This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (5) Citing Articles via Google Scholar Google Scholar Articles by Nishiyama, T. Articles by Hanaoka, K. Search for Related Content PubMed PubMed Citation Articles by Nishiyama, T. Articles by Hanaoka, K. Related Collections Equipment Monitoring (Non-cardiac) Technology

---

TECHNOLOGY, COMPUTING, AND SIMULATION

A Comparison of the Clinical Usefulness of Three Different Electroencephalogram Monitors: Bispectral Index, Processed Electroencephalogram, and Alaris Auditory Evoked Potentials

From the Department of Anesthesiology, The University of Tokyo, Tokyo, Japan, and the *Department of Anesthesia, Yamanashi University, Medical School, Yamanashi, Japan

Address correspondence and reprint requests to Tomoki Nishiyama, MD, PhD, 3–2-6–603, Kawaguchi, Kawaguchi-shi, Saitama, 332–0015, Japan. Address email to nishit-tky{at}umin.ac.jp

	Abstract

Top Abstract Introduction Methods Results Discussion References

 
We compared the usefulness of the Bispectral Index (BIS), Processed electroencephalogram (pEEG), and Alaris auditory evoked potentials (A-AEP). Ninety females scheduled for mastectomy were divided into three groups. Anesthesia was induced with propofol and fentanyl to insert a laryngeal mask airway (LMA) and was maintained by adding nitrous oxide. EEG was monitored by either BIS, spectral edge frequency by pEEG, or A-AEP index by A-AEP. We recorded the number of patients with impedance low enough to extract good EEG signals at the first electrodes application (success rate), the number with an index outside of the range considered appropriate for general anesthesia (inappropriateness rate), changes of the index by LMA insertion or surgical incision (responsiveness), and time to return to good EEG signals after signal disturbance by electric cautery (recovery time). The success rate was larger in BIS A-AEP > pEEG. The inappropriateness rate was smaller in A-AEP BIS pEEG. The A-AEP group showed the largest responsiveness. The recovery time was shorter in pEEG < A-AEP < BIS. In summary, the BIS had the largest success rate, the A-AEP had the least inappropriateness rate and the largest responsiveness, and the pEEG had the fastest recovery time.

IMPLICATIONS: We compared the usefulness of three electroencephalogram monitors. The Bispectral Index was the easiest for obtaining low impedance, the auditory evoked potential index had the least inappropriateness rate for general anesthesia and had the largest responsiveness, and the spectral edge frequency was the fastest in stabilizing measurement after electric cautery.

	Introduction

Top Abstract Introduction Methods Results Discussion References

 
The electroencephalogram (EEG) is an established method for indicating the level of consciousness during anesthesia. Specific EEG indexes, shown as simple numbers, have been developed for clinical application. The 90% spectral edge frequency is the frequency below which 90% of the power in the spectrum resides (1). The Bispectral Index (BIS) depends on the coherence between different frequency components of the frontal EEG waveform (2). Mid-latency auditory evoked potentials (AEP) are small changes noted on EEG caused by discrete auditory stimuli. To extract the AEP from the background EEG activity, 250 to 1000 stimuli are usually necessary. However, an autoregressive model with exogenous input has enabled its extraction within 15 stimuli (3). The three EEG monitors commercially available in clinical anesthesia today include BIS^TM, processed EEG^TM (pEEG, to measure the 90% spectral edge frequency), and Alaris AEP^TM (A-AEP). Many studies have been performed to evaluate the validity of each monitor. However, there has been no comparative study of their practical usefulness. Therefore, in the present study, we compared the practical usefulness of the BIS, pEEG, and A-AEP using the following variables: the number of patients with low enough impedance to extract good EEG signals at first placement of electrodes (success rate); the number of patients with an index outside of the range (shown by the manufacturer) considered appropriate under general anesthesia (inappropriateness rate); changes of the index by laryngeal mask airway (LMA) insertion or surgical incision (responsiveness); the time to return to good EEG signals after signal disturbance by electric cautery (recovery time).

	Methods

Top Abstract Introduction Methods Results Discussion References

 
After approval of the hospital’s research committee and informed patient consent, 90 female patients aged 30 to 60 yr, scheduled for partial mastectomy were randomly divided into 3 groups by an envelope method (30 patients in each group). Those with neurologic disorders, deafness, liver diseases, mental impairment, or who were taking any drugs affecting brain function, such as hypnotics or antidepressants, were excluded. In addition, patients who moved or showed a sudden increase in arterial blood pressure and/or heart rate of more than 30% of the preceding value during maintenance of anesthesia (unexplainable wide fluctuation) were excluded. Their envelopes were reused to maintain an equal number of patients in each group (n = 30).

Atropine 0.5 mg with midazolam 5 mg was administered IM as premedication 30 min before anesthesia induction per our usual practice. Anesthesia was induced with propofol 2 mg/kg and fentanyl 2 µg/kg to permit insertion of LMA (#3, with 20 to 25 mL air in the cuff). Anesthesia was maintained with propofol 2–4 mg · kg^–1 · h^–1 (judged by an anesthesiologist without looking at the EEG monitor), fentanyl 2 µg/kg (4 µg/kg in total), and nitrous oxide 4 L/min in oxygen 2 L/min. No muscle relaxants were used. Ventilation was controlled to keep end-tidal carbon dioxide tension between 30 to 35 mm Hg. All patients were anesthetized by one senior anesthesiologist.

The EEG was monitored continuously using BIS (A1050; Aspect Medical Systems, Framington, MA), 90% spectral edge frequency (pEEG; Draeger, Luebeck, Germany), or A-AEP index (AAI) (A-AEP; Alaris Medical Systems, Hampshire, UK). The BIS uses 4 electrodes. The BIS number was calculated every 0.5 s using the EEG of the preceding 2 s. The mean value of the numbers of the preceding 15 s or 60 s was indicated every second. The pEEG uses 5 electrodes, 3 on the forehead and 2 at the mastoids. The EEG was accumulated at 128 Hz, and 90% spectral edge frequency was indicated every 2 s. The A-AEP uses headphones and 3 electrodes, 2 on the forehead and one at the mastoid. The AEPs were elicited with a binaural click of 65-dB sound intensity, 2-ms duration, delivered at a rate of 9 Hz (one click every 110 ms). The AEP was extracted over 15 sweeps with an update delay of 6 s. The AAI was calculated every 2 to 6 s. All electrodes were applied after rubbing the skin with ethanol. In all monitors, when the impedance of the electrodes is high (5 k for BIS, 10 k for pEEG and A-AEP), or if the electromyographic (EMG) noise exceeds a set percentage, the monitor would alarm or interrupt its calculations. A BIS <60 (4), 90% spectral edge frequency <12 (5), or AAI <30 (Alaris AEP^TM Monitor, Directions for use) is considered adequate sedation under general anesthesia.

We investigated the number of patients with low enough impedance to extract good EEG signals (<5 k in BIS, <10 k in pEEG and A-AEP) at the first electrode application (success rate); the number of patients with an index outside of the range considered appropriate for general anesthesia (BIS <60, 90% spectral edge frequency <12, AAI <30) except when disturbed by LMA insertion, surgical incision, electric cautery, or EMG (inappropriateness rate); changes of the index by LMA insertion or surgical incision (responsiveness); and the time to return to good EEG signals after signal disturbance by electric cautery (recovery time). The recovery time was defined as time from the end of electric cautery to the recovery of good signals and calculated by averaging 5 randomly selected episodes per patient. Arterial blood pressure and heart rate were also monitored.

Data are expressed as mean ± SD. Statistical analysis was performed with analysis of variance (ANOVA) for demographic data and the recovery time, ² test for the number of patients, repeated-measures ANOVA followed by the Student-Newman-Keuls test, a multiple-comparisons correction as a post hoc test for arterial blood pressure, heart rate, and each index. P < 0.05 was considered statistically significant.

	Results

Top Abstract Introduction Methods Results Discussion References

 
Two patients who moved and had sudden arterial blood pressure and heart rate increases unexplainably during maintenance of anesthesia were excluded and 2 other patients were enrolled in their place. Demographic data of the patients were not different among the 3 groups (Table 1). The success rate was the largest in the BIS group, followed by the A-AEP group (Table 2). Each index decreased in all groups by induction of anesthesia. The inappropriateness rate was A-AEP < BIS < pEEG. The index increased significantly with LMA insertion and skin incision in only the A-AEP group (Fig. 1, Table 3). During recovery from anesthesia, 90% spectral edge frequency in the pEEG increased gradually whereas AAI increased rapidly (Fig. 1). The responsiveness was more pronounced in the A-AEP group than in the other 2 groups. The recovery time was 43 ± 14 s in the BIS group, which was longer than 19 ± 6 s in the A-AEP group and 11 ± 6 s in the pEEG group. The latter two times were also significantly different from one another. Arterial blood pressure and heart rate decreased significantly during anesthesia, did not change by LMA insertion or surgical incision, and returned to the control values at the removal of the LMA without any differences among the 3 groups (Table 4).

View larger version (12K): [in this window] [in a new window]   Figure 1. Examples of the changes of each index. These three figures are remade to adapt the x-axis. BIS = bispectral index; SEF = 90% spectral edge frequency; pEEG = processed encephalography; AAI = Alaris auditory potentials index; A-AEP = Alaris auditory evoked potential; A = start of anesthesia induction; B = insertion of laryngeal mask airway; C = start of surgery; D = end of propofol infusion; E = end of surgery and nitrous oxide inhalation; F = respond to verbal command to open eyes; G = removal of laryngeal mask.

	Discussion

Top Abstract Introduction Methods Results Discussion References

The patients who moved or who showed sudden increases in arterial blood pressure and heart rate during maintenance of anesthesia were excluded from the study to assure that all study patients were under adequate anesthesia. However, EEG response to these events is important; therefore this should be investigated in the future with a larger patient population.

We tested each monitor on different patients. To compare the 3 monitors more precisely they should ideally be tested simultaneously in the same patient; however, auditory stimulation may interfere with the BIS and the pEEG. In addition, the electrodes of the 3 monitors should be in the same place, which is impossible.

In recent decades, improved technology has made it possible to monitor EEG in the operating room, where many electrical noises interfere with the EEG reading. Nevertheless, good EEG signals still depend on good electrodes and their placement. Although BIS requires the lowest electrode impedance to silence the alarm and prevent interruption of operation, we still found it to have the most frequent success on first application of electrode. Therefore, in clinical practice, we consider BIS the easiest to use.

The BIS is derived from the bifrontal EEG recordings collected from >5000 subjects who received anesthesia. Glass et al. (6) found a good correlation between BIS and sedation levels during propofol anesthesia. The BIS decreased linearly as propofol blood concentration increased (7). During propofol anesthesia, the BIS at which 50% and 95% of the population lost consciousness were 63 (62–65) and 51 (48–55) (8), respectively. In the present study, those who moved during maintenance of anesthesia were excluded. In addition, after surgery, no patients recalled the events during surgery. Therefore, all patients in this study might be considered to have been adequately anesthetized, although 9 of 30 patients had a BIS of >60 (Table 2). They did not move in response to either LMA insertion or surgical incision.

Care is necessary to obtain AEP in awake patients without premedication because muscle artifacts affect the signal in A-AEP as well as in BIS and pEEG (8). In the present study, we administered midazolam 5 mg with atropine 0.5 mg IM as a premedication in all patients. Nevertheless, the EMG was observed in every patient. In all monitors, when EMG noise was large enough to affect the index, the monitor would indicate alarm and stop calculating the index. These data were excluded from the results in this study. We do not know whether the artifact (EMG) detection function in each monitor was the same; therefore, EMG might have affected the results. In addition, midazolam and atropine might have affected each index. However, all patients were awake when the control values were obtained. Therefore, the data are considered comparable among the three groups.

It has been reported that no significant correlation was observed between the spectral edge frequency and propofol concentration (7). Although the mean values of spectral edge frequency might be well correlated with the depth of sedation, the inappropriateness rate in the pEEG group was the largest in the present study. Therefore, when the BIS <60, 90% spectral edge frequency <12, and AAI <30 are considered appropriate under general anesthesia, 90% spectral edge frequency might be the most inappropriate in propofol-fentanyl-nitrous oxide anesthesia. However, before this could be determined, adequacy of the number of each index should be further investigated.

Propofol produces a marked increase in latencies and decreases in amplitudes up to a complete suppression of the AEP, which indicates that propofol blocks auditory processing at the level of the primary auditory cortex (9). Linear regression analysis showed that both AAI and BIS were linearly related to the observer’s assessment of alertness sedation scale scores (10). In the present study, however, the range of the changes in the index was the largest in the AAI, which suggests that it is the most sensitive to sedation and/or anesthesia.

BIS is considered a useful indicator of anesthetic adequacy, defined as the ability to prevent a 20% or more increase in arterial blood pressure to laryngoscopy and intubation (11). BIS is also reported to be able to predict movement response to LMA insertion (12), intubation (13), and skin incision (14). BIS might be a more accurate predictor of patient movement in response to skin incision during propofol-nitrous oxide anesthesia than are standard power spectrum variables, such as median frequency, relative power, and spectral edge frequency or plasma propofol concentrations (15). In this study, though arterial blood pressure did not change with LMA insertion or surgical incision in any group, only the AAI responded to the stimuli. Surgical stimulation increases the amplitude of AEP, which indicates that the amplitude does not simply reflect graded concentrations of the anesthetic but, more importantly, depicts the balance between surgical stimulation and anesthetic depression (16).

During recovery from an induction dose of propofol, concentration-related effects on the AAI were found (17). In the study by Doi et al. (18), there was no difference between the awake AAI and the values when the patients opened their eyes at command, although they used a different AAI from ours. In our study, BIS, spectral edge frequency, and AAI were not statistically different between the awake state and at arousal from anesthesia. The AAI was considered to be the best at distinguishing the transition from unconsciousness to consciousness during sedation with propofol (19). The AAI was able to follow rapid changes from awake to asleep and to detect short-term periods of consciousness (20). At arousal, in the present study, the AAI increased most rapidly, which might indicate that the AAI is the most useful in distinguishing awake from sleep in propofol-fentanyl-nitrous oxide anesthesia.

The larger response of the AAI to LMA insertion or skin incision, the more rapid changes of the AAI by arousal compared with BIS, and the shorter recovery time in the order of pEEG < A-AEP < BIS might simply have been attributable to the faster response time of the pEEG and A-AEP compared with the BIS, although further studies are necessary to confirm this. The pEEG indicated the spectral edge frequency every 2 seconds. The extraction of the AEP has usually been done by moving time averaging over many sweeps (usually 250 to 1000), which could produce a delay of more than 1 minute (21). However, this problem was resolved by applying an autoregressive model with exogenous input that enables extraction of the AEP within 15 sweeps, which shortens the delay to 6 seconds (3). The BIS value output represents the mean of at least 15, sometimes 60, seconds; therefore a 1-minute delay might be expected. However, in the study of Nieuwenhuijs et al. (22), there were delays of 2 minutes or longer. Our results show a delay of 43 ± 14 seconds for BIS. This was shorter than the data of Nieuwenhuijs et al. (22) but still significantly longer than the delay of the A-AEP and pEEG.

We did not directly compare the indexes because each was derived from a different concept and because the changes in the index according to sedation level were not parallel. However, response to stimuli might be comparable and important in clinical practice.

In conclusion, using the manufacturers’ stated index threshold values considered to be appropriate under general anesthesia, AAI by A-AEP was the least for inappropriateness rate in propofol-fentanyl-nitrous oxide anesthesia. However, the pEEG was the fastest in stabilizing measurement after the noise of electric cautery, and the BIS was the most effective at achieving an impedance low enough to extract good EEG signals at first electrode placement.

	Acknowledgments

 
We thank Professor Chingmuh Lee, MD, Department of Anesthesiology, University of California, Los Angeles for his assistance with English.

	References

Top Abstract Introduction Methods Results Discussion References

 

1. Rampil IJ. A primer for EEG signal processing in anesthesia. Anesthesiology 1998; 89: 980–1002.[Web of Science][Medline]
2. Barnett TP, Johnson LC, Naitoh P, et al. Bispectrum analysis of electroencephalogram signals during waking and sleeping. Science 1971; 172: 401–2.[Abstract/Free Full Text]
3. Jensen EW, Lindholm P, Henneberg SW. Auto regressive modelling with exogenous input of auditory evoked potentials to produce an on-line depth of anaesthesia index. Methods Inf Med 1996; 35: 256–60.[Web of Science][Medline]
4. Kissin I. Depth of anesthesia and Bispectral Index monitoring. Anesth Analg 2000; 90: 1114–7.[Free Full Text]
5. Gurman G. Assessment of depth of anesthesia: observations on processed EEG and spectral edge frequency. Int J Clin Monit Comput 1994; 11: 185–9.[Web of Science][Medline]
6. Glass PS, Bloom M, Kearse L, et al. Bispectral analysis measures sedation and memory effects of propofol, midazolam, isoflurane, and alfentanil in healthy volunteers. Anesthesiology 1997; 86: 836–47.[Web of Science][Medline]
7. Leslie K, Sessler DI, Schroeder M, Walters K. Propofol blood concentration and the bispectral index predict suppression of learning during propofol/epidural anesthesia in volunteers. Anesth Analg 1995; 81: 1269–74.[Abstract]
8. Rundshagen I, Schnabel K, Schulte AM, Esch J. Impaired explicit memory after recovery from propofol/sufentanil anaesthesia is related to changes in the midlatency auditory evoked response. Br J Anaesth 2002; 89: 376–81.[Abstract/Free Full Text]
9. Schwender D, Klasing S, Madler C, et al. Midlatency auditory evoked potentials and cognitive function during general anesthesia. Intern Anesth Clin 1993; 31: 89–106.
10. Ge SJ, Zhuang XL, Wang YT, et al. Changes in the rapidly extracted auditory evoked potentials index and the bispectral index during sedation induced by propofol or midazolam under epidural block. Br J Anaesth 2002; 89: 260–4.[Abstract/Free Full Text]
11. Kearse LA Jr, Manberg P, DeBros F, et al. Bispectral analysis of the electroencephalogram during induction of anesthesia may predict hemodynamic responses to laryngoscopy and intubation. Electroenceph Clin Neurophysiol 1971; 31: 137–48.[Web of Science][Medline]
12. Doi M, Gajraj RJ, Mantzaridis H, Kenny GNC. Prediction of movement at laryngeal mask airway insertion: comparison of auditory evoked potential index, bispectral index, spectral edge frequency, and median frequency. Br J Anaesth 1999; 82: 203–7.[Abstract/Free Full Text]
13. Sleigh JW, Donovan J. Comparison of bispectral index, 95% spectral edge frequency and approximate entropy of the EEG, with changes in heart rate variability during induction of general anaesthesia. Br J Anaesth 1999; 82: 666–71.[Abstract/Free Full Text]
14. Vernon JM, Lang E, Sebel PS, Manberg P. Prediction of movement using bispectral electroencephalographic analysis during propofol/alfentanil or isoflurane/alfentanil anesthesia. Anesth Analg 1995; 80: 780–5.[Abstract]
15. Kearse LA Jr, Manberg P, Chamoun N, et al. Bispectral analysis of the electroencephalogram correlates with patient movement to skin incision during propofol/nitrous oxide anesthesia. Anesthesiology 1994; 81: 1365–70.[Web of Science][Medline]
16. Jones JG. Awareness under anaesthesia. Anaesth Rounds 1988; 21: 1–28.
17. White M, Schenkels MJ, Englbers FH, et al. Effect-site modelling of propofol using auditory evoked potentials. Br J Anaesth 1999; 82: 333–9.[Abstract/Free Full Text]
18. Doi M, Gajraj RJ, Mantzaridis H, Kenny GN. Relationship between calculated blood concentration of propofol and electrophysiological variables during emergence from anaesthesia: comparison of bispectral index, spectral edge frequency, median frequency and auditory evoked potential index. Br J Anasth 1997; 78: 180–4.
19. Gajraj RJ, Doi M, Mantzaridis H, Kenny GNC. Analysis of the EEG bispectrum, auditory evoked potentials and the EEG power spectrum during repeated transitions from consciousness to unconsciousness. Br J Anaesth 1998; 80: 46–52.[Abstract/Free Full Text]
20. Jensen EW, Litvan H. Detection of level of consciousness during propofol anaesthesia by rapidly extracted auditory evoked potentials. Memory and awareness in anaesthesia IV. London 1998 Ed. Jordan C, Newton D.
21. Iselin-Chaves IA, El Moalem HE, Gan TJ, et al. Changes in the auditory evoked potentials and the bispectral index following propofol or propofol and alfentanil. Anesthesiology 2000; 92: 1300–10.[Web of Science][Medline]
22. Nieuwenhuijs D, Coleman EL, Douglas NJ, et al. Bispectral index values and spectral edge frequency at different stages of physiologic sleep. Anesth Analg 2002; 94: 125–9.[Abstract/Free Full Text]

This article has been cited by other articles:

				T. Nishiyama The Effects of Auditory Evoked Potential Click Sounds on Bispectral Index and Entropy Anesth. Analg., August 1, 2008; 107(2): 545 - 548. [Abstract] [Full Text] [PDF]

				J. Bruhn, P. S. Myles, R. Sneyd, and M. M. R. F. Struys Depth of anaesthesia monitoring: what's available, what's validated and what's next? Br. J. Anaesth., July 1, 2006; 97(1): 85 - 94. [Abstract] [Full Text] [PDF]

				H. Schwilden, E. Kochs, M. Daunderer, Ch. Jeleazcov, B. Scheller, G. Schneider, J. Schuttler, D. Schwender, G. Stockmanns, and E. Poppel Concurrent recording of AEP, SSEP and EEG parameters during anaesthesia: a factor analysis Br. J. Anaesth., August 1, 2005; 95(2): 197 - 206. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (5) Citing Articles via Google Scholar Google Scholar Articles by Nishiyama, T. Articles by Hanaoka, K. Search for Related Content PubMed PubMed Citation Articles by Nishiyama, T. Articles by Hanaoka, K. Related Collections Equipment Monitoring (Non-cardiac) Technology