Anesth Analg 2003;97:132-138
© 2003 International Anesthesia Research Society
TECHNOLOGY, COMPUTING, AND SIMULATION
Implicit Memory Varies as a Function of Hypnotic Electroencephalogram Stage in Surgical Patients
Sinikka Münte, MD*,
Thomas F. Münte, MD
,
Jörg Grotkamp, MD
,
Gertrud Haeseler, MD*,
Konstantinos Raymondos, MD*,
Siegfried Piepenbrock, MD*, and
Gabriele Kraus, MD
*Department of Anaesthesiology, Medical School of Hannover, Germany;
Department of Neuropsychology, Otto-von-Guericke University Magdeburg, Germany; and Departments of
Internal Medicine and
Anaesthesiology, Siloah Hospital, Hannover, Germany
Address correspondence and reprint requests to Sinikka Münte, MD, Department of Anesthesiology, Medical School of Hannover, Carl-Neuberg-Str. 1, 30625 Hannover, Germany. Address e-mail to Muente.Sinikka{at}MH-Hannover.de
 |
Abstract
|
|---|
Previous studies have observed a correlation of implicit memory with certain electroencephalogram (EEG) measures during anesthesia. Here, we tested the relationship between hypnotic depth determined by computer system (NarcotrendTM) and implicit memory in anesthetized patients, assessed by a postoperative reading speed test. Thirty-two patients undergoing laparoscopic herniotomy and 30 age-matched volunteer controls were included the study. All patients received IV midazolam 2–3 mg followed by an induction dose of propofol and remifentanil. The anesthesia was maintained with propofol and remifentanil infusions and cisatracurium. Each patient was exposed to 2 of 4 stories, repeated 6 times. The first story was presented during light to moderate hypnotic EEG stages, and the second story was presented during deep hypnosis. Presentation of stories was balanced between patients and hypnotic stages. The controls listened to the two stories without receiving anesthesia. The reading speed for the previously presented stories and two new stories was measured approximately 7 h later with a computer program. No signs of inadequate anesthesia were observed, and no explicit memories of intraoperative events were revealed by a structured interview. No change of reading speed was observed for words presented during deep hypnotic stages. In contrast, an increased reading speed of 20 ms per word was found for content words (i.e., nouns, verbs, and adjectives), but not for function words (conjunctions, prepositions, and so on), presented during light to moderate hypnotic stages. Increased reading speed for semantically rich content words indicates that anesthetized patients are able to process acoustic information during light and moderate, but not deep, hypnosis.
IMPLICATIONS: In this study, implicit memory was observed during general anesthesia at light to moderate, but not deep, hypnotic stages. Hypnotic stages were determined by a commercial electroencephalogram device, and implicit memory was measured by using a postoperative reading speed task. During lighter phases of anesthesia, patients should be protected against acoustic information that could negatively influence their postoperative outcome.
|
Introduction
|
|---|
Conscious (explicit) memory usually is abolished by small doses of hypnotics, whereas unconscious (implicit) memory seems to be more resistant to anesthetic effects (1,2). Thus, in the absence of explicit memories, implicit memories may occur during anesthesia (3). Many questions related to implicit memory during anesthesia remain unanswered. For example, it is not entirely clear how different anesthetics and their doses, surgical stimuli, and depth of hypnosis influence the retention of material presented during anesthesia. Moreover, the possible psychological consequences of implicit memory of intraoperative events are unknown. Therefore, investigators try to determine under which circumstances implicit memory might occur during general anesthesia. The aim of this study was to investigate the relationship between implicit memory and hypnotic depth of anesthesia as determined by one processed electroencephalogram (EEG) device.
Direct tests of memory, such as recognition and free recall, can be used to reveal explicit memories of intraoperative events. In contrast, to investigate implicit memory, indirect test methods have to be used (4). Many indirect test methods rely on "priming" effects, which are defined by faster and/or more accurate performance for stimuli (e.g., words) that have been presented before (5). As an example, consider the reading speed test used in this study (6). We presented short stories during anesthesia. After surgery, the reading speed of previously presented stories, as well as of identically constructed new stories, was measured by using a computer program (Fig. 1). If patients who have no recollection of the previously presented stories show faster reading times for the stories heard during anesthesia compared with the new stories, this indicates that the test material has been processed during anesthesia and has left some memory trace.
View larger version (10K):
[in this window]
[in a new window]
|
Figure 1. The reading-speed task: a participant reads the sentence on the video monitor one word at a time by pressing the space bar on a computer keyboard.
|
|
The EEG has been intensively studied to develop a monitor of anesthetic depth. EEG measurements show a better correlation with sedation, loss of consciousness, and memory performance than with somatic and autonomic responses (7). Therefore, it is well accepted that the processed EEG primarily indicates the hypnotic component of anesthesia (8). A few studies suggest that memory performance declines with increasing hypnotic depth determined by processed EEG variables, such as the bispectral index (BIS) or midlatency auditory evoked responses (9–11). However, up to now, only a very limited range of studies has explored the correlation between EEG variables and implicit memory during anesthesia. In this study, a commercially available EEG device, automatically classifying hypnotic stages on the basis of a system originally developed for visual EEG analysis, was used (12). We hypothesized that memory performance should decline with increasing hypnotic depth, as indicated by the EEG stages.
Postoperative implicit memory of stories played during light to moderate or deep hypnotic stages was compared with memory of previously unheard stories by using the reading-speed task. This test has been used to detect implicit memory in patients with amnesia (13) and in anesthetized patients in experiments of our own group (14,15). The study protocol was restricted to one standardized anesthesia technique (propofol and remifentanil infusions), a single surgical procedure (laparoscopic herniotomy), and a defined study-test interval to reduce variables that might affect implicit memory. An age-matched control group was included in the study because a previous study of our group suggested that the reading-speed test may be sensitive only in relatively young patients (16).
|
Methods
|
|---|
With approval from the IRB and the informed consent of each participant, 32 patients undergoing laparoscopic inguinal herniotomy and 30 volunteers acting as controls were included in the study between April and October 2000 (Table 1). Data from 11 additional patients could not be analyzed because of postoperative vomiting (n = 1), refusal of the postoperative test (n = 1), or other protocol violations (n = 9).
Only patients with ASA physical status I or II and an age between 18 and 50 yr were studied. Non-German speakers were excluded, as were those with hearing impairment, head trauma, known psychiatric or memory disorder, or psychoactive medication. The patients were informed the day before surgery that short stories would be played to them during the surgery and that they would undergo a short interview and a psychological test several hours after surgery. They also were informed that an EEG device would be used to monitor the depth of anesthesia. An age-matched control group consisted of volunteers, who participated in the study in the same format as the surgical group without receiving anesthesia.
A standardized anesthesia protocol was used. No oral premedication was given. All patients received IV midazolam 2–3 mg followed by an induction dose of remifentanil 1 µg/kg and propofol 2.5 mg/kg. After the patients lost consciousness, they were ventilated with 100% oxygen, and 0.15 mg/kg of cisatracurium was given to facilitate tracheal intubation. Immediately after induction, continuous infusion with remifentanil 3 µg · kg-1 · min-1 and propofol 2.5 mg · kg-1 · h-1 was started, and patients were ventilated with an air/oxygen 30% mixture. Propofol (0.5 mg/kg) and remifentanil (0.5 µg/kg) boluses were administered, and infusions were adjusted to achieve light to moderately deep hypnotic stages (see below) before and deep hypnotic stages during the operation. All patients received 6 mg of IV piritramide, a long-acting (4–6 h) opioid agonist, at the end of the operation. Cortical activity was continuously monitored by using the commercially available NarcotrendTM EEG monitor (see below and Appendix 1). Patients’ heart rate, noninvasive blood pressure, and blood oxygen saturation were also monitored.
The depth of hypnosis during anesthesia was determined by using the automatic EEG classification system Narcotrend (http://www.narcotrend.de). This device mimics a visual classification that was first introduced by Loomis et al. (17) and was further elaborated by Kugler (18) and Doenicke et al. (19) (Table 2). The automatic classification of 14 EEG stages ranging from A (fully awake) to F1 (electrical silence) during anesthesia corresponds with visual classification (20) and is useful for clinical settings (21,22). A study that directly compared Narcotrend with BIS showed that BIS values between 100 and 85 (representing awake patients) corresponded with Narcotrend stage A or B and that BIS values between 65 and 40 (representing general anesthesia) corresponded with Narcotrend stages D or E (22).
View this table:
[in this window]
[in a new window]
|
Table 2. Sleep stages by Loomis et al. (17), Kugler (18), and Doenicke et al. (19), Corresponding NarcotrendTM Electroencephalogram Stages, and Depth of Hypnosis
|
|
During anesthesia, each patient was exposed to two of four short stories, each repeated six times (exposure phase). The duration of each story was approximately 2 min 30 s (the length of each story was 199 words: 78 content words and 121 function words). The first story was presented after the induction of anesthesia and before surgery during light to moderate anesthetic depth, corresponding to Narcotrend stages C1 to D2, and the second story was presented during deep anesthesia during surgery, corresponding to stages E0/E1. The controls listened to two stories without receiving anesthesia. Stories were counterbalanced across subjects and anesthetic stages.
In the test phase, all study and control participants were tested 4 to 12 h after the initial stimulus presentation. The implicit memory for the stories was assessed with the reading speed measurement (see below). Explicit memory for intraoperative events and for the stories was assessed by using a structured interview and free recall. The patients were asked about the last thing they remembered before going to sleep, the first thing they remembered when they woke up, and anything in between. Patients were also asked whether they could remember hearing one or two of the stories during anesthesia.
We used the same four stories as in our previous experiments (14–16). Four audiotapes were prepared, each containing six instances of one story read by a male speaker. The stories had identical counts of content words (i.e., nouns, verbs, adjectives, and most adverbs) and function words (prepositions, conjunctions, articles, and some adverbs) and a similar length and structure of sentences.
The implicit memory for the stories was assessed with a reading-speed measurement. We used a text presentation mode called "the moving window," because it has been proven to produce reliable reading time data. In this self-paced reading task, participants have to the push the space bar of a computer keyboard to initiate the presentation of each successive word on a video screen (Fig. 1). Just et al. (6), who primarily used the paradigm for psycholinguistic research, found the reading-time data to be comparable to the more natural eye-fixation data, which are obtained by monitoring the eye fixations of participants during natural reading by infrared or electrophysiological measurements (4). To familiarize the participants with the reading-speed task, they were first given a short, previously not presented practice story to read. Thereafter, the two previously heard and two additional new stories were presented in counterbalanced order on a video monitor, with one word of each sentence visible at a time. The participants were asked to read the four short stories aloud as quickly as they could and were told that they had to answer three questions at the end of each story, which were presented on the computer screen as well. The order of presentation of the four stories was counterbalanced across subjects and anesthetic stages.
The reading time for each word was recorded in a computer file. Median reading times were determined for function words and content words of a story for each subject separately. The statistical comparison was performed in two steps. First, an analysis of variance was conducted on the patient data only with EEG-derived anesthetic depth (light/moderate versus deep EEG stages), repetition (previously presented versus new stories), and word type (function versus content words) as within-subjects factors. A larger reading advantage for previously heard stories was expected for those materials that had been presented during light to moderate anesthesia. With regard to function and content words, it was predicted that any decreases in reading speed due to anesthetic effects should be more pronounced for the content words, which carry the bulk of the semantic information. Second, an additional analysis of variance was performed comparing the data of the patients with those of the control subjects.
|
Results
|
|---|
The results of the reading-speed test are presented in Table 3. No priming of words presented during deep hypnosis (EEG stages E0 and E1) was observed. This was reflected statistically by a nonsignificant main effect of repetition (F1,31 = 0.02; P = 0.8) in an analysis of variance confined to the stories played during deep anesthesia. In contrast, increased reading speed was found for content words presented before surgery during light (C1 and C2) to moderate (D0, D1, and D2) hypnosis, but not for function words, which yielded a repetition x word type interaction (F1,31 = 5.42; P < 0.03). The increased reading speed for content words was examined in more detail in Figure 2: the reading speed advantage for primed content words was 19 ms (range, -38 to 60 ms; SD, 29 ms) during light anesthesia and 0 ms (range, -69 to 41 ms) during deep anesthesia (t31 = 2.42; P = 0.022).
View this table:
[in this window]
[in a new window]
|
Table 3. Mean Reading Speed for Content Words (Nouns, Verbs, and Adjectives) and Function Words (Prepositions, Conjunctions, and Articles)
|
|
View larger version (16K):
[in this window]
[in a new window]
|
Figure 2. The reading-speed advantage for all subjects: primed content words (milliseconds per word) during light to moderate versus deep anesthesia.
|
|
No significant differences between primed and unprimed words were found, if reading times were collapsed over all words (F2,62 = 1.08; P = 0.35). As expected from previous results, reading time for function words was shorter by approximately 100 ms in patients (F1,31 = 145; P < 0.0001; Table 3) and by approximately 75 ms in controls (F1,29 = 53; P < 0.0001; Table 3). The age-matched control group showed a significant priming effect of previously presented stories (F1,29 = 6.75; P < 0.02).
No signs of inadequate anesthesia were observed, and no explicit memories of intraoperative events were reported. None of the patients could remember any dreams or hearing the stories during anesthesia.
|
Discussion
|
|---|
In this study, we found implicit memory, evidenced by increased postoperative reading speed, for content words (verbs, nouns, and adjectives) that had been presented during light to moderate (C and D), but not deep (E0 and E1), hypnotic stages induced by propofol-remifentanil anesthesia. Thus, the automatic EEG classification system used in this study may distinguish between hypnotic stages that allow implicit memory processing and those that do not.
Our data are in line with those few previous studies of anesthetized patients that have used EEG-derived variables of hypnotic depth in conjunction with implicit memory tasks. For example, in the study of Lubke et al. (9), implicit memory varied as a function of hypnotic stage (determined by the BIS) (9). Implicit memory was still present during anesthesia at levels considered to be adequate by BIS monitoring, i.e., values less than 60. No implicit memory was found for words presented during anesthetic phases with a BIS less than 40.
Schwender et al. (10) used the midlatency components of the auditory evoked potentials (AEP) to determine anesthetic depth (10). Midlatency AEPs are suppressed by general anesthetics in a dose-dependent manner. In Schwender et al.’s study, only those patients who had preserved midlatency components showed implicit memory for the story of "Robinson Crusoe," which had been played during flunitrazepam-fentanyl, isoflurane-fentanyl, or propofol-fentanyl anesthesia. In a similar study, Ghoneim et al. (11) also found a relationship between implicit and explicit memory and preserved midlatency AEPs during anesthesia maintained by fentanyl boluses and N2O (11). Although different anesthetic regimens were used in these studies, the results are in agreement with our observation that priming occurred only in relatively lightly anesthetized patients and was absent in deep hypnosis. Apparently patients are able to process acoustic information to some extent after they have lost consciousness and seem to be, according to clinical observation, adequately anesthetized.
The device used in this study, similar to the BIS monitor, uses a multivariate technique to compress the complex EEG information. Rather than providing numbers, the device emulates the classic classification system of Loomis et al. (17), further elaborated by Kugler (18) and Doenicke et al. (23). Few studies are available that have tested the validity of the classification of the sleeping stages used in the Narcotrend in comparison to expert human observers or its stability for different anesthetic regimens (19,20,23–25), and one study shows a correlation between the BIS monitor and Narcotrend (22). Nevertheless, the EEG-derived hypnotic stages proved to reliably distinguish between phases of anesthesia during which patients were susceptible to implicit memory and those phases during which they were not.
Although the overall increase in reading speed (20 ms per content word) may seem comparatively small, we argue that it might be of clinical significance, because of the characteristics of the specific task that was used. To produce an implicit memory effect, i.e., a reading time advantage, the stories must have been processed at what might be called a conceptual level rather than merely a perceptual level, as has been claimed for other implicit memory tests (5,26). In other words, the priming effect emerges if the meaning of the whole story, or, in psycholinguistic terms, the discourse-level message, has been processed by the patient. If, however, stimuli can be processed for their meaning by the patients, they might also do so for other nonexperimental stimuli occurring during surgery.
An increasing literature reports on new applications for EEG monitoring during anesthesia. A smaller consumption of anesthetics, faster recovery, and smaller costs of anesthesia have been stressed as advantages of EEG monitoring (27–29). However, the decreased costs may be achieved at the expense of more postoperative implicit memory, because this study and a few previous reports suggest that implicit memory may occur during EEG stages that indicate light general anesthesia. The presence of implicit memory during anesthesia and its possible influence on postoperative outcome is not well understood. However, if anesthetized patients are able to process acoustic information at a conceptual/meaning level, as occurred in this study, they may have (subconscious) experiences of other intraoperative events as well. There is some anecdotal evidence outside of an experimental context to support the view that these experiences might influence patients’ well-being and postoperative course (30). Clearly, more research is needed to establish a link between postoperative outcome and implicit memory during surgery.
|
Appendix 1
|
|---|
Using the NarcotrendTM system, the electroencephalogram (EEG) is recorded with two bipolar montages. Pregelled, self-adhesive EEG electrodes are placed on the forehead with a distance of at least 10 cm. Biosignals are amplified with a bandpass of 0.5 to 64 Hz and digitized with 128 samples per second. Power spectra are determined for overlapping 20-s epochs (onset delay, 5 s) by using the fast Fourier transform. The hypnotic stage is thus updated every 5 s. Several variables, such as absolute and relative power, in the different frequency bands are determined from the power spectra and fed into a multivariate classification algorithm. In addition, for the F stages that are characterized by burst suppression patterns, special algorithms are used, because burst suppression patterns cannot be classified with spectral analysis alone. For every 5-s period, this analysis supplies a single EEG stage and thus simplifies the interpretation of EEG patterns. The Narcotrend distinguishes 14 EEG stages, ranging from A (fully awake) to F1 (electrical silence), during anesthesia. EEG stage A is characterized by 
activity as seen in normal awake adults. The EEG stages B0 to B2, which correspond to a very light hypnosis, consist of ß activity, 
activity, or both, with rather small amplitudes. From stage C0 to C2, which indicate light hypnosis, an increasing amount of activity can be seen. From stage D0 to D2, the amount of 
activity becomes greater. The EEG stages E0 and E1 represent deep hypnosis. In EEG stage E0, activity is nearly continuously present. EEG stage E1 is characterized by very slow waves and the beginning transition to a burst suppression pattern. In coma stages, F0 and F1 burst suppression patterns appear (F0), and the EEG becomes isoelectric (F1).
|
Acknowledgments
|
|---|
Supported by the Deutsche Forschungsgemeinschaft, Bonn, Germany (TFM).
|
Footnotes
|
|---|
Presented in part at the Fifth International Conference on Memory, Awareness and Consciousness, New York, NY, June 1–3, 2001, and at the German Anaesthesia Congress, Nürnberg, Germany, June 13–14, 2001.
|
References
|
|---|
- Alkire MT, Haier RJ, Barker SJ, et al. Cerebral metabolism during propofol anesthesia in humans studied with positron emission tomography. Anesthesiology 1995; 82: 393–403.[ISI][Medline]
- Alkire MT, Haier RJ, Fallon JH, Barker SJ. PET imaging of conscious and unconscious verbal memory. J Consciousness Stud 1996; 3: 448–62.
- Ghoneim MM, Block RI. Learning and memory during general anesthesia: an update. Anesthesiology 1997; 87: 387–410.[ISI][Medline]
- Richardson-Klavehn A, Bjork RA. Measures of memory. Annu Rev Psychol 1988; 39: 475–543.[ISI]
- Tulving E, Schacter DL. Priming and human memory systems. Science 1990; 247: 301–6.[Abstract/Free Full Text]
- Just MA, Carpenter PA, Woolley JD. Paradigms and processes in reading comprehension. J Exp Psychol Gen 1982; 111: 228–38.[ISI][Medline]
- Jones JG, Aggarwal S. Monitoring the depth of anesthesia. In: Ghoneim MM, ed. Awareness during anesthesia. Oxford: Butterworth-Heinemann, 2001: 69–91.
- Sebel PS, Lang E, Rampil IJ, et al. A multicenter study of bispectral electroencephalogram analysis for monitoring anesthetic effect. Anesth Analg 1997; 84: 891–9.[Abstract]
- Lubke GH, Kerssens C, Phaf H, et al. Dependence of explicit and implicit memory on hypnotic state in trauma patients. Anesthesiology 1999; 90: 670–80.[ISI][Medline]
- Schwender D, Kaiser A, Klasing S, et al. Midlatency auditory evoked potentials and explicit and implicit memory in patients undergoing cardiac surgery. Anesthesiology 1994; 80: 493–501.[ISI][Medline]
- Ghoneim MM, Block RI, Dhanaraj VJ, et al. Auditory evoked responses and learning and awareness during general anesthesia. Acta Anaesthesiol Scand 2000; 44: 133–43.[ISI][Medline]
- Kugler J. Elektroencephalographie in Klinik und Praxis: eine Einführung. Stuttgart: Georg Thieme Verlag, 1981: 53–60.
- Moscovitch M, Winocur G, McLachlan D. Memory as assessed by recognition and reading time in normal and memory-impaired people with Alzheimer’s disease and other neurological disorders. J Exp Psychol Gen 1986; 115: 331–47.[ISI][Medline]
- Münte S, Kobbe I, Demertzis A, et al. Increased reading speed for stories presented during general anesthesia. Anesthesiology 1999; 90: 662–9.[ISI][Medline]
- Münte S, Lüllwitz E, Leuwer M, et al. No implicit memory for stories played during isoflurane/alfentanil/nitrous oxide anesthesia: a reading speed measurement. Anesth Analg 2000; 90: 733–8.[Abstract/Free Full Text]
- Münte S, Münte TF, Mitzlaff B, et al. Postoperative reading speed does not indicate implicit memory in elderly cardiac patients after propofol-remifentanyl anaesthesia. Acta Anaesthesiol Scand 2001; 45: 750–5.[ISI][Medline]
- Loomis AL, Harvey EN, Hobart GA. Cerebral states during sleep, as studied by human brain potentials. J Exp Psychol 1937; 21: 127–44.[ISI]
- Kugler J. Elektroenzephalographie in Klinik und Praxis. Stuttgart: Georg Thieme Verlag, 1963.
- Doenicke A, Löffler B, Kugler J, et al. Plasma concentration and EEG after various regimens of etomidate. Br J Anaesth 1982; 54: 393–400.[Abstract/Free Full Text]
- Schulz A, Schulz B, Grouven U. Validierungsuntersuchungen zum EEG-Monitor Narcotrend. 12. Norddeutsche Anästhesie-Tage, Abstractband, 1999.
- Raymondos K, Münte S, Krauss T, et al. Cortical activity assessed by Narcotrend® in relation to haemodynamic responses to tracheal intubation at different stages of cortical suppression and reflex control. Eur J Anaesthesiol 2002; 19: 1–8.[ISI][Medline]
- Kreuer S, Biedler A, Larsen R, et al. The NarcotrendTM: a new EEG monitor designed to measure the depth of anaesthesia—a comparison with bispectral index monitoring during propofol-remifentanil-anaesthesia. Anaesthesist 2001; 50: 921–5.[ISI][Medline]
- Doenicke A, Kugler J, Schellenbereger A, et al. The use of electroencephalography to measure recovery time after intravenous anaesthesia. Br J Anaesth 1966; 38: 580–90.[Abstract/Free Full Text]
- Pichelmayr I, Lips U, Kunkel H. Das Elektroenzephalogramm in der Anaesthesie. Berlin: Springer, 1983: 85–101.
- Doenicke AW, Roizen MF, Rau J, et al. Pharmacokinetics and pharmacodynamics of propofol in new solvent. Anesth Analg 1997; 85: 1399–403.[Abstract]
- Tardif T, Craik FIM. Reading a week later: perceptual and conceptual factors. J Mem Lang 1989; 28: 107–25.
- Gan TJ, Glass PS, Windsor A, et al. Bispectral index monitoring allows faster emergence and improved recovery from propofol, alfentanil, and nitrous oxide anesthesia. Anesthesiology 1997; 87: 808–15.[ISI][Medline]
- Song D, Joshi GP, White PF. Titration of volatile anesthetics using bispectral index facilitates recovery after ambulatory anesthesia. Anesthesiology 1997; 87: 842–8.[ISI][Medline]
- Yli-Hankala A, Vakkuri A, Annila P, Korttila K. EEG bispectral index monitoring in sevoflurane or propofol anaesthesia: analysis of direct costs and immediate recovery. Acta Anaesthesiol Scand 1999; 43: 545–9.[ISI][Medline]
- Wang M. The psychological consequences of explicit and implicit memories of events during surgery. In: Ghoneim MM, ed. Awareness during anesthesia. Oxford: Butterworth-Heinemann, 2001: 145–53.
Accepted for publication March 10, 2003.
This article has been cited by other articles:
|
|
|
|
 
U. E. G. Dobrunz, K. Jaeger, and G. Vetter
Memory priming during light anaesthesia with desflurane and remifentanil anaesthesia
Br. J. Anaesth.,
April 1, 2007;
98(4):
491 - 496.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
