Patients, as well as anaesthesia practitioners, fear unexpected awareness during general anaesthesia. In that context, a battery of new monitors is currently available for clinical use. Those monitoring devices have been initially developed to evaluate the depth of anaesthesia in an attempt to prevent the occurrence of intraoperative awareness. However, their usefulness, reliability, ease of use and limits remain controversial. The objectives of this review are to critically analyse the information provided by those monitors, define what can be expected from their use and, in light of the most recent literature, determine if they can actually improve the practice of anaesthesia. Understanding those monitors requires some knowledge about the anatomy and physiology of the loss of consciousness, about the type of variable recorded and processed by the monitoring devices, as well as about their physiological basis. The discussion will be limited to the monitors commercially available in Europe and to those which the authors are using clinically. Finally, this review will be completed by giving an overview of the current knowledge and expectations of patients and anaesthetists regarding this issue.
Unexpected awareness: definition, epidemiology, causes and consequences
Unexpected awareness is defined as the unintentional return to consciousness of a patient under general anaesthesia. Under those circumstances, consciousness can be considered as the ability of perceiving and interpreting one's external environment. Intraoperative awareness during general anaesthesia might or might not be associated with post-anaesthesia explicit memory. Its overall incidence is estimated to be 0.1-0.2% [1-4] but depends on several factors. Haemodynamic instability, emergency surgery and Caesarean section are the most frequently encountered favouring circumstances . Unexpected awareness during general anaesthesia will occur in case of insufficient delivery of anaesthetics, which can be absolute or relative. Absolute insufficient delivery may occur intentionally, such as in trauma patients or during Caesarean sections, or incidentally, in case of anaesthetic machine dysfunction or human error. Relative insufficient delivery is mainly encountered in patients addicted to alcohol and in those receiving hepatic enzyme inducers chronically, such as sleep inducers or anticonvulsant medications. The use of muscle relaxants is also generally considered as a factor in anaesthetic awareness in so far as no response to movement is possible in case of too light a level of anaesthesia. Unexpected awareness may provoke patients' movements, haemodynamic modifications and neuropsychological disorders. Movements can impede the work of the surgeon. Haemodynamic modifications can favour bleeding or precipitate cardiovascular disturbances in fragile patients. Unexpected awareness during anaesthesia is a dreadful experience for patients and may result in post-traumatic stress disorder (PTSD). PTSD is characterized by chronic anxiety, irritability, nightmares and fear of death, and is not necessarily associated with explicit memory. The best therapeutic approach to PTSD includes prevention, close dialogue, information and psychological support.
Anaesthesia: definition of depth and mechanisms
Anaesthesia may be regarded as administering anaesthetic agents to allow patients to tolerate painful or unpleasant diagnostic or surgical procedures. The administration of anaesthetic agents triggers a complex variety of pharmacodynamic effects which may classically be divided into alteration of consciousness, amnesia, analgesia, muscle relaxation and autonomic modulation. Each of them is related to certain molecular mechanisms occurring at anatomical sites that are currently in the process of being identified. Strong interaction exists between those effects. Therefore, depth of anaesthesia is a global concept and its definition will depend on the specific component of anaesthesia of concern. However, it must be kept in mind that most of the currently available monitors have been designed to explore only one pharmacodynamic aspect of anaesthesia: the alteration of consciousness.
During anaesthesia, loss of consciousness is defined as the reversible alteration of wakefulness and cognitive function of the brain such as perception of the environment, thinking, attention and memory. Those functions are related to anatomical structures including peripheral sensors, ascending sensory pathways, thalamic relay nuclei, the ascending reticular activating system, as well as functional cortical regions and cholinergic, noradrenergic and serotonergic pathways emerging from the basal forebrain. Electrophysiological reciprocal interactions between those structures sustain consciousness. Anaesthesia induces reversible alterations in synaptic communication between neurons of the above-mentioned functional structures. Anaesthetic agents alter synaptic communication through their binding to specific hydrophobic neuronal membrane sites such as receptors for neurotransmitters, ionophores and second messenger systems. The functional structures involved in the maintenance of consciousness are specifically and reproducibly affected by anaesthetic agents as a function of the nature of the agent and of the administered dose .
Measuring the depth of anaesthesia
Measuring the depth of anaesthesia consists of determining the dose/concentration response relationship between anaesthetic agents and the pharmacodynamic effects of anaesthesia. The effect-site concentration of anaesthetic agents (i.e. their concentration in the target structure of their pharmacodynamic effect) can easily be obtained through the monitoring of the end-expiratory concentration of volatile anaesthetics or approximated through the use of target-controlled infusion (TCI) devices. The study of the pharmacodynamic effects of anaesthesia is more complex because of the following reasons: interactions between those effects (e.g. consciousness is modulated by the level of noxious stimulation), absence of a continuum (e.g. consciousness is more an on/off phenomenon than a continuous variable), variability of pharmacodynamic effects between classes of anaesthetic agents, and interactions between anaesthetic agents (e.g. opioids and hypnotics). When using monitors of the depth of anaesthesia, it is necessary to keep in mind those elements, to know what is really measured, to understand the nature of the given information and to interpret parameters in the light of consciousness physiology.
Clinical methods to evaluate the depth of anaesthesia
The clinical methods for assessing depth of anaesthesia have long been used and remain a reference. The first well established method was described by Guedel in 1937 and concerns a single agent anaesthetic technique, namely ether anaesthesia. Although not applicable blindly to all types of general anaesthetic techniques, it sets up the basis for a good clinical evaluation of anaesthetic depth, which should consider ventilation, pupil diameter, eyelash reflex, swallowing reflex, muscular tone, movements, blood pressure, heart rate, tears, sweating, and response to verbal commands. Although clinical evaluation of the depth of anaesthesia is poorly sensitive and specific, it remains mandatory in current anaesthetic practice. The recent tools have been designed to overcome the problem of specificity and sensitivity but they only partly succeed in this task.
Parameters derived from the electroencephalogram
Anaesthetic agents interfere with electroencephalographic activity. Those interactions are complex, vary according to classes of anaesthetic agents and are not easily detected except by skilled anaesthetists familiar with electrophysiology. These facts have partly motivated the development of the currently available depth of anaesthesia monitors. Most of them display variables derived from the analysis of the electroencephalogram (EEG) as well as tracings of the EEG itself. The EEG signal is weak. Therefore, contact between skull and electrodes must be tight and the signal must be amplified before analysis. As a consequence, although monitors are equipped with sophisticated filters, numerous artefacts can alter EEG recordings and must be recognized by the user for correct interpretation of data. They can arise from electromyographic activity, electrocardiographic activity, movements, swallowing, blinking, power supply or electric devices used concomitantly. It is also important to know that EEG activity is affected by physiologic and pathophysiologic factors such as age, temperature, PaCO2, hyper- or hypoglycaemia, electrolyte disorders, hepatic or renal failure, and endocrine disorders .
In most depth of anaesthesia monitors, the EEG signal is first digitized and then mathematically analysed with a Fourier Transformation. Using a specific algorithm, parameters are then extracted from the generated frequency band power spectrum to produce a final index.
The Bispectral Index
The Bispectral Index™ (BIS™) is an absolute number varying between 0 (flat EEG) and 100 (awake patient) . The mathematical algorithm used to calculate the BIS has been elaborated through the statistical analysis of an EEG data bank that has allowed the identification of parameters significantly correlated with anaesthetic agent concentration and patient reactivity. Those parameters are derived from the spectral, time domain and bispectral analysis of the EEG. They particularly take into account the amount of suppression activity, beta power and slow synchronized activity of the EEG. Each parameter is given weighted coefficients to obtain a linear relationship between BIS on one hand and plasma concentration of the anaesthetic agent and clinical response of patients on the other hand . As a result, a BIS value between 70 and 100 is mostly correlated with beta power, between 40 and 70 with the synchronized fast slow activity, to a quasi-flat EEG activity between 20 and 40 and to suppression ratio between 0 and 20.
The clinical advantages of using BIS include a significant reduction in the amount of general anaesthetic agents administered, a reduction in the frequency of hypnotic anaesthetic agent dosing errors, better haemodynamic stability and improved patient satisfaction as well as faster recovery at the end of the procedure and a shorter stay in the post-anaesthesia care unit [9-11]. Moreover, BIS allows practitioners to provide patients with an individualized and rationalized administration of anaesthetic agents  and this is more efficient when a TCI device is used or when the end-expiratory concentration of volatile anaesthetic agents is measured concomitantly. BIS-guided induction of anaesthesia using propofol is associated with lower doses and better haemodynamic stability than non-BIS-guided inductions . During induction of intravenous anaesthesia using a TCI device, the BIS value recorded upon loss of consciousness provides information about the individual relationship between BIS and level of consciousness. The effect-site concentration of the anaesthetic agent recorded at that time is also informative although it must be kept in mind that it does not necessarily reflect the effect-site concentration at which the patient will return to consciousness during the recovery period . Actually, the TCI devices display an effect-site concentration of return to consciousness usually higher than the concentration needed to lose consciousness. This is probably related to pitfalls in the pharmacokinetic models implemented in TCI devices and to environmental factors. Pharmacokinetic concerns refer to an equilibration between plasma and effect-site concentration that is not instantaneous. The equilibration constant between plasma and effect-site compartment is only an estimate and the bolus effect associated with a rapid induction of anaesthesia is thought to explain the relatively low estimated effect-site concentration at which patients lose consciousness. Environmental factors refer to the fact that, during recovery, the surgical wound and pain may modify the pharmacodynamic response of patients to a given effect-site concentration. However, when reaching a steady state after induction of anaesthesia and before incision, the effect-site anaesthetic agent concentration necessary to obtain a BIS of 60 may be noted and serve as a reference throughout the anaesthetic procedure, in order to limit the risk of explicit awareness. The anaesthetic agent concentration should remain within a safe range around that concentration throughout the procedure. The same rules also apply when using volatile anaesthetics. Finally, the increase of BIS at the time of tracheal intubation can be used to titrate opioid administration, knowing that such an increase in BIS could be observed for each stimulus at least as painful as laryngoscopy .
There are several limits regarding the use of BIS. Despite the fact that BIS value interpretation must take into account the context in which it is recorded, it is worth noting that the delay between EEG acquisition and availability of the BIS value for the user is about 30 s, or even more if artefacts alter the tracing. Therefore, when it is actually displayed, the BIS value reflects the level of consciousness as it actually was 30 s (or more) earlier. This delay has been reduced in the latest version of the BIS monitor (version 4.0). The BIS monitor smoothes the displayed value by averaging data obtained over periods of 15 or 30 s. It is preferable to use a 30 s smoothing period of time in order to limit the variability of BIS. It is likely that the paradoxical increase in BIS reported when the volatile anaesthetic concentration is abruptly increased is not observed when a 30 s smoothing period of time is set up on the monitor . During the washout of high opioid concentrations, paradoxical decreases of BIS can be observed. They are caused by low-amplitude EEG waves interpreted as burst suppression by the BIS monitor . Filters theoretically reject high frequency artefacts but BIS values may be artificially increased when the ranges of filters are overcome. BIS can vary in the absence of any artefact, any modification of the anaesthetic regimen or noxious stimulation such as in the case of the decrease observed during cerebral ischaemia . Attention should be paid on not titrating too tightly the administration of hypnotic anaesthetic agents as a function of BIS. If the anaesthetic agent concentration reaches a level too close to the concentration associated with the loss of consciousness, and if the level of analgesia is too light, awareness could occur as a result of small changes in noxious stimulation or anaesthetic agent concentration. Finally, as unexpected awareness is a rare event, huge cohorts of patients are needed to demonstrate any beneficial effect of BIS in terms of reducing that risk . However, recent studies suggest that BIS is efficient at reducing the incidence of this unpleasant event, at least in high-risk patients .
Used in an efficient way, that is in order to titrate hypnosis according to surgical events and anaesthetic regimen, and keeping a sufficiently safe range in terms of anaesthetic agent concentration based on data recorded at induction of anaesthesia, BIS undoubtedly improves the quality of anaesthesia.
Entropy of the EEG
The Datex-Ohmeda S/5™ Entropy Module (M-Entropy™) is a promising new tool for monitoring the depth of anaesthesia. It computes an estimate of the entropy of the EEG (EE). EE is based on the Kolmogorov-Sinai principle which allows quantification of the amount of regularity in data and was first applied to a power spectrum of a signal by Johnson and Shore in 1984 . The initial hypothesis for developing this tool was that EEG activity would show more regularity in anaesthetized patients than in awake patients . EE quantifies the predictability of subsequent amplitude values of the EEG based on the knowledge of the previous amplitude values by combining time and frequency domain analysis of the signal (time-frequency-balanced spectral entropy). The time window for analysis is chosen in such a way that each frequency component is obtained from a time window that is optimal for that particular frequency. In this way, information is extracted from the signal as fast as possible. The M-Entropy module is original in that it provides two values of entropy, the state entropy (SE) and the responsive entropy (RE). SE and RE are calculated from specific ranges of frequencies and are displayed as numbers varying between 0 and 100. SE is computed over the frequency range from 0.8 to 32 Hz that includes the EEG-dominant part of the spectrum, and, therefore, primarily reflects the cortical state of the patient. RE is computed over a frequency range from 0.8 to 47 Hz and includes both the EEG-dominant and facial EMG-dominant part of the spectrum. RE becomes equal to SE when the EMG power is equal to zero, otherwise it is always higher than SE. It is hypothesized that, in the case of deficient analgesia in a non-paralysed patient, EMG facial activity increases before any change in the EEG activity, leading to an increase in RE before any change in SE. Although this last hypothesis needs to be demonstrated by appropriate studies, EE has proven to be at least as efficient as BIS in predicting changes in the hypnotic component of anaesthesia . Further investigations are needed to know if EE is as efficient as BIS for the individual titration of anaesthesia, and to better define the information it provides for assessing the depth of anaesthesia and its several components.
The auditory evoked potentials
Auditory evoked potentials (AEP) correspond to electric potentials produced in response to auditory stimuli. As the cortical response is of low amplitude (less than 10 μV), it is necessary to repeat auditory stimulations and average multiple EEG recording segments to eliminate background noise. An AEP is made up of positive and negative waves, each being characterized by its latency and amplitude. Each wave corresponds to the transmission of auditory information through specific anatomical structures along the auditory ascending pathways, from peripheral sensors to the cortex. Middle latency AEP (MLAEP, 20-80 ms after stimulation) are related to the transmission of information through the thalamus towards the primary and secondary auditory cortex and have been demonstrated to be modified by anaesthesia. Their amplitude decreases and latency increases with increasing depth of anaesthesia. Monitors have been developed to facilitate tracing analysis and allow the practitioner to have an idea of their modifications at a glance to a unique number. Among them, the Alaris™ A-Line™ monitor is the most largely distributed in Europe. It computes an index (A-Line autoregressive index, AAI™) which varies between 0 and 99 and is based on the amplitude and latency of the MLAEP. An original autoregressive method provides a fast extraction of AEPs and shortens the delay between acquisition and AAI display to 1.7 s, provided that artefacts do not alter the tracing. AAI correlates well with the level of consciousness of patients [23-26]. Awake AAI values range between 50 and 99, while consciousness is lost around 40 and a value of 20 is recommended for surgical anaesthesia. AAI can be altered by the same type of artefacts as BIS, as well as by age, hypoxia, hypothermia, hypercapnia, alcohol and natural sleep. The relationship between AAI and the response to noxious stimulation is not clear as yet . Titration of anaesthetic agent administration using AAI improves emergence and decreases anaesthetic consumption [28,29].
Cost-effectiveness relationship, patients and anaesthetists knowledge and expectations
Depth of anaesthesia monitors are expensive: the initial purchasing cost may be as high as $12 500 and electrode related costs are high. As mentioned above, unexpected awareness is rare and huge cohorts of patients are required to demonstrate any benefit of a given monitor in terms of reducing the risk of such an event. One might therefore question whether developing those tools is useful or not. Considering the anaesthetists' opinion, an Australian study has reported that, although half of the surveyed anaesthetists have experienced a patient with awareness, anaesthetists, especially more senior ones, rate awareness as a moderate problem. However, they are prepared to use depth of anaesthesia monitoring more widely if it were demonstrated to reduce the incidence of awareness . The same authors report that patients' knowledge about the problem is weak. Only half of the patients have already heard about awareness, mainly through the media. Many patients are anxious about it but few would pay for a proven awareness monitor, depending on the perceived risk and on the occurrence of a previous awareness episode . In contrast, another study has shown that, although the amount patients would pay decreases with the risk of awareness, the relationship between the two is not linear, indicating that patients assign an intrinsic base value for the possibility of awareness . Those studies highlight the lack of information and knowledge both of patients and anaesthetists. The financial factor must not be the single criterion to rely on for deciding whether monitoring the depth of anaesthesia is necessary or not. Suffice to say that, if used properly, it largely improves the quality of anaesthesia. Furthermore, avoidance of giving too high a dose of anaesthetic agent limits costs related to anaesthetic agents themselves and to the side-effect of overdosing.
Proper use of depth of anaesthesia monitors improves the quality of anaesthesia. However, for such a purpose, a sufficient knowledge about awareness, physiology of consciousness, pharmacodynamic components of anaesthesia and technical background of the monitors is mandatory. Furthermore, those monitors should be used in combination with devices providing the effect-site concentration of anaesthetic agents. Individual titration of anaesthetic agents according to surgical events and anaesthetic regimen, while keeping a sufficient security range in terms of anaesthetic agent concentrations and based on data recorded at induction of anaesthesia is the best approach to an optimal use. None of the pharmacodynamic components of anaesthesia can be monitored individually, although the hypnotic component is monitored the best by the currently available monitors. The future might see the emergence of new properties of currently available monitors or the emergence of new monitors allowing discrimination between pharmacodynamic components of anaesthesia. Further improvement can also be expected when closed-loop administration systems will be available for current clinical practice .
The authors would like to thank Mark J. Lema, Vinod Malhotra, Thel G. Boyette and Hans J. Priebe for having permitted this work to be presented at a scientific panel of the 57th Postgraduate Assembly in Anesthesiology, held in New York, 12-16 December 2003. Their thanks go also to Carine Vauchel for her help in formatting the manuscript.
1. Liu WH, Thorp TA, Graham SG, Aitkenhead AR. Incidence of awareness with recall during general anaesthesia. Anaesthesia
2. Sandin RH, Enlund G, Samuelsson P, Lennmarken C. Awareness during anaesthesia: a prospective case study. Lancet
3. Leslie K, Lee L, Myles PS, Lerch N, Fiddes C. Patients' knowledge of and attitudes towards awareness and depth of anaesthesia monitoring. Anaesth Intens Care
4. Sebel PS, Bowdle A, Ghoneim MM, Rampil IJ, Padilla RE. The incidence of awareness during anesthesia: a multicenter U.S. study. Anesthesiology
5. Schneider G, Sebel PS. Monitoring depth of anaesthesia. Eur J Anaesthesiol
6. Bonhomme V, Fiset P, Meuret P, et al.
Propofol anesthesia and cerebral blood flow changes elicited by vibrotactile stimulation: a positron emission tomography study. J Neurophysiol
7. Rampil IJ. A primer for EEG signal processing in anesthesia. Anesthesiology
8. Johansen JW, Sebel PS. Development and clinical application of electroencephalographic bispectrum monitoring. Anesthesiology
9. Luginbuhl M, Wuthrich S, Petersen-Felix S, Zbinden AM, Schnider TW. Different benefit of bispectal index (BIS™) in desflurane and propofol anesthesia. Acta Anaesthesiol Scand
10. Lehmann A, Karzau J, Boldt J, Thaler E, Lang J, Isgro F. Bispectral index-guided anesthesia in patients undergoing aortocoronary bypass grafting. Anesth Analg
11. Puri GD, Murthy SS. Bispectral index monitoring in patients undergoing cardiac surgery under cardiopulmonary bypass. Eur J Anaesthesiol
12. Forestier F, Hirschi M, Rouget P, et al.
Propofol and sufentanil titration with the bispectral index to provide anesthesia for coronary artery surgery. Anesthesiology
13. Gurses E, Sungurtekin H, Tomatir E, Dogan H. Assessing propofol induction of anesthesia dose using bispectral index analysis. Anesth Analg
14. Kazama T, Morita K, Ikeda T, Kurita T, Sato S. Comparison of predicted induction dose with predetermined physiologic characteristics of patients and with pharmacokinetic models incorporating those characteristics as covariates. Anesthesiology
15. Guignard B, Menigaux C, Dupont X, Fletcher D, Chauvin M. The effect of remifentanil on the bispectral index change and hemodynamic responses after orotracheal intubation. Anesth Analg
16. Detsch O, Schneider G, Kochs E, Hapfelmeier G, Werner C. Increasing isoflurane concentration may cause paradoxical increases in the EEG bispectral index in surgical patients. Br J Anaesth
17. Muncaster AR, Sleigh JW, Williams M. Changes in consciousness, conceptual memory, and quantitative electroencephalographical measures during recovery from sevoflurane- and remifentanil-based anesthesia. Anesth Analg
18. Guignard B, Chauvin M. Bispectral index increases and decreases are not always signs of inadequate anesthesia. Anesthesiology
19. O'Connor MF, Daves SM, Tung A, Cook RI, Thisted R, Apfelbaum J. BIS monitoring to prevent awareness during general anesthesia. Anesthesiology
20. Myles PS, Leslie K, Forbes A, Chan M, McNeil J. A large randomized trial of BIS monitoring to prevent awareness in high risk patients: The B-Aware Trial. Anesthesiology
21. Johnson RW, Shore JE. Which is the better entropy expression for speech processing: −S log S or log S? IEEE Acoust Speech Signal Proc
22. Bruhn J, Ropcke H, Hoeft A. Approximate entropy as an electroencephalographic measure of anesthetic drug effect during desflurane anesthesia. Anesthesiology
23. Litvan H, Jensen EW, Revuelta M, et al.
Comparison of auditory evoked potentials and the A-line ARX index for monitoring the hypnotic level during sevoflurane and propofol induction. Acta Anaesthesiol Scand
24. Litvan H, Jensen EW, Galan J, et al.
Comparison of conventional averaged and rapid averaged, autoregressive-based extracted auditory evoked potentials for monitoring the hypnotic level during propofol induction. Anesthesiology
25. Ge SJ, Zhuang XL, Wang YT, Wang ZD, Chen SL, Li HT. Performance of the rapidly extracted auditory evoked potentials index to detect the recovery and loss of wakefulness in anesthetized and paralyzed patients. Acta Anaesthesiol Scand
26. 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
27. Struys MM, Jensen EW, Smith W, et al.
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
28. Maattanen H, Anderson R, Uusijarvi J, Jakobsson J. Auditory evoked potential monitoring with the AAITM-index during spinal surgery: decreased desflurane consumption. Acta Anaesthesiol Scand
29. Recart A, Gasanova I, White PF, et al.
The effect of cerebral monitoring on recovery after general anesthesia: a comparison of the auditory evoked potential and bispectral index devices with standard clinical practice. Anesth Analg
30. Myles PS, Symons JA, Leslie K. Anaesthetists' attitudes towards awareness and depth-of-anaesthesia monitoring. Anaesthesia
31. Gan TJ, Ing RJ, de LD, Wright D, El Moalem HE, Lubarsky DA. How much are patients willing to pay to avoid intraoperative awareness? J Clin Anesth
32. Mortier E, Struys M, De Smet T, Versichelen L, Rolly G. Closed-loop controlled administration of propofol using bispectral analysis. Anaesthesia