Annual Scientific Meeting, Belfast; November 2000.

G J McCarthy

In 1847 only months after the demonstration of ether, anaesthesia was already being described in terms of stages or degrees leading to the notion of ‘Light’ and ‘Deep’ anaesthesia. It was immediately appreciated that there could be changes in the ‘Depth’ of anaesthesia which were not always clearly demarcated¹. Successive workers then described a variety of physical signs which were said to characterise the changing state of anaesthesia, many of these depending on muscle tone or reflexes. Of course, with the introduction of muscle relaxants in the mid 1940’s these signs became more difficult to apply. This is reflected in a recent study which reported an incidence of awareness with muscle relaxants of 0.18% but only 0.1% without them.²

Why should this be more of a worry with Total Intravenous Anaesthesia (TIVA) as distinct from inhalational anaesthesia? There are several aspects to the answer. Firstly, it has been observed that nitrous oxide and opiate anaesthesia has the highest incidence of problems³. Secondly, measurement of end-tidal anaesthetic agent concentration is not only a safeguard against failure to actually deliver the agent, but it also gives an indication of agent concentration in the patient’s brain. Contrast this with TIVA, where there is not only a much wider variability in the concentration of the intravenous agent in the plasma, but there is also no immediate way of actually measuring the plasma concentration. This last fact may make intravenous drugs more vulnerable to a ‘dose error’, especially when this is based on crude body weight⁴. A quick calculation, however, reveals that an awareness difference would be difficult to justify. Take the above incidence of awareness of roughly 0.2% and make the unjustified assumption that TIVA might be 50% more likely to result in awareness; with equal numbers of patients in each group and an a of 0.05, a power of 80% and using a two tailed test, it would require a study of roughly 24000 patients in each group to address the question. No such study exists.

One response to this would be to search for some way to measure the degree of the ‘anaesthetic state’ to allow the titration of drug to effect. Since the 1950’s clinical attention has focussed on the use of autonomic reflexes as a guide to depth of anaesthesia, but just as with the muscle relaxants, these reflexes can be suppressed by drugs which do not suppress consciousness. The goal is now to find a monitor which at the very least, measures the reversible drug induced alteration of brain function, which results in the desired lack of awareness to external stimuli but with an absence of any recall when the drug is removed. This last requirement is a much greater task than it first seems, because recall actually consists of at least two components the ‘Explicit’ and ‘Implicit’⁵. The former is where the patient has a clear recollection of intraoperative events, the latter is where the patient has learned under anaesthesia, modifying his behaviour, but in the waking state not consciously remembering events, at least not immediately. Simply adding extra drugs, specifically for amnesia, to the technique does not always help^2,3. Additional problems are also that different drugs may have different modes of action, combinations may interact in different ways and of course the effect being measured may vary unpredictably from patient to patient.

What monitoring techniques are available? The classical method is to use the isolated forearm technique, in which muscle relaxant is excluded from a limb so that the patient can make a motor response to command. This is limited by the duration for which the limb can be isolated and is really useful only for detecting awareness⁶ and not measuring the depth of anaesthesia beyond this point. More recent efforts have concentrated on the availability of computer processing power to manipulate the EEG⁷, evoked potentials of some kind⁸, or heart rate variability⁹ (HRV), each technique having its own advantages and drawbacks. These techniques have been exploited in two commercially available monitors, the ‘Fathom’ (Amtec Medical Limited) which is based on the use of HRV and the ‘Aspect BIS (Bispectral Index)’ monitor (Aspect Medical Systems) based on the EEG.

The Aspect should more properly be regarded as an ‘expert system’ rather than just a monitor. Its output consists of a dimensionless number between 0 and 100 based on the extent of phase coupling between different components in the EEG signal as interpreted in the light of an EEG database of the ‘Sleep/wake’ transition in 2000 patients. A simple electrode is applied to the patients forehead and the time over which the signal is to be averaged is entered. The Aspect can cope with poor amplitude signals, but is susceptible to various types of interference and of course is based on the premise that the patient has a normal EEG before hand. It does not work well with all drug combinations. The Fathom does not use cortical activity directly but depends on the influence of respiration on the brain stem and the resulting change in heart rate. A sinus tachycardia does not invalidate the monitor, but atrial fibrillation, and diabetic neuropathy are among the factors which may do so. Considerable experience is now available with BIS monitor but at this stage experience with the Fathom monitor is very limited.

Avoiding awareness during TIVA depends in the first instance on impeccable technical delivery of anaesthesia, avoidance of the use of muscle relaxants, and an appreciation of the pharmacokinetics and dynamics of the agent being employed. The ideal depth of anaesthesia monitor has yet to be invented but there are exciting new possibilities in titration of dose to effect.

References:

1. Snow J. On the inhalation of Ether. 1847.London:Churchill.3pp.

3. Ghoneim M. Awareness during anesthesia. Anesthesiology 2000;92:597-602.

4. Bouillon T, Shafer S. Does size matter? Anesthesiology 1989; 89:557-559.

5. Veselis R. Memory function during Anesthesia. Anesthesiology 1999;90:648-650

7. Rampil I. A primer for EEG Signal Processing in Anesthesia. Anesthesiology 1998;89:980-1002.