- Email: [email protected]
Midlatency auditory-evoked potentials in the assessment of sedation in cardiac surgery patients
Midlatency Auditory-Evoked Potentials in the Assessment of Sedation in Cardiac Surgery Patients Tadeusz Musialowicz, MD,* Markku Hynynen, MD, PhD,† Heidi Yppa¨rila¨, PhD,‡ Pekka Po¨lo¨nen, MD,* Esko Ruokonen, MD, PhD,* and Stephan M. Jakob, MD, PhD§ Objectives: Midlatency auditory-evoked potentials (MLAEPs) may provide an objective measure of depth of sedation. The aim of this study was to evaluate MLAEPs for measuring sedation in cardiac surgery patients. Design: Prospective study. Setting: Intensive care unit of a university hospital. Participants: Twenty-two patients scheduled for elective coronary artery bypass grafting. Interventions: MLAEPs were obtained at 5 time points: the day before surgery (baseline), 1 hour before surgery, after premedication, postoperatively during deep (Ramsay 6) and moderate (Ramsay 4) sedation, and the day after surgery. Measurements and Main Results: The latency of the Nb MLAEP component increased from 44 ms (38-60 ms; median, range) at baseline to 49 ms (41-64 ms) after premedication (p
ⴝ 0.03) and further to 63 ms (48-80 ms) during deep sedation after surgery (P < 0.01). Although a decreasing clinical level of sedation after rewarming was not associated with a significant change in Nb latency (61 ms [42-78 ms]), the MLAEP NaPa amplitude increased from 0.9 V (0.4-1.6 V) to 1.3 V (0.8-3.9 V; p ⴝ 0.01). Nb latency remained increased the day after surgery (49 ms [37-71 ms]) as compared with baseline (p < 0.01). Conclusions: MLAEP latencies can reflect subtle changes in auditory perception, while amplitudes seem to change with transition between deep levels of sedation. © 2004 Elsevier Inc. All rights reserved.
I
ing data because of technical problems with the data collecting equipment (1 patient). The patients received diazepam, 0.2 mg/kg, orally the night before the operation and 2 hours before surgery. For induction of anesthesia, midazolam, 0.1 mg/kg, alfentanil, 100 g/kg, and pancuronium, 0.1 mg/kg, were used. Anesthesia was maintained with continuous infusions of propofol, 3 mg/kg/h, and alfentanil, 50 g/kg/h. The patients underwent CABG with cardiopulmonary bypass with nonpulsatile pump flow under moderate hypothermia at 34°C. At the end of the operation, alfentanil infusion was discontinued and propofol infusion was continued at a rate of 2 mg/kg/h during transport of the patient to the ICU. A physician and a nurse assessed the Ramsay score immediately before and after measurements of the MLAEPs and consensus was recorded. MLAEP measurements were performed at 5 time points: (1) on the day before surgery (baseline), (2) 1 hour before surgery after premedication, (3) after arrival in the ICU (propofol sedation adjusted to Ramsay score 6 [no response to glabellar tap or loud auditory stimulus]), (4) during maintenance of sedation in the ICU (propofol sedation adjusted to Ramsay score 4 [a brisk response to glabellar tap or loud auditory stimulus]), and (5) on the day after surgery, when the patient was extubated and fully awake. The investigators assessing the Ramsay score were not aware of the MLAEP characteristics.
NAPPROPRIATE SEDATION in cardiac surgery patients may cause cardiorespiratory depression, prolonged recovery times and intensive care unit (ICU) stays, and increased costs.1 Although depth of anesthesia can be monitored by recording an electroencephalogram (EEG) or EEG-derived parameters (spectral edge frequency, approximate entropy, bispectral index [BIS]),2 these parameters have not been proven useful for the monitoring of lighter levels of sedation; for instance, in patients recovering from surgery and anesthesia. In daily practice, the level of sedation is assessed by using clinical scoring systems such as the Ramsay score, which classifies a patient’s response to a command or painful stimulus. However, the Ramsay score has limited value in the prediction of the level of sedation because of many confounding factors, such as motor deficit and inter-rater variation in interpretation. Auditory-evoked potentials measure the output of the central nervous system in response to a controlled input. Most anesthetics depress midlatency auditory-evoked potentials (MLAEPs) in a dose-dependent fashion,3 and the changes are independent of the presence of opioids.4 For these reasons, it has been suggested that MLAEPs measure the hypnotic component of anesthesia and could serve as an effective discriminator between different states of consciousness.5 The purpose of this study was to assess whether graded changes in sedation induced by pre-and postoperative sedative drugs can be detected by any of the MLAEP components. METHODS The local ethics committee approved the study protocol. All patients gave written informed consent. Originally, 30 patients scheduled for elective coronary artery bypass grafting (CABG) were included. Patients were considered eligible for the study if they were less than 75 years old, had a hearing threshold ⱕ40 dB, had no history of neurologic or psychiatric disorders, had no insulin-dependent diabetes mellitus, and had a left ventricular ejection fraction ⬎40%. After inclusion, 8 patients were excluded for the following reasons: need for additional valve replacement (5 patients), ventricular fibrillation during the postoperative period (1 patient), postoperative stroke (1 patient), and miss-
KEY WORDS: auditory-evoked potentials, intensive care unit, sedation, cardiac surgery
From the *Critical Care Research Program, Division of Intensive Care, Department of Anaesthesiology and Intensive Care Medicine and ‡Department of Clinical Neurophysiology, Kuopio University Hospital, Kuopio, Finland; †Department of Intensive Care Medicine, Helsinki University, Central Hospital, Jorvi Hospital, Espoo, Finland; and §Department of Intensive Care Medicine, University Hospital Bern, Bern, Switzerland. Supported by an EU grant (EU-BIOMED 2, BMH4-97-2570) and by an EVO grant from Kuopio University Hospital. Presented in part at the ESICM 13th Annual Congress, Rome, Italy, October 1-4, 2000. Address reprint requests to Stephan M. Jakob, MD, PhD, Department of Intensive Care Medicine, University Hospital Bern, Freiburgstrasse, CH-3010 Bern, Switzerland. E-mail: [email protected] © 2004 Elsevier Inc. All rights reserved. 1053-0770/04/1805-0004$30.00/0 doi:10.1053/j.jvca.2004.07.008
Journal of Cardiothoracic and Vascular Anesthesia, Vol 18, No 5 (October), 2004: pp 559-562
559
560
MUSIALOWICZ ET AL
Table 1. Latencies of Na, Pa, and Nb Components and the Peak-to-Peak Amplitudes of NaPa and PaNb Components of MLAEPs (Median and Range)
Latency Na (ms) Latency Pa (ms) Latency Nb (ms) Amplitude NaPa (V) Amplitude PaNb (V)
Baseline
Preoperative (Ramsay 3)
Deep Sedation (Ramsay 6)
Moderate Sedation (Ramsay 4)
Postoperative (Ramsay 2)
20 (15-23) 31 (27-41) 44 (38-60) 1.7 (0.7-4.9) 0.8 (0.4-1.3)
20 (15-23) 34 (29-41) 49 (41-64)* 1.6 (0.1-2.9) 0.9 (0.4-3.3)
21 (14-24)* 40 (35-51)* 63 (48-80)* 0.9 (0.4-1.6)* 0.4 (0.1-0.9)
20 (14-44) 36 (29-45)* 61 (42-78)* 1.3 (0.8-3.9)† 0.8 (0.2-3.1)
20 (16-23) 35 (27-47)* 49 (37-71)* 1.8 (0.2-3.2) 0.8 (0.0-2.1)
Abbreviations: Baseline, on the day before surgery; preoperative, 1 hour after premedication and prior to surgery, Ramsay score 2; deep sedation, in the ICU, Ramsay score 6; moderate sedation, during maintenance of sedation in the ICU, Ramsay score 4; postoperative, on the first day after surgery, Ramsay score 2. *p ⬍ 0.05 versus baseline. †p ⬍ 0.05 versus deep sedation.
Hearing function was tested in all patients before baseline measurements. Two-channel MLAEP recording was performed by using silversilver-chloride electrodes placed at the vertex and over both mastoids. The grounding electrode was placed at the forehead. Electrode skin impedances were maintained below 5 kOhm. Binaural auditory stimulation was delivered at 70 dB above the hearing level using earphones, and the stimulus was repeated at a rate of 7.1 Hz. The duration of each consecutive click was 100 s. The signal was digitized at 2,400 Hz and band-pass filtered (0.5-1,000 Hz) by using the Datex-Ohmeda AEP/ EEG module included in the CS/3 monitor (Datex-Ohmeda Division, Instrumentarium Corporation, Helsinki, Finland). Two thousand stimuli were averaged with a 100-ms sweep length. Peak identification (Na, Pa, and Nb) and amplitude and latency measurements of all MLAEPs were performed as an off-line validation by a trained hospital physicist and neurophysiologist (auditory-evoked potentials score software, Institute of Information Technology, Tampere, Finland). Descriptive data on event-related potentials used with a standard “oddball” paradigm
and some data on spectral edge EEG frequencies from the same patient group have been published earlier.6,7 Results are expressed as median (range) or number of patients where adequate. Nonparametric analysis of variance for repeated measures (Friedman test) was used to detect significant changes in variables over time. Data at different time points were compared with the Wilcoxon signed rank test, and adjustments were made with the Bonferroni method for multiple testing. Pearson correlation was used to address the association between MLAEP latencies and Ramsay scores. A p value ⬍0.05 was considered significant. RESULTS
Nineteen men and 3 women with a mean age of 55 years (range: 41-75 years) were studied. In 15 patients, all components of the MLAEP recordings could be identified at all time points. At various time points, data of 19 to 22 patients were
Fig 1. Representative mid-latency auditory-evoked potentials (MLAEPs) from one individual. MLAEPs were recorded (from top to bottom) the day before surgery, one hour before surgery (after premedication), postoperatively during deep sedation (Ramsay 6), and during moderate sedation (Ramsay 4), and the day after surgery.
ASSESSMENT OF SEDATION AFTER CARDIAC SURGERY
561
Fig 2. Individual Nb latencies; 1: the day before surgery; 2: 1 hour before surgery (after premedication); 3: postoperatively during deep sedation (Ramsay 6); 4: during moderate sedation (Ramsay 4); and 5: the day after surgery.
available for analyses. Summarized MLAEP data are shown in Table 1, and representative traces from 1 individual in Figure 1. As compared with baseline values, Nb latency was increased after premedication (p ⬍ 0.05, Fig 2). At arrival in the ICU (deep sedation, Ramsay 6), the latencies of all the measured MLAEP components were increased (p ⬍ 0.05, Figs 2 and 3). During moderate sedation (Ramsay 4) and on the first postoperative day, Pa and Nb latencies were increased (p ⬍ 0.05, Figs 2 and 3). NaPa amplitude was decreased at Ramsay 6 as compared both with baseline and to Ramsay 4 (p ⬍ 0.05). Nb latencies after premedication (Ramsay 2) and during deep and moderate sedation correlated well with the respective Ramsay scores (r2: 0.53, p ⫽ 0.01). DISCUSSION
MLAEPs revealed clinically barely detectable alterations in the brain’s perception of auditory stimuli after premedication and at the day after surgery. Deep levels of sedation after surgery were associated with even more pronounced increases
in MLAEP Nb latencies. Graded increases in MLAEP latencies and decreases in amplitudes with increasing concentrations of anesthetic drugs have also been shown by others.3 SchulteTamburen et al8 compared 5 subjective sedation scoring systems with the Nb latency during long-term sedation in ICU patients. Although all scores correlated fairly well (r2: 0.6-0.7) with Nb latency, the performance was best with the Ramsay score. Increasing NaPa amplitudes and decreasing Nb latencies indicate cortical arousal in ICU patients during nursing care and physiotherapy.9 Because the exact source of the MLAEP components is not known, it is difficult to judge which one of the components (Na, Pa, Nb) is clinically most meaningful. As shown by Figures 2 and 3, the behavior of Pa and Nb latencies is similar. The dissociation of MLAEP amplitudes and latencies in the ability to distinguish between very deep (Ramsay 6) and deep level of sedation (Ramsay 4) in these patients deserves further investigation. It is important to realize that a Ramsay score of
Fig 3. Individual Pa latencies; 1: the day before surgery; 2: 1 hour before surgery (after premedication); 3: postoperatively during deep sedation (Ramsay 6); 4: during moderate sedation (Ramsay 4); and 5: the day after surgery.
562
MUSIALOWICZ ET AL
6 is a category bounded on only 1 side. Accordingly, this category may include a great variability in the depth of sedation. The increased latencies of Pa and Nb components on the day after surgery may be explained by the use of analgesic drugs and trace amounts of anesthetics and sedatives used during surgery. Alternatively, sequelae of cardiopulmonary bypass in the central nervous system also may have contributed. Unfortunately, the authors did not perform detailed neuropsychological evaluations to assess whether the observed neurophysiological abnormalities have a clinically measurable correlative. Attenuation of MLAEP amplitudes to the point of near unrecordability early after loss of consciousness and poor differentiation of clinical scoring systems to assess deep sedation may explain the widely scattered distribution of Nb latencies with increasing levels of sedation.8,9 An alternative, nonetheless objective measure of hypnosis is the BIS. Indeed, observational studies found a good correlation between the clinical sedation scores and BIS.10,11 However, other reports have shown no benefit over the standard subjective assessment
scores.12-14 It is obvious that measuring an active signal in response to a stimulus better compares with the clinical assessment of sedation than assessing background activity using processed EEG parameters. MLAEPs show a high intra- and interindividual stability in awake persons, and they have been shown to be affected profoundly by hypnotic drugs in a graded, reversible, and nonagent-specific manner (decrease in amplitudes and increase in latencies).15-20 Few trials have assessed sedation in the ICU using MLAEP, however,8,21 and the correlation with clinically derived sedation scores was rather mediocre.21 Nevertheless, the results suggest that MLAEP latencies can reflect subtle changes in auditory perception, whereas amplitudes seem to change with transition between deep levels of sedation. ACKNOWLEDGMENT This study was completed as a project titled “Improved Monitoring of Brain Function in Intensive Care and Surgery (IBIS),” a part of the Biomed-2 program of the European Union.
REFERENCES 1. Kress JP, Pohlman AS, O’Connor MF, et al: Daily interruption of sedative infusions in critically ill patients undergoing mechanical ventilation. N Engl J Med 342:1471-1477, 2000 2. Bruhn J, Bouillon TW, Radulescu L, et al: Correlation of approximate entropy, bispectral index, and spectral edge frequency 95 (SEF95) with clinical signs of “anesthetic depth” during coadministration of propofol and remifentanil. Anesthesiology 98:621-627, 2003 3. Thornton C: Evoked potentials in anesthesia. Eur J Anaesthesiol 8:89-107, 1991 4. Iselin-Chaves IA, El Moalem HE, Gan TJ, et al: Changes in the auditory-evoked potentials and the bispectral index following propofol or propofol and alfentanil. Anesthesiology 92:1300-1310, 2000 5. Gajraj RJ, Doi M, Mantzaridis H, et al: Analysis of the EEG bispectrum, auditory evoked potentials and the EEG power spectrum during repeated transitions from consciousness to unconsciousness. Br J Anaesth 80:46-52, 1998 6. Yppa¨rila¨ H, Karhu J, Westeren-Punnonen S, et al: Evidence of auditory processing during postoperative propofol sedation. Clin Neurophysiol 113:1357-1364, 2002 7. Yppa¨rila¨ H, Korhonen I, Westeren-Punnonen S, et al: Assessment of postoperative sedation level with spectral EEG parameters. Clin Neurophysiol 113:1633-1639, 2002 8. Schulte-Tamburen A, Scheier J, Briegel J, et al: Comparison of five sedation scoring systems by means of auditory-evoked potentials. Intensive Care Med 25:377-382, 1999 9. Rundshagen I, Schnabel K, Pothmann W, et al: Cortical arousal in critically ill patients: An evoked response study. Intensive Care Med 26:1312-1318, 2000 10. Simmons LE, Riker RR, Prato BS, et al: Assessing sedation during intensive care unit mechanical ventilation with the bispectral index and the sedation-agitation scale. Crit Care Med 27:1499-1504, 1999
11. Riker RR, Fraser GL, Simmons LE, et al: Validating the sedationagitation scale with the bispectral index and visual analog scale in adult ICU patients after cardiac surgery. Intensive Care Med 27:853-858, 2001 12. Ely EW, Truman B, Shintani A, et al: Monitoring sedation status over time in ICU patients: Reliability and validity of the Richmond Agitation-Sedation Scale (RASS). JAMA 289:2983-2991, 2003 13. Walder B, Suter PM, Romand JA: Evaluation of two processed EEG analyzers for assessment of sedation after coronary artery bypass grafting. Intensive Care Med 27:107-114, 2001 14. De Deyne C, Struys M, Decruyenaere J, et al: Use of continuous bispectral EEG monitoring to assess depth of sedation in ICU patients. Intensive Care Med 24:1294-1298, 1998 15. Thornton C, Heneghan CP, James MF, et al: Effects of halothane or enflurane with controlled ventilation on auditory-evoked potentials. Br J Anaesth 56:315-323, 1984 16. Thornton C, Heneghan CP, Navaratnarajah M, et al: Effect of etomidate on the auditory-evoked response in man. Br J Anaesth 57:554-561, 1985 17. Thornton C, Heneghan CP, Navaratnarajah M, et al: Selective effect of althesin on the auditory-evoked response in man. Br J Anaesth 58:422-427, 1986 18. Thornton C, Konieczko KM, Knight AB, et al: Effect of propofol on the auditory-evoked response and oesophageal contractility. Br J Anaesth 63:411-417, 1989 19. Heneghan CP, Thornton C, Navaratnarajah M, et al: Effect of isoflurane on the auditory-evoked response in man. Br J Anaesth 59:277-282, 1987 20. Schwender D, Conzen P, Klasing S, et al: The effects of anesthesia with increasing end-expiratory concentrations of sevoflurane on midlatency auditory-evoked potentials. Anesth Analg 81:817-822, 1995 21. Ceponiene R. Event-related potential (ERP) indices of central auditory development in healthy children and in children with oral clefts (dissertation). Helsinki, Finland, University of Helsinki, 2001
ⴝ 0.03) and further to 63 ms (48-80 ms) during deep sedation after surgery (P < 0.01). Although a decreasing clinical level of sedation after rewarming was not associated with a significant change in Nb latency (61 ms [42-78 ms]), the MLAEP NaPa amplitude increased from 0.9 V (0.4-1.6 V) to 1.3 V (0.8-3.9 V; p ⴝ 0.01). Nb latency remained increased the day after surgery (49 ms [37-71 ms]) as compared with baseline (p < 0.01). Conclusions: MLAEP latencies can reflect subtle changes in auditory perception, while amplitudes seem to change with transition between deep levels of sedation. © 2004 Elsevier Inc. All rights reserved.
I
ing data because of technical problems with the data collecting equipment (1 patient). The patients received diazepam, 0.2 mg/kg, orally the night before the operation and 2 hours before surgery. For induction of anesthesia, midazolam, 0.1 mg/kg, alfentanil, 100 g/kg, and pancuronium, 0.1 mg/kg, were used. Anesthesia was maintained with continuous infusions of propofol, 3 mg/kg/h, and alfentanil, 50 g/kg/h. The patients underwent CABG with cardiopulmonary bypass with nonpulsatile pump flow under moderate hypothermia at 34°C. At the end of the operation, alfentanil infusion was discontinued and propofol infusion was continued at a rate of 2 mg/kg/h during transport of the patient to the ICU. A physician and a nurse assessed the Ramsay score immediately before and after measurements of the MLAEPs and consensus was recorded. MLAEP measurements were performed at 5 time points: (1) on the day before surgery (baseline), (2) 1 hour before surgery after premedication, (3) after arrival in the ICU (propofol sedation adjusted to Ramsay score 6 [no response to glabellar tap or loud auditory stimulus]), (4) during maintenance of sedation in the ICU (propofol sedation adjusted to Ramsay score 4 [a brisk response to glabellar tap or loud auditory stimulus]), and (5) on the day after surgery, when the patient was extubated and fully awake. The investigators assessing the Ramsay score were not aware of the MLAEP characteristics.
NAPPROPRIATE SEDATION in cardiac surgery patients may cause cardiorespiratory depression, prolonged recovery times and intensive care unit (ICU) stays, and increased costs.1 Although depth of anesthesia can be monitored by recording an electroencephalogram (EEG) or EEG-derived parameters (spectral edge frequency, approximate entropy, bispectral index [BIS]),2 these parameters have not been proven useful for the monitoring of lighter levels of sedation; for instance, in patients recovering from surgery and anesthesia. In daily practice, the level of sedation is assessed by using clinical scoring systems such as the Ramsay score, which classifies a patient’s response to a command or painful stimulus. However, the Ramsay score has limited value in the prediction of the level of sedation because of many confounding factors, such as motor deficit and inter-rater variation in interpretation. Auditory-evoked potentials measure the output of the central nervous system in response to a controlled input. Most anesthetics depress midlatency auditory-evoked potentials (MLAEPs) in a dose-dependent fashion,3 and the changes are independent of the presence of opioids.4 For these reasons, it has been suggested that MLAEPs measure the hypnotic component of anesthesia and could serve as an effective discriminator between different states of consciousness.5 The purpose of this study was to assess whether graded changes in sedation induced by pre-and postoperative sedative drugs can be detected by any of the MLAEP components. METHODS The local ethics committee approved the study protocol. All patients gave written informed consent. Originally, 30 patients scheduled for elective coronary artery bypass grafting (CABG) were included. Patients were considered eligible for the study if they were less than 75 years old, had a hearing threshold ⱕ40 dB, had no history of neurologic or psychiatric disorders, had no insulin-dependent diabetes mellitus, and had a left ventricular ejection fraction ⬎40%. After inclusion, 8 patients were excluded for the following reasons: need for additional valve replacement (5 patients), ventricular fibrillation during the postoperative period (1 patient), postoperative stroke (1 patient), and miss-
KEY WORDS: auditory-evoked potentials, intensive care unit, sedation, cardiac surgery
From the *Critical Care Research Program, Division of Intensive Care, Department of Anaesthesiology and Intensive Care Medicine and ‡Department of Clinical Neurophysiology, Kuopio University Hospital, Kuopio, Finland; †Department of Intensive Care Medicine, Helsinki University, Central Hospital, Jorvi Hospital, Espoo, Finland; and §Department of Intensive Care Medicine, University Hospital Bern, Bern, Switzerland. Supported by an EU grant (EU-BIOMED 2, BMH4-97-2570) and by an EVO grant from Kuopio University Hospital. Presented in part at the ESICM 13th Annual Congress, Rome, Italy, October 1-4, 2000. Address reprint requests to Stephan M. Jakob, MD, PhD, Department of Intensive Care Medicine, University Hospital Bern, Freiburgstrasse, CH-3010 Bern, Switzerland. E-mail: [email protected] © 2004 Elsevier Inc. All rights reserved. 1053-0770/04/1805-0004$30.00/0 doi:10.1053/j.jvca.2004.07.008
Journal of Cardiothoracic and Vascular Anesthesia, Vol 18, No 5 (October), 2004: pp 559-562
559
560
MUSIALOWICZ ET AL
Table 1. Latencies of Na, Pa, and Nb Components and the Peak-to-Peak Amplitudes of NaPa and PaNb Components of MLAEPs (Median and Range)
Latency Na (ms) Latency Pa (ms) Latency Nb (ms) Amplitude NaPa (V) Amplitude PaNb (V)
Baseline
Preoperative (Ramsay 3)
Deep Sedation (Ramsay 6)
Moderate Sedation (Ramsay 4)
Postoperative (Ramsay 2)
20 (15-23) 31 (27-41) 44 (38-60) 1.7 (0.7-4.9) 0.8 (0.4-1.3)
20 (15-23) 34 (29-41) 49 (41-64)* 1.6 (0.1-2.9) 0.9 (0.4-3.3)
21 (14-24)* 40 (35-51)* 63 (48-80)* 0.9 (0.4-1.6)* 0.4 (0.1-0.9)
20 (14-44) 36 (29-45)* 61 (42-78)* 1.3 (0.8-3.9)† 0.8 (0.2-3.1)
20 (16-23) 35 (27-47)* 49 (37-71)* 1.8 (0.2-3.2) 0.8 (0.0-2.1)
Abbreviations: Baseline, on the day before surgery; preoperative, 1 hour after premedication and prior to surgery, Ramsay score 2; deep sedation, in the ICU, Ramsay score 6; moderate sedation, during maintenance of sedation in the ICU, Ramsay score 4; postoperative, on the first day after surgery, Ramsay score 2. *p ⬍ 0.05 versus baseline. †p ⬍ 0.05 versus deep sedation.
Hearing function was tested in all patients before baseline measurements. Two-channel MLAEP recording was performed by using silversilver-chloride electrodes placed at the vertex and over both mastoids. The grounding electrode was placed at the forehead. Electrode skin impedances were maintained below 5 kOhm. Binaural auditory stimulation was delivered at 70 dB above the hearing level using earphones, and the stimulus was repeated at a rate of 7.1 Hz. The duration of each consecutive click was 100 s. The signal was digitized at 2,400 Hz and band-pass filtered (0.5-1,000 Hz) by using the Datex-Ohmeda AEP/ EEG module included in the CS/3 monitor (Datex-Ohmeda Division, Instrumentarium Corporation, Helsinki, Finland). Two thousand stimuli were averaged with a 100-ms sweep length. Peak identification (Na, Pa, and Nb) and amplitude and latency measurements of all MLAEPs were performed as an off-line validation by a trained hospital physicist and neurophysiologist (auditory-evoked potentials score software, Institute of Information Technology, Tampere, Finland). Descriptive data on event-related potentials used with a standard “oddball” paradigm
and some data on spectral edge EEG frequencies from the same patient group have been published earlier.6,7 Results are expressed as median (range) or number of patients where adequate. Nonparametric analysis of variance for repeated measures (Friedman test) was used to detect significant changes in variables over time. Data at different time points were compared with the Wilcoxon signed rank test, and adjustments were made with the Bonferroni method for multiple testing. Pearson correlation was used to address the association between MLAEP latencies and Ramsay scores. A p value ⬍0.05 was considered significant. RESULTS
Nineteen men and 3 women with a mean age of 55 years (range: 41-75 years) were studied. In 15 patients, all components of the MLAEP recordings could be identified at all time points. At various time points, data of 19 to 22 patients were
Fig 1. Representative mid-latency auditory-evoked potentials (MLAEPs) from one individual. MLAEPs were recorded (from top to bottom) the day before surgery, one hour before surgery (after premedication), postoperatively during deep sedation (Ramsay 6), and during moderate sedation (Ramsay 4), and the day after surgery.
ASSESSMENT OF SEDATION AFTER CARDIAC SURGERY
561
Fig 2. Individual Nb latencies; 1: the day before surgery; 2: 1 hour before surgery (after premedication); 3: postoperatively during deep sedation (Ramsay 6); 4: during moderate sedation (Ramsay 4); and 5: the day after surgery.
available for analyses. Summarized MLAEP data are shown in Table 1, and representative traces from 1 individual in Figure 1. As compared with baseline values, Nb latency was increased after premedication (p ⬍ 0.05, Fig 2). At arrival in the ICU (deep sedation, Ramsay 6), the latencies of all the measured MLAEP components were increased (p ⬍ 0.05, Figs 2 and 3). During moderate sedation (Ramsay 4) and on the first postoperative day, Pa and Nb latencies were increased (p ⬍ 0.05, Figs 2 and 3). NaPa amplitude was decreased at Ramsay 6 as compared both with baseline and to Ramsay 4 (p ⬍ 0.05). Nb latencies after premedication (Ramsay 2) and during deep and moderate sedation correlated well with the respective Ramsay scores (r2: 0.53, p ⫽ 0.01). DISCUSSION
MLAEPs revealed clinically barely detectable alterations in the brain’s perception of auditory stimuli after premedication and at the day after surgery. Deep levels of sedation after surgery were associated with even more pronounced increases
in MLAEP Nb latencies. Graded increases in MLAEP latencies and decreases in amplitudes with increasing concentrations of anesthetic drugs have also been shown by others.3 SchulteTamburen et al8 compared 5 subjective sedation scoring systems with the Nb latency during long-term sedation in ICU patients. Although all scores correlated fairly well (r2: 0.6-0.7) with Nb latency, the performance was best with the Ramsay score. Increasing NaPa amplitudes and decreasing Nb latencies indicate cortical arousal in ICU patients during nursing care and physiotherapy.9 Because the exact source of the MLAEP components is not known, it is difficult to judge which one of the components (Na, Pa, Nb) is clinically most meaningful. As shown by Figures 2 and 3, the behavior of Pa and Nb latencies is similar. The dissociation of MLAEP amplitudes and latencies in the ability to distinguish between very deep (Ramsay 6) and deep level of sedation (Ramsay 4) in these patients deserves further investigation. It is important to realize that a Ramsay score of
Fig 3. Individual Pa latencies; 1: the day before surgery; 2: 1 hour before surgery (after premedication); 3: postoperatively during deep sedation (Ramsay 6); 4: during moderate sedation (Ramsay 4); and 5: the day after surgery.
562
MUSIALOWICZ ET AL
6 is a category bounded on only 1 side. Accordingly, this category may include a great variability in the depth of sedation. The increased latencies of Pa and Nb components on the day after surgery may be explained by the use of analgesic drugs and trace amounts of anesthetics and sedatives used during surgery. Alternatively, sequelae of cardiopulmonary bypass in the central nervous system also may have contributed. Unfortunately, the authors did not perform detailed neuropsychological evaluations to assess whether the observed neurophysiological abnormalities have a clinically measurable correlative. Attenuation of MLAEP amplitudes to the point of near unrecordability early after loss of consciousness and poor differentiation of clinical scoring systems to assess deep sedation may explain the widely scattered distribution of Nb latencies with increasing levels of sedation.8,9 An alternative, nonetheless objective measure of hypnosis is the BIS. Indeed, observational studies found a good correlation between the clinical sedation scores and BIS.10,11 However, other reports have shown no benefit over the standard subjective assessment
scores.12-14 It is obvious that measuring an active signal in response to a stimulus better compares with the clinical assessment of sedation than assessing background activity using processed EEG parameters. MLAEPs show a high intra- and interindividual stability in awake persons, and they have been shown to be affected profoundly by hypnotic drugs in a graded, reversible, and nonagent-specific manner (decrease in amplitudes and increase in latencies).15-20 Few trials have assessed sedation in the ICU using MLAEP, however,8,21 and the correlation with clinically derived sedation scores was rather mediocre.21 Nevertheless, the results suggest that MLAEP latencies can reflect subtle changes in auditory perception, whereas amplitudes seem to change with transition between deep levels of sedation. ACKNOWLEDGMENT This study was completed as a project titled “Improved Monitoring of Brain Function in Intensive Care and Surgery (IBIS),” a part of the Biomed-2 program of the European Union.
REFERENCES 1. Kress JP, Pohlman AS, O’Connor MF, et al: Daily interruption of sedative infusions in critically ill patients undergoing mechanical ventilation. N Engl J Med 342:1471-1477, 2000 2. Bruhn J, Bouillon TW, Radulescu L, et al: Correlation of approximate entropy, bispectral index, and spectral edge frequency 95 (SEF95) with clinical signs of “anesthetic depth” during coadministration of propofol and remifentanil. Anesthesiology 98:621-627, 2003 3. Thornton C: Evoked potentials in anesthesia. Eur J Anaesthesiol 8:89-107, 1991 4. Iselin-Chaves IA, El Moalem HE, Gan TJ, et al: Changes in the auditory-evoked potentials and the bispectral index following propofol or propofol and alfentanil. Anesthesiology 92:1300-1310, 2000 5. Gajraj RJ, Doi M, Mantzaridis H, et al: Analysis of the EEG bispectrum, auditory evoked potentials and the EEG power spectrum during repeated transitions from consciousness to unconsciousness. Br J Anaesth 80:46-52, 1998 6. Yppa¨rila¨ H, Karhu J, Westeren-Punnonen S, et al: Evidence of auditory processing during postoperative propofol sedation. Clin Neurophysiol 113:1357-1364, 2002 7. Yppa¨rila¨ H, Korhonen I, Westeren-Punnonen S, et al: Assessment of postoperative sedation level with spectral EEG parameters. Clin Neurophysiol 113:1633-1639, 2002 8. Schulte-Tamburen A, Scheier J, Briegel J, et al: Comparison of five sedation scoring systems by means of auditory-evoked potentials. Intensive Care Med 25:377-382, 1999 9. Rundshagen I, Schnabel K, Pothmann W, et al: Cortical arousal in critically ill patients: An evoked response study. Intensive Care Med 26:1312-1318, 2000 10. Simmons LE, Riker RR, Prato BS, et al: Assessing sedation during intensive care unit mechanical ventilation with the bispectral index and the sedation-agitation scale. Crit Care Med 27:1499-1504, 1999
11. Riker RR, Fraser GL, Simmons LE, et al: Validating the sedationagitation scale with the bispectral index and visual analog scale in adult ICU patients after cardiac surgery. Intensive Care Med 27:853-858, 2001 12. Ely EW, Truman B, Shintani A, et al: Monitoring sedation status over time in ICU patients: Reliability and validity of the Richmond Agitation-Sedation Scale (RASS). JAMA 289:2983-2991, 2003 13. Walder B, Suter PM, Romand JA: Evaluation of two processed EEG analyzers for assessment of sedation after coronary artery bypass grafting. Intensive Care Med 27:107-114, 2001 14. De Deyne C, Struys M, Decruyenaere J, et al: Use of continuous bispectral EEG monitoring to assess depth of sedation in ICU patients. Intensive Care Med 24:1294-1298, 1998 15. Thornton C, Heneghan CP, James MF, et al: Effects of halothane or enflurane with controlled ventilation on auditory-evoked potentials. Br J Anaesth 56:315-323, 1984 16. Thornton C, Heneghan CP, Navaratnarajah M, et al: Effect of etomidate on the auditory-evoked response in man. Br J Anaesth 57:554-561, 1985 17. Thornton C, Heneghan CP, Navaratnarajah M, et al: Selective effect of althesin on the auditory-evoked response in man. Br J Anaesth 58:422-427, 1986 18. Thornton C, Konieczko KM, Knight AB, et al: Effect of propofol on the auditory-evoked response and oesophageal contractility. Br J Anaesth 63:411-417, 1989 19. Heneghan CP, Thornton C, Navaratnarajah M, et al: Effect of isoflurane on the auditory-evoked response in man. Br J Anaesth 59:277-282, 1987 20. Schwender D, Conzen P, Klasing S, et al: The effects of anesthesia with increasing end-expiratory concentrations of sevoflurane on midlatency auditory-evoked potentials. Anesth Analg 81:817-822, 1995 21. Ceponiene R. Event-related potential (ERP) indices of central auditory development in healthy children and in children with oral clefts (dissertation). Helsinki, Finland, University of Helsinki, 2001