Tailor-made Anesthesia

Interpretation of anesthetic adequacy - A clinical experience with spectral entropy

B. Rehberg-Klug, MD, PD, Dr
Charité - Universitätsmedizin
Berlin, Germany

The article also available in PDF: 72KB

Introduction

Although the primary target of anesthetic drugs is the CNS, most anesthesiologists confine monitoring during anesthesia to cardiocirculatory and respiratory parameters. Among other reasons, this is because the measurable signals coming from the CNS such as the electroencephalogram (EEG) or evoked potentials are difficult to interpret.

Researchers have therefore attempted to condense the complex information of the EEG into a single parameter. Examples of the such parameters tested over many years are the spectral edge frequency or the median frequency of the power spectrum of the EEG [1,2]. However, these parameters have never entered widespread clinical practice for two reasons: they are sensitive to artifact, and they are also difficult to interpret due to the biphasic concentration-response relations with some drugs. Further developments have resulted in parameters derived as a weighted combination of several sub-parameters. The most familiar of these is the Bispectral Index (BIS), now widely accepted and supported by many experimental and clinical studies [3].

New mathematical developments and advances in computer technology have now made several different approaches available to EEG analysis, especially those derived from information theory. Several new parameters derived from the EEG have been suggested such as approximate entropy, correlation dimension, and spectral entropy [4-6].

Recently, spectral entropy of the EEG has become commercially available in the Datex-Ohmeda S/5 Anesthesia Monitor using the M-Entropy module. In addition to a new computational approach of EEG analysis, this monitor also takes into account new developments in the concept of “adequacy of anesthesia”.

Contrary to previous assumptions, the anesthetic state is now understood to consist of several components, including unconsciousness, amnesia, antinociception, immobility and autonomic stability. Since only unconsciousness and amnesia are considered to be of cortical origin, parameters of the EEG are likely to correlate only with these cortical components of anesthesia. In contrast, spontaneous muscle activity measured as the frontal electromyogram (FEMG) is influenced by subcortical structures. The M-Entropy module, using a single sensor, gives two separate readings: one for the spectral entropy of the pure EEG signal (state entropy) and the other for the spectral entropy of the combined EEG-FEMG signal (response entropy).

Methods

In a clinical evaluation trial, we tested the entropy measurement with the new M-Entropy module during routine anesthesia cases in the urological and gynecological OR suites of the Charité university hospital. As a training hospital, staff vary from first year trainees to senior anesthesiologists. A wide range of procedures is performed, from short diagnostic ones to renal tumor resections under cardiac bypass.

Our evaluation of the entropy measurement focused on three points:

• First, we were interested to know whether the anesthesiologist in charge thought the entropy values correlated with their personal judgment of anesthetic adequacy.
• Second, we asked whether they considered the additional information helpful, useless or disturbing in daily clinical practice.
• As a third point, we looked at the stability of the signal in the awake state (baseline stability) [7] and at stable anesthetic concentrations. In addition, we evaluated the stability of the signal to artifacts such as electrocautery.

Results

The overall response of the participating anesthesiologists was positive. Entropy values generally correlated well with the clinical evaluation of anesthetic adequacy by the anesthesiologist in charge. This correlation seemed to be better for experienced anesthesiologists.

Conversely, especially colleagues in training thought that the additional monitoring information was helpful to countercheck their clinical assessment. Nevertheless, even experienced anesthesiologists admitted that the entropy readings were reassuring in cases where interpretation of anesthetic adequacy was difficult. Specifically mentioned were cases where ardiocirculatory parameters were largely influenced by factors other than anesthetic depth (e.g. adrenalectomy), and patients whose anesthetic requirements were difficult to assess (e.g. morbidly obese patients).

Evaluation of the signal stability revealed low variability of the baseline signal, estimated as less than 5 % (median baseline value for state entropy = 88.5). Eye movements, which have been a problem with some previous devices, did not appear to alter the entropy measurement. During surgery, the entropy signal was stable even during electrocautery, which traditionally has been a problem with neuromonitoring. During anesthesia considered as adequate for surgery by the anesthesiologist, entropy values were in the range of 40-60, and appeared to be stable at constant levels of surgical stimulus and anesthetic concentration.

Discussion

In a clinical pilot evaluation, we looked at different practical aspects of monitoring entropy with the Datex-Ohmeda S/5 Entropy module. Our results suggest that spectral entropy of the EEG correlates well with anesthetic adequacy as judged by experienced anesthesiologists. In addition, the monitor signal proved to be stable at baseline and constant anesthetic depth, and was resistant to artifacts.

These results must be considered preliminary and do not stem from a controlled or blind study. Since there is no “gold standard” of anesthetic depth with which to compare the entropy values, the evaluation is inevitably subjective. There were times where clinical judgment and entropy differed even among experienced anesthesiologists. We cannot say whether anesthesia was truly adequate or inadequate at that time; that question must be addressed in outcome-oriented studies.

Nor did we specifically examine the different components of anesthetic depth (cortical vs. subcortical). The monitor differentiates between state entropy”, stemming from cortical EEG signals, and “response entropy”, including EMG activity. In our evaluation, we did not ask the anesthesiologists to specify the adequacy of different components of anesthesia. This is another issue that needs to be addressed in further studies.

However, it did become clear from this preliminary evaluation that monitoring the effects of anesthetics on the central nervous system is a welcome addition to the armory of the practicing anesthesiologist. This requires monitoring devices which are easy to use and easy to interpret as well being resistant to artifacts. The entropy monitor appears to be promising on all these points.

Although it is difficult to prove that entropy monitoring prevents unwanted outcome, it can be assumed that it helps as an “aid to vigilance”, offering reassurance especially in difficult situations.

References

1. Rampil IJ, Matteo RS: Changes in EEG spectral edge frequency correlate with the hemodynamic response to laryngoscopy and intubation. Anesthesiology 1987; 67: 139-42
2. Schwilden H, Stoeckel H: [Investigations on several EEG-parameters as indicators of the state of anesthesia the median - a quantitative measure of the depth of anesthesia (author's translation)]. Anesth Intensivther Notfallmed 1980; 15: 279-86
3. Sebel PS, Lang E, Rampil IJ, White PF, Cork R, Jopling M, Smith NT, Glass PS, Manberg P: A multicenter study of bispectral electroencephalogram analysis for monitoring anesthetic effect. Anesth Analg 1997; 84: 891-9
4. Bruhn J, Ropcke H, Rehberg B, Bouillon T, Hoeft A: Electroencephalogram approximate entropy correctly classifies the occurrence of burst suppression pattern as increasing anesthetic drug effect. Anesthesiology 2000; 93: 981-5
5. Widman G, Schreiber T, Rehberg B, Hoeft A, Elger CE: Quantification of depth of anesthesia by nonlinear time series analysis of brain electrical activity. Phys Rev E Stat Phys Plasmas Fluids Relat Interdiscip Topics. 2000; 62: 4898-903
6. Steyn-Ross DA, Steyn-Ross ML, Wilcocks LC, Sleigh JW: Toward a theory of the general-anesthetic-induced phase transition of the cerebral cortex. II. Numerical simulations, spectral entropy, and correlation times Phys Rev E Stat Nonlin Soft Matter Phys. 2001; 64: 011918
7. Bruhn J, Bouillon TW, Hoeft A, Shafer SL: Artifact robustness, inter- and intra-individual baseline stability, and rational EEG parameter selection. Anesthesiology 2002
   

Last updated: 1 September 2003Created
Legal notice	© GE Healthcare 2008 ISSN 1795-6269 (Web) ISSN 1795-6277 (CD)	Webmaster