DEPTH OF ANAESTHESIA

This invention generally relates to depth of anaesthesia, and in particular methods of assessing the depth of anaesthesia to assist the administration of hypnosis (sleep inducing) and analgesic (pain relief) agents, which form part of the overall assessment of anaesthesia, using Auditory Evoked Potential (AEP) signals.

In WO 98/10701 (The University Court of the University of Glasgow), there is disclosed a method of obtaining a patient's AEP from his/her EEG signal and deriving an index indicative of the depth of anaesthesia, being a combination of hypnotic, analgesic and surgical stimulation. In this prior specification, the AEP index has been defined as a signal corresponding to the coarseness of the monitored AEP signal and said index is expressed on a scale of 0-100. The disclosure of WO 98/10701 is deemed incorporated herein by way of this reference thereto.

The AEP is a portion of a patient's full EEG signal, being a response from the nervous system and brain stem to stimulation of the auditory system (auditory stimulation). The depth of anaesthesia is the combination of the effect of any hypnotic agents inducing an unconscious state, the effect of analgesic agents reducing the sensation of pain and surgical stimulation which cause arousal in the patient offsetting the hypnotic agents. WO 98/10701 also discloses the use of the AEP index to control the hypnotic element of the anaesthesia by controlling the system delivering the hypnotic agent.

This invention builds upon known existing methods of determining the depth of anaesthesia to produce a more sophisticated and comprehensive monitor and control system for anaesthesia and analgesia.

According to one aspect of this invention, there is provided a method of monitoring a patient's EEG signal while subjecting said patient to auditory stimulation said method including extracting the Auditory Evoked Potential (AEP) waveform from the said EEG signal by means of a computing system and further determining an AEP index based on the said waveform which is in turn further analysed to determine the deviation in said AEP index, this

deviation being indicative of the effectiveness of the analgesic being administered with respect to the requirement for analgesia to overcome stimulation from trauma or surgical stimulation.

Further, in a preferred form the invention also analyses from the AEP index a trend in the change of said index and that trend is used to determine the effectiveness of the hypnotic agent being administered with respect to the requirement for maintaining the conscious state of a patient.

In WO 98/10701 the AEP waveform is produced by averaging 256 consecutive EEG waveforms obtained following auditory stimulation and therefore including the AEP. The specification describes no means of testing that the waveform had been derived from auditory stimulation. Therefore in a case where no auditory stimulation is present, a waveform and associated index could still be produced. In this invention it is proposed that the change in amplitude of the derived waveform in the first third of the waveform be compared to that in the last third when viewing typically a 144 millisecond window following auditory stimulation. If the ratio of the rate of change of amplitude in the first third is less than 130% of the rate of change in the final third, as shown in Figure 5 herein, then the waveform will be rejected as not being typical of an AEP waveform. This would therefore reject waveforms where there has been no auditory stimulation to form the typical peaks in the first third of the waveform and also reject waveforms with interference where peaks due to external electrical interference or muscle movement appear throughout the waveform and form peaks in the final third as well as the first third. As an alternative the waveform can be analysed to determine the location of the peaks and compare this to the typical timing of an AEP waveform following auditory stimulation.

The AEP index fluctuates with the intensity of surgical stimulation; the greater the stimulation, the greater the fluctuation in said index unless the pain is suppressed by analgesic. The more analgesic agent administered, the less the effect of surgical stimulation with respect to arousal and therefore the deeper the anaesthesia for the same level of hypnotic agent. Consequently, the less the effect of surgical stimulation, the smaller the effect on said AEP index.

This invention therefore utilises computer equipment and software to analyse the fluctuations which can be identified as deviation in the AEP index and expresses this as a second index which preferably has a value on a scale of 0 to 100, where the extreme of 0 represents a full analgesic and minimal deviation and the other extreme of 100 represents no analgesic and equating to the maximum deviation clinically experienced in the AEP index. This indicates the level of perceived surgical stimulation and therefore the effective level of analgesic and allowing adjustment of the level of analgesic either manually or via an automatic interface to the delivery system, to arrive at an optimal level for the patient and procedure.

This additional index derived from the deviation in the AEP index can also be referred to as the level of analgesic/stimulation and can expressed as a number in the range of 0-100 which can be suitably displayed for informing the anaesthetist. In one form of the invention outlined here a computer system determines the fluctuation (deviation) in the AEP index and compares this to a preset or patient-specific value, and then adjusts the analgesic level to maintain the analgesic within preset or patient-specific levels. Therefore if there are large variations in the AEP index indicating excessive surgical stimulation then the system may increase the level of analgesia to reduce the effect of surgical stimulation and reduce the variation in the AEP index, which acts as a feedback loop to the analgesic delivery equipment.

Following or simultaneously with the adjustment of the analgesic, if the level of anaesthesia is still not meeting the patient's requirements the level of hypnotic agent can be adjusted to obtain and maintain a stable level required for the surgical procedure. This decision to adjust the hypnotic element can be made by reviewing the AEP index and additionally the gradient of a plot of AEP index over a period of time.

The AEP index can be determined over a period of time and the gradient of the graph produced can be said to represent the trend of the anaesthesia. If the gradient is horizontal or close to zero, the anaesthesia is stable. If the gradient is positive, then the anaesthesia is becoming lighter and additional

- A -

anaesthetic or hypnotic agent is required to be administered to maintain a constant level of anaesthesia. If the gradient is negative, then the anaesthesia is becoming deeper and the amount of anaesthetic or hypnotic agent should be reduced to maintain a constant level of consciousness. These parameters can be used to adjust the level of anaesthetic or hypnotic agent being delivered by interfacing with the syringe pump or other delivery system.

The gradient of the AEP trend can be expressed as an index. This gradient or hypnotic index can be displayed to the anaesthetist for information; the mid point of the index range would indicate stable anaesthesia. Additionally the gradient or hypnotic index and the deviation or analgesic index can be adjusted proportionately so that they form a balance in keeping with the AEP index, representing an overall index of anaesthesia.

It can therefore be seen that when the graph of the AEP index is analysed a short-term variation or deviation is normally indicative of stimulation and analgesic, whereas a longer-term trend or gradient is indicative of the hypnotic element of the anaesthesia. The additional indexes described herein can be calculated and displayed to communicate the information in a simple format to the anaesthetist while either the indexes or the raw data can be used by the system automatically to control the delivery systems for hypnotic and analgesic agents.

A number of alternative ways could be devised to indicate visually or numerically the relationships between the hypnotic and analgesic elements of the overall AEP index, once they have been calculated.

A constant auditory stimulation of around 7 Hz would be too intensive for long term sedated patients such as may occur during so called intensive care or intensive therapy, where monitoring could be required for periods of days or weeks with a relatively stable level of sedation or light anaesthesia, or in the case of unconsciousness due to coma for example resulting from viral or traumatic cause. It is therefore proposed to introduce periods without stimulation and to monitor the depth of sedation only periodically, dependent upon the longer-term stability of the results, which may be determined by reference to the short term deviation of the graph and the longer term gradient

of the AEP index relative to time as illustrated in Figures 2 and 3. This therefore extends the current technology from the area of anaesthesia to that of sedation, such as may be required in intensive care.

In one embodiment of the invention outlined here, the computer system monitoring the patient would determine alarm or action levels above and below the AEP graph with respect to time. Over an initial fifteen minute period, for example, the system may determine a weighted average value and then set alarm levels typically either 5 points (on a scale of 100) above and below the average or 2 points above and below the maximum and minimum values within the fifteen minute period, whichever gives the widest range requesting confirmation from the clinician. Alternatively, the clinician could set his/her own alarm or action levels. If the AEP index is well within the alarm or action limits, the frequency of stimulation or the frequency of the bursts of the stimulation may be reduced. If the AEP index approaches within one point of the alarm or action levels, then the frequency or duration of periods without stimulation and subsequent monitoring of the depth of sedation may be reduced progressively until a level of constant monitoring is achieved. During this period an alarm could be sounded to attract attention to the change in clinical conditions. If the actual AEP index returns to normal or the alarm level is moved to accept the current clinical conditions then the alarm would cease and the level of auditory stimulation may be reduced, either by reducing the frequency of 7 Hz or by introducing periods of silence between bursts of stimulation; such bursts typically may last for 37 seconds with 256 readings taken at 7 Hz.

In an embodiment of the invention outlined here, following a pause in the auditory stimulation, the stimulation may commence at a low level and build up to a normal operating level, to reduce the arousal effect of a sudden noise.

Systems are known for controlling the flow or volume of hypnotic or analgesic agent based on the known pharmokinetic response of the anaesthesia or analgesic agents, as well as on patient details such as age, weight and gender; these systems are referred to as Target Controlled Infusion (TCI) systems. In another embodiment of this invention, a TCI is adjusted by the computing system based on the analysis of the AEP to produce the indexes,

as described above. It is proposed that the adjustment of the pharmokinetic derived dose within preset safety limits allows an element of closed loop control in conjunction with TCI, avoiding the dangers associated with total closed loop control and the inaccuracy of general pharmokinetic derived TCI dosage due to ethnic and individual patient variations. This invention thus allows the benefit of adjustment of the hypnotic or analgesic dispensing system to allow response to patient, trauma or surgical needs in a better manner than that normally associated with TCI.

In a typical application of this technology by way of explanation, it is proposed that the signals generated from the monitoring system adjust within preset safe limits the flow to the patient of hypnotic and/or analgesic agents, to maintain a stable hypnotic and analgesic state.

In the embodiment outlined here by way of example only (see Figure 1 ), the equipment 1 generates auditory stimulation. This is played to the patient 2 via earphones or electrodes for bone stimulation, in response to which the patient generates an Auditory Evoked Potential within an EEG signal.

The patient's EEG is filtered at stage 3 to reduce noise and is then processed at stage 4 by circuitry or software to obtain an AEP signal from which the AEP index is derived by software at stage 5 by a rolling average process, typically averaging the results of the last 256 EEG waveforms.

Figure 3 shows a typical AEP index plotted with respect to time. The trend of the AEP index based on the gradient of the graph as shown in Figure 3 is determined by circuitry or software at stage 6 (Figure 1 ). If the trend is negative, the level of anaesthesia is deepening and the volume of hypnotic agent can be reduced; if the trend is positive (section B), then the level of anaesthesia is lightening and the volume of hypnotic agent may need to be increased.

The deviation in the AEP index over time is determined by software at stage 7. If the deviation is high then the level of analgesia can be considered to be too low and can be increased. This will in effect deepen the level of anaesthesia and therefore the computing system could trade-off adjustments to the level of hypnotic agent against adjustments to the level of analgesic agent.

Conversely, if the deviation is small the level of analgesia is suitable for the current stimulation and need not be increased, and therefore the level of anaesthesia may have to be balanced by the level of hypnotic agent with no trade-off between the two elements being possible. The deviation in the AEP index is fed to an analgesia controller 9 to adjust the flow of analgesic agent. A low deviation will reduce the flow while a higher deviation will increase the flow, within defined limits compared to the theoretical TCI value.

Software analysis at stage 8 of the AEP index based on the actual level of the index and the gradient of the graph (trend) as illustrated in Figure 3 would adjust the flow of anaesthetic agent 10, from a predetermined TCI value.

The trend and deviation derived from stages 6 and 7 are fed to software in stage 1 1 , which if set to intensive care mode by the operator will process the data to determine the level of auditory stimulation and change the actual frequency of stimulation or the time between bursts of stimulation to meet the patient's need for monitoring, as shown in stage 12. The computer system monitoring the patient would determine preset levels above and below the AEP graph with respect to time and over an initial fifteen minute period determine a weighted average value and then set alarm levels either 5 points (on a scale of 0-100) above and below the average, or 2 points above and below the maximum and minimum values within the fifteen minute period, whichever gives the widest range. Alternatively, a clinician could set his/her own alarm levels. If the AEP index is well within the alarm limits, the frequency of stimulation or the frequency of the bursts of stimulation may be reduced. If the AEP index approaches within one point of the alarm levels then the frequency of periods without stimulation and subsequent monitoring of the depth of sedation could be reduced progressively until a state of continuous monitoring is achieved. During this period an alarm could be sounded to attract attention for manually changing the clinical conditions, if required. If the actual AEP Index returns to normal or the alarm level is moved to accept the current clinical conditions, then the alarm would cease and the level of auditory stimulation would reduce either by reducing the frequency of 7 Hz or by introducing periods of silence between

bursts of stimulation, such bursts typically lasting for 37 seconds so allowing the taking of 256 readings at 7 Hz. Additionally, the automatic system for adjustment as described above would continue within the constraints of the allowable TCI limits. If selection at step 1 1 is set to operating room mode, the device will generate auditory stimulation at the nominal frequency of typically 7 Hz. This in turn triggers the equipment 1 to produce the stimulation.

The AEP waveform is displayed for reference and is scaled against the largest acceptable amplitude of an AEP waveform. The scale may further be adjusted if the AEP index at that time was not at maximum value possible for an awake patient.

In this embodiment of the invention, the software in stages 3 and 4 determines the amplitude of the incoming signal and rejects signals which are known to be larger than those which are typically produced by a patient. Additionally, signals of a frequency not typical of EEG are filtered out. Also, the resulting AEP waveform produced by averaging the subsequent 256 EEG waveforms is analysed. If the rate of change in amplitude of the waveform in the first third is similar to that in the final third, as shown in Figure 5, then the waveform should be rejected from inclusion in the calculations as not being typical of an AEP waveform; generally, such a waveform has a greater rate of change of amplitude in the first third of the waveform compared to the last third, when viewing a 144 millisecond window following auditory stimulation, as shown in Figure 4. If the ratio of the rate of change of amplitude in the first third is less than 130% of the rate of change in the final third, the waveform should be rejected. This would also allow the system to determine an optimum value to achieve a recognisable AEP index waveform, thus allowing it to operate at volumes lower than might otherwise be required.

The system may be configured to allow adjustment of syringe pumps for intravenous medication and also of vaporisers for volatile hypnotic and analgesic agents.

The computing system may also be able to validate the indexes produced against a patient's autonomic reactions, such as temperature, pulse, oxygen

saturation, spontaneous breathing, skin resistance (sweating) and blood pressure, as well as feed back from the anaesthesia system such as the concentration of expired anaesthetic agents.

The computing system may be configured for computing the AEP index typically over 256 stimulations (37 seconds) to produce a stable figure but also over a shorter period of typically 64 stimulations (9 seconds) to produce an indication of changes in events and to allow the system to set alarms and indicators at an earlier stage.

The scaling of the AEP waveform on the anaesthetist's display can be initially based on the amplitude of the signal but then adjusted on the basis of the AEP index. Therefore, a signal which computes to an index of 50 could be scaled so that an index of 100 based on a geometrically similar waveform would achieve full scale deflection. This allows an understanding of the waveform to be more intuitive.