By: 7 January 2015
EEG – Anaesthesia by numbers

NICE recommended the use of EEG-based depth of anaesthesia monitoring in 2012 and yet less than 2% of anaesthetists in the UK use these devices regularly. Sam McCaffery investigates why the uptake of these monitors is so poor in clinical practice.

For a patient, the idea of being awake and aware during surgery is a terrifying one, even though it is uncommon, and can have serious and long-lasting psychological effects [1,2]. In 2012, the National Institute for Health and Care Excellence (NICE) published guidance recommending the use of EEG-based depth-of-anaesthesia (DOA) monitors for patients receiving total intravenous anaesthesia and in patients considered at higher risk of adverse outcomes during any type of general anaesthesia [3]; however, in a survey conducted since then, it was found that one-third of hospitals have no such monitors and fewer than 2% of anaesthetists in the UK use them regularly [1].

Why are anaesthetists so reluctant to make use of this technology? One factor may be the rarity of accidental awareness during anaesthesia. A study into patients’ spontaneous reports of awareness found that awareness happened in only one in 15,000 cases. However, when patients were proactively asked about awareness after general anaesthesia using the Brice questionnaire, the incidence was between one and two per 1000 cases [4]. The majority of these are brief recollections and do not involve painful experiences. Professor Jaideep Pandit, Consultant Anaesthetist at Oxford University Hospitals and clinical lead for the NAP5 study of accidental awareness during surgery, commented: “One of the things we are looking at closely in the NAP5 study is the proportion of patients who reported pain or distress as part of their experience, and in the baseline study we found only one-third of those who reported awareness reported pain or distress. The other two-thirds reported it in a very neutral way. That means that the incidence of the awareness of experience that is detrimental is less likely. This may explain the difference between spontaneous reporting and the Brice questions; it could be that those who don’t report [awareness] spontaneously find it such a neutral experience that they don’t bother to tell anyone about it unless asked.”

A spontaneous reporting rate of one in 15,000 means that awareness may be expected to occur about once in a clinician’s career, and so it is possible that many anaesthetists believe this is simply not a problem in their practice.

When it comes to DOA monitors, NICE recommended that there was evidence to support the use of three different types of monitor: the E-Entropy, Narcotrend-Compact M and the Bispectral Index (BIS). Of the three it is the BIS monitor that has received the higher recommendation and has become by far the most widely used in hospitals. The other two devices have not received as much clinical validation or testing, and the E-Entropy especially suffers from compatibility issues with other monitoring systems already used in operating theatres.

The BIS system uses a disposable four-electrode sensor on the patient’s forehead to measure the electrical activity in the brain. This EEG data is then processed using a proprietary algorithm to calculate a number between 0 and 100, with 0 representing an absence of electrical activity in the brain and 100 meaning the patient is fully awake. Through clinical research Covidien, who design and produce the BIS, have identified a target range of 40–60 for general anaesthesia. BIS monitors also provide anaesthetists access to the raw EEG waveform in real time should they wish to study this as well.

An important point to note, said Scott Kelley, Chief Medical Officer for Covidien, is that clinical guidance for these devices states that a BIS output reading only reflects the ‘probability’ of unconsciousness, with a reading of between 40–60 indicating a low probability of awareness. This is where many anaesthetists’ issues with these devices start, as it has never been specified what ‘low’ actually means. Jaideep Pandit was critical of NICE in this regard: “Nobody at all is telling us, especially not NICE, what that probability actually is. In other words, what we don’t know and they’re not saying is if the monitor reads 80 the probability of [the patient] being anaesthetised is X% and if reads 70 the probability is Y%, and so on. We’re completely in the dark there.”

While NICE and Covidien do not state what the probabilities relating to the numerical values are, part of Covidien’s core educational strategy for clinicians using BIS is that if the numerical value does go above 60, the first thing the anaesthetist should do is assess the patient, as one of the main issues the monitor could be alerting to is a problem with anaesthetic delivery.

Scott Kelley uses BIS regularly in his anaesthetics practice at Brigham & Women’s Hospital in Boston and thinks that a BIS reading above 60 could be the first sign that the patient is receiving an insufficient dose of anaesthetic, particularly if using IV anaesthetics, which have a higher potential for disruption of anaesthetic delivery. “In that situation, particularly if I’m using an IV anaesthetic, I always check whether my medication is being delivered into the patient’s IV,” he said. “This is one way the technology can further enhance safety: by alerting anaesthesiologists to a disruption of anaesthetic delivery.”

The specific numerical output of the BIS monitors is not the only area causing concern. As Pandit pointed out, the BIS monitor is ‘blind’ to several anaesthetic drugs. Nitrous oxide (NO) and ketamine are two well-established anaesthetic agents but their anaesthetising influence does not show up on the BIS monitor’s numerical value; and, paradoxically, it can actually cause the value to increase.

According to Kelley, this happens because both NO and ketamine have atypical effects on a patient’s EEG: “Those particular two anaesthetics have unique and different EEG signatures. Therefore, because they impact the EEG in a different way, the BIS value may not reflect their contribution to the anaesthetic. Take ketamine for example: when we look at the EEG in those patients it is atypical of the dose-response effects for traditional anaesthetics; it actually activates the EEG and makes it very high frequency and it’s almost erratic. Consequently, the way the BIS system analyses it is as an atypical BIS response.”

When using anaesthetics such as NO and ketamine, Kelley uses the EEG waveform along with the BIS value, and even if the value is above 60 he is often happy with the level of anaesthesia because of the EEG signature.

While there are serious questions about the practical usefulness of DOA monitors in detecting and preventing accidental awareness, one area where many anaesthetists are finding they prove useful is in helping avoid an overdose of anaesthetic agent. Before the introduction of these monitors there was no other device that could measure the effect an anaesthetic drug was having on a patient so anaesthetists might err on the side of caution by giving “too much”, said Kelley. Recently, however, it has started to become clear that too much anaesthetic can be harmful for the patient, leading to an increase in side-effects such as delirium. Kelley believes that monitors like the BIS system will help anaesthestists to be confident that the anaesthetic agent is having an effect at lower concentrations and that this could be the future for such devices.


Sam McCaffery spoke to Scott Kelley, medical director at Covidien and Professor Jaideep Pandit, Consultant Anaesthetist at Oxford University Hospitals and clinical lead for NAP5.

  1. Pandit, J.J., Cook, T.M., Jonker, W.R. & O’Sullivan, E. (2014) A national survey of anaesthetists (NAP5 Baseline) to estimate an annual incidence of accidental awareness during general anaesthesia in the UK on behalf of the 5th National Audit Project (NAP5) of the RCA and the AAGBI. Anaesthesia, in press
  2. Bruchas, R.R., Kent, C.D., Wilson, H.D. & Domino, K.B. (2011)
  3. Clin. Psych. Med. Set. 18, 257–267
  4. (Accessed 23 September 2014)
  5. Avidan, M.S., Zhang, L., Burnside, B.A., et al. (2008) N. Engl. J. Med. 358, 1097–1108