with CPB were randomly assigned to the group with a burst suppression pattern in the EEG during surgery or to the control group. The burst suppression pattern in the EEG did not significantly reduce the incidence or severity of neurologic or neuropsychological dysfunction compared to the control group. The study of Morimoto et al. (2005) suggested that the use of BIS and raw EEG may be useful for detecting brain ischemia during cardiac surgery, especially when the insult is sudden and located in the frontal area. BIS was used the detect the cerebral hypoperfusion and ischemia undergoing the awake carotid endarterectomy suggesting that sudden BIS decrease may correlate with clinical cerebral ischemia (Estruch-Perez et al. 2010a and 2010b) but another study (Deogaonkar et al. 2005) showed no relationship between low BIS and cerebral ischemia. Bonhomme at al. (2007a) in the study in patients undergoing carotid endarterectomy with TIVA and BIS monitoring showed that carotid cross-clamping under constant level of anaesthesia may increase, decrease or remain unchanged the BIS values. For this reason, the use of the EEG and BIS during cardiac surgery to monitor brain ischemia is only employed for research purposes, and techniques that are more sophisticated are used in practice, such as near-infrared spectroscopy, which can monitor the haemoglobin saturation in the brain tissues during anaesthesia for cardiac surgery (Nollert et al. 1998).

2.6.1 Hypothermia, cardiopulmonary bypass and EEG-based monitoring

The institution of cardiopulmonary bypass during cardiac surgery cause changes in the CBF and in cerebral metabolism. The priming solution use in the CPB machine led to thermodilution of blood, reduction in the blood’s viscosity, and increase in the CBF (Stump et al. 1999), and decreases the delivery of oxygenated haemoglobin to the brain. The CPB is a non-physiological procedure, and non-pulsatile pump perfusion is associated with an increase in the vascular resistance and results in microcirculatory collapse in the cerebral cortex. The use of extracorporeal circulation and blood contact with artificial surface initiated activation of systemic inflammatory response. This may cause microemboli formulation, which leads to neurological complications (Stump et al. 1999). The hypo- and hypercapnia causes vasoconstriction and vasodilatation in the brain vessels and changes in the EEG (Hänel et al. 1998). CPB alters the pharmacokinetics and pharmacodynamics of drugs resulting in drug concentrations changes (Hammaren et al. 1996, Bailey et al. 2009). Thiopentone (Hynynen et al. 1989) and propofol sequestration within the extracorporeal circuit in vitro was detected which may affect plasma propofol concentration in vivo (Hynynen et al. 1994a). The changes in the unbound fraction associated with significant changes in plasma total and unbound concentrations of alfentanil during CPB was also

38

reported (Hynynen et al. 1994b).

The priming solution and thermodilution may decrease blood anaesthetic concentration as well as BIS values. The sudden elevation of BIS after initiation and termination of CPB were reported by Puri and Murthy (2003) and by Vretzakis et al. (2006). In 11 patients scheduled to undergo heart surgery under extracorporeal circulation, the anaesthesia was maintained by intravenous administration of propofol combined with remifentanil and a BIS monitor (target BIS value: 40-45). Raw EEG signals were collected through the BIS monitor (Hayashi et al. 2010). In the late hypothermic CPB period, the EEG often expresses a slow EEG rhythm accompanied by disproportional de-synchronous characteristics, which are not quantifiable by the BIS algorithm. The surgical procedures, such as aorta cross clamping, result in the EEG slowing due to hypotension (Puri and Murthy 2003). All of these factors may produce alterations in the EEG and in the EEG-based monitoring of anaesthesia.

Hypothermia is used during the cardiac surgery in order to protect the brain. The use of mild hypothermia during cardiac surgery procedures evokes changes in the cerebral metabolic rate and in the EEG pattern (Doi et al. 1997a, Gugino and Yli-Hankala 2004). Temperature changes during CPB can result in EEG changes (Hett et al. 1995), and at 18° C, EEG becomes isoelectric with BIS values of 0 (Hayashida et al. 2007). The opposite effect was found during the re-warming phase of CPB (Hayashida et al. 2007, Ziegler et al. 2010). A one Celsius decrease in body temperature decreases BIS by 1.12 units (Mathew et al. 2001). On the other hand, BIS does not change during normothermic cardiopulmonary bypass (Hirschi at al. 2000, Schmidlin et al. 2001). BIS-guided anaesthesia during CABG reduces the propofol consumption by 30% without effecting stress response measured as a blood concentration of catecholamines and interleukin 6 (Bauer et al. 2004). The findings in this paragraph pertain only to cardiac surgery features effecting anaesthesia depth monitoring. These features can lead to changes in the EEG indices without changes in the anaesthetic state. Active scientific research is required in this field to recognise all those specific situations.

2.6.2 Postoperative sedation after cardiac surgery

The primary reasons for measuring depth of sedation in the postoperative period after cardiac surgery are to improve patient care and to reduce recovery time. Sedation is used after cardiac surgery to allow haemodynamic stabilisation, and to avoid agitation with increased oxygen consumption during the period of re-warming. Any sudden increases in heart rate and blood pressure may have damaging consequences for the heart, as myocardial oxygen supply can cause arrhythmias, acute ischemia, and heart failure. Pain after cardiac surgery impairs coughing and thus leads to atelectasis. Cardiac surgery procedures with CPB involve mild hypothermia, and patients need time for re-warming after surgery. Too deep level of sedation can lead to prolonged recovery, circulatory problems with hypotension, and respiratory problems. Excessively deep sedation can also cause a delay in extubation and ICU discharge. In order to achieve the goal of adequate

39

sedation, anaesthesiologists must constantly evaluate level of sedation and adjust the therapy as necessary. Plaschke et al. (2010) detected delirium in 28% of 114 patients after cardiac surgery. The mean BIS during postoperative sedation was significantly lower in delirious patients (72 vs. 85) compared to non-delirious patients. In the raw BIS-EEG data, lower alpha and higher theta power was detected in the delirious patients. Moreover, a significant correlation was found between the plasma stress hormone cortisol level and the BIS.

2.7 REFRACTORY STATUS EPILEPTICUS IN THE ICU

Status epilepticus is a rare syndrome with a prevalence of 18 patients per 10000 within a population. It is associated with high mortality rates, if not treated aggressively (Legriel et al. 2008, Drislane et al. 2009). The symptoms of status epilepticus are unconsciousness, tonic-clonic seizures, and urinary incontinence. If the seizures persist for longer than 30 minutes or come in repeated intervals when the patient is not conscious and not responding to the first and second line epileptic drugs, then emergency treatment in the ICU is necessary (Lowenstein et al. 1990, Kälviäinen et al. 2005, Meierkord et al. 2006). During the persistent seizures, the brain blood pressure regulation, intracranial pressure, and CBF are disturbed, so oxygen delivery to brain tissues may be insufficient and can lead to brain damage (Treiman et al. 1990). The mortality associated with status epilepticus is about 10% (Hauser 1990, Towne at al. 1994). Continuous EEG (cEEG) monitoring is required to assess the diagnosis and to detect non-convulsive status epilepticus, complex partial or subtle status epilepticus (Meierkord et al. 2006). During treatment, cEEG is also indispensable for informing the anaesthesiologists about the depth of anaesthesia and duration of burst suppression. The availability of cEEG may be limited and the interpretation of cEEG requires the service of a neurophysiologist (Leira et al. 2004).

2.7.1 Anaesthesia and goal of the treatment

Treatment of refractory status epilepticus consists of general anaesthesia that aims to stop visible convulsions and epileptic activity in the EEG. The goal of anaesthesia is a deep coma with a burst suppression pattern in the EEG (Chen and Wasterlain 2006). The barbiturates (thiopental, phenobarbital, and pentobarbital) were the first used for anaesthesia in the treatment of RSE (Rashkin et al. 1987, Osario and Reed 1989, Parviainen et al. 2002, Claassen et al. 2002). The problem with treating RSE with barbiturates is that the barbiturates cumulate during long administration and cause prolonged emergency from coma and higher infection risk. Short-acting propofol and midazolam are also used for pharmacological comas in the treatment of RSE (Claassen et al. 2002, Rossetti et al. 2004, Parviainen et al. 2006).

The workshop with the first London Colloquium on Status Epilepticus (Shorvon et al. 2008) provides the guidelines for RSE treatment. The document was approved by the Commission on European Affairs of the International League against Epilepsy. In this

40

document, the protocol for the in-hospital treatment of tonic-clonic RSE in adults is divided into three stages. In Stage I, a 4-mg i.v. bolus of lorazepam is recommended for the early status (0-10/30 min.), and this administration can be repeated. In Stage II, the phenobarbital i.v. infusion of 10mg/kg, the phenytoin i.v. infusion of 15mg/kg, the fosphenytoin i.v. infusion of 15 mg PE/kg, or the valproate i.v. infusion 25mg/kg can be used for established status (30/60-90 min.). If seizures continue after the first stage of treatment for 30-90 min the refractory status epilepticus treatment should be started in the ICU (Stage III). Propofol i.v. bolus of 2mg/kg is repeated if seizures persist, followed by continuous infusion of 5-10 mg/kg/hour to maintain a burst suppression pattern in the EEG (usually 3-5 mg/kg/hour). To control the seizures in the ICU setting, the clinician can administer a thiopental i.v. bolus of 100-250 mg, followed by i.v. infusion to maintain BS on the EEG. Alternatively, the clinician can administer a midazolam i.v. bolus of 0.1-0.3 mg/kg, followed by continuous i.v. infusion (usually 0.05-0.4 mg/kg/hour). When seizures have been controlled with BS pattern on the EEG for 12 hours, the drugs’ dosages should be reduced over the next 12 hours. If epileptic activity recurs in the EEG, the general anaesthetic agent i.v. infusion should be given again for another 12 hours. This cycle may need to be repeated every 24 hours until seizure control is achieved.

In the recent literature review by Meierkord et al. (2010), they discussed the degree of the evidence of the management of status epilepticus in adults. The first line drug for the treatment of RSE is i.v. lorazepam at 4-8 mg or diazepam at 10 mg, directly followed by 18 mg/kg of phenytoin. If the seizure continues, then clinicians can repeat the i.v. drugs. In the treatment of RSE, the barbiturates thiopental, propofol, and midazolam are recommended, but it is difficult to achieve burst suppression with midazolam (Meierkord et al. 2010). The propofol i.v. should be administered at an initial dose of 2-3 mg/kg, followed by continuous infusion at 4-10 mg/kg. Thiopental should begin with an i.v. bolus of 3-5 mg/kg and then be delivered in further boluses of 1-2 mg/kg until seizures are controlled. Once they are under control, a continuous infusion of thiopental at a rate of 3-7 mg/kg should achieve burst suppression on the EEG. The dose of midazolam required to achieve burst suppression is 0.2 mg/kg with continuous infusion at rates of 0.05-0.4 mg/kg/h.

2.7.2 Monitoring sedation depth during status epilepticus treatment

Seizures result from imbalances between excitation and inhibition in brain cells as well as imbalances between neuronal synchrony and dys-synchrony. Cerebral cortex and thalamic reticular formation are involved in the synchronisation of cortex neurons and are implicated in the generation of seizures (Chen and Wasterlain 2006). Continuous EEG monitoring is required for diagnosis of RSE and during ICU treatment to guarantee electrophysiological suppression. Unfortunately, cEEG monitoring is not always available outside normal working hours. Also, a placement of EEG electrodes is a time-consuming procedure and requires the presence of specialised staff from the neurophysiological laboratory. Automatic classification and analysing method of epileptic EEG spikes has also studied but not yet incorporated in the EEG analysis (Kortelainen et al. 2010).

41

Krishnamurthy and Drislane (1999) determined the depth of EEG suppression in 35 patients with persistent seizures or who went into relapse of RSE after the taper of pentobarbital. Twelve patients were under burst suppression, and persistent seizure control was achieved in six of them. Twenty patients were under a very deep coma, and 17 of those 20 patients exhibited a flat isoelectric EEG. However, isolated epileptiform discharges during the barbiturate coma did not correlate with the outcome in that study. In a study of 154 patients with RSE, epileptic activity persisted on the EEGs in 48% of the patients even after their clinical seizures were abolished (Delorenzo et al. 1999). In 117 children admitted to the ICU with persistent seizures or unexplained decreases in mental status, 44% had seizures on cEEG (Jette et al. 2006). Of this group, 39% had non-convulsive seizures. Twenty-three per cent of all 117 children had status epilepticus which was in the 89% of cases non-convulsive. The recent guidelines on the management of RSE in adults recommend administration of anaesthetic agents to achieve the burst suppression pattern in the EEG or even to achieve an isoelectric EEG (Meierkord et al. 2010). In the literature, there are a couple of case reports (Jaggi et al. 2003, Chinzei et al. 2004) in which the BIS was also used for the monitoring of anaesthesia depth during the treatment of status epilepticus.

42

43

3 Aims of the study

The aim of this study was to evaluate and compare the utility of different EEG-based neuromonitoring methods during general anaesthesia for cardiac surgery, postoperative sedation in ICU settings and during anaesthesia for the treatment of status epilepticus.

The purposes of the individual studies were:

1. To evaluate the utility and applicability of MLAEPs during short-term sedation after cardiac surgery in the ICU, when the level of sedation was assessed by the Ramsay score. It was to assess whether graded changes in sedation induced by pre- and postoperative sedative drugs can be detected by any of the MLAEP components (study I).

2. To test whether the parameters derived from EEG (MLAEPs and BIS) indicate similar states of anaesthesia during cardiac surgery with two different anaesthetic techniques, TIVA with propofol and alfentanil or balanced inhalation anaesthesia with isoflurane and alfentanil (study II).

3. To evaluate the reliability of BIS in recognising the burst suppression pattern in the EEG in the treatment of RSE in emergency situations in the ICU (study III).

4. To compare the primary metrics of the spectral entropy M-ENTROPY™ module and BIS VISTA™ monitor—i.e., BIS, SE, and RE—in terms of agreement and correlation during general anaesthesia for cardiac surgery. It was also hypothesised that RE would better reflect responses to surgical noxious stimulation (skin incision, sternotomy) than BIS, EMG, or haemodynamic parameters (study IV).

44

45

4 Patients and methods

4.1 STUDY PATIENTS

All four studies were approved by the Ethics Committee of Kuopio University Hospital according to the latest version of the Helsinki Declaration. Studies I and III were carried out in the Department of Intensive Care Medicine and studies II and IV in the Department of Anaesthesiology, Kuopio University Hospital. Informed written consent was obtained from all the patients in studies I, II, and IV or next of kin in study III. Inclusion criteria in studies I, II, and IV were as follows: age over 18 years, hearing threshold < 40 dB (study I and II), no history of neurologic or psychiatric disorders, no insulin-dependent diabetes mellitus, left ventricular ejection fraction < 40%, and no alcohol abuse or use of drugs affecting CNS. In study III, there were no exclusion criteria; all patients who were admitted to the ICU for the treatment of refractory status epilepticus were included. In studies I and II, only CABG patients were included. Study IV had the same inclusion criteria as I and II with the addition of cardiac valve surgery patients. Altogether, 103 patients were enrolled in the study, and after exclusions, data from 88 patients was analysed. The characteristics of the studies are shown in Table 12.

Table 12. The characteristics of the study.

Study I Study II Study III Study IV

Procedure Sedation after CABG Anaesthesia for

CABG Anaesthesia for RSE

Anaesthesia for cardiac surgery

Monitoring MLAEP MLAEP and BIS EEG and BIS Entropy and BIS

Patients included (n) 30 28 12 33

Patients excluded (n) 8 4 2 1

Evaluated patients (n) 22 24 10 32

Age, range (years) 41-75 47-79 32-66 46-78

Anaesthesia propofol propofol, isoflurane, alfentanil

propofol propofol, sufentanil

46

4.2 ANAESTHESIA, SEDATION PROTOCOL AND CARDIOPULMONARY BYPASS

All cardiac surgery patients from studies I, II, and IV received premedication of diazepam at 0.2 mg/kg before surgery in addition to their routine β-blocker and nitrate medications. Anaesthesia induction and maintenance protocol for cardiac surgery was standardised in three studies (I, II, IV). For the induction of anaesthesia propofol (1.0-1.5 mg/kg), midazolam (0.1mg/kg), pancuronium (0.1 mg/kg) and alfentanil (100 μg/kg or 50 μg/kg) were used in studies I and II, and sufentanil (1.5 μg/kg) in the study IV. The TIVA was maintained with continuous infusion of propofol (2-8 mg/kg/hour) and opioids, alfentanil 50 μg/kg/hour (Study I, II) and sufentanil 1.5 μg/kg/hour (Study IV). For the infusion of alfentanil , an open-loop plasma target controlled infusion pump system (RUGLOOP®, Ghent University, Ghent, Belgium) programmed via a microcomputer to control the infusion pump (JMS Syringe pump, SP-100, Hiroshima, Japan) was used in this study. The target plasma concentration of alfentanil was 300 ng mL -1

Isoflurane was given to one of two sub-groups of patients in study II at a minimum end-tidal concentration of 0.2% to achieve a BIS value below 60. During the CPB the isoflurane was delivered by a gas vaporiser connected to the fresh gas flow of the cardiopulmonary bypass machine with a minimum end-tidal concentration of also 0.2%.

In study IV, the level of the neuromuscular blockade was tested with the train-of-four simulation mode, and the target was one, with an additional dose of pancuronium at 0.02 mg/kg.

For the sedation protocol in study I, continuous infusion of propofol was used with an initial dose of 2 mg/kg/h, and the level of sedation was assessed with a Ramsay sedation score. Study III applied a predefined sedation protocol for RSE treatment in the ICU. Anaesthesia was induced via a bolus dose of 2-3 mg/kg of propofol until clinical seizures stopped. After the patient was intubated and mechanical ventilation was established, the continuous infusion of propofol at 4 mg/kg/h was started, and the bolus of 1-2 mg/kg was given every 3-5 min. until a BS pattern was achieved in the cEEG. Thereafter, the dose of propofol was maintained and continued for 12 hours. If the burst suppression pattern was not maintained, a bolus of propofol was given (1 mg/kg), and the infusion rate was increased by 1 mg/kg/h. After 12 hours of burst suppression in the EEG, the propofol infusion was tapered off by a 20% decrease every third hour. If epileptic activity again occurred in the EEG, the treatment protocol was started and maintained for another 12 hours. If needed, patients received oxycodone (0.05 mg/kg) intravenously for analgesia.

The CPB procedure was standardised under mild hypothermia at 33-34 ° C according to our hospital guidelines. Non-pulsatile pump flow and membrane oxygenation (D903,

47

Avant®, Dideco, Italy) were used for CPB. The pump flow rate was kept at 2.5 l�min-1�m -2

under moderate hypothermia at 34 °C, the mean arterial pressure was between 60 and 80 mmHg, and the patients received phenylephrine as needed. Myocardial protection during aortic cross clamping was achieved with a cold crystalloid cardioplegia solution infused intermittently and anterogradely. Systemic anticoagulation was produced with an initial dose of heparin at 3 mg/kg and with subsequent additional doses of heparin to maintain blood activated clotting time for a duration of > 480s. Afterwards, CPB heparin coagulation was neutralised with protamine sulphate 3 mg/kg.

For studies II and IV, all patients were interviewed on the first and second postoperative day in order to evaluate intraoperative awareness and recall. The interview featured a modified version of the Brice questionnaires (Ranta et al. 1996). Patients were asked the following five questions:

1. What is the last thing that you remember before going to sleep for your operation?

2. What is the first thing that you remember on waking up after your operation?

3. Do you remember anything in between?

4. Did you have any dreams during the operation?

5. What was the most unpleasant thing you remember from you operation and anaesthesia?

4.3 DATA COLLECTION

To record the MLAEPs, in the study I, the AEP/EEG module included in the anaesthesia CS/3 monitor by Datex – Ohmeda Division, Finland was used. The Ag/AgCl electrodes were placed on the scalp according to the international 10-20 system, and the electrode places Fz, Cz, A1, A2, C3, C4, P3, and P4 were used. EEG was recorded in a bipolar fashion between electrodes A1-Fz, A2-Cz, C3-P3, and C4-P4. MLAEP was recorded from A1 and A2, which were both referred to Cz. EEG, MLAEP, haemodynamic data, and clinical annotations (sedation level and nursing procedures during recordings) were stored in an IBIS workstation (Eindhoven, Belgium). The workstation was connected to AS/3 anaesthesia monitor, but not directly to the patient. Hearing function was tested in all patients with the audiometric test before baseline measurement. A two-channel MLAEP recording (Cz-M1 and Cz-M2) was performed using Ag/AgCl electrodes that were placed at the vertex (Cz) and on both mastoids (M1 and M2). The grounding electrode was placed on the forehead (Fpz). Binaural auditory stimulation was given at 70 dB above the hearing threshold using earphones with a stimulation rate of 7.1 Hz. The duration of each condensation click was 100 �s, the signal was digitised at 2400 Hz, and the band pass was filtered (0.5-1000 Hz). A total of two hundred responses were recorded and averaged. Electrode skin impedances were maintained below 5 kΏ. The AEP/EEG module has a built-

48

in artefact rejection algorithm that precludes the display of the AEP curves if the algorithm detects significant artefact. During the baseline measurement, the recording was performed in steady conditions (patients’ eyes closed, quiet room, and relaxed yet not allowed to fall asleep). The neuromonitoring measurements included the two EEG channel MLAEP recordings and were performed five times: the day before operation, one hour before operation after premedication, after operation in the ICU during deep level of sedation (Ramsay score 6), during moderate sedation (Ramsay score 4), and on the day after surgery. The identification of AEP peaks and measurements of latencies and amplitudes were performed by Heidi Yppärillä. Latencies of Na, Pa and Nb were marked as well as peak-to-peak amplitudes of Na-Pa and Pa-Nb by using the AEP score software by the Institute of Information Technology, Tampere, Finland. Picture 1 shows the Datex-Ohmeda module use in the study protocol.

Picture 1. 1. The Datex-Ohmeda EEG and AEP module used in the AS/3 anaesthesia monitor. 2.

Headbox ; 3. Leadset; 4. AEP Earphones. The measurement electrodes were placed at Cz

(central electrode) and over the left (A1) and right (A2) mastoid. Ground electrode is also

shown. Reprinted with the permission of General Electric, Healthcare, Helsinki, Finland.

In study II, the same EEG Datex – Ohmeda module from study I was used for MLAEP monitoring. Continuous BIS recording was started before the induction of anaesthesia using a BIS A-2000 XP monitor (Aspect Medical System�, Newton, MA, USA). Recording was continued until the operation ended. After skin preparation, a BIS electrode was placed on the patient’s forehead, and the BIS electrode impendence and signal quality were checked automatically so that the impedance was maintained below 3 kΩ. A 15-second

1

2 3 4

49

smoothing window with updates every 2 seconds was used. BIS values were stored on a personal computer (Hewlett Packard, USA). In the statistical analysis, we used the BIS mean values averaged over exactly the same period when the MLAEPs measurement was performed. The MLAEPs were obtained seven times: before induction of anaesthesia, 10 min after intubation, 30 min after sternotomy, during CPB at the cross-clamping of the aorta, during hypothermia, after de-clamping of the aorta, and after the operation. In the operating room, two anaesthetists and a research nurse were involved in clinical patient care and activities related to the study recordings. One anaesthetist was responsible for administering the anaesthetic and for neurophysiologic monitoring and measurements. The other anaesthetist was responsible for the cardiopulmonary bypass, and the research nurse ensured proper functioning of the microcomputers. BIS and MLAEP data were stored on a personal computer for off-line analysis. Heidi Yppärillä, who was blinded to the group allocation of patients, performed the MLAEPs data analysis as an off-line validation.

In study III, we evaluated the utility of the BIS monitor to measure the depth of the propofol coma during the treatment of refractory status epilepticus. We used the same BIS settings as in study II; the BIS index and SR were downloaded and stored every 5 seconds on computer’s hard disk. For the EEG recordings, we used a full-channel EEG digital device called the Grass-Telefactor Twin 2.6. A total of 192 hours of cEEG monitoring were recorded and analysed by an experienced clinical neurophysiologist (Esa Mervaala), who was blinded to the BIS index data. The EEG was divided into three periods: 1) the start of anaesthesia before the burst suppression pattern on the EEG (21 hours), 2) the burst suppression period on the EEG (126 hours), and 3) recovering from anaesthesia while tapering off propofol (45 hours). In each patient, the mean values of BIS index and SR were calculated for every hour during each period and these values were used for the analyses. In addition, 120 burst suppression periods of 3-6 minutes each were identified in the raw EEG in all patients (12 per patient in a one-hour interval), and the number of bursts per minute was measured. Altogether, 1506 bursts were identified and used in the comparison of BIS index value and suppression ratio with bursts per minute recorded by the EEG.

In the last study, the newest BIS monitor, BIS VISTA™, was used with the BIS Quattro sensor, which offered enhanced artefact and EMG filtering. For the entropy monitoring, the M-Entropy module of the AS/5 Anaesthesia monitor was used with an entropy sensor placed on the forehead just below the BIS sensor (Picture 2).

50

Picture 2. BIS and entropy sensors placed on the patient’s forehead before induction of

anaesthesia in the operating room. The entropy sensor is placed below the BIS sensor. The

picture is published with the patient’s permission.

The sensor impendence was checked automatically and kept under 7.5 kΏ. The average window for SE calculation was set at 15s. The set of data included SE, BIS, RE, BIS-EMG, and haemodynamic variables. These variables were all obtained and analysed 1 minute before and after the following time points: induction of anaesthesia, intubation, skin incision, sternotomy, cannulation of the aorta, cardiopulmonary bypass, cross clamping of the aorta, de-clamping of the aorta, and at the end of CPB. They were also gathered at the re-warming phase and at 10, 20, 30, and 40 min after. Neuromuscular blockade was tested with train-of-four (TOF) stimulation mode. Routine monitoring included the five-channel electrocardiogram, radial artery pressure, central venous pressure, capnography, pulse oximetry, central and peripheral body temperature, end-tidal carbon dioxide concentrations were continuously monitored and recorded during all studies.

4.4 STATISTICAL ANALYSIS

In study I, the non-parametric analysis of variance for repeated measures was used to detect significant changes in variables over time (Friedman test). If significant changes occurred, the data between different measurement points was compared with the Wilcoxon signed rank test with Bonferroni correction for multiple testing. The Pearson correlation was used to address the association between MLAEP parameters and Ramsay sedation score. In study II, the non-parametric analysis of variance for repeated measures was used to detect significant changes in variables over time between and within the groups. If significant changes occurred, differences between the groups at different time points were compared by using the Mann-Whitney U-test. Within the groups, differences between each time point were compared by using the Wilcoxon signed rank test with adjustment to the Bonferroni correction for multiple comparisons. In study III, data are reported as mean and at a 95% confidence interval. The abilities of the BIS index and the suppression ratio to

51

discriminate between the no-burst and burst suppression patterns in the EEG were evaluated using receiver operating characteristic (ROC) curves and the respective areas under the curve (AUC). Due to concerns that the data may not have been normally distributed, the non-parametric Spearman coefficient was used to test the correlation between variables, and the Wilcoxon test was used to compare data before and during a burst suppression pattern as well as before and during epileptic activity in EEG. In study IV, a difference of 10 units between the BIS index and entropy was considered clinically significant. Based on the pilot study (5 patients), in order to achieve a power of 90% with an α value of 0.05, a sample size of 32 patients was calculated as being appropriate. The assumption of data normality was tested using the Kolmogorov-Smirnov test. Normally distributed data are reported as mean and standard deviations. A Bland-Altman analysis and Pearson correlation coefficient were used to compare the two measurement systems. The differences before and after time points were compared using paired T-tests. To find the surgical stimulation’s effect on the EEG-parameters, EMG, and haemodynamic parameters, the Cohen´s d effect size test was used. A p value of < 0.05 was considered significant. All statistical tests were performed using the SPSS statistical software (SPSS Inc., Chicago, IL, U.S.A.).

52

5 Results

5.1 MLAEPS DURING ICU SEDATION

The data analysis features 110 neurophysiologic AEP measurements from 22 patients. During deep sedation, the Nb peak in the MLAEP waveforms was sometimes impossible to identify. In only 15 patients all components of MLAEP were identified at all-time points. The latency of the Nb component increased significantly from the baseline median of 44 ms (range 38-60 ms) to 49 ms (41-64 ms) after premedication. It increased to 63 ms (48-80 ms) during deep sedation (Ramsay 6). It also increased in the next day after surgery when patients awoke (49 ms [range 37-71 ms]). The Nb latency correlated well (r2 = 0.53; p= 0.01) with Ramsay scores after premedication, during deep, and under moderate sedation. The amplitude of Na-Pa was decreased at Ramsay 6 as compared to the baseline and light sedation (Ramsay 4). Figure 4A shows the baseline MLAEP measurement after auditory stimulation in the awake individual, and figure 4B during sedation in the ICU.

Figure 4A. The baseline (patient awake) MLAEP measurement to an auditory stimulus at the

left (dashes lines) and at the right ear (continuous line). The measurement electrodes were

placed at Cz (central electrode) and over the right (M2) and left (M1) mastoid. The identified

MLAEP-peaks Na, Pa, Nb and Pb peaks are labelled over the MLAEP-curve. The artefact arising

from the activation of musculus auricularis posterior is labelled as (I). Not published previously.

53

Figure 4B. The MLAEP measurements at baseline (awake), during moderate sedation (continuous line; Ramsay score 4) and during deep sedation (dashes line; Ramsay score 6). The patient is the same as in Figure 3A. The auditory stimulus was applied at the right ear (Cz-M2). With the deeping of sedation the latencies are increasing and the amplitudes are decreasing. Not published previously.

5.2 MLAEPS VERSUS BIS DURING ANAESTHESIA FOR CARDIAC SURGERY

In study II, 168 MLAEPs measurements were obtained in the 24 patients who were randomised to propofol (Group P; n=12) or isoflurane (Group I; n=12) maintained anaesthesias. Eight hundred and eleven AEP parameters were detected, and 29 AEP waveforms were excluded from the analysis, because we were not able to detect all MLAEP peaks during deep anaesthesia. Within the groups, significant changes over time were observed in BIS (p< 0.001), in the latencies of the Pa (p< 0.001) and Nb (p< 0.001) components, and in the amplitude of the Na-Pa (p< 0.001) components of MLAEP. The BIS values were similar in both groups at all-time points. Ten minutes after intubation and 30 minutes after sternotomy, the Nb latency was significantly increased in group I compared to group P. There were no differences between the groups in other measures of MLAEP at any time point. BIS increased significantly after intubation and skin incision, but there was no difference between the groups regarding the BIS values. No patient reported recall of intraoperative events.

54

5.3 BIS MONITORING DURING THE TREATMENT OF STATUS EPILEPTICUS IN ICU

The bispectral index (r2 = 0.9; p< 0.001) and suppression ratio (r2 = 0.88; p< 0.001) correlated well with the electroencephalogram burst rate per minute. The sensitivity and specificity of bispectral index score 30 for detecting burst suppression in EEG were 99% and 98%, respectively. The AUC for the BIS’s ability to detect burst suppression patterns in the EEG was 0.99. There was a significant difference in the bispectral index during anaesthesia before burst suppression (mean 65; [95% CI 64-66]; p< 0.001) and during the burst suppression pattern in electroencephalogram (mean 19; [95% CI 19-20]; p=0.001). The BIS value was significantly higher (p = 0.001) during epileptic activity in the EEG (mean 64; CI 53-74) than the value during the burst suppression pattern. In Figure 5, typical epileptic discharges in the EEG are seen in one patient admitted to the ICU before burst suppression pattern.

Figure 5. The raw 12-channel EEG waveforms of one patient with RSE admitted to the ICU for

propofol anaesthesia. In the 1-min EEG waveforms regional epileptic discharges are shown in

channels T3-PO3, AF3-C3 ja C3-T3. Not published previously.

55

5.4 BIS VERSUS SPECTRAL ENTROPY DURING ANAESTHESIA FOR CARDIAC SURGERY

The data used for Bland-Altman and correlation analysis included 704 paired values of BIS and spectral entropy. The mean difference between BIS and SE was significantly different from zero, being 2.14 (95% CI 1.59-2.67) with an upper limit of agreement of 16 and a lower limit of agreement of -11.8. The mean difference between BIS and RE was 0.02 (95% CI 0.01-0.06) with an upper limit of agreement of 14.3 and a lower limit of agreement of -14.2. The BIS and SE differed by more than 10 units in 16.6 % of measurements, BIS, and RE in 15.4%, respectively. The overall correlations during anaesthesia between BIS and SE (r2 = 0.66) and BIS and RE (r2 = 0.7) were good. All three EEG parameters, EMG, and blood pressure significantly increased after intubation, skin incision, and sternotomy. The effect of surgical stimulation on parameters, as measured by Cohen’s d, was highest for RE after skin incision (-0.71) and sternotomy (-0.91). Systolic and mean blood pressure increased after surgical stimulation. Heart rate did not change significantly at any time point during the study. There were no differences in TOF values between patients during the points of measurements. The cannulation of the aorta induced changes in all three indices. During the re-warming phase of the CPB, all three indices increased 10, 20, 30, and 40 minutes after the start of the patient’s re-warming. No patient reported any recall of intraoperative events.

56

6 Discussion

In the present study EEG-based monitoring during general anaesthesia for cardiac surgery as well as during ICU sedation and RSE treatment was evaluated.

6.1 ASSESSMENT OF SEDATION BY MLAEPS AFTER CARDIAC SURGERY

6.1.1 MLAEPs and anaesthesia

In the first study, the employment of auditory-evoked potentials in the assessment of postoperative sedation after cardiac surgery was investigated. The AEPs indicate the CNS’s responses to external stimulation (click auditory stimulus derived through headphones), whereas the EEG reflects only the resting level of the EEG signal (Thornton 1991). The AEP signals accomplish the criterion of monitoring anaesthetic depth: they show graded changes with increasing anaesthetic concentration and a pattern that is similar to the patterns created by different anaesthetic agents. AEPs also show changes after surgical stimulation and indicate awareness (Thornton et al. 1989a). In the present study, the MLAEPs were evaluated, because the brainstem auditory evoked potentials show no changes under some anaesthetic agents (Thornton 1991). The study was strictly neurophysiological, using full-range EEG monitoring and AEP analyses performed by a neurophysiologist. A pattern of graded changes in the MLAEPs with increases in the latencies of Na, Pa, and Nb components and a decrease in the amplitude of Na-Pa during deepening levels of sedation were observed.

The results supported that changes in the AEP amplitudes and latencies are dependent on the level of sedation and are parallel with changes during general anaesthesia. Similar changes were reported during anaesthesia with four different anaesthetics: halothane, enflurane (Thornton et al. 1984), isoflurane (Heneghan et al. 1987), and propofol (Thornton et al. 1989b, Iselin–Chaves et al. 2000). An opposite change was found with lightening of sedation and after regaining consciousness (Thornton 1991 and 1993). Na-Pa amplitude and Nb latency indicate cortical arousal in ICU patients during nursing care and physiotherapy (Rundshagen et al. 2000). In the present study only Na-Pa amplitude was able to distinguish between deep (Ramsay 6) and moderate (Ramsay 4) levels of sedation. On the first day after surgery, the latencies of the Pa and Nb waves stayed over their baseline measurements, even when the patients were awake. This might have been due to analgesic drugs used in the postoperative period and due to trace amounts of the anaesthetics and sedatives used for surgery. Because this increase in the latencies in the postoperative period has not been reported in non-cardiac surgery patients, squeal of CPB and brain oedema (Harris et al. 1993)) after CPB have to be taken into account.

57

6.1.2 MLAEPs and Ramsay score

The Ramsay sedation score was used to assess the level of sedation, and a good correlation (r2 = 0.53) was found between the score and Nb latency after premedication (Ramsay 2) and moderate and deep sedation (Ramsay 4 and 6, respectively). Schulte-Tamburen et al. (1999) compared five subjective sedation scoring systems with the Nb latency during long-term sedation in ICU patients. While all scores correlated fairly well with Nb latency (r2=0.6-0.7), the performance was best with the Ramsay score, which parallels the findings in this study.

6.1.3 MLAEPs after premedication

Benzodiazepine premedication increased the Nb latency significantly in our study. The effects of benzodiazepines on MLAEP have been studied in only a couple of trials. Schwender et al. (1993) examined the effects of intravenous induction of general anaesthesia with midazolam, diazepam, and flunitrazepam on MLAEP in patients scheduled for minor gynaecological procedures. In this study the MLAEPs did not change markedly in amplitudes or latencies during induction of general anaesthesia with benzodiazepines. Only the patient’s LOC with administration of midazolam was associated with an increase in mean Nb latency. The effect of bolus dose of midazolam on the AEPs was studied in the 9 patients (Brunner et al. 1999). Mean Nb latency in awake patients was 44.3 ms and when the eyelash reflex was lost, Nb latency was 55.7 ms (p< 0.001). Brunner et al. (2002) also studied the effect of midazolam (given before anaesthesia) on the amount of propofol needed to achieve LOC. The auditory evoked responses were compared in these patients with those in a group of control patients who were not given midazolam. Following LOC, Na-Pa amplitude decreased by 60% in the control group and by 54% in the midazolam group, whereas Nb latency increased by 24% in the control group and by 32% in the midazolam group. These differences were not statistically significant. In spite of the automatic AEP analysis techniques employed in the AEP monitors, the use of AEP monitoring during sedation and anaesthesia has yet to become popular and is not routinely used in clinical practice as of January 2013. The obstacles to its popularity are flattening of the waveforms during deep anaesthesia and sedation and thereby preventing the detection of all the MLAEP peaks, which makes the interpretation of data difficult. Otherwise, the AEPs are still a promising method of measuring external input impact on the EEG during anaesthesia and sedation, for which reason they are under active scientific research (Jindestål et al. 2011).

58

6.2 MLAEPS IN BISPECTRAL INDEX-GUIDED ANAESTHESIA DURING CARDIAC SURGERY

In the second study, the differences in the level of anaesthesia during cardiac surgery, as measured by AEP and BIS under two anaesthesia regimes, volatile-based and TIVA-based, were assessed. When BIS indicated a similar level of hypnosis with both anaesthesia techniques, there was a difference in both Nb and Pa latency of MLAEP during the pre-bypass period in the cardiac surgery patients. One explanation of our results is that propofol anaesthesia caused less suppression of the Nb and Pa latencies of MLAEP than isoflurane anaesthesia. Liu et al. (2005), who utilised somatosensory-evoked potentials (SSEP) (which cannot be used in the monitoring of anaesthesia depth), results are in parallel with the findings of the present study. In their study of SSEP during anaesthesia with propofol and isoflurane (which they monitored with BIS as in present study), the changes in the SSEP latency were similar to those observed in AEPs in the present trial. At an equivalent depth of hypnosis as indicated by BIS (BIS level 40-50), isoflurane caused a greater decrease than propofol in the amplitude of the SSEP, and the former also caused a greater increase in the latency of the SSEP. It is not known whether the differences in the Nb and Pa latencies between the groups are indicative of a difference in the state of anaesthesia during the early phases of the CABG procedure. Propofol with alfentanil and isoflurane with alfenatnil were compared. A number of previous studies (Iselin-Chaves et al. 1998, Glass et al. 1997) have revealed a good correlation between BIS and the level of sedation with propofol. The use of alfentanil in addition to propofol did not influence the sedation produced by propofol alone (Iselin-Chaves et al. 2000, Glass et al. 1997). A dynamic relationship between the end-tidal isoflurane concentration and BIS has been described (Olofsen and Dahan 1999). The interaction between opioids and propofol in reducing the Cp50 (the plasma concentration that will prevent a response in 50% of patients for a given stimulus) of propofol and the interaction between opioids and isoflurane in reducing minimum alveolar concentration has been studied (Glass 1995) and is very similar for both propofol and isoflurane and the opioids. Therefore an interaction between alfentanil and propofol or alfentanil and isoflurane may not be the explanation for observations in the present study. 6.2.1 MLAEPs and BIS after noxious stimulation

The first differences in the AEP values between the groups were noticed after induction of anaesthesia and endotracheal intubation. Shinner et al. (1999) studied the effect of bolus doses of alfentanil on AEP and the effect of response to intubation on AEPs. The Pa amplitude increased after intubation in the group that did not receive an alfentanil bolus before intubation. This result suggests that the effect of an antinociceptive drug on AEP could be reversed with intubation. In the present study, BIS increased significantly for a short period after intubation, but there were no significant differences in BIS between the groups. BIS was below 60 while MLAEP measurements were taken. However, the changes caused by intubation persisted in MLAEP and actually overlapped the baseline values with

59

a wide inter-individual variability. As indicated with BIS, there was a constant and similar level of hypnosis in both groups, and no BIS values overlapped with pre-induction values at any time. This may imply that MLAEPs seem to indicate other components of anaesthesia, rather than only the state of hypnosis after induction of anaesthesia.

An explanation for why the Nb latency decreased at 10 minutes after intubation in the propofol group may be that lightening in the level of anaesthesia occurred in some patients due to stimulation after intubation in spite of the constant level of hypnosis that was indicated by BIS. The MLAEP may have still been reflecting that fact when recordings were taken. That not all patients responded to intubation is reflected in the wide variation in Nb latency at this stage. The ability of the AEP to discriminate the response to a noxious stimulus between the groups is explained by the fact that it reflects not only cortical, but also sub-cortical brain activities. Increase in heart rate in the propofol group after intubation may also indicate stress response to the noxious stimulus. The second measurement, in which we found differences in both Nb and Pa latencies between the groups, was performed after surgical stimulation (sternotomy). BIS increased after sternotomy in both groups but was below 60 during the MLAEP measurements. Thornton and colleagues (1988) showed that the effect of anaesthesia on AEP could be reversed by pain stimuli. In addition, De Beer et al. (1996b) reported that after sternotomy, Na and Nb amplitudes were higher in responsive patients and concluded that AEP is more sensitive than the spectral features of the EEG to pain stimuli. The results of present study seem to be in line with those of the study by Bonhomme et al. (2006a), where during a constant level of hypnosis (as monitored by BIS), AEP responses to surgical stimulation were significantly different according to the analgesic regime. When the BIS was constant at a level of 50, they observed an increase in the A-line AAI index after surgical stimulation in the patient group that received only normal saline via the epidural catheter as compared to the group that received ropivacaine. The AAI index is based on parameters derived from MLAEP and is a linear mapping of the MLAEP amplitudes and latencies. Bonhomme et al. concluded that the combined use of two different electrophysiological monitors (AEP and BIS) provided information about the anti-nociceptive component of anaesthesia. 6.2.2 TIVA versus volatile anaesthesia

When interpreting results of the present study, it is crucial to note that BIS and AEPs measure different aspects of brain activity. The AEPs measure the output of the CNS to a controlled input, and thus AEP and BIS may contain complementary information. Because BIS only measures cortical function, it is only able to monitor the sedative or hypnotic state, whereas the AEP provides information about the function of the brainstem as well as sub-cortical and cortical components, making AEP better at tracking the overall anaesthetic state, including anti-nociception/nociception and probability of response to pain (Bonhomme et al. 2006a). BIS values are indicators of hypnotic effects and are poor indicators of responsitivity to noxious stimuli (Mi et al. 1998).

60

Positron emission tomography (PET) has also been used in studies on the effect of anaesthesia. Studies with PET have shown differences in regional brain activity between propofol and volatile anaesthetics (Kaisti et al. 2002, Jeong et al. 2006). Considering that propofol and isoflurane have different regional effects on the brain and that BIS and MLAEPs measure different aspects of brain activity, it was not surprising that we found different responses in the indices between the two anaesthesia methods. In the excellent study by Långsjö et al. (2012) the arousal-induced brain activations after propofol and dexmedetomidine anaesthesia was visualised with PET. Emergence of consciousness, assessed by the motor response to a spoken command, was associated with activation of deep brain structures involving subcortical and limbic regions functionally connected with parts of frontal and inferior parietal cortices. The emergency of propofol anaesthesia, which was also used in our study, showed even less neocortical activation compared with dexmedetomidine results. The BIS monitor, which is based on EEG measurement, provides information only on brain cortical function, whereas the AEP measure the output of the EEG to the stimulation and thus information from deeper parts of the brain (brainstem and subcortical components). According the results of Långsjö’s study the deep, phylogenetically old brain structures, are involved during arousal, which can be better traced by AEP measurements than BIS. The Nb latency of MLAEP was different between the groups before CPB, but not during or after CPB. One explanation for this finding could be that a more stable state of anaesthesia had been achieved. Alternative explanations are that many confounding factors, such as extracorporeal circulation, hypothermia, and altered cerebral blood flow, may interfere with the action of anaesthetics during CPB. In the present study, the neurophysiological parameters (i.e., the AEP and BIS) were obtained by using a separate electrode system for each. The system used for measuring the amplitudes and latencies of the AEP utilised methodology similar to that used also by many (Shinner et al. 1999, Iselin-Chaves et al. 2000), whereas others have used different, more novel methodology, such as the AAI index or the auditory evoked potential index (Doi et al. 1997b, Struys et al. 2002). The AEP/2 monitor, which is available on the market, uses the AAI index. The AAI index combines information from two different sources: MLAEPs and EEG. Active measurement is based on MLAEP and passive measurement during deep anaesthesia is based on spontaneous EEG. The present study was designed as a neurophysiologic study, with particular interest in AEP waveforms instead of one index number based on a mixed value of all the components of AEP or a combined index obtained from AEP and spontaneous EEG. The results of this study also supported the hypothesis that haemodynamic parameters like blood pressure and heart rate do not correlate with arousal or the patients’ consciousness level. There were no differences between groups in heart rate and blood pressure at any measurement time point, probably due to the presence of β-blockers and antihypertensive drugs in all study patients. According to results of present study, the parallel use of different anaesthesia monitors can provide better information of the global level of anaesthesia, including the

61

analgesic component. Our results also supported the proposition that volatile anaesthetic affects the CNS in different ways than intravenous propofol (Campagna et al. 2003).

6.3 BISPECTRAL INDEX MONITORING DURING THE TREATMENT OF STATUS EPILEPTICUS

The third study investigated the use of BIS in the treatment of refractory status epilepticus under propofol anaesthesia. The main finding of this study was that BIS monitoring can be used to guide the depth of anaesthesia in the treatment of RSE in times when cEEG monitoring is not available, such as during “out-of-office” hours and weekends. The values of BIS and BIS-suppression ratio (SR) correlated extremely well with the burst rate per minute as monitored by cEEG. The results demonstrate that a BIS index value of less than 30 corresponds to a burst suppression pattern in the EEG with a sensitivity of 99% and a specificity of 98%. If the BIS index is between 0 and 30, the number of bursts per minute in the EEG is between 0 and 10. There is no consensus in the literature regarding the appropriate number of bursts per minute when aiming at burst suppression in the treatment of RSE (Krishnamurthy and Drislane 1999), but usually the target is 3-5 bursts per minute (Riker et al. 2003, Cottenceau et al. 2008), which corresponds to a BIS number of around 15. Even though BIS monitoring can be used to guide the depth of anaesthesia, EEG monitoring will remain the primary method for monitoring treatment response, and it is absolutely essential in the definitive diagnosis of subtle status epilepticus, non-convulsive status epilepticus, and after monitoring with BIS index in cases where there is no full and rapid recovery.

6.3.1 BIS and burst suppression in the EEG

Riker at al. (2003) compared cEEG with the BIS monitor in the treatment of elevated intracranial pressure with pentobarbital infusion. A burst suppression pattern of 3-5 bursts/minute in EEG corresponded to a mean BIS index of 15 and a SR of 71%. They concluded that the BIS index and SR values correlated well with the number of bursts/minute on the raw EEG. The mean BIS index was 18 when there was a burst suppression pattern with a SR of 56%. In the present study, five bursts/minute corresponded to a mean BIS index of 15 and a suppression ratio of 61% (95% CI 60-63), and the mean BIS index was 19 during a burst suppression pattern with the mean SR value being 54%. The results of present study bear similarities with those by Riker, suggesting that monitoring the depth of anaesthesia during a burst suppression pattern in EEG with the BIS index are independent of which intravenous anaesthetic is being used (thiopental versus propofol).

62

Tobias (2008) evaluated the association of the bispectral index and SR with a burst suppression pattern in the EEG of patients with elevated intracranial pressure treated with pentobarbital coma. He demonstrated that a BIS index between 10 and 20 corresponded to 3-5 bursts/min in the EEG with a sensitivity of 96% and specificity of 92%. The sensitivity and specificity of SR in the range 65% to 85% were 89% and 88%, respectively. In addition, Cottenceau et al. (2008) demonstrated that a BIS index in the range between 6 and 15 predicted a burst rate of 2-5/min in EEG in brain injury patients with intracranial hypertension undergoing barbiturate coma.

One case report (Jaggi et al. 2003) described a patient with generalised status epilepticus treated with a barbiturate-induced coma and monitored with BIS and cEEG. A good correlation between the BIS index and the SR with bursts/min in EEG was noted. The mean hourly BIS was seven (95% CI 4-10) with a SR value of 85% (95% CI 78-92) when bursts were less than 5/min. BIS and SR correlated linearly when BIS values were in the range between 0-30, which is in agreement with our findings. It appears that the BIS becomes totally controlled by SR when the SR increases above a critical threshold of 50%. In healthy volunteers, SR values over 40% have been shown to correlate linearly with BIS index values between 0 and 30 during propofol infusion (Bruhn et al. 2000).

In the study evaluating the usefulness of the BIS monitor to detect BS during barbiturate-induced coma in children, the average correlation (r2=0.5) between the BS of the BIS and suppression ratio of EEG was only moderate (Prins at al. 2007). However, when the BIS-EEG traces were visually inspected, it was found that they did correspond rather well with real time EEG in all patients.

6.3.2 BIS and epileptiform activity in the EEG

A study on the effect of sevoflurane anaesthesia on regional blood flow in the brain reported two cases of EEG verified epileptiform activity (Kaisti et al. 1999). During deep sevoflurane anaesthesia, the EEG displayed a burst suppression pattern, and the BIS decreased to zero. Subsequently, the EEG recording revealed a rhythmic epileptiform discharge that was fully compatible with the clinical seizure type (partial motor seizure). During the epileptiform discharges, the BIS value increased markedly from zero to 44 in one patient and up to 73 in another patient. In the case report by Guerrero et al. (2008), a patient with epilepsy received vagal neurostimulation under general anaesthesia. The EEG on the BIS monitor screen abruptly changed from low-frequency waves to high-voltage, high-frequency waves, suggestive of epileptic activity, which was followed by increase in the values of the BIS index.

In present study, when the patient experienced epileptiform activity during burst suppression in the EEG, the value of the BIS increased from under 30 to a mean of 64 (95% CI 53-74). This phenomenon can reflect sudden transitions between low amplitude and slow, negative-wave EEG activity, which are typical of propofol-induced burst suppression

63

and abnormal epileptiform EEG (slow 1.5-2.5 Hz high-voltage spike-and-wave complex). During sevoflurane anaesthesia, Särkelä et al. (2007) have reported that significantly higher BIS values during epileptic discharges were recorded (median 57; interquartile range 46-69) compared with the BIS values during burst suppression pattern in EEG (median 25; interquartile range 7-45).

In one of patients in the present study, the bursts observed during a suppression EEG pattern were epileptic with spikes. The BIS algorithm has been developed in healthy volunteers and cannot necessarily be extrapolated to situations with different pathological conditions. The BIS monitor also includes the real time raw EEG waveforms, but distinguishing between epileptic and normal bursts on the small EEG screen is difficult. It was not the objective of our study to trace the bursts suppression on the BIS monitor EEG screen but probably is not a problem to see BS pattern on the screen. For the epileptic spikes we calculated the BIS/SR ratio, for example for 3 bursts/minute. The BIS/SR ratio was higher in the patient with epileptic bursts (0.52) compared to that patient’s periods with a non-epileptic burst (0.08). What can be the reason for this? During deep anaesthesia, episodes of suppression alternate with high frequency, high amplitude electrical activity (bursts). Epileptic bursts usually have higher frequencies and amplitudes, which may account for the higher BIS values and lower SR values than during non-epileptic burst suppression. High frequency bursts increase the high frequency components of EEG similarly to the effects seen during light anaesthesia (Bruhn et al. 2000). Clinicians require the services of a neurophysiologist to undertake the diagnosis of these epileptic bursts from the full-channel EEG. In addition, in two patients in the present study, a burst suppression EEG pattern was observed, but a regional epileptic discharge was seen in the full-channel EEG. BIS monitoring records frontal brain structures and does not necessarily react to the changes in other parts of the brain (Hamada et al. 2008). Therefore further EEG analyses are necessary in RSE.

6.4 COMPARISON OF BIS VISTA™ AND ENTROPY MONITORING DURING CARDIAC SURGERY

Bispectral index and spectral entropy are the two methods for monitoring the depth of anaesthesia during cardiac surgery. In fourth study of this dissertation, the agreement between those two methods was poor. Agreement was considered to be good if the values were in the range between 10 units above and below the mean difference (Bonhomme and Hans 2007). In the current study, a disagreement between BIS and SE of more than 10 units was observed in 16.6% of the measurements, with an upper limit of agreement of 16 and a lower limit of agreement of -11. The disagreement between BIS and RE of more than 10 units was observed in 15.4% of the measurements. We consider this degree of agreement to be unsatisfactory. It is clear that these two monitors of the depth of anaesthesia record different levels of hypnosis in the same patient.

64

6.4.1 Correlation and agreement

The results of current study are in parallel with (2006b) study by Bonhomme et al. (2006b which found a good global correlation between BIS and SE (0.87), but a poor agreement during propofol anaesthesia, with differences of more than 20 units. Also, in a study with sufentanil-sevoflurane–nitrous oxide anaesthesia (Leffol-Masson et al. 2007), the Bland–Altman analysis revealed a bias of -2 between BIS and SE with a limit of agreement in the range between -18 and 9. On the other hand, in the study by Iannuzzi et al. (2005), the Bland-Altman analysis between BIS and SE revealed a good comparability between the two indices with a mean difference of only 0.1. In that study, BIS and SE differed by more than 20% in only 4.3% of over 140 measurements as compared to 704 in the current study. However, agreement between BIS and SE was good in their awake patients and in their paralysed patients, though during the other conditions, differences of more than 20 units were observed, which is in agreement with findings of the present study. The agreement between BIS and SE was moderate in 102 patients receiving propofol-sufentanil anaesthesia and was affected by patients age (Aime et al. 2012).

In patients undergoing CPB (Maybohm et al. 2010), a good agreement between BIS and state entropy was observed during normothermic conditions but not under hypothermic conditions. The decrease in SE and RE was more pronounced than the corresponding decline in BIS index values during the cooling phase of CPB. In present study, remarkably higher correlations between BIS and entropy parameters and better agreement between BIS and SE were observed under hypothermic conditions compared with the findings of Maybohm. There are some differences between studies that can explain the results. First, milder hypothermia (33-35°C) was used in the present study than in that by Maybohm et al. (31-34°C); in their study, the difference between BIS and entropy parameters was largest during the lowest temperature. Second, there were more patients and more paired values that were analysed in the present study, than in the study by Maybohm et al. which may affect results. In addition, the decreased bias between BIS and SE during CPB with mild hypothermia compared to normothermia before CPB may be due to more stable anaesthesia without major surgical stimulation. Moreover, the BIS monitors used were different from those used by Maybohm et al. (VISTA™ vs. BIS XP™, respectively).

The explanation of these discrepancies may lie in the differences in the algorithms used in BIS and entropy monitors or in the time delay in the index calculation. Vakkuri et al. (2004) reported that entropy parameters and BIS show no linear correlation outside the BIS range 40-60, which parallels findings in the present study. The entropy monitor uses the same algorithm to calculate the indices throughout all phases of general anaesthesia, but the BIS algorithm is calculated from the BetaRatio algorithm in light sedation, from SyncFastFlow (bispectral component) during a surgical level of anaesthesia (BIS 40-60), and from the Burst Suppression Ratio/Quazi-algorithm (BIS under 40; Rampil 1998a). The entropy index is based on physical principals. Entropy is independent of the frequency and amplitude

65

scales of EEG, which are patient dependent. Entropy reflects the irregularities in the EEG signal. In the entropy monitor, the SE index data was obtained from a 15-second EEG window. The BIS EEG length of the smoothing window was also 15s (Freye and Levy 2005, Vakkuri et al. 2004). The data was averaged over one minute, which should attenuate the effect of time delay on agreement (Bonhomme et al. 2006b), and thus the time delay of index calculation does not account for findings in this study.

6.4.2 EMG activity, BIS and RE

In the present study, the relationship between BIS and RE was even better than between BIS and SE. This may be because there is an effect of EMG activity on calculations of the BIS, even in the BIS VISTA™ monitor, which should eliminate artefacts from EMG activity. In Feld’s study (2006), the RE activation during anaesthesia was associated with an increase in BIS and EMG activity in the BIS monitor. The values of RE and BIS were highly correlated during the entire anaesthesia, in agreement with results in the current study. A strong correlation between BIS and frontal EMG was found in a study comparing BIS and spectral entropy during CABG surgery (Baulig et al. 2009). The limits of agreement between BIS and RE were also poor (+20.5/-24.1) in this study.

When the patient was conscious, there were differences between BIS and SE, but not between BIS and RE because of the different scale that SE uses. The BIS index scale is 0-100 and SE is 0-91, because SE measures the lower frequency of the EEG signal (32 Hz), whereas the RE utilises the lower frequency of the EEG plus higher frequency signal from the EMG, leading up to 47 Hz. Therefore, RE’s scale is the same as BIS (0-100). SE and RE declined more rapidly after induction of anaesthesia than did BIS. In this light state of anaesthesia, the characteristic increase in the β EEG activity (13-30 Hz) would be suggestive of arousal being detected by the monitoring. In the BIS algorithm, a β activity sub-variable “β ratio” is included, but the Entropy module does not have such a variable for high β activity. Spectral entropy monotonically declines with increasing anaesthetic drug concentration, and during induction of anaesthesia, it will decrease more rapidly than BIS (Vakkuri et al. 2004). Baulig et al. (2009) and Venluchene et al. (2004) noticed that RE and SE tended to be lower than BIS during general anaesthesia.

6.4.3 RE and noxious stimulation

In our study, RE was the best predictor of noxious stimulation compared to other EEG parameters, EMG, and haemodynamic variables. In a previous report (Wheeler et al. 2005), RE increased after pain stimulation, indicating arousal even in paralysed patients. The RE calculation of the Entropy module suggested a more rapid rise of the RE because of the EMG frequencies that update more rapidly than the SE (2 s versus 15 s, respectively). This outcome indicates that RE is an early indicator of frontal EMG activity caused by stimulation. It also suggests that frontal muscles may be more resistant to paralysis than the hand muscle where TOF is measured, because there were no differences between patients in TOF in the present study.

66

6.4.4 Entropy and muscle relaxation

Many studies suggest that muscle relaxation may confound entropy monitoring. In the study of Hans et al. (2006), a dose of rocuronium before laryngoscopy altered the RE and RE-SE gradient. In those patients who did not receive rocuronium, RE and RE-SE increased more markedly than in patients who received rocuronium after noxious stimulation. Takamatsu et al. (2006) described an increase in the RE-SE difference after noxious stimulation in patients undergoing gynaecological surgery. The RE-SE difference is an indicator for EMG activation during arousal, and RE becomes equal to SE when the EMG power (sum of spectral power between 32Hz and 47 Hz) is equal to zero (Viertiö-Oja et al. 2004). In the present study, the RE and RE-SE response to surgical stimulation persisted even at a surgical level of muscle relaxation, and the RE and SE-RE values were significantly higher after skin incision and sternotomy. This suggests that the RE reflects nociception in response to painful stimulation even in a paralysed patient, so it is useful for identifying inadequate analgesia.

The increase in RE in the present study was followed by a rapid increase in SE. The Cohen’s d effect size for the RE-SE difference was only moderate, probably caused by immediate rise in SE. In the study of Aho et al. (2009), the increase in the RE after nociceptive stimulation was followed by increased SE in patients not receiving neuromuscular blockers. These increases were associated with the appearance of EMG in the EEG raw signal. The recovery from succinylcholine paralysis increased RE and EMG but not the SE values (Baughman et al. 2005). Their findings and those of our study suggest the increase in the difference between RE and SE is not a good indicator of nociception.

Blood pressure and heart rate were poor indicators of painful stimulation; the heart rate did not significantly rise at any measurement time point. This result is probably attributable to the use of β-blockers in all patients. Intubation evoked an increase in all three indices, but BIS was more affected than SE and RE. BIS proved to be a good predictor of arousal after intubation stimulation, but BIS was found to be a poor indicator of nociception, as indicated by increased systolic blood pressure (Driessen et al. 1999).

This is the first study to reveal that aortic cannulation induced a decrease in the EEG-based hypnosis indices, such as BIS and spectral entropy. The manipulation of the aorta evokes a reduction in blood pressure, leading to reduced cerebral blood flow. EEG changes may occur in this phase of cardiac surgery, with progressive EEG slowing and resulting in a decline in the EEG-based indices without any changes in the level of hypnosis. A gradual increase in all three indices during the re-warming phase of CPB from moderate hypothermia back to 37°C was also observed. The increases in BIS, SE, and RE probably reflect the increase in the brain metabolic rate and propofol requirement during the re-warming phase and the loss of the CNS depressant effect that is caused by hypothermia. Maybohm et al. (2010) noticed an increase in the BIS and entropy indices during the re-warming phase of CPB despite an unchanged anaesthesia regimen.

67

6.5 LIMITATIONS OF THE STUDY

The present study has several limitations. In study I, neuropsychological evaluations of the patients were not performed before or after cardiac surgery to assess whether the observed changes in the MLAEPs after operation have clinical implications. Moreover, the propofol blood concentration during the different measurements was not assessed, nor the correlation of propofol concentration with the changes in the MLAEP parameters.

In study II, the BIS test variable was used as a clinical endpoint to achieve a standardised depth of anaesthesia. Ideally, it should not differ between the groups at any time point. This appears to result in an unintentional bias against AEP. A third variable as a control parameter could have been used to standardise the anaesthetic depth. However, it is of note that there is no generally accepted and validated gold standard to monitor all the components of anaesthesia for the type of patient group studied. There is also a lack of information about the plasma propofol concentrations in our patients at the time when we performed the AEP and BIS recordings. There is no data regarding MLAEPs immediately before or immediately after intubation or commencement of surgery. The MLAEP measurement after skin incision and sternotomy is difficult to obtain because of the liberal use of diathermy at that stage of surgery. Use of diathermy interferes with the recordings of AEPs and also BIS.

BIS monitoring has several limitations, especially regarding the monitoring of status epilepticus patients. General muscle activity or rigidity as well as head or body motion may cause artefacts during BIS monitoring. Artefacts and poor signal quality may be sources of unreliable BIS values. However, the signal quality was excellent during the BIS monitoring in our study. For about 75% per cent of the monitoring time, the mean SQI was above 80 (63-100), and during deep anaesthesia with burst suppression, the value was 80-100. If BIS values suddenly increase during treatment, we cannot be sure if this increase is due to lightening of anaesthesia or due to the presence of on-going epileptic activity in the EEG. Thus, cEEG is needed to diagnose possible epileptic activity, as proposed by Chamorro et al. (2008). During burst suppression, bursts may be epileptic and cannot necessarily be diagnosed solely by BIS monitoring. In one of our patients, the bursts observed during a suppression EEG pattern were epileptic with spikes. In addition, two other patients exhibited a burst suppression EEG pattern, but local epileptic discharge was seen in the full-channel EEG. Therefore, further EEG analyses are necessary to refine RSE. The sample of present study is small, but the results are unambiguous and clear.

When comparing two different measurement methods in the study IV, they should ideally use the same units. Unfortunately, there are no exact details of how the BIS algorithm is calculated, since the algorithm is not freely available. In this study plasma concentrations of propofol at different time points were not measured.

68

6.6 GENERAL CONCLUSIONS

In spite of the rapid development in neuromonitoring in this field, there is still no gold standard for monitoring the depth of anaesthesia, and especially sedation during ICU treatment. There are many different monitors of anaesthetic hypnosis components available commercially, but only the bispectral index and spectral entropy are well evaluated and investigated. Spectral entropy is also a very promising monitoring technique, and it offers the possibility to monitor the analgesic component of anaesthesia assessment through response entropy. The anaesthetists who assess the depth of hypnosis should analyse information not only from EEG-based monitors but also from haemodynamic signs like blood pressure and pulse frequency. The level of muscular block should be taken into account. Many studies have shown that the level of muscular block can interfere with the values of EEG-based monitoring (Hans et al. 2006, Aho et al. 2011).

The neuronal mechanism of human consciousness formulation in the brain is not well known in spite of active research. The findings of the Långsjö et al. (2012) study were the first to reveal that not only cortical brain structures are involved in the human consciousness and arousal, but also deep subcortical structures. Maybe this is a reason why BIS monitoring, based on the cortex EEG, failed in the prevention of awareness during anaesthesia (Avidan et al. 2011). According to the new study (Långsjö et al. 2012) on the neuronal mechanism of human consciousness and arousal during the anaesthesia, we need new research in this field.

The anaesthesiologist has new tools based on EEG-based monitoring to measure the hypnotic level of anaesthesia and to improve the patient care, recovery, and outcome by keeping the anaesthesia stable, avoiding excessively deep levels of hypnosis, and reducing the incidence of perioperative awareness. In spite of ambivalences, the results of some studies (Avidan et al. 2008 and 2011) as well as the existing literature as a whole suggest that the level of hypnosis during general anaesthesia at least in patients with a high risk of awareness should be guided by EEG-based monitoring (Apfelbaum et al. 2006)). EEG-based monitoring also helps clinicians to understand better the effect of anaesthetic action on humans and the relationship between different aspects of anaesthesia (hypnosis, antinociception, and immobility). Future large trials are necessary to evaluate information from different monitors during different levels of hypnosis.

There is still no good monitor for ICU sedation, for a light state of sedation with a tranquil patient (and co-operative patient, if the procedure requires the patient’s involvement). Available anaesthesia monitors have not been shown to be useful during ICU sedation because of the artefacts from the EMG during the analysis of the EEG signal. On the other hand, the BIS monitor was successfully used in the present study, during deep propofol-induced anaesthesia in the RSE treatment in the ICU. The development and research of new monitors only for sedation monitoring, or with improved performance during the sedation monitoring, is required.

69

7 Conclusions

1. The latencies and the amplitudes of the MLAEPs are suitable to be used for objective monitoring of short-term sedation after CABG surgery.

2. Parallel use of two EEG-based monitoring methods, BIS and AEPs, can show differences in the components of anaesthesia between various anaesthesia regimes in cardiac surgical patients.

3. Continuous EEG can be considered as the primary monitoring technique in the assessment of the depth and effect of anaesthesia in the treatment of RSE. If cEEG is not available, a BIS monitor can be used to guide the level of anaesthesia in the treatment of RSE with propofol.

4. Agreement was poor between the BIS index and SE values measured by Bland-Altman analysis at critical anaesthesia time points in patients who were undergoing cardiac surgery. RE was a good predictor of arousal after surgical stimulation regardless of the surgical level of muscle relaxation. Index differences most likely resulted from different algorithms for calculating consciousness levels.

70

8 References

Agarwal J, Puri GD, Mathew PJ. Comparison of closed loop vs. manual administration of propofol using the Bispectral index in cardiac surgery. Acta Anaesthesiol Scand 2009;53:390-397.

Ahlgren E, Aren C. Cerebral complication after coronary artery bypass and heart valve surgery: risk factors and onset of symptoms. J Cardiothorac Vasc Anesth 1998;12:270-273.

Aho AJ, Yli-Hankala A, Lyytikäinen LP, Jäntti V. Facial muscle activity, Response Entropy, and State Entropy indices during noxious stimuli in propofol-nitrous oxide or propofol-nitrous oxide-remifentanil anaesthesia without neuromuscular block. Br J Anaesth 2009;102:227-233.

Aho AJ, Lyytikäinen LP, Yli-Hankala A, Kamata K, Jäntti V. Explaining Entropy responses after a noxious stimulus, with or without neuromuscular blocking agents, by means of the raw electroencephalographic and electromyographic characteristics. Br J Anaesth 2011;106:69-76.

Ahonen J, Jokela R, Uutela K, Huiku M. Surgical stress index reflects surgical stress in gynaecological laparoscopic day-case surgery. Br J Anaesth 2007;98:456-461.

Ahonen J, Olkkola K, Hynynen M, Seppälä T, Ikävalko H, Remmerie B, Salmenperä M. Comparison of alfentanil, fentanyl and sufentanil for total intravenous anaesthesia with propofol in patients undergoing coronary artery bypass surgery. Br J Anaesth 2000;85:533- 540.

Aimé I, Verroust N, Masson-Lefoll C, Taylor G, Laloë PA, Liu N, Fischler M. Does monitoring bispectral index or spectral entropy reduce sevoflurane use? Anesth Analg 2006;103:1469-1477.

Aimé I, Gayat E, Fermanian C, Cook F, Peuch C, Laloë PA, Mantz J, Fischler M. Effect of age on the comparability of bispectral and state entropy indices during the maintenance of propofol-sufentanil anaesthesia. Br J Anaesth 2012;108:638-643.

Anderson RE, Jakobsson JG. Entropy of EEG during anaesthetic induction: a comparative study with propofol or nitrous oxide as sole agent. Br J Anaesth 2004;92:167-170.

Apfelbaum JL, James F. Arens JF, Cole DJ, Connis RT, Domino KB, Drummond JC, Kalkman CJ, Miller RD, Nickinovich DG,Todd MM. American Society of Anesthesiologists Task Force on Intraoperative Awarenes. Practice advisory for intraoperative awareness and brain function monitoring: a report by the American Society of Anesthesiologists task force on intraoperative awareness. Anesthesiology 2006;104:847-864.

Arbour R, Waterhouse J, Seckel MA, Bucher L. Correlation between the Sedation-Agitation Scale and the Bispectral Index in ventilated patients in the intensive care unit. Heart Lung 2009;38:336-345.

71

Arroliga A, Frutos-Vivar F, Hall J, Esteban A, Apezteguía C, Soto L, Anzueto A. International Mechanical Ventilation Study Group. Use of sedatives and neuromuscular blockers in a cohort of patients receiving mechanical ventilation. Chest 2005;128:496-506.

Avidan MS, Zhang L, Burnside BA, Finkel KJ, Searleman AC, Selvidge JA, Saager L, Turner MS, Rao S, Bottros M, Hantler C, Jacobsohn E, Evers AS. Anesthesia awareness and the bispectral index. N Engl J Med 2008;358:1097-1108.

Avidan MS, Jacobsohn E, Glick D, Burnside BA, Zhang L, Villafranca A, Karl L, Kamal S, Torres B, O'Connor M, Evers AS, Gradwohl S, Lin N, Palanca BJ, Mashour GA. BAG-RECALL Research Group. Prevention of intraoperative awareness in a high-risk surgical population. N Engl J Med 2011;365:591-600.

Barnard JP, Bennett C, Voss LJ, Sleigh JW. Can anaesthetists be taught to interpret the effects of general anaesthesia on the electroencephalogram? Comparison of performance with the BIS and spectral entropy. Br J Anaesth 2007;99:532-537.

Bashein G, Nessly ML, Bledsoe SW, Townes BD, Davis KB, Coppel DB, Hornbein TF. Electroencephalography during surgery with cardiopulmonary bypass and hypothermia. Anesthesiology 1992 ;76:878-891.

Bauer M, Wilhelm W, Kraemer T, Kreuer S, Brandt A, Adams HA, Hoff G, Larsen R. Impact of bispectral index monitoring on stress response and propofol consumption in patients undergoing coronary artery bypass surgery. Anesthesiology 2004;101:1096-1104.

Baughman VL, Hoffman WE, Koenig HM, Wheeler PL, Ananda RC, Wang M. Recovery from paralysis with succinylcholine increased response entropy and EMG but not the state entropy. J Clin Monit Comput 2005;19:201-205.

Baulig W, Seifert B, Schmid ER, Schwarz U. Comparison of spectral entropy and bispectral index electroencephalography in coronary artery bypass graft surgery. J Cardiothorac Vasc Anesth 2009;24:544-549.

Bailey JM, Mora CT, Shafer SL. Pharmacokinetics of propofol in adult patients undergoing coronary revascularization. The Multicenter Study of Perioperative Ischemia Research Group. Anesthesiology 1996;84:1288-1297.

Bonhomme V, Hans P. Monitoring depth of anaesthesia: is it worth the effort? Eur J Anaesthesiol 2004;21:423-428.

Bonhomme V, Llabres V, Dewandre PY, Brichant JF, Hans P. Combined use of Bispectral Index and A-Line Autoregressive Index to assess anti-nociceptive component of balanced anaesthesia during lumbar arthrodesis. Br J Anaesth 2006;96:353-360 (a).

Bonhomme V, Deflandre E, Hans P. Correlation and agreement between bispectral index and state entropy of the electroencephalogram during propofol anaesthesia. Br J Anaesth 2006;97:340-346 (b).

Bonhomme V, Desiron Q, Lemineur T, Brichant JF, Dewandre PY, Hans P. Bispectral index profile during carotid cross clamping. J Neurosurg Anesthesiol 2007;19:49-55 (a).

72

Bonhomme VL, Hans PC. Bispectral index and state entropy of the electroencephalogram during propofol anaesthesia (letter). Br J Anaesth 2007;98:145-146 (b).

Bonhomme V, Uutela K, Hans G, Maquoi I, Born JD, Brichant JF, Lamy M, Hans P. Comparison of the surgical Pleth Index™ with haemodynamic variables to assess nociception-anti-nociception balance during general anaesthesia. Br J Anaesth 2011;106:101-111.

Brice DD, Hetherington RR, Utting JE. A simple study of awareness and dreaming during anaesthesia. Br J Anaesth 1970;42:535-542.

Brook AD, Ahrens TS, Schaiff R, Prentice D, Sherman G, Shannon W, Kollef MH. Effect of a nursing-implemented sedation protocol on the duration of mechanical ventilation. Crit Care Med 1999;27:2609-2615.

Brown EN, Lydic R, Schiff ND. General anesthesia, sleep, and coma. N Engl J Med 2010;363:2638-2650.

Bruhn J, Bouillon TW, Shafer SL. Bispectral index (BIS) and burst suppression: revealing a part of the BIS algorithm. J Clin Monit Comput 2000;16:593-596.

Bruhn J, Myles PS, Sneyd R, Struys MMRF. Depth of anaesthesia monitoring: what’s available, what’s validated and what`s next? Br J Anaesth 2006;97:85-94.

Brunner MD, Umo-Etuk J, Sharpe RM, Thornton C. Effect of a bolus dose of midazolam on the auditory evoked response in humans. Br J Anaesth 1999;82:633-634.

Brunner MD, Nel MR, Fernandes R, Thornton C, Newton DE. Auditory evoked response during propofol anaesthesia after pre-induction with midazolam. Br J Anaesth 2002;89:325-327.

Campagna JA, Miller KW, Forman SA. Mechanisms of actions of inhaled anesthetics. N Engl J Med 2003;22:2110-2124.

Chamorro C, Romera MA, Balandín B, Valdivia M. Nonconvulsive status and bispectral index. Crit Care Med 2008;36:2218-2219.

Chen X, Thee C, Gruenewald M, Wnent J, Illies C, Hoecker J, Hanss R, Steinfath M, Bein B. Comparison of surgical stress index-guided analgesia with standard clinical practice during routine general anesthesia: a pilot study. Anesthesiology 2010;112:1175-1183.

Chen JWY, Wasterlain CG. Status epilepticus: pathophysiology and management in adults. Lancet Neuro 2006;5:246-256.

Cheng DC, Karski J, Peniston C, Raveendran G, Asokumar B, Carroll J, David T, Sandler A. Early tracheal extubation after coronary artery bypass graft surgery reduces costs and improves resource use. A prospective, randomized, controlled trial. Anesthesiology 1996;85:1300-1310 (a).

73

Cheng DC, Karski J, Peniston C, Asokumar B, Raveendran G, Carroll J, Nierenberg H, Roger S, Mickle D, Tong J, Zelovitsky J, David T, Sandler A. Morbidity outcome in early versus conventional tracheal extubation after coronary artery bypass grafting: a prospective randomized controlled trial. J Thorac Cardiovasc Surg 1996;112:755-764 (b).

Cheng DC. Fast-track cardiac surgery: economic implications in postoperative care. J Cardiothorac Vasc Anesth 1998;12:72-79.

Chinzei M, Sawamura S, Hayashida M, Kitamura T, Tamai H, Hanaoka K. Change in Bispectral Index During Epileptiform Electrical Activity Under Sevoflurane Anesthesia in a Patient with Epilepsy. Anesth Analg 2004;98:1734-1736.

Chortkoff BS, Eger EI, Crankshaw DP, Gonsowski CT, Dutton RC, Ionescu P. Concentrations of desflurane and propofol that suppress response to command in humans. Anesth Analg 1995;81:737-743.

Claassen J, Hirsch LJ, Emerson RG, Mayer SA. Treatment of Refractory Status Epilepticus with Pentobarbital, Propofol, or Midazolam: A Systematic Review. Epilepsia 2002;43:146-153.

Cottenceau V, Petit L, Masson F, Guehl D, Asselineau J, Cochard J-F, Pinaquy C, Leger A, Sztrak F. The Used of Bispectral Index to Monitor Barbiturate Coma in Severely Brain-Injured Patients with Refractory Intracranial Hypertension. Anesth Analg 2008;107:1676-1682.

Dahaba AA. Different conditions that could result in the bispectral index indicating an incorrect hypnotic state. Anesth Analg 2005;101:765-773.

Dahaba AA, Bornemann H, Hopfgartner E, Ohran M, Kocher K, Liebmann M, Wilfinger G, Metzler H. Effect of sugammadex or neostigmine neuromuscular block reversal on bispectral index monitoring of propofol/remifentanil anaesthesia. Br J Anaesth 2012;108:602-606.

Davydow DS, Gifford JM, Desai SV, Needham DM, Bienvenu OJ. Posttraumatic stress disorder in general intensive care unit survivors: a systematic review. Gen Hosp Psychiatry 2008;30:421-434.

De Beer NA, van Hooff JC, Brunia CH, Cluitmans PJ, Korsten HH, Beneken JE. Midlatency auditory evoked potentials as indicators of perceptual processing during general anaesthesia. Br J Anaesth 1996;77:617-624 (a).

De Beer NA, van Hooff JC, Cluitmans PJ, Korsten HH, Grouls RJ. Haemodynamic responses to incision and sternotomy in relation to the auditory evoked potential and spontaneous EEG. Br J Anaesth 1996;76:685-693 (b).

De Deyne C, Struys M, Decruyenaere J, Creupelandt J, Hoste E, Colardyn F. Use of continuous bispectral EEG monitoring to assess depth of sedation in ICU patients. Intensive Care Med 1998;24:1294-1298.

74

De Jonghe B, Cook D, Appere-De-Vecchi C, Guyatt G, Meade M, Outin H. Using and understanding sedation scoring systems: a systematic review. Intensive Care Med 2000;26:275-285.

De Jonghe B, Cook D, Griffith L, Appere-de-Vecchi C, Guyatt G, Théron V, Vagnerre A, Outin H. Adaptation to the Intensive Care Environment (ATICE): development and validation of a new sedation assessment instrument. Crit Care Med 2003;31:2344-2354.

de Lemos J, Tweeddale M, Chittock D. Measuring quality of sedation in adult mechanically ventilated critically ill patients. the Vancouver Interaction and Calmness Scale. Sedation Focus Group. J Clin Epidemiol 2000;53:908-919.

DeLorenzo RJ, Garnett LK, Towne AR, Waterhouse EJ, Boggs JG, Morton L, Choudhry MA, Barnes T, Ko D. Comparison of status epilepticus with prolonged seizure episodes lasting from 10 to 29 minutes. Epilepsia 1999;40:164-169.

Deogaonkar A, Vivar R, Bullock RE, Price K, Chambers I, Mendelow AD. Bispectral index monitoring may not reliably indicate cerebral ischaemia during awake carotid endarterectomy. Br J Anaesth 2005;94:800-804.

Doi M, Gajraj RJ, Mantzaridis H, Kenny GN. Effects of cardiopulmonary bypass and hypothermia on electroencephalographic variables. Anaesthesia 1997;52:1048-1055 (a).

Doi M, Gajraj RJ, Mantzaridis H, Kenny GN. Relationship between calculated blood concentration of propofol and electrophysiological variables during emergence from anaesthesia: comparison of bispectral index, spectral edge frequency, median frequency and auditory evoked potential index. Br J Anaesth 1997;78:180-184 (b).

Domino KB, Posner KL, Caplan RA, Cheney FW. Awareness during anesthesia: a closed claims analysis. Anesthesiology 1999;90:1053-1061.

Dowd NP, Cheng DC, Karski JM, Wong DT, Munro JA, Sandler AN. Intraoperative awareness in fast-track cardiac anesthesia. Anesthesiology 1998;89:1068-1073.

Driessen JJ, Harbers JBM, van Egmond J, Booij LHDJ. Evaluation of the electroencephalographic bispectral index during fentanyl-midazolam anaesthesia for cardiac surgery. Does it predict haemodynamic responses during endotracheal intubation and sternotomy? Eur J Anaesthesiol 1999;16:622-627.

Drislane FW, Blum AS, Lopez MR, Gautam S, Schomer DL. Duration of refractory status epilepticus and outcome: Loss of prognostic utility after several hours. Epilepsia 2009;50:1566-1571.

Edmonds HL, Rodrigues RA, Audenaert SM, Austin EH, Pollock SB, Ganzel BL. The role of neuromonitoring in cardiovascular surgery. J Cardiothorac Vasc Anesth 1996;10:15-23.

Ellerkmann RK, Liermann VM, Alves TM, Wenningmann I, Kreuer S, Wilhelm W, Roepcke H, Hoeft A, Bruhn J. Spectral entropy and bispectral index as measures of the electroencephalographic effects of sevoflurane. Anesthesiology 2004;101:1275-1282.

75

Ellerkmann RK, Soehle M, Alves TM, Liermann VM, Wenningmann I, Roepcke H, Kreuer S, Hoeft A, Bruhn J. Spectral entropy and bispectral index as measures of the electroencephalographic effects of propofol. Anesth Analg 2006;102:1456-1462.

Ekman A, Lindholm ML, Lennmarken C, Sandin R. Reduction in the incidence of awareness using BIS monitoring. Acta Anaesthesiol Scand 2004;48:20-26.

Ely EW, Inouye SK, Bernard GR, Gordon S, Francis J, May L, Truman B, Speroff T, Gautam S, Margolin R, Hart RP, Dittus R. Delirium in mechanically ventilated patients: validity and reliability of the confusion assessment method for the intensive care unit (CAM-ICU). JAMA 2001;286:2703-2710.

Ely EW, Truman B, Shintani A, Thomason JW, Wheeler AP, Gordon S, Francis J, Speroff T, Gautam S, Margolin R, Sessler CN, Dittus RS, Bernard GR. Monitoring sedation status over time in ICU patients: reliability and validity of the Richmond Agitation-Sedation Scale (RASS). JAMA 2003;289:2983-2991.

Estruch-Pérez MJ, Barberá-Alacreu M, Ausina-Aguilar A, Soliveres-Ripoll J, Solaz-Roldán C, Morales-Suárez-Varela MM. Bispectral index variations in patients with neurological deficits during awake carotid endarterectomy. Eur J Anaesthesiol 2010;27:359-363 (a).

Estruch-Pérez MJ, Ausina-Aguilar A, Barberá-Alacreu M, Sánchez-Morillo J, Solaz-Roldán C, Morales-Suárez-Varela. Bispectral index changes in carotid surgery. Ann Vasc Surg 2010;24:393-399 (b).

Feld J, Hoffman W, Park H. Response entropy is more reactive than bispectral index during laparoscopic gastric banding. J Clin Monit Comput 2006;20:229-234.

Fraser GL, Prato BS, Riker RR, Berthiaume D, Wilkins ML. Frequency, severity, and treatment of agitation in young versus elderly patients in the ICU. Pharmacotherapy 2000;20:75-78.

Freye E, Levy JV. Cerebral monitoring in the operating room and the intensive care unit: an introductory for clinician and guide for the novice wanting to open a window to the brain. J Clin Monit Comput 2005;19:1-76.

Gan TJ, Glass PS, Windsor A, Payne F, Rosow C, Sebel P, Manberg P. Bispectral index monitoring allows faster emergence and improved recovery from propofol, alfentanil, and nitrous oxide anesthesia. BIS Utility Study Group. Anesthesiology 1997;87:808-815.

Ge SJ, Zhuang XL, Wang YT, Wang ZD, Li HT. Changes in the rapidly extracted auditory evoked potentials index and the bispectral index during sedation induced by propofol or midazolam under epidural block. Br J Anaesth 2002;89:260-264.

Gibbs FA, Gibbs LE, Lennox WG. Effects on the electroencephalogram of certain drugs which influence nervous activity. Arch Intern Med 1937;60:154-156.

Girard TD, Kress JP, Fuchs BD, Thomason JW, Schweickert WD, Pun BT, Taichman DB, Dunn JG, Pohlman AS, Kinniry PA, Jackson JC, Canonico AE, Light RW, Shintani AK, Thompson JL, Gordon SM, Hall JB, Dittus RS, Bernard GR, Ely EW. Efficacy and safety of a

76

paired sedation and ventilator weaning protocol for mechanically ventilated patients in intensive care (Awakening and Breathing Controlled trial): a randomised controlled trial. Lancet 2008;371:126-134.

Glass PS. Pharmacokinetic and pharmacodynamic principles in providing "fast-track" recovery. J Cardiothorac Vasc Anesth 1995;9:16-20.

Glass PS, Bloom M, Kearse L, Rosow C, Sebel P, Manberg P. Bispectral analysis measures sedation and memory effects of propofol, midazolam, isoflurane, and alfentanil in healthy volunteers. Anesthesiology 1997;86:836-847.

Ghoneim MM, Block RI, Haffarnan M, Mathews MJ. Awareness during anesthesia: risk factors, causes and sequelae: a review of reported cases in the literature. Anesth Analg 2009;108:527-535.

Groesdonk HV, Pietzner J, Borger MA, Fassl J, Haentschel D, Paarmann H, Ender J. The incidence of intraoperative awareness in cardiac surgery fast-track treatment. J Cardiothorac Vasc Anesth 2010;24:785-789.

Gruenewald M, Zhou J, Schloemerkemper N, Meybohm P, Weiler N, Tonner PH, Scholz J, Bein B. M-Entropy guidance vs standard practice during propofol-remifentanil anaesthesia: a randomised controlled trial. Anaesthesia 2007;62:1224-1229.

Gugino LD, Aglio LS, Yli-Hankala A. Monitoring the electroencephalogram during bypass procedures. Semin Cardiothorac Vasc Anesth 2004;8:61-83

Guedel A. Inhalation anesthesia. A fundamental guide. New York. Macmillan Company. 1937.

Guerrero JL, Alsina E, Del Blanco B, Matute E. Clinical utility of the bispectral index for detecting epileptic seizures. Rev Esp Anestesiol Reanim 2008;55:127-128.

Hadzidiakos D, Petersen S, Baars J, Herold K, Rehberg B. Comparison of a new composite index based on midlatency auditory evoked potentials and electroencephalographic parameters with bispectral index (BIS) during moderate propofol sedation. Eur J Anaesthesiol 2006;23:931-936.

Haenggi M, Yppärilä H, Hauser K, Caviezel C, Korhonen I, Takala J, Jakob SM. The effects of dexmedetomidine/remifentanil and midazolam/remifentanil on auditory-evoked potentials and electroencephalogram at light-to-moderate sedation levels in healthy subjects. Anesth Analg 2006;103:1163-1169.

Haenggi M, Yppärilä-Wolters H, Bieri C, Steiner C, Takala J, Korhonen I, Jakob SM. Entropy and bispectral index for assessment of sedation, analgesia and the effects of unpleasant stimuli in critically ill patients: an observational study. Crit Care 2008;12(5):R119. Epub ahead of print.

Haenggi M, Yppärilå-Wolters H, Hauser K, Caviezel C, Takala J, Korhonen I, Jakob SM Intra- and inter-individual variation of BIS-index and Entropy during controlled sedation with midazolam/remifentanil and dexmedetomidine/remifentanil in healthy volunteers: an interventional study. Crit Care 2009;13(1):R20. Epub ahead of print (a).

77

Haenggi M, Yppärilä-Wolters H, Buerki S, Schlauri R, Korhonen I, Takala J, Jakob SM. Auditory event-related potentials, bispectral index, and entropy for the discrimination of different levels of sedation in intensive care unit patients. Anesth Analg 2009;109:807-816 (b).

Hamada S, Laloë P-A, Hausser-Hauw C, Fischler M. Seizure After Aortic Clamp Release: A Bispectral Index Pitfall. 2008. J Cardiothorac Vasc Anesth 22:119-121.

Hammaren E, Yli-Hankala A, Rosenberg PH, Hynynen M. Cardiopulmonary bypass-induced changes in plasma concentrations of propofol and in auditory evoked potentials. Br J Anaesth 1996;77:360-364.

Hans P, Giwer J, Brichant JF, Dewandre P-Y, Bonhomme V. Effect of an intubation dose of rocuronium on Spectral Entropy and Bispectral Index™ responses to laryngoscopy during propofol anaesthesia. Br J Anaesth 2006;97:842-847.

Hans P, Verscheure S, Uutela K, Hans G, Bonhomme V. Effect of a fluid challenge on the Surgical Pleth Index during stable propofol-remifentanil anaesthesia. Acta Anaesthesiol Scand 2012; Jan. Epub ahead of print.

Hansen-Flaschen JH, Brazinsky S, Basile C, Lanken PN. Use of sedating drugs and neuromuscular blocking agents in patients requiring mechanical ventilation for respiratory failure. A national survey. JAMA 1991;266:2870-2875.

Harris DN, Bailey SM, Smith PL, Taylor KM, Oatridge A, Bydder GM. Brain swelling in first hour after coronary artery bypass surgery. Lancet 1993;4:586-587.

Hauser WA. Status epilepticus: epidemiologic considerations. Neurology 1990;40:9-13.

Hayashi K, Mita K, Sawa T. Electroencephalographic changes in the late cardiopulmonary bypass period are not reflected in the bispectral index. Clin Neurophysiol 2010;121:1198-1204.

Hayashida M, Sekiyama H, Orii R, Chinzei M, Ogawa M, Arita H, Hanaoka K, Takamoto S. Effects of deep hypothermic circulatory arrest with retrograde cerebral perfusion on electroencephalographic bispectral index and suppression ratio. J Cardiothorac Vasc Anesth 2007;21:61-67.

Hemmerling TM, Charabati S, Zaouter C, Minardi C, Mathieu PA. A randomized controlled trial demonstrates that a novel closed-loop propofol system performs better hypnosis control than manual administration. Can J Anaesth 2010;57:725-735.

Heneghan CP, Thornton C, Navaratnarajah M, Jones JG. Effect of isoflurane on the auditory evoked response in man. Br J Anaesth 1987;59:277-282.

Hernández-Gancedo C, Pestaña D, Pérez-Chrzanowska H, Martinez-Casanova E, Criado A. Comparing Entropy and the Bispectral index with the Ramsay score in sedated ICU patients. J Clin Monit Comput 2007;21:295-302.

Hett DA, Smith DC, Pilkington SN, Abbot TR. Effect of temperature and cardiopulmonary bypass on the auditory evoked response. Br J Anaesth 1995;75:293-296.

78

Hirschi M, Meistelman C, Longrois D. Effects of normothermic cardiopulmonary bypass on bispectral index. Eur J Anaesthesiol 2000;17:499-505.

Huiku M, Uutela K, van Gils M, Korhonen I, Kymäläinen M, Meriläinen P, Paloheimo M, Rantanen M, Takala P, Viertiö-Oja H, Yli-Hankala A. Assessment of surgical stress during general anaesthesia. Br J Anaesth 2007;98:447-455.

Hynynen M, Korttila K, Wirtavuori K, Lehtinen AM. Comparison of alfentanil and fentanyl as supplements to induction of anaesthesia with thiopentone. Acta Anaesthesiol Scand 1985;29:168-174.

Hynynen M, Takkunen O, Salmenperä M, Haataja H, Heinonen J. Continuous infusion of fentanyl or alfentanil for coronary artery surgery. Plasma opiate concentrations, haemodynamics and postoperative course. Br J Anaesth 1986;58:1252-1259 (a).

Hynynen M, Lehtinen AM, Salmenperä M, Fyhrquist F, Takkunen O, Heinonen J. Continuous infusion of fentanyl or alfentanil for coronary artery surgery. Effects on plasma cortisol concentration, beta-endorphin immunoreactivity and arginine vasopressin. Br J Anaesth 1986;58:1260-1266 (b).

Hynynen M, Olkkola KT, Näveri E, Palojoki R, Neuvonen PJ, Heinonen J. Thiopentone pharmacokinetics during cardiopulmonary bypass with a nonpulsatile or pulsatile flow. Acta Anaesthesiol Scand 1989;33:554-560.

Hynynen M, Hammarén E, Rosenberg PH. Propofol sequestration within the extracorporeal circuit. Can J Anaesth 1994;41:583-588 (a).

Hynynen M, Hynninen M, Soini H, Neuvonen PJ, Heinonen J. Plasma concentration and protein binding of alfentanil during high-dose infusion for cardiac surgery. Br J Anaesth 1994 ;72:571-576 (b).

Hänel F, von Knobelsdorff G, Werner C, Schulte am Esch J. Hypercapnia prevents jugular bulb desaturation during rewarming from hypothermic cardiopulmonary bypass. Anesthesiolog 1998;89:19-22.

Iannuzzi M, Iannuzzi E, Rossi F, Berrino L, Chiefari M. Relationship between Bispectral Index, electroencephalographic state entropy and effect-site EC50 for propofol at different clinical endpoints. Br J Anaesth 2005;94:492-495.

Ilies C, Gruenewald M, Ludwigs J, Thee C, Höcker J, Hanss R, Steinfath M, Bein B. Evaluation of the surgical stress index during spinal and general anaesthesia. Br J Anaesth 2010;105:533-537.

Iselin-Chaves IA, Flaishon R, Sebel PS Howell S, Gan TJ, Sigl J, Ginsberg B, Glass PS. The effect of the interaction of propofol and alfentanil on recall, loss of consciousness, and the bispectral index. Anesth Analg 1998;87:949-955.

Iselin-Chaves IA, El Moalem HE, Gan TJ, Ginsberg B, Glass PS. Changes in the auditory evoked potentials and the bispectral index following propofol or propofol and alfentanil. Anesthesiology 2000;92:1300-1310.

79

Jacobi J, Fraser GL, Coursin DB, Riker RR, Fontaine D, Wittbrodt ET, Chalfin DB, Masica MF, Bjerke HS, Coplin WM, Crippen DW, Fuchs BD, Kelleher RM, Marik PE, Nasraway SA Jr, Murray MJ, Peruzzi WT, Lumb PD. Task Force of the American College of Critical Care Medicine (ACCM) of the Society of Critical Care Medicine (SCCM), American Society of Health-System Pharmacists (ASHP), American College of Chest Physicians. Clinical practice guidelines for the sustained use of sedatives and analgesics in the critically ill adult. Crit Care Med 2002;30:119-141.

Jaggi P, Schwabe MJ, Gill K, Horowitz IN. Use of an Anesthesia Cerebral Monitor Bispectral Index to Assess Burst-Suppression in Pentobarbital Coma. Pediat Neurol 2003;28:219-222.

Jameson LC, Sloan TB. Using EEG to monitor anesthesia drug effects during surgery. J Clin Monit Comput 2006;20:445-472.

Jakob SM, Lubszky S, Friolet R, Rothen HU, Kolarova A, Takala J. Sedation and weaning from mechanical ventilation: effects of process optimization outside a clinical trial. J Crit Care 2007;22:219-228.

Jeong YB, Kim JS, Jeong SM, Park JW, Choi IC. Comparison of the effects of sevoflurane and propofol anaesthesia on regional cerebral glucose metabolism in humans using positron emission tomography. J Int Med Res 2006;34:374-384.

Jette N, Claassen J, Emerson RG, Hirsch LJ. Frequency and predictors of nonconvulsive seizures during continuous electroencephalographic monitoring in critically ill children. Arch Neurol 2006;63:1750-1755.

Jildenstål PK, Hallén JL, Rawal N, Gupta A, Berggren L. Effect of auditory evoked potential-guided anaesthesia on consumption of anaesthetics and early postoperative cognitive dysfunction: a randomised controlled trial. Eur J Anaesthesiol 2011;28:213-219.

Johansen JW. Update on Bispectral Index monitoring. Best Pract Research Clin Anaesth 2006;1:81-99.

John ER, Prichep LS. The anesthetic cascade: a theory of how anesthesia suppresses consciousness. Anesthesiology 2005;102:447-471.

Jones J, Konieczko K. Hearing and memory in anaesthetised patients. Br Med J 1986;292:1291-1293.

Jäntti V, Yli-Hankala A. Correlation of instantaneous heart rate and EEG suppression during enflurane anaesthesia: synchronous inhibition of heart rate and cortical electrical activity? Electroencephalogr Clin Neurophysiol 1990;76:476-479.

Kaisti KK, Jääskeläinen SK, Rinne JO, Metsähonkala L, Scheinin H. Epileptiform Discharges during 2 MAC Sevoflurane Anesthesia in Two Healthy Volunteers. Anesthesiology 1999;91:1952-1955.

Kaisti KK, Metsähonkala L, Teräs M, Oikonen V, Aalto S, Jääskeläinen S, Hinkka S, Scheinin H. Effects of surgical levels of propofol and sevoflurane anesthesia on cerebral blood flow in healthy subjects studied with positron emission tomography. Anesthesiology 2002;96:1358-1370.

80

Kaskinoro K, Maksimow A, Långsjö J, Aantaa R, Jääskeläinen S, Kaisti K, Särkelä M, Scheinin H. Wide inter-individual variability of bispectral index and spectral entropy at loss of consciousness during increasing concentrations of dexmedetomidine, propofol, and sevoflurane. Br J Anaesth 2011;107:573-580.

Karamchandani K, Rewari V, Trikha A, Batra RK. Bispectral index correlates well with Richmond agitation sedation scale in mechanically ventilated critically ill patients. J Anesth 2010;24:394-398.

Kawaguchi M, Takamatsu I, Kazama T. Rocuronium dose-dependently suppresses the spectral entropy response to tracheal intubation during propofol anaesthesia. Br J Anaesth 2009;102:667-672.

Kertai MD, Pal N, Palanca BJ, Lin N, Searleman SA, Zhang L, Burnside BA, Finkel KJ, Avidan MS. B-Unaware Study Group Association of perioperative risk factors and cumulative duration of low bispectral index with intermediate-term mortality after cardiac surgery in the B-Unaware Trial. Anesthesiology 2010;112:1116-1127.

Kertai MD, Whitlock EL, Avidan MS Review article: brain monitoring with electroencephalography and the electroencephalogram-derived bispectral index during cardiac surgery. Anesth Analg 2012;114:533-546.

Kollef MH, Levy NT, Ahrens TS, Schaiff R, Prentice D, Sherman G. The use of continuous i.v. sedation is associated with prolongation of mechanical ventilation. Chest 1998;114:541-518.

Korhonen I, Yli-Hankala A. Photoplethysmography and nociception. Acta Anaesthesiol Scand 2009;53:975-985.

Kortelainen J, Silfverhuth M, Suominen K, Sonkajarvi E, Alahuhta S, Jantti V, Seppanen T. Automatic classification of penicillin-induced epileptic EEG spikes. Conf Proc IEEE Eng Med Biol Soc 2010;6674-6677.

Kress JP, Pohlman AS, O’Connor MF, Hall JB. Daily interruption of sedative infusions in critically ill patients undergoing mechanical ventilation. N Engl J Med 2000;342:1471-1477.

Krishnamurty KB, Drislane FW. Depth of EEG Suppression and Outcome in Barbiturate Anesthetic Treatment for Refractory Status Epilepticus. Epilepsia 1999;40:759-762.

Kälviäinen R, Eriksson K, Parviainen I. Refractory Generalised Convulsive Status Epilepticus: A Guide to Treatment. CNS Drugs 2005;19:759-768.

Lamas A, López-Herce J, Sancho L, Mencía S, Carrillo A, Santiago MJ, Martínez V. Assessing sedation in critically ill children by bispectral index, auditory-evoked potentials and clinical scales. Intensive Care Med 2008;34:2092-2099.

Långsjö JW, Alkire MT, Kaskinoro K, Hayama H, Maksimow A, Kaisti KK, Aalto S, Aantaa R, Jääskeläinen SK, Revonsuo A, Scheinin H. Returning from oblivion: imaging the neural core of consciousness. J Neurosci 2012;32:4935-4943.

LeBlanc JM, Dasta JF, Kane-Gill SL. Role of the bispectral index in sedation monitoring in the ICU. Ann Pharmacother 2006;40:490-500.

81

Lefoll-Masson C, Fermanian C, Aimē I, Verroust N, Taylor G, Laloë PA, Liu N, Aegerter P, Fischler M. The comparability of bispectral index and state entropy index during maintenance of sufentanil-sevoflurane–nitrous oxide anesthesia. Anesth Analg 2007;105:1319-1325.

Legriel S, Mourvillier B, Bele N, Amaro J, Fouet P, Manet P, Hilpert F. Outcomes in 140 critically ill patients with status epilepticus. Intensive Care Med 2008;34:476-480.

Leira EC, Bertrand ME, Hogan ER, Cruz-Flores S, Wyrwich KW, Albaker OJ, Holzemer EM. Continuous or emergent EEG: can bedside caregivers recognize epileptiform discharges? Intensive Care Med 2004;30:207-212.

Leslie K, Myles PS, Forbes A, Chan MT. The effect of bispectral index monitoring on long-term survival in the B-aware trial. Anesth Analg 2010;110:816-822.

Levy WJ, Shapiro HM, Marachuk G, Meathe E. Automated EEG processing for the intraoperative monitoring: a comparison of techniques. Anesthesiology 1980;53:223-236.

Lindholm ML, Träff S, Granath F, Greenwald SD, Ekbom A, Lennmarken C, Sandin RH. Mortality within 2 years after surgery in relation to low intraoperative bispectral index values and preexisting malignant disease. Anesth Analg 2009;108:508-512.

Liu SS. Effects of Bispectral Index monitoring on ambulatory anesthesia: a meta-analysis of randomized controlled trials and a cost analysis. Anesthesiology 2004;101:311-315.

Liu N, Chazot T, Huybrechts I, Law-Koune J-D, Barvais L, Fischler M. The influence of a muscle relaxant on bispectral and Datex-Ohmeda entropy values during propofol-remifentanil induced loss of consciousness. Anesth Analg 2005;101:1713-1718.

Liu EH, Wong HK, Chia CP, Lim HJ, Chen ZY, Lee TL. Effects of isoflurane and propofol on cortical somatosensory evoked potentials during comparable depht of anaesthesia as guided by bispectral index. Br J Anaest 2005;94:193-197.

Liu N, Chazot T, Genty A, Landais A, Restoux A, McGee K, Laloë PA, Trillat B, Barvais L, Fischler M. Titration of propofol for anesthetic induction and maintenance guided by the bispectral index: closed-loop versus manual control: a prospective, randomized, multicenter study. Anesthesiology 2006;104:686-695.

Lowenstein E, Hallowell P, Levine FH, Daggett WM, Austen WG, Laver MB. Cardiovascular response to large doses of intravenous morphine in man. N Engl J Med 1969;18:1389-1393.

Lowenstein DH, Alldredge B. Managing status epilepticus. Lancet 1990;336:1451.

Laureys S. The neural correlate of (un)awareness: lessons from the vegetative state. Trends Cogn Sci 2005;9:556-559.

Lu CH, Man KM, Ou-Yang HY, Chan SM, Ho ST, Wong CS, Liaw WJ. Composite auditory evoked potential index versus bispectral index to estimate the level of sedation in paralyzed critically ill patients: a prospective observational study. Anesth Analg 2008;107:1290-1294 (a).

82

Lu CH, Ou-Yang HY, Man KM, Hsiao PC, Ho ST, Wong CS, Liaw WJ. Relative reliability of the auditory evoked potential and Bispectral Index for monitoring sedation level in surgical intensive care patients. Anaesth Intensive Care 2008;36:553-559 (b).

Lysakowski C, Dumont L, Pellegrini M, Clergue F, Tassonyi E. Effects of fentanyl, alfentanil, remifentanil and sufentanil on loss of consciousness and bispectral index during propofol induction of anaesthesia. Br J Anaesth 2001;86:523-527.

Lysakowski C, Elia N, Czarnetzki C, Dumont L, Haller G, Combescure C, Tramèr MR. Bispectral and spectral entropy indices at propofol-induced loss of consciousness in young and elderly patients. Br J Anaesth 2009;103:387-393.

Mahon P, Kowalski RG, Fitzgerald AP, Lynch EM, Boylan GB, McNamara B, Shorten GD. Spectral entropy as a monitor of depth of propofol induced sedation. J Clin Monit Comput 2008;22:87-93.

Maksimow A, Kaisti K, Aalto S, Mäenpää M, Jääskeläinen S, Hinkka S, Martens S, Särkelä M, Viertiö-Oja H, Scheinin H. Correlation of EEG spectral entropy with regional cerebral blood flow during sevoflurane and propofol anaesthesia. Anaesthesia 2005;60:862-869.

Maksimow A, Särkelä M, Långsjö JW, Salmi E, Kaisti KK, Yli-Hankala A, Hinkka-Yli-Salomäki S, Scheinin H, Jääskeläinen SK. Increase in high frequency EEG activity explains the poor performance of EEG spectral entropy monitor during S-ketamine anesthesia. Clin Neurophysiol 2006;117:1660-1668.

Maksimow A, Snapir A, Särkelä M, Kentala E, Koskenvuo J, Posti J, Jääskeläinen SK, Hinkka-Yli-Salomäki S, Scheinin M, Scheinin H. Assessing the depth of dexmedetomidine-induced sedation with electroencephalogram (EEG)-based spectral entropy. Acta Anaesthesiol Scand 2007;51:22-30.

Mangano DT, Siliciano D, Hollenberg M, Leung JM, Browner WS, Goehner P, Merrick S, Verrier E. Postoperative myocardial ischemia. Therapeutic trials using intensive analgesia following surgery. The Study of Perioperative Ischemia (SPI) Research Group. Anesthesiology 1992;76:342-345.

Martin JT, Faulconer A Jr, Bickford RG. Electroencephalography in anesthesiology. Anesthesiology 1959;20:359-376.

Mascia MF, Koch M, Medicis JJ. Pharmacoeconomic impact of rational use guidelines on the provision of analgesia, sedation, and neuromuscular blockade in critical care. Crit Care Med 2000;28:2300-2306.

Mashour GA, Esaki RK, Tremper KK, Glick DB, O'Connor M, Avidan MS. A novel classification instrument for intraoperative awareness events. Anesth Analg 2010;110:813-815.

Mathew JP, Weatherwax KJ, East CJ, White WD, Reves JG. Bispectral analysis during cardiopulmonary bypass: the effect of hypothermia on the hypnotic state. J Clin Anesth 2001;13:301-315.

83

Meybohm P, Gruenewald M, Höcker J, Renner J, Graesner J-T, Ilies C, Scholz J, Bein B. Correlation and agreement between the bispectral index vs. state entropy during hypothermic cardio-pulmonary bypass. Acta Anaesthesiol Scand 2010;54:169-175.

Meierkord H, Boon P, Engelsen B, Göcke K, Shorvon S, Tinuper P, Holtkamp M. EFNS guideline on the management of status epilepticus. Eur J Neurol 2006;13:445-450.

Meierkord H, Boon P, Engelsen B, Göcke K, Shorvon S, Tinuper P, Holtkamp M. European Federation of Neurological Societies (EFNS) guideline on the management of status epilepticus in adults. Eur J Neurol 2010;17:348-355.

Mi WD, Sakai T, Takahashi S, Matsuki A. Haemodynamic and electroencephalograph responses to intubation during induction with propofol or propofol/fentanyl. Can J Anaesth 1998;45:19-22.

Mi WD, Sakai T, Singh H, Kudo T, Kudo M. Hypnotic endpoints vs. bispectral index, 95% spectral edge frequency and median frequency during propofol infusion with or without fentanyl. Eur J Anaesthesiol 1999;16:47-52.

Moerman N, Bonke B, Oosting J. Awareness and recall during general anesthesia. Facts and feelings. Anesthesiology 1993;79:454-464.

Moller DH, Rampil IJ. Spectral entropy predicts auditory recall in volunteers. Anesth Analg 2008;106:873-879.

Monk TG, Saini V, Weldon BC, Sigl JC. Anesthetic management and one-year mortality after noncardiac surgery. Anesth Analg 2005;100:4-10.

Morimoto Y, Monden Y, Ohtake K, Sakabe T, Hagihira S. The detection of cerebral hypoperfusion with bispectral index monitoring during general anesthesia. Anesth Analg 2005;100:158-161.

Myles PS, Leslie K, McNeil J, Forbes A, Chan MT. Bispectral index monitoring to prevent awareness during anaesthesia: the B-Aware randomised controlled trial. Lancet 2004;29;1757-1763.

Möllhoff T, Herregods L, Moerman A, Blake D, MacAdams C, Demeyere R, Kirnö K, Dybvik T, Shaikh S. Remifentanil Study Group. Comparative efficacy and safety of remifentanil and fentanyl in 'fast track' coronary artery bypass graft surgery: a randomized, double-blind study. Br J Anaesth 2001;87:718-722.

Nasraway SA Jr, Jacobi J, Murray MJ, Lumb PD. Sedation, analgesia and neuromuscular blockade of critically ill adult: revised clinical practice guidelines for 2002. Crit Care Med 2002;30:117-118 (a).

Nasraway SA Jr, Wu EC, Kelleher RM, Yasuda CM, Donnelly AM. How reliable is the Bispectral Index in critically ill patients? A prospective, comparative, single-blinded observer study. Crit Care Med 2002;30:1483-1487 (b).

84

Nelskylä KA, Yli-Hankala AM, Puro PH, Korttila KT. Sevoflurane titration using bispectral index decreases postoperative vomiting in phase II recovery after ambulatory surgery. Anesth Analg 2001;93:1165-1169.

Nishiyama T, Matsukawa T, Hanaoka K. Is the ARX index a more sensitive indicator of anesthetic depth than the bispectral index during sevoflurane/nitrous oxide anesthesia? Acta Anaesthesiol Scand 2004;48:1028-1032.

Nishiyama T, Hanaoka K. The A-line ARX index may be a more sensitive detector of arousal than the bispectral index during propofol-fentanyl-nitrous oxide anesthesia: a preliminary investigation. Can J Anaesth 2004;51:539-544.

Nordström O, Engström AM, Persson S, Sandin R. Incidence of awareness in total i.v. anaesthesia based on propofol, alfentanil and neuromuscular blockade. Acta Anaesthesiol Scand 1997;41:978-984.

Nollert G, Shin'oka T, Jonas RA. Near-infrared spectrophotometry of the brain in cardiovascular surgery. Thorac Cardiovasc Surg 1998;46:167-175.

Ogilvie MP, Pereira BM, Ryan ML, Gomez-Rodriguez JC, Pierre EJ, Livingstone AS, Proctor KG. Bispectral index to monitor propofol sedation in trauma patients. J Trauma 2011;71:1415-1422.

Olofsen E, Dahan A. The dynamic relationship between end-tidal sevoflurane and isoflurane concentrations and bispectral index and spectral edge frequency of the electroencephalogram. Anesthesiology 1999;90:1345-1453.

Orser BA. Depth-of anesthesia monitor and the frequency of intraoperative awareness. N Engl J Med 2008;358:1189-1191.

Osorio I, Reed RC. Treatment of refractory generalized tonic-clonic status epilepticus with pentobarbital anesthesia after high-dose phenytoin. Epilepsia 1989;30:464-471.

Ostermann ME, Keenan SP, Seiferling RA, Sibbald WJ. Sedation in the intensive care unit. A systematic review. JAMA 2000;283:1451-1459.

Parviainen I, Uusaro A, Kälviäinen R, Kaukanen E, Mervaala E, Ruokonen E. High-dose thiopental in the treatment of refractory status epilepticus in intensive care unit. Neurology 2002;59:1249-1251.

Parviainen I, Uusaro A, Kälviäinen R, Mervaala E, Ruokonen E. Propofol in the treatment of refractory status epilepticus. Intensive Care Med 2006;32:1075-1079.

Payen JF, Chanques G, Mantz J, Hercule C, Auriant I, Leguillou JL, Binhas M, Genty C, Rolland C, Bosson JL. Current practices in sedation and analgesia for mechanically ventilated critically ill patients: a prospective multicenter patient-based study. Anesthesiology 2007;106:687-695.

Phillips AA, McLean RF, Devitt JH, Harrington EM. Recall of intraoperative events after general anaesthesia and cardiopulmonary bypass. Can J Anaesth 1993;40:922-926.

85

Plaschke K, Fichtenkamm P, Schramm C, Hauth S, Martin E, Verch M, Karck M, Kopitz J. Early postoperative delirium after open-heart cardiac surgery is associated with decreased bispectral EEG and increased cortisol and interleukin-6. Intensive Care Med 2010;36:2081-2089.

Pomfrett CJ, Sneyd JR, Barrie JR, Healy TE. Respiratory sinus arrhythmia: comparison with EEG indices during isoflurane anaesthesia at 0.65 and 1.2 MAC. Br J Anaesth 1994;72:397-402.

Pomfrett CJ. Heart rate variability, BIS and “depth of anaesthesia”. Br J Anaesth 1999;82:659-662.

Punjasawadwong Y, Boonjeungmonkol N, Phongchiewboon A. Bispectral index for improving anaesthetic delivery and postoperative recovery. The Cochrane Library 2010;1-70.

Prins SA, de Hoog M, Blok JH, Tibboel D, Visser GH. Continuous noninvasive monitoring of barbiturate coma in critically ill children using the Bispectral™ index monitor. Critical Care 2007;11:R108. Epub ahead of print.

Puri GD, Murthy SS. Bispectral index monitoring in patients undergoing cardiac surgery under cardiopulmonary bypass. Eur J Anaesthesiol 2003;20:451-456.

Rampil IJ. A primer for EEG signal processing in anesthesia. Anesthesiology 1998;89:980- 1002 (a).

Rampil IJ, Kim JS, Lenhardt R, Negishi C, Sessler DI. Bispectral EEG index during nitrous oxide administration. Anesthesiology 1998;8 9:671-677 (b).

Ramsay MAE, Savege TM, Simpson BRJ, Goodwin R. Controlled sedation with Alphaxalone- Alphadalone. Br Med J 1974;2:656-659.

Ranta S, Jussila J, Hynynen M. Recall of awareness during cardiac anaesthesia: influence of feedback information to the anaesthesiologist. Acta Anaesthesiol Scand 1996;40:554-560.

Ranta SO, Laurila R, Saario J, Ali-Melkkilä T, Hynynen M. Awareness with recall during general anesthesia: incidence and risk factors. Anesth Analg 1998;86:1084-1089.

Ranta SO-V, Herranen P, Hynynen M. Patients` conscious recollections from cardiac anesthesia. J Cardiothorac Vasc Anesth 2002;16:426-430.

Ranta S. Awareness with recall during general anesthesia. Academic dissertation. Helsinki University Publication 2002;104 p.

Rantanen M, Yli-Hankala A, van Gils M, Yppärilä-Wolters H, Takala P, Huiku M, Kymäläinen M, Seitsonen E, Korhonen I . Novel multiparameter approach for measurement of nociception at skin incision during general anaesthesia. Br J Anaesth 2006;96:367-376.

Rashkin MC, Youngs C, Penovich P. Pentobarbital treatment of refractory status epilepticus. Neurology 1987;37:500-503.

Recart A, Gasanova I, White PF, Thomas T, Ogunnaike B, Hamza M, Wang A. The effect of cerebral monitoring on recovery after general anesthesia: a comparison of the auditory

86

evoked potential and bispectral index devices with standard clinical practice. Anesth Analg 2003;97:1667-1674 (a).

Recart A, White PF, Wang A, Gasanova I, Byerly S, Jones SB. Effect of auditory evoked potential index monitoring on anesthetic drug requirements and recovery profile after laparoscopic surgery: a clinical utility study. Anesthesiology 2003;99:813-818 (b).

Riad W, Schreiber M, Saeed AB. Monitoring with EEG entropy decreases propofol requirement and maintains cardiovascular stability during induction of anaesthesia in elderly patients. Eur J Anaesthesiol 2007;24:684-688.

Rinaldi S, Consales G, De Gaudio AR. State entropy and bispectral index: correlation with end tidal sevoflurane concentrations. Minerva Anestesiol 2007;73:39-48.

Riker RR, Picard JT, Fraser GL. Prospective evaluation of the Sedation-Agitation Scale for adult critically ill patients. Crit Care Med 1999;27:1325-1329.

Riker RR, Fraser GL, Wilkins ML. Comparing the Bispectral Index and Suppression Ratio with Burst Suppression of the Electroencephalogram During Pentobarbital Infusion in Adult Intensive Care Patients. Pharmacotherapy 2003;23:1087-1093.

Roach GW, Kanchuger M, Mangano CM, Newman M, Nussmeier N, Wolman R, Aggarwal A, Marschall K, Graham SH. Adverse cerebral outcomes after coronary bypass surgery. Multicenter Study of Perioperative Ischemia Research Group and the Ischemia Research and Education Foundation Investigators Engl J Med 1996;19:1857-1863.

Roach GW, Newman MF, Murkin JM, Martzke J, Ruskin A, Li J, Guo A, Wisniewski A, Mangano DT. Ineffectiveness of burst suppression therapy in mitigating perioperative cerebrovascular dysfunction. Multicenter Study of Perioperative Ischemia (McSPI) Research Group. Anesthesiology 1999;90:1255-1264.

Rossetti AO, Reichhart MD, Schaller MD, Despland PA, Bogousslavsky J. Propofol treatment of refractory status epilepticus: a study of 31 episodes. Epilepsia 2004;45:757-63.

Rundshagen I, Schnabel K, Pothmann W, Schleich B, Schulte am Esch J. Cortical arousal in critically ill patients: an evoked response study. Intensive Care Med 2000;26:312-318.

Ruokonen E, Parviainen I, Jakob SM, Nunes S, Kaukonen M, Shepherd ST, Sarapohja T, Bratty JR, Takala J. "Dexmedetomidine for Continuous Sedation" Investigators. Dexmedetomidine versus propofol/midazolam for long-term sedation during mechanical ventilation. Intensive Care Med 2009;35:282-290.

Sackey PV, Radell PJ, Granath F, Martling CR. Bispectral index as a predictor of sedation depth during isoflurane or midazolam sedation in ICU patients. Anaesth Intensive Care 2007;35:348-356.

Sakai T, Singh H, Mi WD, Kudo T, Matsuki A. The effect of ketamine on clinical endpoints of hypnosis and EEG variables during propofol infusion. Acta Anaesthesiol Scand 1999;43:212-216.

87

Sandin RH, Enlund G, Samuelsson P, Lennmarken C. Awareness during anaesthesia: a prospective case study. Lancet 2000;26:707-711.

Sarén-Koivuniemi TJ, Yli-Hankala AM, van Gils MJ. Increased variation of the response index of nociception during noxious stimulation in patients during general anaesthesia. Comput Methods Programs Biomed 2011;104:154-160.

Savoia G, Esposito C, Belfiore F, Amantea B, Cuocolo R. Propofol infusion and auditory evoked potentials. Anaesthesia 1988;43:46-49.

Schlünzen L, Juul N, Hansen KV, Gjedde A, Cold GE. Regional cerebral glucose metabolism during sevoflurane anaesthesia in healthy subjects studied with positron emission tomography. Acta Anaesthesiol Scand 2010;54:603-609.

Schmidlin D, Hager P, Schmid ER. Monitoring level of sedation with bispectral EEG analysis: comparison between hypothermic and normothermic cardiopulmonary bypass. Br J Anaesth 2001;86:769-776.

Schmidt GN, Bischoff P, Standl T, Hellstern A, Teuber O, Schulte J. Comparative evaluation of the Datex-Ohmeda S/5 Entropy Module and the Bispectral Index monitor during propofol-remifentanil anesthesia. Anesthesiology 2004;101:1283-1290.

Schulte-Tamburen AM, Scheier J, Briegel J, Schwender D, Peter K. Comparison of five sedation scoring system by means of auditory evoked potentials. Intensive Care Med 1999;25:377-382.

Schwender D, Klasing S, Madler C, Pöppel E, Peter K. Effects of benzodiazepines on mid-latency auditory evoked potentials. Can J Anaesth 1993;40:1148-1154.

Sebel PS, Lang E, Rampil IJ, White PF, Cork R, Jopling M, Smith NT, Glass PS, Manberg P.

A multicenter study of bispectral electroencephalogram analysis for monitoring anesthetic effect. Anesth Analg 1997;84:891-899.

Sebel PS, Bowdle TA, Ghoneim MM, Rampil IJ, Padilla RE, Gan TJ, Domino KB. The incidence of awareness during anesthesia: a multicenter United States study. Anesth Analg 2004;99:833-839 (a).

Sebel PS. Comfortably numb? Editorial. Acta Anaesthesiol Scand 2004;48:1-3 (b).

Seitsonen ER, Korhonen IK, van Gils MJ, Huiku M, Lötjönen JM, Korttila KT, Yli-Hankala AM. EEG spectral entropy, heart rate, photoplethysmography and motor responses to skin incision during sevoflurane anaesthesia. Acta Anaesthesiol Scand 2005;49:284-292.

Serfontein L. Awareness in cardiac anesthesia. Curr Opin Anaesthesiol 2010;23:103-108.

Sessler CN, Gosnell MS, Grap MJ, Brophy GM, O'Neal PV, Keane KA, Tesoro EP, Elswick RK. The Richmond Agitation-Sedation Scale: validity and reliability in adult intensive care unit patients. Am J Respir Crit Care Med 2002;166:1338-1344.

Shapiro BA, Warren J, Egol AB, Greenbaum DM, Jacobi J, Nasraway SA, Schein RM, Spevetz A, Stone JR. Practice parameters for intravenous analgesia and sedation for adult patients in the intensive care unit: an executive summary. Crit Care Med 1995;23:1596-1600.

88

Shinner G, Sharpe RM, Thornton C, Doré CJ, Brunner MD. Effect of bolus doses of alfentanil on the arousal response to intubation, as assessed by the auditory evoked response. Br J Anaesth 1999;82:925-928.

Shorvon S, Baulac M, Cross H, Trinka E, Walker M. The drug treatment of status epilepticus in Europe: Consensus document from a workshop at the first London Colloquium on Status Epilepticus. Task Force on Status Epilepticus of the ILAE Commission for European Affairs. Epilepsia 2008;49:1277-1284.

Sigl JC, Chamoun NG. An introduction to bispectral analysis for the electroencephalogram. J Clin Monit 1994;10:392-404.

Silbert BS, Santamaria JD, O'Brien JL, Blyth CM, Kelly WJ, Molnar RR. Early extubation following coronary artery bypass surgery: a prospective randomized controlled trial. The Fast Track Cardiac Care Team. Chest 1998;113:1481-1488.

Simmons LE, Riker RR, Prato BS, Fraser GL. Assessing sedation during intensive care unit mechanical ventilation with the Bispectral Index and the Sedation-Agitation Scale. Crit Care Med 1999;27:1499-1504.

Sneyd JR, Wang DY, Edwards D, Pomfrett CJ, Doran BR, Healy TE, Pollard BJ. Effect of physiotherapy on the auditory evoked response of paralysed, sedated patients in the intensive care unit. Br J Anaesth 1992;68:349-351.

Solanki A, Puri GD, Mathew PJ. Bispectral index-controlled postoperative sedation in cardiac surgery patients: a comparative trial between closed loop and manual administration of propofol. Eur J Anaesthesiol 2010;27:708-713.

Soliman HM, Mélot C, Vincent JL. Sedative and analgesic practice in the intensive care unit: the results of a European survey. Br J Anaesth 2001;87:186-192.

Steyn-Ross ML, Steyn-Ross DA, Sleigh JW, Liley DT. Theoretical electroencephalogram stationary spectrum for a white-noise-driven cortex: evidence for a general anesthetic-induced phase transition. Phys Rev E Stat Phys Plasmas Fluids Relat Interdiscip Topics 1999;60:7299-7311.

Storm H, Myre K, Rostrup M, Stokland O, Lien MD, Raeder JC. Skin conductance correlates with perioperative stress. Acta Anaesthesiol Scand 2002;46:887-895.

Struys MM, De Smet T, Versichelen LF, Van De Velde S, Van den Broecke R, Mortier EP. Comparison of closed-loop controlled administration of propofol using Bispectral Index as the controlled variable versus "standard practice" controlled administration. Anesthesiology 2001;95:6-17.

Struys MM, Jensen EW, Smith W, Smith NT, Rampil I, Dumortier FJ, Mestach C, Mortier EP. Performance of the ARX-derived auditory evoked potential index as an indicator of anesthetic depth: a comparison with bispectral index and hemodynamic measures during propofol administration. Anesthesiology 2002;96:803-816.

89

Struys MM, Vereecke H, Moerman A, Jensen EW, Verhaeghen D, De Neve N, Dumortier FJ, Mortier EP. Ability of the bispectral index, autoregressive modelling with exogenous input-derived auditory evoked potentials, and predicted propofol concentrations to measure patient responsiveness during anesthesia with propofol and remifentanil. Anesthesiology 2003;99:802-812.

Struys MM, Vanpeteghem C, Huiku M, Uutela K, Blyaert NB, Mortier EP. Changes in a surgical stress index in response to standardized pain stimuli during propofol-remifentanil infusion. Br J Anaesth 2007;99:359-367.

Stump DA, Jones TJ, Rorie KD Neurophysiologic monitoring and outcomes in cardiovascular surgery. J Cardiothorac Vasc Anesth 1999;13:600-613.

Svircevic V, Nierich AP, Moons KG, Brandon Bravo Bruinsma GJ, Kalkman CJ, van Dijk D. Fast-track anesthesia and cardiac surgery: a retrospective cohort study of 7989 patients. Anesth Analg 2009;108:727-733.

Särkelä M, Mustola S, Seppänen T, Koskinen M, Lepola P, Suominen K, Juvonen T, Tolvanen-Laakso H, Jäntti V. Automatic analysis and monitoring of burst suppression in anesthesia. J Clin Monit Comput 2002;17:125-134.

Särkelä MKO, Ermes MJ, van Gils MJ, Yli-Hankala AM, Jäntti VH, Vakkuri AP. Quantification of Epileptiform Electroencephalographic Activity during Sevoflurane Mask Induction. Anesthesiology 2007;107:928-938.

Takamatsu I, Ozaki M, Kazama T. Entropy indices vs. the bispectral index™ for estimating nociception during sevoflurane anaesthesia. Br J Anaesth 2006;96:620-626.

Thornton C, Catley DM, Jordan C, Lehane JR, Royston D, Jones JG. Enflurane anaesthesia causes graded changes in the brainstem and early cortical auditory evoked response in man. Br J Anaesth 1983;55:479-488.

Thornton C, Heneghan CP, James MF, Jones JG. Effects of halothane or enflurane with controlled ventilation on auditory evoked potentials. Br J Anaesth 1984;56:315-323.

Thornton C, Heneghan CP, Navaratnarajah M, Bateman PE, Jones JG. Effect of etomidate on the auditory evoked response in man. Br J Anaesth 1985;57:554-561.

Thornton C, Konieczko K, Jones JG, Jordan C, Doré CJ, Heneghan CP. Effect of surgical stimulation on the auditory evoked response. Br J Anaesth 1988;60:372-378.

Thornton C, Barrowcliffe M, Konieczko K. The auditory evoked response as a indicator of awareness. Br J Anaesth 1989;63:113-115 (a).

Thornton C, Konieczko KM, Knight AB, Kaul B, Jones JG, Dore CJ, White DC. Effect of propofol on the auditory evoked response and oesophageal contractility. Br J Anaesth 1989;63:411-417 (b).

Thornton C. Evoked potentials in anaesthesia. Eur J Anaesthesiol 1991;8:89-107.

90

Thornton C, Jones JG. Evaluating depth of anesthesia: review of methods.

Int Anesthesiol Clin 1993;31:67-88.

Tobias JD. Bispectral Index Monitoring Documents Burst Suppression During Pentobarbital Coma. J Intensive Care Med 2008;23:258-262.

Toner I, Taylor KM, Lockwood G, Newman S, Smith PL. EEG changes during cardiopulmonary bypass surgery and postoperative neuropsychological deficit: the effect of bubble and membrane oxygenators Eur J Cardiothorac Surg 1997;11:312-329.

Towne AR, Pellock JM, Ko D, DeLorenzo RJ. Determinants of mortality in status epilepticus. Epilepsia 1994;35:27-34.

Treiman DM, Walton NY, Kendrick C. A progressive sequence of electroencephalographic changes during generalized convulsive status epilepticus. Epilepsy Res 1990;5:49-60.

Vakkuri A, Yli-Hankala A. Talja P, Mustola S, Tolvanen-Laakso H, Sampson T, Viertiö-Oja H. Time – frequency balanced spectral entropy as a measure of anesthetic drug effect in central nervous system during sevoflurane, propofol, and thiopental anesthesia. Acta Anaesthesiol Scand 2004;48:145-153.

Vakkuri A, Yli-Hankala A, Sandin R, Mustola S, Høymork S, Nyblom S, Talja P, Sampson T, van Gils M, Viertiö-Oja H. Spectral entropy monitoring is associated with reduced propofol use and faster emergence in propofol-nitrous oxide-alfentanil anesthesia. Anesthesiology 2005;103:274-279.

Valjus M, Ahonen J, Jokela R, Korttila K. Response Entropy is not more sensitive than State Entropy in distinguishing the use of esmolol instead of remifentanil in patients undergoing gynaecological laparoscopy. Acta Anaesthesiol Scand 2006;50:32-39.

Vanluchene ALG, Struys MMRF, Heyse BEK, Mortier EP. Spectral entropy measurement of patient responsiveness during propofol and remifentanil. A comparison with the bispectral index. Br J Anaesth 2004;93:645-654.

Velly LJ, Rey MF, Bruder NJ, Gouvitsos FA, Witjas T, Regis JM, Peragut JC, Gouin FM. Differential dynamic of action on cortical and subcortical structures of anesthetic agents during induction of anesthesia. Anesthesiology 2007;107:202-212.

Vivien B, Di Maria S, Ouattara A, Langeron O, Coriat P, Riou B. Overestimation of Bispectral Index in sedated intensive care unit patients revealed by administration of muscle relaxant. Anesthesiology 2003;99:9-17.

Viertiö-Oja H, Maja V, Särkelä M, Talja P, Tenkanen N, Tolvanen-Laakso H, Paloheimo M, Vakkuri A, Yli-Hankala A, Meriläinen P. Description of the Entropy™ algorithm as applied in the Datex-Ohmeda S/5™ Entropy Module. Acta Anaesthesiol Scand 2004;48:154-161.

Vretzakis G, Dragoumanis C, Argiriadou H, Papaziogas B, Konstantinou F, Tsakiridis K. Inaccuracy of BIS values produced by the cardiopulmonary bypass machine during the operative repair of an aortic dissection. J Cardiothorac Vasc Anesth 2006;20:68-70.

91

von Dossow V, Moshirzadeh M, Kastrup M, Wernecke KD, Konertz W, Spies C. Performance of the A-line Autoregressive Index (AAI) and of the Bispectral Index (BIS) at assessing depth of short-term sedation following cardiac surgery. J Int Med Res 2009;37:611-612.

Whitlock EL, Villafranca AJ, Lin N, Palanca BJ, Jacobsohn E, Finkel KJ, Zhang L, Burnside BA, Kaiser HA, Evers AS, Avidan MS. Relationship between bispectral index values and volatile anesthetic concentrations during the maintenance phase of anesthesia in the B-Unaware trial. Anesthesiology 2011;115:1209-1218.

Walsh TS, Ramsay P, Lapinlampi TP, Särkelä MO, Viertiö-Oja HE, Meriläinen PT. An assessment of the validity of spectral entropy as a measure of sedation state in mechanically ventilated critically ill patients. Intensive Care Med 2008;34:308-315.

Walsh TS, Lapinlampi TP, Ramsay P, Särkelä MO, Uutela K, Viertiö-Oja HE. Responsiveness of the frontal EMG for monitoring the sedation state of critically ill patients. Br J Anaesth 2011;107:710-718.

Wennervirta J, Hynynen M, Koivusalo AM, Uutela K, Huiku M, Vakkuri A. Surgical stress index as a measure of nociception/antinociception balance during general anesthesia. Acta Anaesthesiol Scand 2008;52:1038-1045.

Weatherburn C, Endacott R, Tynan P, Bailey M. The impact of bispectral index monitoring on sedation administration in mechanically ventilated patients. Anaesth Intensive Care 2007;35:204-208.

Weil G, Passot S, Servin F, Billard V. Does spectral entropy reflect the response to intubation or incision during propofol –remifentanil anesthesia? Anesth Analg 2008;106:152-159.

Weinert C, McFarland L. The state of intubated ICU patients: development of a two-dimensional sedation rating scale for critically ill adults. Chest 2004;126:1883-1890.

Wheeler P, Hoffman WE, Baughman V, Koeing H. Response entropy increases during painful stimulation. J Neurosurg Anesthesiol 2005;17:86-90.

White PF, Ma H, Tang J, Wender RH, Sloninsky A, Kariger R. Does the use of electroencephalographic bispectral index or auditory evoked potential index monitoring facilitate recovery after desflurane anesthesia in the ambulatory setting? Anesthesiology 2004;100:811-817.

Wolman RL, Nussmeier NA, Aggarwal A, Kanchuger MS, Roach GW, Newman MF, Mangano CM, Marschall KE, Ley C, Boisvert DM, Ozanne GM, Herskowitz A, Graham SH, Mangano DT.Cerebral injury after cardiac surgery: identification of a group at extraordinary risk. Multicenter Study of Perioperative Ischemia Research Group (McSPI) and the Ischemia Research Education Foundation (IREF) Investigators. Stroke 1999;30:514-522.

Xu L, Wu AS, Yue Y. The incidence of intra-operative awareness during general anesthesia in China: a multi-center observational study. Acta Anaesthesiol Scand 2009;53:873-882.

92

Yli-Hankala A. The effect of nitrous oxide on EEG spectral power during halothane and isoflurane anaesthesia. Acta Anaesthesiol Scand 1990;34:579-584 (a).

Yli-Hankala A, Heikkilä H, Värri A, Jäntti V. Correlation between EEG and heart rate variation in deep enflurane anaesthesia. Acta Anaesthesiol Scand 1990;34:138-143 (b).

Yli-Hankala A, Lindgren L, Porkkala T, Jäntti V. Nitrous oxide-mediated activation of the EEG during isoflurane anaesthesia in patients. Br J Anaesth 1993;70:54-57.

Yli-Hankala A, Vakkuri A, Annila P, Korttila K. EEG bispectral index monitoring in sevoflurane or propofol anaesthesia: analysis of direct costs and immediate recovery. Acta Anaesthesiol Scand 1999;43:545-549.

Young GB. The EEG in coma. J Clin Neurophysiol 2000;17:473-485.

Ziegeler S, Buchinger H, Wilhelm W, Larsen R, Kreuer S. Impact of deep hypothermic circulatory arrest on the BIS index. J Clin Anesth 2010;22:340-345.

ORIGINAL PUBLICATIONS (I- IV)

I

Midlatency auditory-evoked potentials in the assessment of sedation in cardiac surgery patients

Musialowicz T, Hynynen M, Yppärilä H, Pölönen P, Ruokonen E, Jakob SM

Journal of Cardiothoracic and Vascular Anesthesia 18:559-562, 2004.

Reprinted with the permission of the copyright owner Elsevier Incorporation

II

Auditory evoked potentials in bispectral index-guided anaesthesia for cardiac surgery

Musialowicz T, Niskanen M, Yppärilä-Wolters H, Pöyhönen M, Pitkänen O,

Hynynen M.

European Journal of Anaesthesiology 24:571-579, 2007.

Reprinted with the permission of the copyright owner European Society of Anaesthesiology (Wolters Kluwer Health)

III

Can BIS monitoring be used to assess the depth of propofol anesthesia in the treatment of refractory status epilepticus?

Musialowicz T, Mervaala E, Kälviäinen R, Uusaro A, Ruokonen E, Parviainen I.

Epilepsia 51:1580-1586, 2010.

Reprinted with the permission of the copyright owner International League Against Epilepsy (John Wiley and Sons)

IV

Comparison of spectral entropy and BIS VISTA™ monitor during general

anesthesia for cardiac surgery.

Musialowicz T, Lahtinen P, Pitkänen O, Kurola J, Parviainen I.

Journal of Clinical Monitoring and Computing 25: 95-103, 2011.

Reprinted with the permission of the copyright owner Springer Science

Publications of the University of Eastern Finland

Dissertations in Health Sciences

isbn 978-952-61-1006-6

Publications of the University of Eastern FinlandDissertations in Health Sciences

dissertatio

ns | 149 | T

ad

eusz M

usia

low

icz | EE

G-based M

onitoring durin

g Gen

eral An

aesthesia an

d Sedation

Tadeusz MusialowiczEEG-based Monitoring

during General Anaesthesia and Sedation

Studies on Cardiac Surgery

and Status Epilepticus Patients

Tadeusz Musialowicz

EEG-based Monitoring during General Anaesthesia and SedationStudies on Cardiac Surgery and Status Epilepticus

Patients

The use of anaesthetic agents

during general anaesthesia

and sedation in the intensive

care unit induce changes in

patient’s electroencephalographic

(EEG) waveforms of the brain

cortex. The utility of EEG-based

neuromonitoring during general

anaesthesia for patients undergoing

cardiac surgery and postoperative

sedation and during the treatment

of status epilepticus in the intensive

care unit was evaluated in this

study. The dissertation provides new

information about the differences

between EEG-based monitoring

techniques and their feasibility to

assess hypnosis in cardiac surgery

and in intensive care settings.