Anaesthetists use monitors to measure and control the depth of anaesthesia. Two of the most commonly used monitoring systems are electroencephalographic state entropy and bispectral index (BIS). Both systems measure the effect of an anaesthetic agent on neurophysiological activity in the frontal cortex. The BIS system, which has been available since the mid-1990s, is a composite of different electroencephalogram (EEG) processing techniques, including power spectral analysis, bispectral analysis and time domain analysis. The EEG signal is converted by mathematical analyses to a numerical value between 0 (isoelectric EEG) and 100 (fully awake). Even though there is a wide variation of the conscious state in BIS values around 80,1 the target range of BIS for surgical anaesthesia seems to be between 40 and 60.2 Burst suppression should correlate with a BIS value below 20.2
The National Institute for Clinical Excellence recommends that monitors be used to measure depth of anaesthesia in patients with a higher risk of awareness or excessively deep anaesthesia and in all patients receiving total intravenous anaesthesia.3
Studies have shown that monitoring consciousness can lead to earlier recovery because of a reduction in drug consumption. In addition, some studies indicate that BIS monitoring reduces the risk of intraoperative awareness.4,5
However, more recent and larger studies have not been able to reproduce these results, finding that BIS is not more effective at reducing the incidence of awareness than measuring end-tidal anaesthetic agents.6 This was confirmed by Avidan and colleagues7 in 2011 and by Whitlock and colleagues,8 who also found that BIS correlates poorly with end-tidal concentrations. Others had similar findings with other monitors.9 Also, data indicate that processed EEG monitoring systems may not be suitable for assessing unconsciousness. With isoflurane-based anaesthesia, a BIS value of 70 correlates with a 50% probability of unconsciousness, whereas the same BIS value in a patient anaesthetised with propofol correlates with a 15% probability of unconsciousness. The usefulness and cost-effectiveness of depth of anaesthesia monitoring has, therefore, been questioned.3
The aim of this study was to evaluate the reliability of BIS monitoring for assessing changes in levels of consciousness, specifically during the transition from consciousness to unconsciousness, in patients undergoing total intravenous anaesthesia with propofol and remifentanil. BIS was compared with the gold standard for assessing neurophysiological activity in the brain, continuous multiple-lead electroencephalography and to clinical loss of consciousness (LOC).
The study was approved by the Regional Ethics Committee University Hospital in Linköping, Sweden on 6 October 2008 (M184–2007). Written informed consent was obtained from all participants. We intended to recruit 35 American Society of Anesthesiologists’ (ASA) physical status 1 patients aged 18 to 49 years, with a BMI of 20 to 30 kg m−2, scheduled for day-case surgery under general anaesthesia at Linköping University Hospital, Finspång Hospital and Kalmar Hospital between 5 October 2011 and 16 April 2013.
Patients with central nervous system disorders, a history of smoking, psychiatric diseases, alcohol or drug abuse, analgesic use within 12 h prior to surgery, pregnancy or allergies to soy beans or peanuts were excluded.
All patients received paracetamol and/or a NSAID prior to induction. No other premedication was administered. A peripheral venous cannula was inserted into the forearm. Standard monitoring included pulse oximetry, noninvasive blood pressure (BP), three-lead electrocardiography and capnography. A second peripheral intravenous line was used to collect blood samples during the anaesthetic period.
Bispectral monitoring and analysis
Following preparation of the skin with alcohol, a BIS sensor was placed on the forehead according to the manufacturer's instructions. The depth of anaesthesia was registered as BIS values (Aspect BIS 14, Aspect Medical Systems, Inc., Norwood, Massachusetts, USA) at the time of clinical LOC.
Electroencephalographic monitoring and analysis
EEG recordings were performed on a Nicolet One Neurodiagnostic system (Viasys, CareFusion Inc., San Diego, California, USA). A bipolar montage of four active electrodes, F3-T3 and F4-T4, was used. Electrode impedance was less than 10 kΩ and the low-pass filter was set at 70 Hz. All recordings were later analysed by the same clinical neurophysiologist, who was blinded to the events at the time of surgery (when the anaesthetic was injected and when signs of clinical anaesthesia were confirmed). The EEG was manually scored in 10-s epochs and classified into five different stages: stage 1: awake; stage 2: paradoxical excitation, dominating β activity (indicating the first effect of the anaesthetic) during at least 50% of the recording time in three 10-s epochs; stage 3: dominating δ activity (frequency 1 to 3 Hz) (anaesthetic state) during at least 50% of the recording time in three 10-s epochs; stage 4: dominating sub-δ activity (<1 Hz) during at least 50% of the recording time in three 10-s epochs; and stage 5: first instance of a burst suppression pattern.10 The longest silent period during the burst suppression pattern was calculated and noted. The EEG ran for 20 min.
Following preoxygenation and during continuous EEG and BIS monitoring, anaesthesia was induced with 10 mg ml−1 of propofol (Propofol Lipuro B. Braun AG, Melsungen, Germany) and 50 μg ml−1 of remifentanil (Remifentanil Actavis, Elaiapharm, Valbonne, France).
To obtain constant effect-site concentrations, the infusions of propofol and remifentanil were administered by two computer-controlled infuser systems: Alaris TIVA 1000 (LB 0029 ISS4) in Finspång/Linköping, and Perfusor Space, Braun, in Kalmar. The effect-site concentrations of propofol and remifentanil were estimated using the three-compartment pharmacokinetic models developed by Schneider et al.11 and Minto et al.12, respectively. The estimated effect-site concentration during induction was set to 6 μg ml−1 for propofol and 6 ng ml−1 for remifentanil.
Clinical LOC was based on the disappearance of a patient's eyelash reflex. During induction, the patients were asked repeatedly every 15 to 30 s to open their eyes. When there was no response to auditory commands, the eyelid reflex was tested. The time of clinical LOC was defined as the moment when patients did not respond to either stimulus.
All patients received a laryngeal mask and the lungs were mechanically ventilated.
Blood samples for the plasma concentration of propofol were obtained at clinical LOC and at 20 and 30 min after clinical LOC. At these points, BIS values and EEG signs of anaesthetic depth were recorded.
Neurophysiological activity was monitored by EEG and BIS during the first 30 min of anaesthesia. The attending anaesthetist and anaesthetic nurse were blinded to the BIS values and EEG recordings.
All patients received bolus doses of morphine or fentanyl at the end of surgery prior to discontinuation of anaesthetic drugs.
The following data were documented: age, BMI, SBP and DBP, times of clinical LOC, induction doses and total drug doses for propofol and remifentanil, BIS values at the time of clinical LOC, and the durations of anaesthesia and surgery. S-albumin was also analysed.
Plasma concentration of propofol
Blood samples for analysing the plasma concentration of propofol were frozen at −70°C. The plasma concentration of propofol was analysed using a validated liquid chromatography method on a Nova-pak C18 column (39 mm × 150 mm). Isocratic elution was performed using acetonitrile and a sodium acetate buffer (55 : 45 v/v) at a flow rate of 1.1 ml min−1. Propofol was detected using fluorescence (with excitation and emission wavelengths of 276 and 310 nm, respectively), and quantification was performed using a calibration curve in plasma between 150 and 10 000 ng ml−1. The intra and inter-batch accuracy, and precision were between 97 and 105%, and below 12%, respectively.
The significance of the relationship between BIS values and EEG patterns was assessed using a Spearman correlation analysis. A P value less than 0.05 was considered statistically significant. All graphs were made using Statistica software (version 13.1, Stat Soft. Inc., Tulsa, OK, USA) and Microsoft Excel.
A total of 41 patients were enrolled in this study; four were excluded because of unreadable EEG registrations, and two were excluded because of problems with time synchronisation between the monitors, leaving 35 for the final analysis. In all, 15 patients (43%) were men. The mean age was 33 ± 9.6 years, and the mean BMI was 24 ± 4 kg m−2. The median time from start of anaesthesia to EEG stage 3 (δ activity) was 3 min, and the median time to clinical LOC was 4 min. The median difference between EEG δ activity and clinical LOC was 1 ± 4 min (see Supplemental Table 1, http://links.lww.com/EJA/A104). One patient received ephedrine because of a drop in blood pressure during induction. The baseline characteristics of the included patients are shown in Table 1.
Analysis of plasma propofol concentrations
The plasma concentrations at clinical LOC varied, ranging from 1787 to 15 134 ng ml−1 (6803 ± 3543 ng ml−1). The propofol concentration in venous plasma reached a steady state within 20 min, as indicated by minor changes in intra-individual concentration (Table 1).
Analysis of bispectral monitoring
A large proportion of patients had low BIS values at the induction of anaesthesia. In total, 54% of the patients had BIS values below 40 at clinical LOC, with a median of 38 (IQR 30 to 43) and a range from 16 to 50 (Fig. 1). At EEG stage 3, BIS values had a wide range (30 to 97), with a median of 53 (IQR 41 to 69; Fig. 2). The median BIS value at baseline was 97 (IQR 95 to 98, range 93 to 99), β activity 95 (IQR 71 to 97, range 24 to 98), sub-δ activity 37 (IQR 30 to 44, range 16 to 94) and at burst suppression 25.5 (IQR 23 to 29, range 18 to 45).
Analysis of continuous electroencephalogram monitoring
The EEG was registered during anaesthesia with the BIS values for all patients. The EEG recordings for all 35 patients were readable and lacked artefacts.
At clinical LOC, one patient (3%) was at EEG stage 2, 15 patients (43%) were at EEG stage 3, 13 patients (37%) were at EEG stage 4 and six patients (17%) were at stage 5 (Fig. 3).
There was no statistically significant correlation between the BIS value and EEG stage at clinical LOC (r = −0.1, P = 0.064; Fig. 1).
This study aimed to assess the depth of anaesthesia in 35 ASA 1 patients undergoing total intravenous anaesthesia with propofol and remifentanil from target-controlled infusion systems for elective day-case surgery.
There was no correlation between anaesthetic depth, as measured by BIS values and clinical LOC. The BIS value at clinical LOC was lower than 38 (IQR 30 to 43) in 54% of patients. At clinical LOC, there was no correlation between BIS and EEG values. These results lead to uncertainty in the ability of BIS values to measure anaesthetic depth. Clinical LOC lags 1 min behind LOC at EEG stage 3.
The depth of anaesthesia is of great importance for patients and much research has been done on awareness. A wide variation in the incidence of awareness has been reported; large studies have reported an incidence of 1 to 2 : 1000 to 1 : 15 000.13–15 Anaesthesia is associated with postoperative dysfunction, especially in elderly patients, but there is no evidence of an association with deep anaesthesia.16,17 However, in 2005, Monk and colleagues18 found that low blood pressure and deep anaesthesia are independently associated with increased postoperative mortality. In addition, accumulated time with low BIS values is an independent predictor of negative postoperative outcomes.19,20 Sessler and colleagues21 have also shown an increased risk of mortality when low values of mean arterial pressure, BIS and minimum alveolar concentration occur simultaneously, called a triple low state. However, recent studies could not find an association between a triple low state and postoperative mortality.22,23 In our study, we found low BIS values in combination with rather high induction doses of propofol. However, we found only one drop in blood pressure during induction in our ASA 1 patients.
In an earlier pharmacogenetic study,24 we investigated the reason for the large inter-individual variation in plasma concentrations of propofol at clinical LOC. However, the pharmacokinetic variation in propofol could not be explained by polymorphism in the metabolism of enzymes or receptors.
Other studies have demonstrated that BIS values correlate well with increasing propofol doses, even if BIS levels at induction are lower than expected.25 In 2013, Martín-Mateos and colleagues26 showed that BIS values fluctuate around an index of 50 and that this fluctuation varies among patients. In our study, approximately 50% of patients demonstrated low BIS values, and in five of these patients, the EEG showed a burst suppression pattern, which signifies overly deep anaesthesia. The mean induction dose of propofol was 3.1 mg kg−1, which is higher than usual and could be because of an absence of premedication or variability in the determination of clinical LOC.
There is a time delay for withdrawing blood for propofol plasma concentrations, both in this study and in our previous studies. This might affect the pharmacokinetics and pharmacodynamics. However, in this study, we only evaluated the plasma concentration and the aim was not to do a pharmacokinetic–pharmacodynamic model as this has already been well reported by several research groups. A two-compartment effect-site model describes the BIS after different rates of propofol infusion.27
Despite the need to achieve and monitor a well tolerated, balanced anaesthetic depth, we still lack the clinical system to do this in a well tolerated and reliable way.28 This might be because such a measurement is influenced by too many confounding factors, providing an even greater reason for continued study in this area. In this study, clinical LOC was based on responses to verbal commands and eyelash reflexes. The patients were repeatedly asked during induction to open their eyes. Our results show a delay of clinical LOC of 1 ± 4 min from EEG stage 3, which is the stage at which surgical anaesthesia takes place. This delay occurs because it takes time to evaluate patients’ auditory responses and eyelash reflexes. Beyond the usual clinical signs of LOC, it can be determined by the ability to place a laryngeal mask.29 Ryu and colleagues also showed that no patient needed more than 0.3 μg kg−1 of remifentanil for a successful laryngeal mask insertion. In our study, a minimum of 0.31 μg kg−1 of remifentanil was administered before clinical LOC.
Other studies have found large inter-individual variations in BIS values measured at clinical LOC in healthy individuals.30 Kaskinoro and colleagues29 described the difficulties of monitoring unconsciousness using an anaesthetic depth monitor applied to the forehead. The thalamus, which is not reflected in BIS values, is involved in the regulation of consciousness.31 Moreover, muscle activity can cause the EEG patterns in the BIS monitor to tend towards higher values, resulting in higher BIS values for patients that are unconscious but who have not received muscle relaxant.32,33 Neuromuscular blockade was not used in our study, leading to possible interference from muscle activity, and consequently unreliable BIS values.
In a recent study, a large variation among EEG patterns measured at clinically evaluated LOCs was found. The EEG patterns varied from β activity to burst suppression.24 Some anaesthetics, including propofol, induce CNS excitation during the anaesthetic initiation phase with increased oscillatory activity in the higher β frequency bands (12.5 to 25 Hz) and decreased activity in slower frequency bands (3.5 to 12.5 Hz).34 This state is marked by a lack of inhibition and a loss of both motor and affective control.35 When patients are more deeply sedated, EEG patterns become slower with increased delta (1 to 3 Hz) and sub-δ (<1 Hz) activity. In some instances, burst suppression patterns also develop.36 Low-frequency Electromyograph (EMG) signals may simulate high EEG signals associated with being awake and superficial anaesthesia (ca. 30 Hz).37 However, the BIS algorithm includes a reduction in the impact of EMG contamination, both in sedation ranges and anaesthesia, and a false elevation of BIS is highly unlikely.
The induction phase of propofol is rapid and clearly visible on the EEG as faster β frequencies. However, the slower θ and δ waves emerge gradually as sedation deepens and they are highly variable between patients during and after LOC.38,39
In this study, EEG activity was reduced with clinical LOC. However, there was no statistically significant correlation between EEG and BIS values. This could be explained by different values of BIS depending on the spectral decomposition of the waves. At clinical LOC, 54% of patients had BIS values below 40, and the majority of patients had BIS values lower than 45 at 20 and 30 min after clinical LOC. The optimal BIS value for well tolerated anaesthesia is between 40 and 60.1 This means that deep anaesthesia was reached in more than 50% of our patients. However, deep anaesthesia could not be verified by EEG.
Despite the fact that EEG is the gold standard for monitoring brain activity, it is not fully clear how EEG algorithms can be used to determine anaesthetic depth. Moreover, even if the National Institute for Clinical Excellence recommends anaesthetic depth monitors, there are still many unanswered questions about the reliability of BIS monitoring.3
The study was small, which made it difficult to draw any definite conclusions. Arterial plasma concentrations of propofol might be more appropriate than venous concentrations and this could explain the somewhat higher concentration of propofol.40 However, we had no ethical approval for an arterial line.
Another limitation is that most indicators such as BIS, EEG patterns and clinical LOC are dependent on a time delay and, therefore, the response to physiological unconsciousness cannot be registered immediately. If this time delay is different between the variables this can become a confounding factor.
There was no statistically significant correlation between BIS value and EEG at clinical LOC. Our findings suggest that BIS monitoring may not be reliable and is in need of further development.
Acknowledgements relating to this article
Assistance with the study: the authors would like to thank research assistant Lena Sundin, Department of Anaesthesia and Intensive Care, University Hospital, Linköping, Sweden.
Financial support and sponsorship: this work was financially supported by grants from the Medical Research Council of Southeast Sweden, the Swedish Research Council and the County Council in Östergötland.
Conflicts of interest: none.
1. Schneider G, Gelb AW, Schmeller B, et al. Detection of awareness in surgical patients with EEG-based indices: bispectral index and patient state index. Br J Anaesth
2. Punjasawadwong Y, Phongchiewboon A, Bunchungmongkol N. Bispectral index for improving anaesthetic delivery and postoperative recovery. Cochrane Database Syst Rev
2014; (6):CD003843. doi: 10.1002/14651858.CD003843.pub3.
3. Pandit JJ, Cook TM. National Institute for Clinical Excellence guidance on measuring depth of anaesthesia: limitations of EEG-based technology. Br J Anaesth
4. Myles PS, Leslie K, McNeil J, et al. Bispectral index monitoring to prevent awareness during anaesthesia: the B-Aware randomised controlled trial. Lancet
5. Ekman A, Lindholm ML, Lennmarken C, Sandin R. Reduction in the incidence of awareness using BIS monitoring. Acta Anaesthesiol Scand
6. Avidan MS, Zhang L, Burnside BA, et al. Anesthesia awareness and the bispectral index. N Engl J Med
7. Avidan MS, Jacobsohn E, Glick D, et al. Prevention of intraoperative awareness in a high-risk surgical population. N Engl J Med
8. Whitlock EL, Villafranca AJ, Lin N, et al. Relationship between bispectral index values and volatile anesthetic concentrations during the maintenance phase of anesthesia in the B-Unaware trial. Anesthesiology
9. Enlund M, Jansson P. A comparison of auditory evoked potentials and spectral EEG in the ability to detect marked sevoflurane concentration alterations and clinical events. Ups J Med Sci
10. Brown EN, Lydic R, Schiff ND. General anesthesia, sleep, and coma. N Engl J Med
11. Schneider TW, Minto CF, Gambus PL, et al. The influence of method of administration and covariates on the pharmacokinetics of propofol in adult volunteers. Anesthesiology
12. Minto CF, Schneider TW, Egan TD, et al. Influence of age and gender on the pharmacokinetics and pharmacodynamics of remifentanil. I. Model development. Anesthesiology
13. Sebel PS, Bowdle TA, Ghoneim MM, et al. The incidence of awareness during anesthesia: a multicenter United States study. Anesth Analg
14. Sandin RH, Enlund G, Samuelsson P, et al. Awareness during anaesthesia: a prospective case study. Lancet
15. Pollard RJ, Coyle JP, Gilbert RL, et al. Intraoperative awareness in a regional medical system: a review of 3 years’ data. Anesthesiology
16. Monk TG, Weldon BC, Garvan CW, et al. Predictors of cognitive dysfunction after major noncardiac surgery. Anesthesiology
17. Krenk L, Rasmussen LS, Kehlet H. New insights into the pathophysiology of postoperative cognitive dysfunction. Acta Anaesthesiol Scand
18. Monk TG, Saini V, Weldon BC, et al. Anesthetic management and one-year mortality after noncardiac surgery. Anesth Analg
19. Leslie K, Myles PS, Forbes A, et al. The effect of bispectral index monitoring on long-term survival in the B-aware trial. Anesth Analg
20. Lindholm ML, Traff S, Granath F, et al. Mortality within 2 years after surgery in relation to low intraoperative bispectral index values and preexisting malignant disease. Anesth Analg
21. Sessler DI, Sigl JC, Kelley SD, et al. Hospital stay and mortality are increased in patients having a ‘triple low’ of low blood pressure, low bispectral index, and low minimum alveolar concentration of volatile anesthesia. Anesthesiology
22. Willingham MD, Karren E, Shanks AM, et al. Concurrence of intraoperative hypotension, low minimum alveolar concentration, and low bispectral index is associated with postoperative death. Anesthesiology
23. Kertai MD, White WD, Gan TJ. Cumulative duration of ‘triple low’ state of low blood pressure, low bispectral index, and low minimum alveolar concentration of volatile anesthesia is not associated with increased mortality. Anesthesiology
24. Khan MS, Zetterlund EL, Green H, et al. Pharmacogenetics, plasma concentrations, clinical signs and EEG during propofol treatment. Basic Clin Pharmacol Toxicol
25. Puri GD, Mathew PJ, Sethu Madhavan J, et al. Bi-spectral index, entropy and predicted plasma propofol concentrations with target controlled infusions in Indian patients. J Clin Monit Comput
26. Martin-Mateos I, Mendez Perez JA, Reboso JA, et al. Modelling propofol pharmacodynamics using BIS-guided anaesthesia. Anaesthesia
27. Bjornsson MA, Norberg A, Kalman S, et al. A two-compartment effect site model describes the bispectral index after different rates of propofol infusion. J Pharmacokinet Pharmacodyn
28. Bruhn J, Myles PS, Sneyd R, et al. Depth of anaesthesia monitoring: what's available, what's validated and what's next? Br J Anaesth
29. Ryu J, Oh AY, Baek JS, et al. Remifentanil dose for laryngeal mask airway insertion with a single standard dose of propofol during emergency airway management in elderly patients. Korean J Anesthesiol
30. Kaskinoro K, Maksimow A, Langsjo J, et al. 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
31. Hudetz AG. General anesthesia and human brain connectivity. Brain Connect
32. Aho AJ, Kamata K, Yli-Hankala A, et al. Elevated BIS and Entropy values after sugammadex or neostigmine: an electroencephalographic or electromyographic phenomenon? Acta Anaesthesiol Scand
33. LeBlanc JM, Dasta JF, Pruchnicki MC, et al. Bispectral index values, sedation-agitation scores, and plasma Lorazepam concentrations in critically ill surgical patients. Am J Crit Care
34. Gugino LD, Chabot RJ, Prichep LS, et al. Quantitative EEG changes associated with loss and return of consciousness in healthy adult volunteers anaesthetized with propofol or sevoflurane. Br J Anaesth
35. Fulton SA, Mullen KD. Completion of upper endoscopic procedures despite paradoxical reaction to midazolam: a role for flumazenil? Am J Gastroenterol
36. Mustola ST, Baer GA, Toivonen JK, et al. Electroencephalographic burst suppression versus loss of reflexes anesthesia with propofol or thiopental: differences of variance in the catecholamine and cardiovascular response to tracheal intubation. Anesth Analg
37. Baldesi O, Bruder N, Velly L, et al. Spurious bispectral index values due to electromyographic activity. Eur J Anaesthesiol
38. Lewis LD, Weiner VS, Mukamel EA, et al. Rapid fragmentation of neuronal networks at the onset of propofol-induced unconsciousness. Proc Natl Acad Sci U S A
39. Sleigh JW, Steyn-Ross DA, Steyn-Ross ML, et al. Comparison of changes in electroencephalographic measures during induction of general anaesthesia: influence of the gamma frequency band and electromyogram signal. Br J Anaesth
40. Wang YP, Cheng YJ, Fan SZ, et al. Arteriovenous concentration differences of propofol during and after a stepdown infusion. Anesth Analg