BJA Advance Access originally published online on July 6, 2007
British Journal of Anaesthesia 2007 99(4):547-551; doi:10.1093/bja/aem189
New parameters of skin conductance compared with bispectral index® monitoring to assess emergence from total intravenous anaesthesia
1 Department of Anaesthesia and Pain Medicine, Royal Perth Hospital, Wellington Street Campus, Perth, WA 6000, Australia
2 Department of Anaesthesia and Intensive Care Medicine, University Hospital Schleswig-Holstein, Campus Kiel, Germany
3 School of Medicine and Pharmacology, The University of Western Australia, Perth, Australia
* Corresponding author: Department of Anaesthesia and Pain Medicine, Royal Perth Hospital, Wellington Street Campus, Perth, WA 6000, Australia. E-mail: thomas.ledowski{at}health.wa.gov.au
Accepted for publication May 29, 2007.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Background: Arousal after total i.v. anaesthesia (TIVA) has been reported to be detectable by monitoring the number of fluctuations per second (NFSC), a parameter of skin conductance (SC). However, compared with monitoring of the bispectral index (BIS®), the predictive probability of NFSC was significantly lower. The aim of this study was to determine the value of the two new, not yet published parameters of SC, area under the curve (AUC) methods A and B, for monitoring emergence from TIVA compared with monitoring of NFSC and BIS®.
Methods: Twenty-five patients undergoing surgery were investigated. NFSC, AUC A, AUC B, BIS®, and haemodynamic parameters (mean arterial pressure and heart rate) were recorded simultaneously. The performance of the monitoring devices in distinguishing between the clinical states ‘steady-state anaesthesia’, ‘first clinical reaction’, and ‘extubation’ were compared using the method of prediction probability (Pk) calculation.
Results: BIS® showed the best performance in distinguishing between ‘steady-state anaesthesia’ vs ‘first reaction’ (Pk BIS® 0.95; NFSC 0.73; AUC A 0.54; AUC B 0.62) and ‘steady-state anaesthesia’ vs ‘extubation’ (Pk BIS® 0.99; NFSC 0.73; AUC A 0.71; AUC B 0.67). However, the time from first BIS®>60/SC>0 to a first clinical reaction was significantly shorter for BIS® (median BIS® 180 s; NFSC 780 s; AUC A 750 s; AUC B 690 s; P < 0.001).
Conclusions: AUC A and AUC B did not improve accuracy of SC monitoring in patients waking after TIVA.
Keywords: anaesthetics i.v.; monitoring, bispectral index; monitoring, depth of anaesthesia; sympathetic nervous system
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Neurophysiological arousal has been shown to coincide with an increase in sympathetic tone, finally leading to an increased firing rate of sympathetic, post-ganglionic cholinergic neurones.1 2 The resulting increase in sweat gland filling can be measured in terms of skin conductance (SC).3
The SC software used in this study allows almost online measurement of sympathetic activity (see Methods) and, as fast assessment is crucial for depth of anaesthesia monitoring, SC could potentially be superior when compared with other estimators of sympathetic tone, such as heart rate (HR) variability.
Emergence after total i.v. anaesthesia (TIVA) has been reported to be detectable by monitoring the number of fluctuations per second (NFSC), a parameter of SC.4 5 However, NFSC was less accurate to predict awakening when compared with bispectral index (BIS®)4 and slower compared with response entropy.5
Although NFSC correlates with sympathetic activity,6 the parameter does not provide information regarding the amplitude of the fluctuations (other than exceeding a certain pre-set threshold), which has also been proven to reflect different states of sympathetic tone.7
We hypothetized that a combined SC parameter may be more precise in the prediction of awakening than the previously tested parameter NFSC.4
Therefore, the aim of this study was to investigate the two new parameters of SC, area under the curve (AUC) methods A and B, that combine the information of NFSC and the amplitude of SC fluctuations.
We compared monitoring of AUC A, AUC B, NFSC, and BIS® to assess emergence from TIVA, with the hypothesis that AUC A and AUC B would be a more accurate parameter compared with NFSC.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
After approval by the ethics committee of Royal Perth Hospital, 25 patients undergoing minor elective surgery gave consent to participate in the trial. None of the participants received premedication on the day of surgery. The BIS® and SC monitors were connected when the patient was on the table in a supine position.
BIS® monitoring was performed using the BIS XP A 2000TM monitor (Aspect Medical Systems, Norwood, MA, USA) with BIS QUATTROTM single-use electrodes (Aspect Medical Systems) and a smoothing rate of 15 s. BIS® values displayed with a negative signal quality index (SQI) were not used for further analysis. During anaesthesia until cessation of remifentanil, a BIS value of 40–55 was maintained. A BIS® value <60 was considered an indicator of light anaesthesia.
The SC monitoring was achieved using the MEDSTORM AS 2005 monitor (Medstorm Innovations, Oslo, Norway) with three single-use Ag/AgCl paediatric ECG electrodes (NEOTRODE®, ConMed Corp., Rome, USA) attached to the palmar surface of the hand. The exosomatic electrodermal activity was measured in terms of conductance. The equipment used an alternating current of 88 Hz and an applied voltage of 50 mV (highest density 2.5 µA). A three-electrode system (measuring, counter, and reference electrodes) was used for unipolar measurement with a constant voltage applied to the stratum corneum beneath the measuring electrode. The monitor was connected to a laptop computer via a standard serial port connection to visualize and process the obtained data.
The SC software was able to define peaks and troughs to determine the amplitude of fluctuations within the mean SC, and from this count the NFSC. To reduce the electronic noise, the minimum amplitude was set at 0.02 µS. According to the study by Storm and colleagues,8 an NFSC of >0 s–1 in a period of 15 s, equalling NFSC >0.066 s–1, can be counted as a significant change and indicates an increase in the sympathetic outflow/arousal. As with the pre-set time window of 15 s, the next obtainable value for NFSC is 0.07 s–1, the term NFSC >0 (to indicate awakening) is used in the following text as a simplification. During deep anaesthesia, in contrast, NFSC is normally 0.
AUC A was calculated by measuring the AUC of the mean SC from the level of the first defined trough (Fig. 1A). This method was designed to monitor very broad SC fluctuations as seen during awakening. In order to receive a result >0, at least one trough and one peak needed to be defined within the analysing window. AUC B was calculated by measuring the sum of the AUC of each individual fluctuation (Fig. 1B) within the given time window. This method was designed to measure smaller, sharper sympathetic bursts, as described in situations with sufficient anaesthesia, but insufficient analgesia.8 An AUC A, AUC B>0 µSiemens x s was considered as an indication of increased sympathetic tone and therefore potentially indicating arousal.
|
Anaesthesia was induced with propofol 2 mg kg–1 and remifentanil 0.5 µg kg–1 min–1. If required, muscle relaxation was achieved with rocuronium 0.6 mg kg–1. After placement of the airway device, either a tracheal tube or a laryngeal mask, anaesthesia was maintained using propofol and remifentanil, as clinically appropriate. Ten minutes before the anticipated end of surgery, the remifentanil infusion was stopped and fentanyl was given in a dose considered appropriate for postoperative analgesia (dose range 0–125 µg). Propofol was stopped at the end of surgery. Neuromuscular block was not antagonized.
At the time of stopping the remifentanil infusion, a stopwatch was started and mean arterial pressure (MAP), HR, BIS®, NFSC, AUC A, AUC B, and a clinical score of depth of sedation, the observer alertness assessment scale (OAAS),9 were recorded every 2.5 min and at defined time points (‘first clinical reaction’, ‘extubation’). The times from cessation of remifentanil and propofol to the first BIS® value >60 and NFSC, AUC A, AUC B>0, and from a BIS®>60 and NFSC, AUC A, AUC B>0 to the time of first clinical reaction (defined as a first reaction of the patient indicating a light anaesthesia, such as coughing, movement, and eye opening) and extubation were recorded. The patients were extubated as soon as they were considered clinically suitable and showed a minimum OAAS of 3 points.
Statistical analysis
The accuracy in distinguishing between the anaesthetic states of ‘steady state’ vs ‘first clinical reaction’ and ‘steady state’ vs ‘extubation’ was assessed by means of prediction probability (Pk). This method was originally described by Smith and colleagues.10
For the calculation of the Pk values, a custom spreadsheet macro (PkMACRO) as described by Smith and colleagues10 was used. A second spreadsheet, PkDMACRO, was used to compute the t-value for a comparison between Pk values of the different monitors. The standard error of the estimate was computed by the jackknife method.10 A Pk value of 1 means a 100% correct prediction of a certain clinical state by a specific monitor, whereas a value of 0.5 represents only a 50:50 chance.
Correlations of NFSC, AUC A, AUC B, and BIS® were estimated using the Spearman rank correlation coefficient (
). The probability for first reaction (OAAS>0) compared with steady-state anaesthesia was calculated for certain values of BIS®, NFSC, AUC A, and AUC B.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
The data of 25 patients [five female, 20 male; age 45 (SD 17) yr; weight 82 (23) kg; height 172 (15) cm; nine ASA I, 14 ASA II, two ASA III] were included in the analysis. A laryngeal mask was used as an airway device in 19 patients and a tracheal tube in six patients. Fentanyl was given in a mean dose of 34 µg (0–125 µg), at the time the remifentanil infusion {0.27 (0.17) µg kg–1 min–1 [mean (SD)]} was stopped.
Median (25th/75th percentiles) values for NFSC [0 (0/0.27)], AUC A [0 (0/0.15)], AUC B [0 (0/0.05)], and BIS® [38 (34/44.5)] at steady-state anaesthesia were significantly different (Wilcoxon test) from those at the time of first clinical reaction for NFSC [0.27 (0.1/0.6); P < 0.001], AUC B [0 (0/0.25); P = 0.034], and BIS® [65 (57.5/77); P < 0.001], but not for AUC A [0 (0/0.1)]. Only BIS® changed significantly from the time of first clinical reaction to the time of extubation [77 (68.5/85); P < 0.001], whereas values for all other parameters did not change significantly [NFSC 0.2 (0.2/0.55), AUC A 0.2 (0/0.6), AUC B 0.1 (0/0.25)].
The investigated response times ‘cessation of remifentanil to first BIS®>60, NFSC, AUC A, AUC B>0’, ‘cessation of propofol to first BIS®>60, NFSC, AUC A, AUC B>0’, ‘first BIS®>60, NFSC, AUC A, AUC B>0 to first clinical reaction’, and ‘first BIS®>60, NFSC, AUC A, AUC B>0 to extubation’ were significantly different (Table 1), with NFSC, AUC A, and AUC B reacting faster after cessation of remifentanil and giving a longer pre-warning time from its change to a clinical first reaction.
|
There was no significant difference between patients with a laryngeal mask or with a tracheal tube. When comparing the response times of patients who did (n = 10) and did not (n = 15) receive fentanyl before waking up, no significant effect was seen for any of the SC parameters (ANOVA).
BIS® performed significantly better than all other investigated parameters regarding the differentiation between ‘steady-state anaesthesia’ and ‘first clinical reaction’ and ‘steady-state anaesthesia’ and ‘extubation’ (Table 2). The investigated parameters of SC were not better in distinguishing the investigated clinical states than MAP and HR. NFSC performed better than AUC A and B regarding the prediction of the different clinical states (Table 2).
|
There was no significant correlation between points on the OAAS scale at the time of ‘first clinical reaction’ and ‘extubation’ and the BIS® and NFSC, AUC A, AUC B readings at these times. No significant correlation was found between BIS® and NFSC readings at any of the investigated time points. The probability of an OAAS scale value of >0 as a function of BIS®, NFSC, AUC A, and AUC B at all 308 readings was calculated and is shown in Figure 2.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
The notion of using changes in the electrogalvanic activity of the skin for assessment of sympathetic activity and, related to that, sedation was raised almost 40 yr ago by Nisbet and colleagues11 and reinvestigated by Goddard12 in 1982.
NFSC as one parameter of SC was introduced to assess responses to noxious and awakening stimuli in anaesthetized patients by Storm and colleagues.8 More recently, Gjerstad and colleagues5 compared NFSC and entropy, and concluded that both modalities were able to predict emergence at the end of anaesthesia. Ledowski and colleagues13 demonstrated that NFSC performed similarly to BIS® in patients waking after general anaesthesia using sevoflurane and remifentanil. In an identical protocol, but using TIVA with propofol and remifentanil, the authors found NFSC, compared with BIS® to be markedly less accurate in the prediction of different clinical states of arousal, but faster regarding the investigated response times.4 Therefore, it has been suggested that NFSC alone may not be the ideal parameter of SC to predict awakening.4
Storm and colleagues8 suggested that a combination of the two parameters of SC, NFSC and mean SC, could have the potential to detect inadequate hypnotic states. We tested this hypothesis by combining NFSC and an increase in mean SC, AUC A, with the same protocol used by Ledowski and colleagues4 13 comparing NFSC, BIS®, and haemodynamic parameters. As mean SC has been questioned to be of clinical use in stress assessment,14 the AUC of NFSC not including the mean SC, AUC B, was additionally assessed.
Although we hypothesized that AUC A and AUC B could have potential benefits for the assessment of awakening, the results obtained in this study disproved this theory: both methods were less accurate in predicting a first clinical reaction and awakening after general anaesthesia when compared with NFSC and BIS®. The design of this study and the fact that SC and BIS® measure completely different parameters may have contributed to the results: in our protocol, the cessation of remifentanil defined the start of the period of awakening. A monitor that reacts not only to awakening (BIS®), but also to noxious stimuli during a period with decreasing analgesia (SC) to react faster, but for the same reason is likely to be less accurate, if compared with a clinical score of awareness (OAAS). Despite this potential limitation, we opted to use the same protocol as in our previous study4 as it best reflects a normal clinical situation.
Retrospectively, we tried to identify reasons that may have contributed to the poor outcome of AUC A and AUC B, when compared with NFSC: as per definition of our software, at least one peak and one trough was needed within the pre-set time window to define a value of >0 for AUC A, some fluctuations might have been missed, whenever they were broader than the selected time window (15 s). In addition, it has been mentioned that with anxiety, the amplitude of the SC response is initially increased, but decreases with extreme anxiety levels.11 Furthermore, during states of drowsiness, SC amplitude has been shown to be greater than with complete awareness.11 One suggestion that may explain lower SC amplitudes during extreme stress may be a desensitization of muscarinic G-protein-coupled receptors due to receptor internalization.15 As these receptors are essential for signal transduction and hence sweat gland filling, a desensitization would result in lower sweat gland filling and therefore lower SC amplitudes during extreme stress. Our results, however, do not allow a conclusion about that matter.
In the current study, NFSC showed lower Pk values compared with previously published results obtained with the same study protocol.4 As both investigations recruited only a limited number of subjects, this difference could be coincidental. Although in this study BIS® was proven to be the superior predictor of awakening, other parameters of SC may be more accurate than those studied by us. Gjerstad and colleagues16 found the derivative of the SC and entropy showed similar discrimination between sound responses at different sedation levels.
All investigated parameters of SC reacted significantly faster and provided a longer pre-warning time before a first clinical reaction than BIS®. This would allow the anaesthetist more time to deepen an unwanted light anaesthetic. This result matches with those of our previously published study comparing NFSC and BIS® during awakening after TIVA.4
The faster refreshing rate of the SC monitor (1 s), compared with BIS®, which is known to represent the clinical state with a delay of approximately 30 s17 could be a potential explanation. In addition, the choice of cut-off values for BIS® and SC parameters used in this study may as well have contributed to the difference in reaction times and may therefore be seen as a potential limitation of our results. Although a BIS® higher than 60 is widely accepted as a value indicating arousal18 19 and recommended for this purpose by the manufacturer, it is unclear, whether or not it reflects the same state of arousal as an SC parameter >0, a value that has been reported to indicate awakening.5 6 8
Compared with SC, BIS® showed the highest accuracy to predict the investigated clinical states during awakening from TIVA. Both methods involving the amplitude of SC fluctuations, AUC A and AUC B, performed worse than NFSC.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
This study was funded by the Department of Anaesthesia and Pain Medicine, Royal Perth Hospital. Each monitor was a research-bound loan device from the manufacturers: the device for skin conductance by Medstorm Innovation, Oslo, Norway; and the Bispectral Index® Monitor by Aspect Medical Systems Inc., Norwood, MA, USA.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
1 Storm H, Fremming A, Oedegaard S, Martinsen OG, Moerkrid L. The development of a software program for analysing spontaneous and elicited skin conductance changes in infants and adults. Clin Neurophysiol (2000) 111:1889–98.[CrossRef][Web of Science][Medline]
2 Critchley HD, Elliott R, Mathias CJ, Dolan RJ. Neural activity relating to generation and representation of galvanic skin conductance responses: a functional magnetic resonance imaging study. J Neurosci (2000) 20:3033–40.
3 Edelberg A. Electrical properties of the skin. In: Methods in Psychophysiology—Brown CC, ed. (1967) Baltimore: Williams & Wilkins. 1–53.
4 Ledowski T, Bromilow J, Paech MJ, Storm H, Hacking R, Schug SA. Skin conductance monitoring compared with bispectral index to assess emergence from total i.v. anaesthesia using propofol and remifentanil. Br J Anaesth (2006) 97:817–21.
5 Gjerstad AC, Storm H, Hagen R, Huiku M, Qvigstad E, Raeder J. Comparison of skin conductance with entropy during intubation, tetanic stimulation and emergence from general anaesthesia. Acta Anaesthesiol Scand (2007) 51:8–15.[CrossRef][Web of Science][Medline]
6 Storm H, Myre K, Rostrup M, Stokland O, Lien MD, Raeder JC. Skin conductance correlates with perioperative stress. Acta Anaesthesiol Scand (2002) 46:887–95.[CrossRef][Web of Science][Medline]
7 Lindberg L, Wallin G. Sympathetic skin nerve discharges in relation to amplitude of skin resistance responses. Psychophysiology (1981) 18:268–70.[Web of Science][Medline]
8 Storm H, Shafiei M, Myre K, Raeder J. Palmar skin conductance compared to a developed stress score and to noxious and awakening stimuli on patients in anaesthesia. Acta Anaesthesiol Scand (2005) 49:798–803.[CrossRef][Web of Science][Medline]
9 Chernik DA, Gillings D, Laine H, et al. Validity and reliability of the observer's assessment of alertness/sedation scale: study with intravenous midazolam. J Clin Psychopharmacol (1990) 10:244–51.[Web of Science][Medline]
10 Smith WD, Dutton RC, Smith NT. Measuring the performance of anesthetic depth indicators. Anesthesiology (1996) 84:38–51.[CrossRef][Web of Science][Medline]
11 Nisbet HIA, Norris W, Brown J. Objective measurement of sedation, IV. The measurement and interpretation of electrical changes in the skin. Br J Anaesth (1967) 39:798–805.
12 Goddard GF. A pilot study of the changes of skin conductance in patients undergoing general surgery. Anaesthesia (1982) 37:408–15.[Web of Science][Medline]
13 Ledowski T, Paech MJ, Storm H, Jones R, Schug SA. Skin conductance monitoring compared with bispectral index monitoring to assess emergence from general anaesthesia using sevoflurane and remifentanil. Br J Anaesth (2006) 97:187–91.
14 Harrison D, Boyce S, Loughnan P, Dargaville P, Storm H, Johnston L. Skin conductance as a measure of pain and stress in hospitalized infants. Early Hum Dev (2006) 82:603–8.[CrossRef][Web of Science][Medline]
15 Ferguson SS. Evolving concepts in G protein-coupled receptor endocytosis: the role in receptor desensitization and signaling. Pharmacol Rev (2001) 53:1–24.
16 Gjerstad AC, Storm H, Hagen R, Huiku M, Qvigstad E, Raeder J. Skin conductance or entropy for the detection of non-noxious stimulation during different levels of sedation. Acta Anaesthesiol Scand (2007) 51:1–7.[CrossRef][Web of Science][Medline]
17 Jensen EW, Litvan H. Rapid extraction of mid-latency auditory-evoked potentials (letter). Anesthesiology (2001) 94:718.[Web of Science][Medline]
18 Myles PS, 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.[CrossRef][Web of Science][Medline]
19 Ekman A, Lindholm ML, Lennmarken C, et al. Reduction in the incidence of awareness using BIS monitoring. Acta Anaesthesiol Scand (2004) 48:20–6.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||