Comparative Analysis of Phase Lag Entropy and Bispectral Index as Anesthetic Depth Indicators in Patients Undergoing Thyroid Surgery with Nerve Integrity Monitoring

Abstract

Background: Most depth of anesthesia (DOA) monitors rely on the temporal characteristics of a single-channel electroencephalogram (EEG) and cannot provide spatial or connectivity information. Phase lag entropy (PLE) reflects DOA by calculating diverse connectivity from temporal patterns of phase relationships. The aim of this study was to compare the performance of PLE and bispectral index (BIS) monitors for assessing DOA during anesthesia induction, nerve integrity monitoring (NIM), and anesthesia emergence.

Methods: Thirty-five patients undergoing elective thyroid surgery with recurrent laryngeal nerve NIM received propofol and remifentanil via target-controlled infusion. After applying PLE and BIS monitors, propofol infusion was initiated at a calculated effect site concentration (Ce) of 2 μg/mL and then increased in 1-μg/mL Ce increments. After propofol Ce reached 5 μ/mL, a remifentanil infusion was begun, and anesthesia induction was considered complete. During NIM, PLE and BIS values were compared at a specific time points from platysma muscle exposure to subcutaneous tissue closure. PLE and BIS values were recorded continuously from preanesthetic state to full recovery of orientation; bias and limits of agreement between monitors were calculated.

Results: PLE and BIS values decreased progressively with increasing propofol Ce during anesthetic induction and increased by stages during emergence. The prediction probabilities of PLE and BIS for detecting propofol Ce changes were 0.750 and 0.756, respectively, during induction and 0.749 and 0.746, respectively, during emergence. No aberrant PLE or BIS values occurred during NIM. Correlation coefficients for BIS and PLE were 0.98 and 0.92 during induction and emergence, respectively. PLE values were significantly higher than BIS values at full recovery of orientation. Estimated bias between monitors was -4.16 ± 8.7, and 95% limits of agreement were -21.21 to 12.89.

Conclusion: PLE is a reasonable alternative to BIS for evaluating consciousness and DOA during general anesthesia and during NIM.

Trial registration: Clinical Research Information Service Identifier: KCT0003490.

Keywords: Consciousness Monitors; Electroencephalography; Entropy; General Anesthesia; Propofol.

Conflict of interest statement

The authors have no potential conflict of interest to disclose. The present study was supported from InBody to use the PLEM100 device and PLEM-ES100 electrode. However, their support did not affect the results of the present study.

© 2019 The Korean Academy of Medical Sciences.

Figures

Fig. 1. Placement of BIS and PLE electrodes. BIS electrodes are located on the upper forehead (lower area of the photograph) and correspond to the 10–20 International System bipolar montages of Fz (Fpz or Afz: lead 1) to F7 (F9 or FT9: lead 3) and F3 (AF3: lead 2) to F7. Lead 4 is the ground electrode. PLE electrodes are located on the lower forehead (upper area of the photograph): Fp1 (L1), AF5 (L2), Fp2 (R1), AF6 (R2), and ground (C).

BIS = bispectral index, PLE = phase lag entropy.

Fig. 2. Box and whiskers plots of PLE and BIS values during anesthesia induction and emergence. (A) Both PLE and BIS values decrease progressively with increasing propofol Ce during anesthetic induction. (B) Both PLE and BIS values increase by stages during emergence from anesthesia. The upper and lower limits of the box indicate the 75th and 25th percentiles of the sample, and the horizontal line inside each box indicates the median. The upper and lower notches indicate the 90th and 10th percentiles of the sample. Values below and above the notches are drawn as individual points.

PLE = phase lag entropy, BIS = bispectral index.

Fig. 3. Bland-Altman plot for (BIS+PLE)/2 vs. (BIS−PLE) containing 70,282 paired-index points recorded for 34 patients, reflecting the agreement between BIS and PLE values. The thin dotted line indicates the similarity of the 2 indices, the thick solid line indicates the estimated bias of the 2 indexes, and the thick dotted lines show the 95% limits of agreement.

BIS = bispectral index, PLE = phase lag entropy.