Last updated 2nd Nov 2022 - Tom Heaton
Measurement of the depth of anaesthesia is of great interest to anaesthetists.
A number of non-specific features are commonly used (physiological parameters, MAC), but have significant limitations.
It is generally accepted that consciousness is an emergent property of the brain’s activity, and therefore it would be logical that EEG information can be used to assess the depth of anaesthesia.
Interpretation of raw data requires specialist knowledge, but can be useful for anaesthetists clinically.
In addition, there are forms of processed EEG data which can be more practical.
A number of non-specific features are commonly used (physiological parameters, MAC), but have significant limitations.
It is generally accepted that consciousness is an emergent property of the brain’s activity, and therefore it would be logical that EEG information can be used to assess the depth of anaesthesia.
Interpretation of raw data requires specialist knowledge, but can be useful for anaesthetists clinically.
In addition, there are forms of processed EEG data which can be more practical.
EEG
A basic understanding of the EEG is useful to aid understanding of the different parts of depth of anaesthesia monitoring.
It is covered in more detail in the notes here:
This video provides a useful introduction to EEG overall: https://www.youtube.com/watch?v=XMizSSOejg0
In essence, there are a broad number of different waveforms of differing frequencies that can be categorised into 4 main groups:
Fourier analysis allows the conversion of the raw EEG data into these component parts.
This allows analysis of the different component waveforms which provide information on the underlying brain state.
Lower frequencies of increased amplitude tend to predominate with increasing depth of anaesthesia.
As a rough guide to EEG features:
Awake - high beta (20-30 Hz)
Light GA - Shift to low beta, alpha, non-slow wave anaesthesia
Adequate GA - low beta/alpha/theta wave and slow delta waves
Excessive anaesthesia - burst suppression and persistent suppression
This is a good video on the basics of anaesthetic implementation: https://www.youtube.com/watch?v=K2dMqs7GfuI
The common EEG changes relate mainly to propofol and volatile anaesthesia.
Some of the other agents can produce different EEG features which may complicate interpretation.
It is covered in more detail in the notes here:
This video provides a useful introduction to EEG overall: https://www.youtube.com/watch?v=XMizSSOejg0
In essence, there are a broad number of different waveforms of differing frequencies that can be categorised into 4 main groups:
- Beta (12-30)
- Alpha (8-12 hz)
- Theta (4-8 Hz)
- Delta (0-4 Hz)
Fourier analysis allows the conversion of the raw EEG data into these component parts.
This allows analysis of the different component waveforms which provide information on the underlying brain state.
Lower frequencies of increased amplitude tend to predominate with increasing depth of anaesthesia.
As a rough guide to EEG features:
Awake - high beta (20-30 Hz)
Light GA - Shift to low beta, alpha, non-slow wave anaesthesia
Adequate GA - low beta/alpha/theta wave and slow delta waves
Excessive anaesthesia - burst suppression and persistent suppression
This is a good video on the basics of anaesthetic implementation: https://www.youtube.com/watch?v=K2dMqs7GfuI
The common EEG changes relate mainly to propofol and volatile anaesthesia.
Some of the other agents can produce different EEG features which may complicate interpretation.
Processed EEG
There is probably little scope for using a full EEG setup in theatre, or to have anaesthetists fully trained in EEG interpretation.
As such, there are a number of processed EEG technologies that are used instead.
These minimise both the EEG intake (usually just a forehead strip) and the output to make it much more practical for general theatre use.
As such, there are a number of processed EEG technologies that are used instead.
These minimise both the EEG intake (usually just a forehead strip) and the output to make it much more practical for general theatre use.
Proprietary Algorithms
There are a number of proprietary forms of EEG analysis that are available for depth of anaesthesia monitoring.
BIS is the one that seems to be most widely used in my practice and will be the focus here.
BIS is the one that seems to be most widely used in my practice and will be the focus here.
Bispectral Index (BIS)
Bispectral index (BIS) uses a mathematical algorithm to convert the raw EEG data into a single numerical value (although additional information is available).
It is dual channel and is set up to filter out much of the interference from theatre equipment.
This video provides a good overview: https://www.youtube.com/watch?v=096eAC1WMnM
The monitoring is achieved through a single forehead strip.
There is some degree of lag (up to 30s) for the numerical change, although changes should be immediately visible in the EEG signal.
The main numerical output is on a scale of 0-100.
When looking at BIS interpretation in more detail it is useful to look at the 4 S’s
Signal Quality
This is the percentage of EEG signals that goes into calculating processed variables.
This should be close to 100 (may be displayed as green bars) and if the signal quality isn’t adequate then it may not display a number.
BIS has an EMG assessment built in that detects the higher frequency activity of frontalis muscle activity.
This may indicate frowning or the like, but can just be related to other electrical interference and so could be a clue to interference.
Of note, the ‘raw’ EEG that the monitor displays is actually filtered.
If you wished to do more interpretation of the EEG waveform yourself you would need to turn the filtering off through the settings.
Suppression Ratio
This is a calculation of the proportion of time in which the EEG is suppressed i.e. flat.
This should be 0 as more implies excess anaesthesia.
Spectral Edge Frequency
EEG is the summation of lots of different waves.
Fourier transformation converts this into constituent waveforms.
The waves below the SEF (a wave frequency value provided by the machine) make up 95% of the total waveforms.
As such it tells which wave frequencies are dominating the signal.
A SEF of 12 tells us that 95% of the waveforms are below a frequency of 12 Hz.
Sweep Speeds
Slower speeds will give an appreciation of slower waves e.g. 6.25mm/s - therefore giving a better appreciation of the slow wave anaesthesia patterns
Faster speeds give a better appreciation of the dominant waves,
BIS scores seem to correlate well with cerebral metabolic rate.
As such, factors which affect this can correspondingly affect the BIS score.
Increases may be seem with fever and ketamine, whilst hypoxia or hypothermia can drop the score.
It is dual channel and is set up to filter out much of the interference from theatre equipment.
This video provides a good overview: https://www.youtube.com/watch?v=096eAC1WMnM
The monitoring is achieved through a single forehead strip.
There is some degree of lag (up to 30s) for the numerical change, although changes should be immediately visible in the EEG signal.
The main numerical output is on a scale of 0-100.
- 90s-100 - Awake
- > 70 - light sedation
- 60-70 - deep sedation
- 40-60 - General anaesthesia
- 40 - deep hypnotic state with burst suppression
- 0 - No electrical activity
When looking at BIS interpretation in more detail it is useful to look at the 4 S’s
- Signal quality
- Suppression ratio
- Spectral edge frequency
- Sweep speeds
Signal Quality
This is the percentage of EEG signals that goes into calculating processed variables.
This should be close to 100 (may be displayed as green bars) and if the signal quality isn’t adequate then it may not display a number.
BIS has an EMG assessment built in that detects the higher frequency activity of frontalis muscle activity.
This may indicate frowning or the like, but can just be related to other electrical interference and so could be a clue to interference.
Of note, the ‘raw’ EEG that the monitor displays is actually filtered.
If you wished to do more interpretation of the EEG waveform yourself you would need to turn the filtering off through the settings.
Suppression Ratio
This is a calculation of the proportion of time in which the EEG is suppressed i.e. flat.
This should be 0 as more implies excess anaesthesia.
Spectral Edge Frequency
EEG is the summation of lots of different waves.
Fourier transformation converts this into constituent waveforms.
The waves below the SEF (a wave frequency value provided by the machine) make up 95% of the total waveforms.
As such it tells which wave frequencies are dominating the signal.
A SEF of 12 tells us that 95% of the waveforms are below a frequency of 12 Hz.
Sweep Speeds
Slower speeds will give an appreciation of slower waves e.g. 6.25mm/s - therefore giving a better appreciation of the slow wave anaesthesia patterns
Faster speeds give a better appreciation of the dominant waves,
BIS scores seem to correlate well with cerebral metabolic rate.
As such, factors which affect this can correspondingly affect the BIS score.
Increases may be seem with fever and ketamine, whilst hypoxia or hypothermia can drop the score.
Ketamine
Ketamine has a different pharmacological and clinical profile from the volatile anaesthetic agents and propofol, and the EEG changes reflect this.
Whilst it does have GABAa potentiation, it is also a major NMDA antagonist and have different anatomical foci of action.
At induction it induces high frequency beta waves with subsequent increase in theta and delta activity.
This increased high frequency activity can lead to processed EEG technology ascribing a higher number than is clinically the case.
Additional classes of drugs that may interfer with processed EEG interpretation include NO2 and the alpha 2 adrenoceptor agonist (e.g. clonidine).
Ketamine has a different pharmacological and clinical profile from the volatile anaesthetic agents and propofol, and the EEG changes reflect this.
Whilst it does have GABAa potentiation, it is also a major NMDA antagonist and have different anatomical foci of action.
At induction it induces high frequency beta waves with subsequent increase in theta and delta activity.
This increased high frequency activity can lead to processed EEG technology ascribing a higher number than is clinically the case.
Additional classes of drugs that may interfer with processed EEG interpretation include NO2 and the alpha 2 adrenoceptor agonist (e.g. clonidine).
Entropy
This monitor uses both raw EEG data and information from frontalis EMG.
The output is a similar scale to that seen in BIS.
A difference is the separate outputs of response entropy and state entropy.
The output is a similar scale to that seen in BIS.
A difference is the separate outputs of response entropy and state entropy.
Cautions
There are some understandable cautions that need to be taken into account when deciding on using depth of anaesthesia monitoring.
Some of the more adventurous anaesthetists have demonstrated some of the limitations of the technology previously: http://www.rapidsequence.org.uk/blog/the-bis-is-dead-long-live-the-bis
- There is a lag time in creation of the processed number.
- Caution needed with N2O, ketamine or dexmedetomidine - they increase high frequency power signals and so can lead to a higher BIS.
- Artefact can interfere with analysis.
- It does not take into account baseline brain status.
- Monitor strip can cause pressure damage if used prone
Some of the more adventurous anaesthetists have demonstrated some of the limitations of the technology previously: http://www.rapidsequence.org.uk/blog/the-bis-is-dead-long-live-the-bis
Advantages
The use of such monitoring does appear to have some clear benefits beyond other methods for monitoring depth of anaesthesia.
It is clearly an additional guard against awareness under anaesthesia, and this is reflected in the guidance in its use.
This is particularly relevant when neuromuscular blocking agents are used with TIVA. https://anaesthetists.org/Portals/0/PDFs/Guidelines%20PDFs/Recommendations%20for%20standards%20of%20monitoring%20during%20anaesthesia%20and%20recovery%202021.pdf?ver=2021-05-26-141701-007
In addition, the ability to know the ‘end organ’ degree of anaesthesia seems to increase the confidence of anaesthetists to reduce excess anaesthetic dosing.
This appears to reduce the total dose of anaesthetic given (volatile and TIVA) with corresponding reductions in some of the expected adverse effects.
The causal relationship of this has been a focus of research because of the many confounding factors e.g. increasing age, hypotension.
Regardless, it seems likely that excessive anaesthesia brings little benefits with clear associated detriments, and being able to titrate more effectively is an advantage.
Chan et al (2013) of the CODA was a major study that demonstrated a reduced postoperative delirium risk in higher risk patients.
This was suppported by a following Cochrane review - https://www.cochrane.org/CD011283/ANAESTH_optimized-anaesthesia-depth-guided-brain-electrical-activity-protection-against-postoperative
Previous work has shown that excessive anaesthesia compounds the adverse effects associated with hypotension (the triple low) to the extent that it actually increases mortality.
There has been some discussion about the causal nature of this with the argument that low BIS is reflective of a relative anaesthetic overdose which leads to some organ dysfunction.
This was not clearly demonstrated in the BALANCE study however, suggesting that the range of ‘safe’ anaesthetic depth may be broader than first worried.
This seems to provide some reassurance, but it is clear that even the ‘deeper’ anaesthesia group in this study were not dramatically over-anaesthetised (average BIS 38).
In addition, death is a pretty concerning endpoint and perhaps not the first marker of harm to be alert to.
It is clearly an additional guard against awareness under anaesthesia, and this is reflected in the guidance in its use.
This is particularly relevant when neuromuscular blocking agents are used with TIVA. https://anaesthetists.org/Portals/0/PDFs/Guidelines%20PDFs/Recommendations%20for%20standards%20of%20monitoring%20during%20anaesthesia%20and%20recovery%202021.pdf?ver=2021-05-26-141701-007
In addition, the ability to know the ‘end organ’ degree of anaesthesia seems to increase the confidence of anaesthetists to reduce excess anaesthetic dosing.
This appears to reduce the total dose of anaesthetic given (volatile and TIVA) with corresponding reductions in some of the expected adverse effects.
The causal relationship of this has been a focus of research because of the many confounding factors e.g. increasing age, hypotension.
Regardless, it seems likely that excessive anaesthesia brings little benefits with clear associated detriments, and being able to titrate more effectively is an advantage.
Chan et al (2013) of the CODA was a major study that demonstrated a reduced postoperative delirium risk in higher risk patients.
This was suppported by a following Cochrane review - https://www.cochrane.org/CD011283/ANAESTH_optimized-anaesthesia-depth-guided-brain-electrical-activity-protection-against-postoperative
Previous work has shown that excessive anaesthesia compounds the adverse effects associated with hypotension (the triple low) to the extent that it actually increases mortality.
There has been some discussion about the causal nature of this with the argument that low BIS is reflective of a relative anaesthetic overdose which leads to some organ dysfunction.
This was not clearly demonstrated in the BALANCE study however, suggesting that the range of ‘safe’ anaesthetic depth may be broader than first worried.
This seems to provide some reassurance, but it is clear that even the ‘deeper’ anaesthesia group in this study were not dramatically over-anaesthetised (average BIS 38).
In addition, death is a pretty concerning endpoint and perhaps not the first marker of harm to be alert to.
Links & References
- ICE-TAP - international consortium. http://icetap.org/
- Nimmo, A. et al. Guidelines for the safe practice of total intravenous anaesthesia (TIVA): Joint Guidelines from the Association of Anaesthetists and the Society for Intravenous Anaesthesia. Anaesthesia. 2019. 74(2):211-224. https://www.ncbi.nlm.nih.gov/pubmed/30378102
- Moeller, J. Introduction to EEG. Youtube. 2014. https://www.youtube.com/watch?v=XMizSSOejg0
- Stumpfle, R. Sharp, R. BIS, Entropy and evoked potentials. e-LFH. 2012.
- EEG for Anesthesia. EEG for Anesthesiology - Part 1: Introduction to the EEG for Anesthesiology. Youtube. 2018.
- EEG for anesthesia. EEG for anesthesiology. Youtube. 2018. https://www.youtube.com/watch?v=Alp7n5IKTmU
- Magee, P. Physics for anaesthesia. BJA Education. 2018. 18(4):102-108. https://bjaed.org/article/S2058-5349(18)30018-0/fulltext
- Avidan, M. EEG anaesthesia clinical decision making. Youtube. 2016. https://www.youtube.com/watch?v=JJw_xA-glAA
- Learn with NUH. Consciousness monitoring during anaesthesia (BIS) technology. Youtube. 2019. https://www.youtube.com/watch?v=096eAC1WMnM
- Chan, M. et al. BIS-guided anesthesia decreases postoperative delirium and cognitive decline. Journal of neurosurgical anesthesiology. 2013. 25(1):33-42. https://pubmed.ncbi.nlm.nih.gov/23027226/
- Punjasawadwong, Y. et al. Optimized anaesthesia depth guided by brain electrical activity for protection against postoperative delirium and cognitive dysfunction in adults. Cochrane. 2018. https://www.cochrane.org/CD011283/ANAESTH_optimized-anaesthesia-depth-guided-brain-electrical-activity-protection-against-postoperative
- Chambler, D. BALANCE. The Bottom Line. 2019. https://www.thebottomline.org.uk/summaries/pom/balance/
- Association of Anaesthetists. Recommendations for standards of monitoring during anaesthesia and recovery. 2021. https://anaesthetists.org/Portals/0/PDFs/Guidelines%20PDFs/Recommendations%20for%20standards%20of%20monitoring%20during%20anaesthesia%20and%20recovery%202021.pdf?ver=2021-05-26-141701-007
- Leslie, K. Low bispectral index values and death. Anesthesia and analgesia. 2011. 113(3):160-63. https://journals.lww.com/anesthesia-analgesia/fulltext/2011/09000/low_bispectral_index_values_and_death__the.35.aspx
