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 (7) Citing Articles via Google Scholar Google Scholar Articles by Soto, R. Articles by Smith, R. A. Search for Related Content PubMed PubMed Citation Articles by Soto, R. Articles by Smith, R. A. Related Collections Equipment Monitoring (Non-cardiac)

---

TECHNOLOGY, COMPUTING, AND SIMULATION

A Comparison of Bispectral Index and Entropy, or How to Misinterpret Both

Department of Anesthesiology, University of South Florida College of Medicine. Tampa, Florida

Address correspondence and reprint requests to Roy G. Soto, MD, 12901 Bruce B. Downs Blvd, MDC 59, USF Department of Anesthesiology, Tampa, FL 33612. Address e-mail to rsoto{at}hsc.usf.edu.

	Abstract

Top Abstract Introduction Case Report Discussion References

 
Consciousness monitoring has become increasingly popular in general anesthesia cases, and a new technology has recently been introduced with potential advantages over the other available products. In this case report, we discuss a patient who was monitored simultaneously with Bispectral Index and Entropy and evaluate the differences between the two. More importantly, we emphasize the importance of vigilance when using new technologies and discuss the potential impact of lack of vigilance on patient outcome.

	Introduction

Top Abstract Introduction Case Report Discussion References

 
Coverage of level of consciousness (LOC) monitoring has had a resurgence in both the medical and lay press recently, specifically as a result of the recent publication of large trials suggesting that the Bispectral Index (BIS) can prevent or reduce the incidence of awareness during anesthesia (1–3). In addition, Entropy, a new technology, has recently been introduced as a competitive product by Datex-Ohmeda (Helsinki, Finland). Unlike BIS, which relies on a library of data that is compared with the patient's electroencephalogram (EEG), Entropy relies on the extent of disorder in both EEG and electromyograph (EMG) signals as measured from the patient's forehead. We present a case of a patient monitored with both in whom a rapidly increasing LOC was detected quickly by one monitor but slowly by the anesthesia provider.

	Case Report

Top Abstract Introduction Case Report Discussion References

 
A 66-yr-old woman scheduled to undergo elective total abdominal hysterectomy was monitored intraoperatively with both BIS and Entropy. Because the BIS XP sensor is applied diagonally, the Entropy sensor fits very easily on the forehead as recommended by the manufacturer. The patient had no significant medical history, was not ß blocked, and was not taking pain or sedative medications. The patient did not receive preoperative sedation and underwent an uneventful induction of anesthesia and tracheal intubation with propofol and rocuronium. Towards the end of the case, the anesthesia resident came to the room and checked the trends (Fig. 1) and noted that both the Entropy and BIS had transiently increased rapidly. The anesthesia resident stated that he only noted the changes when the BIS level was approximately 80, the state entropy (SE) was approximately 65, and the response entropy (RE) was approximately 85, at which time he increased the patient's depth of consciousness with sevoflurane (from an end-tidal concentration of 1.8% to 2.5%). The heart rate and arterial blood pressure did not increase during this episode (which occurred during fascial closure), and the patient showed no evidence of intraoperative awareness immediately or at 2 or 10 days postoperatively.

View larger version (165K): [in this window] [in a new window]   Figure 1. Thirty-min trend of Bispectral Index and entropy. White line = state entropy; darker line = response entropy.

	Discussion

Top Abstract Introduction Case Report Discussion References

 
Estimating LOS is an important element of current monitoring technology. Although the benefits to patient safety, hospital cost, and recovery have been, and will continue to be, debated, these devices monitor the brain, the end-organ of consciousness. For more than 150 years we have relied on movement, breathing patterns, and hemodynamics to indirectly aid in our assessment of anesthetic brain effect. Imperfect though they may be, the new monitors of consciousness are an important step in our understanding of the state we call "general anesthesia."

The BIS monitors electrocortical activity and compares it with a database to produce a number that reflects level of hypnosis. In effect, the technology is "reverse engineered" to compare data with a static dataset. By design, therefore, patterns not incorporated into the data library will not be accurately reflected by the BIS algorithm. Values below 60 are typically considered to correlate with adequate anesthetic level, although case reports have suggested that patients can have recall at lower numbers and can be "asleep" at higher numbers (4,5). This is not surprising, as a certain percentage of patients will be outside of the expected norm, which is inherent in a database generated from healthy subjects and based on standard anesthetic regimens. For the majority of patients, BIS seems to reflect LOC very well; however, patients with preexisting brain injury, those receiving nonstandard anesthetics, and simple statistical outliers may render the monitor less reliable.

The Entropy algorithm, in contrast, as applied in the Datex-Ohmeda Entropy Module (available with the S/5 Anesthesia Monitor using current software), analyzes the complexity of signals of both the EEG and frontalis EMG. In signal analysis, entropy describes the irregularity or unpredictability characteristic of a signal. When entropy is used as an analytic technique in describing EEG signals it is able to describe the complexity or "order" in the EEG. As the anesthetic depth is increased, the EEG data become more predictable or contain more "order." When interpreted in terms of entropy, more order is less complex or lower entropy. As the anesthetic depth is decreased, less order or more irregularity is seen in EEG data (6). Entropy is independent of absolute frequency and amplitude scales of EEG, which are patient dependent. In this way, each patient behaves as his or her own control, presumably independent of preexisting brain injury or anesthetic technique (7). In contrast to the BIS algorithm, therefore, the Entropy algorithm does not weight parameters according to prior values of the derived index and is based on physical principles.

As alluded to, the Entropy module reports two scales: the SE (values from 0–91) and RE (values from 0–100). Detected scalp signal reflects differential frequencies. The SE measures predominantly the lower frequency, EEG signal (up to 32 Hz), whereas the RE measures the lower frequency plus the higher frequency signal (up to 47 Hz), which is mainly predominated by frontalis EMG. Therefore, SE is a measure of EEG, and RE is a measure of EEG plus EMG activity (hence the maximum value of RE is higher than that of SE). With deep neuromuscular paralysis the RE may be blunted, rendering it equivalent in absolute number to the SE; however, RE persists even with typical surgical levels of muscle relaxation. Although the BIS monitor (and the PSA4000 monitor as well) attempts to filter EMG signal from their data interpretation, the Entropy algorithm posits that EMG data are useful and may in some circumstances be more sensitive to light LOC or analgesia than EEG.

As shown in our case report, this unparalyzed patient showed a remarkable increase in her RE that preceded and predicted the increase in both SE and BIS by approximately 4 minutes. The increasing difference between RE and SE revealed an increase in EMG tone before EEG. What this suggests is that frontalis muscle tone in an unparalyzed patient may be a quicker indicator of return of wakefulness than EEG, or at least it leads EEG signal by a certain time factor that has yet to be determined. Perhaps RE should be considered an ideal indicator of LOC in unparalyzed patients and SE or BIS in paralyzed patients in whom RE is presumably less accurate (although frontalis EMG has been shown to be more resistant to neuromuscular blockade than the abductor pollicis).

Regardless of the impact of LOC monitoring on future anesthetic practice, this case report illustrates the importance of the preeminence of vigilance over technology. Although multiple monitors were warning of impending wakefulness to varying degrees of success, none was detected until late in the course of the event. A patient's quality of care can potentially be improved with LOC monitoring, but caregivers must be open to accept and use the information the monitors provide in clinical assessment.

	Footnotes

 
Accepted for publication September 14, 2004.

	References

Top Abstract Introduction Case Report Discussion References

 

1. Sandin R, Enlund G. Awareness during anaesthesia: a prospective case study. Lancet 2000;355:707.[Web of Science][Medline]
2. Myles P, Leslie K, McNeil J, et al. Bispectral index monitoring to prevent awareness during anaesthesia: the B-Aware randomised controlled trial. Lancet 2004;363:1757–63.[Web of Science][Medline]
3. Sebel P, Bowdle A. The incidence of awareness during anesthesia: a multicenter US study [abstract]. Anesthesiology 2003;99:360.[Medline]
4. Vuyk J, Lichtenbelt B, Vieveen J, et al. Low Bispectral index values in awake volunteers receiving a combination of propofol and midazolam. Anesthesiology 2004;100:179–81.[Web of Science][Medline]
5. 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]
6. Viertiö-Oja H, Maja V, Särkelä M, et al. Description of the Entropy algorithm as applied in the Datex-Ohmeda S/5TM Entropy Module. Acta Anesthesiol Scand 2004;48:154–61.[Web of Science][Medline]
7. Anderson R, Jakobsson J. Entropy of EEG during anaesthetic induction: a comparative study with propofol or nitrous oxide as sole agent. Br J Anaesth 2004;92:159–61.[Free Full Text]

This article has been cited by other articles:

				G. Weil, S. Passot, F. Servin, and V. Billard Does Spectral Entropy Reflect the Response to Intubation or Incision During Propofol-Remifentanil Anesthesia? Anesth. Analg., January 1, 2008; 106(1): 152 - 159. [Abstract] [Full Text] [PDF]

				C. Lefoll-Masson, C. Fermanian, I. Aime, N. Verroust, G. Taylor, P.-A. Laloe, N. Liu, P. Aegerter, and M. Fischler The Comparability of Bispectral Index and State Entropy Index During Maintenance of Sufentanil-Sevoflurane-Nitrous Oxide Anesthesia Anesth. Analg., November 1, 2007; 105(5): 1319 - 1325. [Abstract] [Full Text] [PDF]

				M. Vassiliadis, D. Geros, and K. Maria Awareness Despite Low Spectral Entropy Values Anesth. Analg., August 1, 2007; 105(2): 535 - 535. [Full Text] [PDF]

				V. Bonhomme, E. Deflandre, and P. Hans Correlation and agreement between bispectral index and state entropy of the electroencephalogram during propofol anaesthesia Br. J. Anaesth., September 1, 2006; 97(3): 340 - 346. [Abstract] [Full Text] [PDF]

				G. Litscher Electroencephalogram-entropy and acupuncture. Anesth. Analg., June 1, 2006; 102(6): 1745 - 1751. [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 (7) Citing Articles via Google Scholar Google Scholar Articles by Soto, R. Articles by Smith, R. A. Search for Related Content PubMed PubMed Citation Articles by Soto, R. Articles by Smith, R. A. Related Collections Equipment Monitoring (Non-cardiac)