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 Google Scholar Google Scholar Articles by Kreuer, S. Articles by Wilhelm, W. Search for Related Content PubMed PubMed Citation Articles by Kreuer, S. Articles by Wilhelm, W. Related Collections Clinical Pharmacology Technology Pharmacology

---

ANESTHETIC PHARMACOLOGY

Comparative Pharmacodynamic Modeling Using Bispectral and Narcotrend-Index With and Without a Pharmacodynamic Plateau During Sevoflurane Anesthesia

Sascha Kreuer, MD^*, Jörgen Bruhn, MD, Elisabeth Walter, MD^*, Reinhard Larsen, MD^*, Christian C. Apfel, MD, Ulrich Grundmann, MD^*, Andreas Biedler, MD^*, and Wolfram Wilhelm, MD

From the *Department of Anesthesiology and Intensive Care Medicine, University of Saarland, Homburg/Saar, Saarland, Germany; Department of Anesthesiology, UMC St. Radboud, Nijmegen, The Netherlands; Department of Anesthesia and Perioperative Care, University of California San Francisco, San Francisco; and Department of Anesthesiology and Intensive Care Medicine, St. Marien-Hospital, Luenen, NRW, Germany.

Address correspondence and reprint requests to Sascha Kreuer, MD, Department of Anesthesiology and Intensive Care Medicine, University of Saarland, 66421 Homburg/Saar, Germany. Address e-mail to sascha.kreuer{at}uniklinik-saarland.de.

	Abstract

Top Abstract Introduction METHODS RESULTS DISCUSSION REFERENCES

 
BACKGROUND: We compared two pharmacodynamic models, one with and one without a plateau effect. Bispectral indices (BIS, Aspect Medical Systems, Natick, MA, version XP) and Narcotrend (NCT, MonitorTechnik, Bad Bramstedt, Germany, version 4.0) were used as an electroencephalographic measure of sevoflurane drug effect. In addition, we tried to analyze the origin of the plateau.

METHODS: We investigated 26 adult patients scheduled for radical prostatectomy. At least 45 min after induction of general anesthesia, end-tidal sevoflurane concentrations were varied between 1 vol% and 4 vol%. To evaluate the relationship between concentrations and electroencephalographic indices, two different pharmacodynamic models were applied: a conventional model based on a single sigmoidal curve, and a novel model based on two sigmoidal curves for BIS and NCT values with and without burst suppression. The parameters of the models were estimated by NONMEM V (GloboMax, Hanover) by minimizing log likelihood. Statistical significance between the two models was calculated by the likelihood ratio test.

RESULTS: The end-tidal sevoflurane concentration ranged from 1.04 ± 0.17 vol% to 4.43 ± 0.43 vol%. The difference between the log likelihood values of the new pharmacokinetic/pharmacodynamic model with two connected sigmoidal curves and the classical E_max model with one sigmoidal curve is 396 (P < 0.001) for the BIS monitor and 1121 (P < 0.001) for the NCT. The plateau is positioned at the change between the maximum power and the increase of burst suppression ratio.

CONCLUSION: A pharmacokinetic/pharmacodynamic model consisting of two sigmoid curves with an intervening plateau describes the effect of sevoflurane on BIS and NCT indices better than a model with a single sigmoid curve.

	Introduction

Top Abstract Introduction METHODS RESULTS DISCUSSION REFERENCES

 
For inhaled anesthetics, a hysteresis, or time lag, between end-tidal concentration and the calculated effect compartment concentration has been recognized and physiologically modeled.1 The speed of onset of anesthetic action can be classified by quantitative analysis of this hysteresis. Simultaneous pharmacokinetic-pharmacodynamic (PK/PD) modeling using the Hill equation2 enables the determination of the parameters E₀, E_max, C₅₀, , and k_e0.3 The E₀ and E_max values describe the minimum and maximum drug effects, whereas the drug concentration that causes 50% of the maximum effect is given by the C₅₀ value. The assigns the steepness of the concentration-response curve, and the k_e0 value is the first-order rate constant determining the efflux from the effect compartment.

In a previous investigation we compared the applicability of the Bispectral Index (BIS-XP, Aspect Medical Systems Inc., Natick, MA) and the Narcotrend index (MonitorTechnik, Bad Bramstedt, Germany, version 4.0) for quantification of the electroencephalographic (EEG) effects of isoflurane4 with a newly developed model using two connected sigmoidal curves, with the first sigmoidal curve describing the EEG response before the onset of burst suppression, and the second curve describing the EEG response with burst suppression.

The BIS5,6 and the Narcotrend index7,8 are multiparameter EEG indices approved for clinical use. Both monitoring systems use algorithms that display a value from 99 (awake) to 0 (no electrical brain activity). Both indices were based on changes in the EEG and were developed to measure the changing status of sedation and consciousness at low anesthetic concentrations. At higher anesthetic concentrations than the minimum alveolar anesthetic concentration value, the indices could not be correlated with clinical pharmacodynamic changes. Thus, the indices values were empirically derived and may not reflect true physiological changes. This could lead to a lack of precision at higher volatile anesthetic concentrations.

In the present investigation we compared the applicability of the new model with two connected sigmoidal curves with the classical E_max model with one sigmoidal curve during sevoflurane anesthesia. We hypothesized that the new model with two connected sigmoidal curves yielded into a better population fit than the classical E_max model with one sigmoidal curve. In addition, we tried to analyze the origin of the observe plateau.

	METHODS

Top Abstract Introduction METHODS RESULTS DISCUSSION REFERENCES

 
Patients and Anesthesia
With local ethics committee (Ärztekammer des Saarlandes, Saarbrücken, Saarland, Germany) approval and written informed consent, we investigated 26 adult patients scheduled for radical prostatectomy. All patients were premedicated with midazolam 7.5 mg orally on the morning before surgery. After the skin of the forehead had been degreased with 70% isopropanol, the BIS (BIS-XP sensor, Aspect Medical Systems, Natick, MA) and the Narcotrend (Blue sensor, Medicotest, Olstykke, Denmark) electrodes were positioned as recommended by the manufacturers. For the Narcotrend, two electrodes were placed on the patients’ foreheads with a minimum distance of 8 cm, and a third electrode was positioned laterally serving as a reference electrode. Impedances were measured for each set of electrodes to ensure adequate electrode contact defined as 6 k for the Narcotrend and 7.5 k for the BIS as required by the manufacturers.

While oxygen 10 L/min was given via a facemask for administration of oxygen, induction of anesthesia was started with a remifentanil infusion at 0.4 µg • kg^–1 • min^–1. Five minutes later the patients received 2.0 mg/kg propofol. After loss of consciousness and facemask ventilation with oxygen, patients received 0.5 mg/kg of atracurium. Three minutes later the trachea was intubated, and the lungs of the patients were ventilated to an end-tidal carbon dioxide concentration of 35 mm Hg (4.6 kPA). Immediately after intubation, the remifentanil infusion was stopped and sevoflurane in 1.5 L/min fresh gas flow (0.5 L/min O₂ and 1 L/min air) was given for hypnosis.

After induction of anesthesia, patients received 12 mL bupivacaine 0.5% epidurally for intraoperative analgesia. Complete neuromuscular blockade was obtained by repetitive injections of 0.25 mg/kg atracurium using train-of-four and double burst stimulation monitoring. All patients were visited on the postanesthesia care unit and on the first and third postoperative day and interviewed about intraoperative recall.

Study Measurements
To eliminate residual effects from propofol or remifentanil and to ensure a condition of constant surgical stimulation, study measurements were performed during dissection of the prostate, a minimum of 45 min after induction of anesthesia. To obtain concentration response curves, end-tidal concentrations were subsequently increased and decreased two times. Starting at an end-tidal sevoflurane concentration of 1%, the vaporizer was opened to the maximum of 8% sevoflurane concentration until the end-tidal sevoflurane concentration reached 4%. Thereafter, the vaporizer was closed (0% sevoflurane concentration) until the end-tidal sevoflurane concentration had decreased to 1% or a BIS value of 60 had been reached. This sequence was repeated a second time. After the final suture was placed, the sevoflurane was discontinued and patients were allowed to awaken from anesthesia. The measurements recorded during emergence from anesthesia were also included in the pharmacodynamic model.

End-tidal sevoflurane concentrations were measured using infrared absorption technology (pm 8050, Dräger, Lübeck, Germany). The precision of the end-tidal measurements was 0.1%. Study data were automatically recorded using the software programs "Proto 99" (version 1.0.2.0, Dräger, Lübeck, Germany) for the end-tidal sevoflurane concentrations and "Hyperterminal" (Microsoft, Redmond, VA) for BIS values. The Narcotrend index was recorded using a modified research laptop version with the Narcotrend software 4.0. The power was calculated and recorded with the Narcotrend monitor. The burst suppression ratio (BSR) was recorded from the BIS A-2000 monitor. The synchronization of the data was performed automatically by the software Access 2000 (Microsoft) by the time given by each device for each individual patient.

Pharmacodynamic Analysis
Using the program system NONMEM V (GloboMax LLC, Hanover, MD) we modeled the dose–response relationship between sevoflurane and the EEG indices. EEG parameters were related to sevoflurane effect compartment concentrations using a classical first-order effect site:

where C_et is the end-tidal concentration, C_eff is the effect compartment concentration, and k_e0 is the first-order rate constant determining the equilibration between the two.

The relationship between effect compartment concentrations and EEG without the plateau effect was modeled with a fractional sigmoid E_max model (Hill equation)2:

where 99 is the baseline effect, C_eff is the apparent effect compartment concentration, C₅₀ is the concentration that causes 50% of the maximum effect, and describes the slope of the concentration-response relation.

In the next step, we visually investigated the dose–response relationship without any underlying model. An individual k_e0 value was chosen which collapsed the hysteresis loop adequately. Sevoflurane revealed a pronounced plateau leading to a biphasic dose–response curve. The relationship between effect compartment concentrations and EEG with the plateau effect was modeled with a variation of the classical fractional sigmoid E_max model with two linked sigmoidal curves describing the EEG effect at sevoflurane concentrations without burst suppression (noBSR) and after the onset of burst suppression (BSR).

For C_eff C_plateau:

For C_eff > C_plateau:

Both sigmoidal curves have their own parameters. C_eff is the apparent effect compartment concentration. The first curve goes from 99, the presumed value in the absence of an anesthetic, to E_plateau, where a transition occurs to the second curve, which extends from E_plateau to E_max, the presumed maximum drug effect. C_plateau is the effect site concentration at E_plateau. C_50noBSR is the effect site concentration associated with 50% decrease from 99 to E_plateau, and _{no BSR} is the steepness of that relationship. C_50BSR is the effect site concentration associated with 50% decrease from E_plateau to 0, and _BSR is the steepness of that relationship.

The relationship between effect compartment concentrations and power was modeled with a variation of the classical fractional sigmoid E_max model with two linked sigmoidal curves describing the power effect at increasing sevoflurane concentrations with a unimodal form.

For C_eff C_deltamax:

For C_eff > C_deltamax:

The first curve goes from 0, the presumed power in the absence of anesthesia, to _max, where the maximum power is reached. The second decreasing curve occurs, which extends from _max to 0. C_deltamax is the anesthetic concentration at _max. C_50increase is the concentration associated with 50% increase from 0 to _max, and _increase is the steepness of that relationship. C_decrease is the concentration associated with 50% decrease from _max to 0, and _decrease is the steepness of that relationship.

The relationship between effect compartment concentrations and BSR was modeled with an inverted sigmoid E_max model.

where 0 is the baseline BSR in the absence of an anesthetic and 100 the maximum BSR.

All parameters were calculated as a population fit by NONMEM V in one step by minimizing log likelihood, which maximizes the likelihood between the measured and the predicted EEG parameters. For the calculation the connect-the-dots approach was used.9,10 By using this algorithm the first effect compartment concentration normally has to be zero but, due to our study design, the first effect compartment concentration was set to the end-tidal concentration. This was possible because the end-tidal concentration was constant before increasing the sevoflurane concentration.

In addition, we calculated the correlation between the measured EEG parameters and the predicted concentrations for the individual patient. These computations were performed using Excel 2000 software (Microsoft) and the parameters were optimized with the Solver tool within Excel.

Statistical Analysis
For testing statistical significance between the NONMEM population fits of the two models with and without the plateau effect, the likelihood ratio test was used. The difference between log likelihood values follows a ² distribution. For the pharmacodynamic model with the plateau effect we added four parameters to the model without the plateau effect. With a probability of 0.05 and 1 degree of freedom, the value of the ² distribution is 9.49. If the difference in the NONMEM objective functions for the two models differs by the four parameters exceeds 9.49, then the parameters are significant at P < 0.05.9 Residual intra-individual variability was modeled using a standard additive error model. The residual error describes the accuracy of the indices scale.

For the individual patients, we calculated the correlation coefficient between the calculated effect compartment concentrations and measured EEG index values versus the predicted PK/PD model. A value of R close to one means that all changes in the EEG measurement can be entirely explained by changes in the effect compartment concentrations. A value of R close to zero means that there is no relationship between effect compartment concentration and EEG effect. The R values calculated for the BIS were compared with those calculated for the Narcotrend indices and tested with the Student’s t-test. All tests were two-tailed with statistical significance defined as P < 0.05.

Data are presented as mean and standard deviation. Statistical calculations were performed using SigmaStat^TM 2.03, SigmaPlot^TM 2000 and SPSS^TM (version 12; SPSS GmbH, Erkrath, NRW, Germany) computer software.

	RESULTS

Top Abstract Introduction METHODS RESULTS DISCUSSION REFERENCES

 
Twenty-six male ASA physical status II patients (64 ± 8 yr, 83 ± 8 kg, 173 ± 5 cm) were enrolled in this study. No patient complained of postoperative recall of intraoperative awareness.

The maximum end-tidal sevoflurane concentrations during the two concentrations ramps were 4.43% ± 0.43%. Study measurement time was 131 ± 15 min. A total of 150575 data pairs were collected. Because of artifact detection no parameter value could be calculated by the Narcotrend monitor in 12% of EEG-epochs versus 3% of EEG-epochs detected by the A-2000 BIS monitor.

To verify if a bisigmoidal model is adequate to describe the relationship between sevoflurane effect compartment concentration versus BIS and Narcotrend index, we visually investigated the dose–response relationship without any underlying model. An individual k_e0 value was chosen which collapsed the hysteresis loop adequately. Sevoflurane revealed a pronounced plateau leading to a biphasic dose–response curve.

The NONMEM population fits with (right graphs) and without (left graphs) the plateau effect of the measured BIS (upper graph) and Narcotrend index (lower graph) values versus the calculated sevoflurane effect compartment concentrations of all patients are shown in Figure 1. The pharmacodynamic parameters of the NONMEM population fits are shown in Table 1. The difference between the log likelihood values of the new PK/PD model with two connected sigmoidal curves and the classical E_max model with one sigmoidal curve is 396 (P < 0.001) for the BIS monitor and 1121 (P < 0.001) for the Narcotrend.

View larger version (44K): [in this window] [in a new window]   Figure 1. Comparisons between the dose–response curves calculated with the new bisigmoidal model (right graphs) and the classical Emax model with one sigmoidal curve (left graphs) of the sevoflurane effect compartment concentration versus Bispectral (upper graphs) and Narcotrend (lower graphs) index.

Figures 2 and 3 shows the best fit (a), median fit (b), and worst fit (c) relations between the BIS (Fig. 2) or the Narcotrend index (Fig. 3) and the sevoflurane effect compartment concentration of both models for individual patients. The correlation of BIS and Narcotrend indices to sevoflurane effect compartment concentrations including the plateau effect (r = 0.86 ± 0.09 for BIS and r = 0.92 ± 0.06 for Narcotrend) was significantly better than the correlation of both indices to sevoflurane effect compartment concentrations without the plateau effect (r = 0.78 ± 0.12 for BIS and r = 0.80 ± 0.18 for Narcotrend) (P < 0.05). The PK/PD parameters of the individual fits of both models are given in Table 2. The mean k_e0 value of the individual fits including the plateau effect derived from the Narcotrend index was 0.17 ± 0.06 min^–1 and that from the BIS data 0.19 ± 0.08 min^–1. The change between the first and the second sigmoidal curve was positioned at a mean BIS value of 40.2 ± 5.9 and a mean Narcotrend index of 47.6 ± 7.8. The respective sevoflurane effect compartment concentrations were 1.8 ± 0.6 vol % for the Narcotrend monitor and 2.04 ± 0.39 vol % for the BIS monitor. The C₅₀ values for the EEG response prior to the onset of burst suppression were 1.1% ± 0.2% for the Narcotrend index and 1.0% ± 0.1% for the BIS. The C₅₀ values with burst suppression were significantly higher than those without burst suppression, i.e., for Narcotrend index 2.5% ± 0.4% and for BIS 3.4% ± 2.6%.

View larger version (23K): [in this window] [in a new window]   Figure 2. Individual dose–response curves between Bispectral Index and the sevoflurane effect compartment concentrations for best fit (a), median fit (b), and worst fit (c) calculated with the new bisigmoidal model (right graphs) and the classical Emax model with one sigmoidal curve (left graphs). Dots are the respective electroencephalogram parameter values of a 5-s epoch.

View larger version (23K): [in this window] [in a new window]   Figure 3. Individual dose–response curves between Narcotrend index and the sevoflurane effect compartment concentrations for best fit (a), median fit (b), and worst fit (c) calculated with the new bisigmoidal model (right graphs) and the classical Emax model with one sigmoidal curve (left graphs). Dots are the respective electroencephalogram parameter values of a 5-s epoch.

Delta Power and BSR
The NONMEM population fit of the unimodal form dose–response curves of the sevoflurane effect compartment concentration versus the power is shown in Figure 4 (upper graph). Figure 5 shows the best fit (a), median fit (b), and worst fit (c) relations between power (left graph) or the BSR (right graph) and the sevoflurane effect compartment concentration for individual patients. The mean k_e0 value derived from the power was 0.19 ± 0.08 min^–1 for the individual fits and 0.11 min^–1 for the population fit. The maximum power (_max) was 63.9 ± 8.2 (individual fits) and 61.4 (population fit), and the respective sevoflurane effect compartment concentration were 1.75% ± 0.36% (individual fits) and 1.64% (population fits). The C_50increase value for the increasing part of the dose– response curve was 1.50% ± 0.13% (individual fits) and 0.91% (population fit), and for the decreasing part 3.45% ± 1.05% (individual fits) and 3.72% (population fit). The accuracy of the power scale was = 19.9 (population fit).

View larger version (27K): [in this window] [in a new window]   Figure 4. Unimodal form dose–response curves of the sevoflurane effect compartment concentration versus the power NONMEM V population fit (upper graph). The sevoflurane effect compartment concentration was calculated with a ke0 value of 0.11 min–1. Inverted dose–response curves of the sevoflurane compartment concentration versus burst suppression ratio (lower graph).

View larger version (23K): [in this window] [in a new window]   Figure 5. Unimodal form dose–response curves between power (left graphs) and the sevoflurane effect compartment concentrations and the inverted dose–response curves of the sevoflurane compartment concentration versus burst suppression ratio (right graphs) for best fit (a), median fit (b), and worst fit (c).

The NONMEM population fit of the inverted dose– response curves of the sevoflurane effect compartment concentration versus BSR is shown in Figure 4 (lower graph). The C₅₀ value for the BSR was 2.74% ± 0.35% (individual fits) and 3.5% (population fit), and was 3.15 ± 2.94 (individual fits) and 5.25 (population fit). The accuracy of the BSR scale was = 10.2 (population fit).

	DISCUSSION

Top Abstract Introduction METHODS RESULTS DISCUSSION REFERENCES

 
The aim of the study was to calculate dose–response curves between BIS and Narcotrend index and sevoflurane concentration with a classical sigmoidal model. After these calculations were determined, we recognized that this classical dose–response curve did not describe the relationship adequately (Figs. 2 and 3). In a next step, we visually investigated the dose–response relationship without any underlying model. An individual k_e0 value was chosen which collapsed the hysteresis loop adequately. Sevoflurane revealed a pronounced plateau leading to a biphasic dose–response curve between effect compartment concentrations versus BIS and Narcotrend index. The plateau indicates unchanged EEG parameter values despite changing sevoflurane concentrations. These visual evaluations led to the development of the bi-sigmoidal model. We found that a two sigmoidal curves model including a plateau effect describes EEG effects measured with BIS and Narcotrend index of sevoflurane significantly better than a single sigmoidal curve model.

Pharmacodynamic modeling is performed to better understand drug behavior. There are various classical models describing the pharmacodynamics of drugs. In most cases a single sigmoidal curve has been used to describe the dose–response relation between anesthetic drugs and EEG data. A sigmoidal function per se has the advantage of high flexibility, i.e., it is suitable to describe data following near a step function as well as data following near a linear function. The administration of anesthetics such as volatile anesthetics, barbiturates, etomidate, and propofol produces comparable and dose-dependent changes of the raw EEG signals according to the following pattern. In small doses, anesthetics lead to a short-term desynchronization, predominantly showing a high frequency β rhythm. With increasing anesthetic effect, a slowing of the EEG frequency is observed. The EEG is again synchronized with increasing and then activity with a simultaneous increase in amplitude. A further increase of the anesthetic dose results in an increasing inhibition of the electrical activity until a flat EEG (suppression) is registered, which is merely disturbed by short-term activity bursts. This is called a "burst suppression pattern." Finally, a flat EEG prevails, indicating cortical silence. After stopping the anesthetic, these changes are observed in reverse order until the patient awakens.

The BIS is a multiparameter EEG index that integrates bispectral, spectral, quasisuppression, and time domain (BSR) parameters.5,6 The BIS is derived by statistical correlation from these EEG parameters aimed to create a linear correlation with clinical end-points. This could lead to a lack of precision at higher volatile anesthetic concentrations. Previous studies have demonstrated that the BIS exhibits a plateau,11 which takes place at the transition to burst suppression.12,13

The Narcotrend monitor uses a classification developed for visual assessment of the EEG based on a six-letter ordinal scale ranging from A (awake) to F (general anesthesia with increasing burst suppression). The Narcotrend analysis integrates this scale using a multivariate statistical algorithm that yields into the index value.

Fitting this dose–response curve including a substantial plateau with a single sigmoidal curve would result in a model misspecification. This would influence not only the value for the goodness of the fit but also the estimation of the pharmacodynamic parameter, e.g., a falsely high C₅₀ value and a falsely low . Others have handled this problem by limiting drug doses to avoid the appearance of a burst suppression pattern14 or by using prediction probability P_K to measure the performance of anesthetic depth indicators15,16 which is model-independent. In another study,4 we developed a new model for this purpose with two subsequent sigmoidal curves. The first describes the EEG effect before the appearance of a burst suppression pattern; the second describes the effect resulting in a burst suppression pattern. With this new model the dose–response curve for Narcotrend and BIS during isoflurane anesthesia has been adequately described.

We used the population pharmacokinetics based NONMEM software to directly compare the single sigmoidal curve model and the two sigmoidal curves model for describing the effects of sevoflurane. The plateau of the BIS and Narcotrend indices is positioned at the change between the maximum power and the increase of BSR (Fig. 6). Both might contribute to the occurrence of the plateau effect. Methodological issues inherent to the monitor algorithms or modeling factors might add to the plateau effect. It remains unclear whether the power in this range is really unchanged or whether the monitor algorithm is just too insensitive in this range. For the BIS monitor, a BSR of 40% has been selected to transfer the algorithm to another dimension using a linear correlation between BSR and BIS.4 A different algorithm for example with a change at a BSR of 10% might result in a smaller plateau effect.

View larger version (24K): [in this window] [in a new window]   Figure 6. Comparison of the population fits for Bispectral Index (BIS), Narcotrend-index, power and burst suppression ratio versus sevoflurane effect site concentration.

The plateau effect is clinically important. For BIS or Narcotrend-guided titration of anesthetic drugs, the plateau is a disadvantage because changing drug concentrations are not reflected by both EEG parameters. This is of less importance if an EEG target value has been chosen at the steep part of the first sigmoidal curve, but is of more importance at the lower inflection point, respectively, the beginning of the plateau. The beginning of the plateau effect varies interindividually and may be dependent upon the drugs used.

While the interpretation algorithms of the BIS and the Narcotrend monitor are completely different, both monitors show a similar performance in measuring the drug effect during sevoflurane anesthesia. Schmidt et al.16 studied 25 patients after the end of elective spine surgery without any surgical stimulation. During step-by-step reduction of propofol, starting from a target concentration of 3 µg/mL, BIS and Narcotrend (Version 2.0 AF/F) performed equally in predicting the propofol target concentration. In a previous study, we investigated 18 patients undergoing radical prostatectomy with combined epidural–propofol general anesthesia and found that the Narcotrend index (Version 4.0) and the BIS were comparable in predicting propofol effect site concentrations.16 Similarly, BIS and Narcotrend index (Version 4.0) detected the EEG effects of isoflurane equally during combined epidural–isoflurane anesthesia.4 In addition, we can confirm the finding, previously reported for propofol anesthesia, that the Narcotrend index and BIS values cannot be interchanged 1:1. Deviation from the line of identity, especially in the small- and the large-dose range, must be considered.17

In conclusion, the new two sigmoidal curves model including a plateau effect describes the EEG effect measured with BIS and Narcotrend index of sevoflurane better than a single sigmoidal curve model. The Narcotrend index has a similar performance compared with BIS as a measure of anesthetic drug effect. The advantages of this new model might be relevant for future studies including higher dose ranges with appearance of burst suppression pattern.

	Footnotes

 
Accepted for publication December 18, 2007.

Supported by departmental funding. This study was presented in part at the American Society of Anesthesiologists meeting October 20, 2005 Atlanta.

	REFERENCES

Top Abstract Introduction METHODS RESULTS DISCUSSION REFERENCES

 

1. Rehberg B, Bouillon T, Zinserling J, Hoeft A. Comparative pharmacodynamic modeling of the electroencephalography-slowing effect of isoflurane, sevoflurane, and desflurane. Anesthesiology 1999;91:397–405[Web of Science][Medline]
2. Hill AV. The possible effect of the aggregation of the molecules of hemoglobin on its dissociation curves. J Physiol 1910;40:iv–vii
3. Sheiner LB, Stanski DR, Vozeh S, Miller RD, Ham J. Simultaneous modeling of pharmacokinetics and pharmacodynamics: application to d-tubocurarine. Clin Pharmacol Ther 1979;25:358–71[Web of Science][Medline]
4. Kreuer S, Bruhn J, Larsen R, Grundmann U, Wilhelm W. Application of Narcotrend index and bispectral index to the measurement of the EEG effects of isoflurane with and without burst suppression. Anesthesiology 2004;101:847–54[Web of Science][Medline]
5. Johansen JW, Sebel PS. Development and clinical application of electroencephalo-graphic bispectrum monitoring. Anesthesiology 2000;93:1336–44[Web of Science][Medline]
6. Rampil IJ. A primer for EEG signal processing in anesthesia. Anesthesiology 1998;89:980–1002[Web of Science][Medline]
7. Kreuer S, Biedler A, Larsen R, Altmann S, Wilhelm W. Narcotrend monitoring allows faster emergence and a reduction of drug consumption in propofol-remifentanil anesthesia. Anesthesiology 2003;99:34–41[Web of Science][Medline]
8. Schultz B, Kreuer S, Wilhelm W, Grouven U, Schultz A. The Narcotrend monitor. Development and interpretation algorithms. Anaesthesist 2003;52:1143–8[Web of Science][Medline]
9. Shafer SL. Pharmacokinetic and Pharmacodynamic Analysis with NONMEM – Basic Concepts. Workshop, Palo Alto, California, USA, 2004;9:17
10. Schnider TW, Minto CF, Shafer SL, Gambus PL, Andresen C, Goodale DB, Youngs EJ. The influence of age on propofol pharmacodynamics. Anesthesiology 1999;90:1502–16[Web of Science][Medline]
11. Olofsen E, Dahan A. The dynamic relationship between end-tidal sevoflurane and isoflurane concentrations and bispectral index and spectral edge frequency of the electroencephalogram. Anesthesiology 1999;90:1345–53[Web of Science][Medline]
12. Bruhn J, Bouillon T, Shafer SL. Bispectral index (BIS) and burst suppression: revealing a part of the BIS algorithm. J Clin Monitor Comp 2001;16:593–6
13. Bruhn J, Bouillon T, Shafer SL. The onset of propofol-induced burst suppression may be correctly detected by approximate entropy but not by the bispectral index. Br J Anaesth 2001;87:505–7[Abstract/Free Full Text]
14. Vanluchene ALG, Vereecke H, Thas O, Mortier EP, Shafer SL, Struys MM. Spectral entropy as an electroencephalographic measure of anesthetic drug effect: a comparison with bispectral index and processed midlatency auditory evoked response. Anesthesiology 2004;101:34–42[Web of Science][Medline]
15. Kreuer S, Wilhelm W, Grundmann U, Larsen R, Bruhn J. Narcotrend index versus bispectral index as electroencephalogram measures of anesthetic drug effect during propofol anesthesia. Anesth Analg 2004;98:692–7[Abstract/Free Full Text]
16. Schmidt GN, Bischoff P, Standl T, Voigt M, Papavero L, Schulte am Esch J. Narcotrend, bispectral index, and classical electroencephalogram variables during emergence from propofol/remifentanil anesthesia. Anesth Analg 2002;95:1324–30[Abstract/Free Full Text]
17. Kreuer S, Bruhn J, Larsen R, Bialas P, Wilhelm W. Comparability of Narcotrend index and bispectral index during propofol anaesthesia. Br J Anaesth 2004;93:235–40[Abstract/Free Full Text]

This article has been cited by other articles:

				R. K. Ellerkmann, J. Bruhn, M. Soehle, M. Kehrer, A. Hoeft, and S. Kreuer Maximizing Prediction Probability PK as an Alternative Semiparametric Approach to Estimate the Plasma Effect-Site Equilibration Rate Constant ke0 Anesth. Analg., November 1, 2009; 109(5): 1470 - 1478. [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 Google Scholar Google Scholar Articles by Kreuer, S. Articles by Wilhelm, W. Search for Related Content PubMed PubMed Citation Articles by Kreuer, S. Articles by Wilhelm, W. Related Collections Clinical Pharmacology Technology Pharmacology