- Email: [email protected]
Electroencephalogram monitoring facilitates sedation with propofol for routine ERCP: A randomized, controlled trial
Electroencephalogram monitoring facilitates sedation with propofol for routine ERCP: a randomized, controlled trial Till Wehrmann, MD, Jörg Grotkamp, MD, Nikos Stergiou, MD, Andrea Riphaus, MD, Annegret Kluge, MD, Bernhard Lembcke, MD, Arthur Schultz, MD Hanover and Gladbeck, Germany
Background: Endoscopy with the patient under sedation with propofol requires careful monitoring of patient consciousness and vital signs to achieve the desired hypnotic effect without overdosage. Because level of consciousness is difficult to judge by clinical observation alone, electroencephalogram monitoring has been used to guide general anesthesia. Methods: Eighty consecutive patients (mean [SD] age 62 [14] years) undergoing ERCP for various indications were randomly allocated to 2 groups. In group A (n = 40), propofol sedation was guided by conventional monitoring (heart rate, blood pressure, oxygen saturation, electrocardiogram), whereas electroencephalogram monitoring was performed but not displayed to the physician who administered the drug. In group B (n = 40), electroencephalogram monitoring was displayed online and used to guide propofol administration to maintain a preselected sedation level. Procedurerelated parameters, recovery time, and quality (postanesthesia recovery score), as well as patient cooperation and tolerance for the procedure (visual analog scale) were prospectively assessed. Results: The groups were comparable with regard to demographic, clinical, and procedure-related parameters. Mean propofol dose was significantly lower in group B than in group A (respectively, 290 [158] mg vs. 374 [166] mg; p = 0.02). The mean decrease in arterial blood pressure was significantly lower in group B than in group A (respectively, 11 [6] mm Hg vs. 14 [7] mm Hg; p < 0.05). Clinically relevant changes in vital signs to below critical values were observed in both groups, albeit infrequently. Efficacy of sedation was also rated similar in both groups, whereas mean recovery time was significantly faster in group B than in group A (respectively, 16 [7] minutes vs. 20 [8] minutes; p = 0.02). Accordingly, the recovery score tended to be better in group B compared with group A (respectively, 8.4 [1.0] points vs. 8.0 [0.9] points; p = 0.06). The predefined target level of sedation was maintained during 75% of the procedure time in group B but in only 58% of the time in group A (p < 0.05), and deeper sedation levels were achieved significantly more often in group B patients compared with group A patients (p < 0.05). Conclusion: Electroencephalogram monitoring enables more effective titration of propofol dosage for sedation during endoscopy and is, therefore, associated with faster patient recovery. (Gastrointest Endosc 2002;56:817-24.)
There is growing interest in sedation by intravenous administration of propofol during endoscopic procedures.1-11 Propofol, a substituted phenol derivative, is an established anesthetic agent in clinical practice.12 It has a rapid onset of action and short half-life that results in rapid recovery from anesthesia.13,14 However, it has pronounced cardiorespiratory depressant effects.12,13 Therefore, Received November 8, 2001. For revision February 20, 2002. Accepted June 19, 2002. Current affiliations: Department of Internal Medicine I, Academic Hospital Hanover-Siloah, Hanover, Germany, Department of Internal Medicine, St. Barbara Hospital, Gladbeck, Germany, and Department of Anesthesiology IV, Medical University of Hanover, Hanover, Germany. Reprint requests: Prof. Dr. Till Wehrmann, Department of Internal Medicine I, Academic Hospital Hanover-Siloah, Roesebeckstr. 15, 30449 Hanover, Germany. Copyright © 2002 by the American Society for Gastrointestinal Endoscopy 0016-5107/2002/$35.00 + 0 37/1/129603 doi:10.1067/mge.2002.129603 VOLUME 56, NO. 6, 2002
propofol administration has to be titrated to maintain the desired hypnotic effect, usually by continuous intravenous infusion or intermittent bolus injection. Because propofol has a narrow therapeutic range and patients’ transition from conscious to deep sedation is easily obtained, the intermittent bolus technique is preferred. Administration of propofol doses is guided by observation of level of consciousness and vital signs (e.g., automated registration of blood pressure, electrocardiogram [ECG], oxygen saturation). The main goal of titration is to obtain that level of sedation during the entire endoscopic procedure that allows maximum patient tolerance without inducing cardiorespiratory side effects. However, level of consciousness cannot be reliably judged by somatic or hemodynamic responses alone, especially when the drug is used intravenously for anesthesia. In this context, it is helpful that alterations of cerebral function caused by anesthetic agents are reflected in the electroencephalogram GASTROINTESTINAL ENDOSCOPY
817
T Wehrmann, J Grotkamp, N Stergiou, et al.
EEG monitoring for sedation during ERCP
Table 1. Demographic and clinical data Age, y (SD) (range) Gender (male:female) ASA class I (n) II (n) III (n) IV (n)
Group A (n = 40)
Group B (n = 40)
61 (15) (26-91) 15:25
62 (14) (24-92) 14:26
3 8 20 9
3 8 21 8
Study covers 80 patients who underwent ERCP under sedation with propofol based on clinical monitoring with (group A) or without the aid of additional EEG monitoring (group B). Differences between both groups are not significant. ASA, American Society of Anesthesiology.
(EEG) rhythm.16-21 Increasing hypnotic effect is associated with a slowing of the EEG. Thus, evaluation of the EEG provides important and objective information about the hypnotic state of the patient. A novel EEG monitor for guiding sedation and anesthesia, developed by a research group at Hanover Medical School, incorporates algorithms for an automatic assessment of the EEG.22 Therefore, the efficacy of EEG monitoring with this device during propofol sedation for ERCP was evaluated in a prospective, randomized, and controlled study. PATIENTS AND METHODS A total of 90 adult patients undergoing ERCP under sedation with propofol were randomly assigned by computer-generated list to have conventional monitoring without (group A) or with additional EEG monitoring (group B). Exclusion criteria were age less than 18 years, pregnancy, inability to obtain informed consent because of medical condition (e.g., septic cholangitis, acute biliary pancreatitis), partial or total gastrectomy, or critical illness (ASA-class V). Each patient participated only once, that is, re-examinations were not included. After randomization, 4 patients had to be excluded because they refused participation. In 4 patients cannulation was unsuccessful, and in 2 patients malignant gastric outlet obstruction hindered positioning of the duodenoscope. The latter 6 patients were excluded because the short duration of the procedure (median 6 minutes) did not allow meaningful comparisons. Clinical data for the remaining 80 patients are shown in Table 1. Written informed consent was obtained from all patients, and the study protocol was approved by the ethics committee of our institution. Parameters assessed All endoscopic examinations were performed by a single experienced investigator (T.W.) who used standard techniques and duodenoscopes with the patient always positioned on the left side. Heart rate (3-lead ECG), oxygen saturation (pulse oxymetry with continuous oxygen supplementation with 2 L per minute through nasal prongs) and blood pressure (automated measurement every 5 min818
GASTROINTESTINAL ENDOSCOPY
Table 2. EEG stages A-F according to Kugler24 Stage
State
Leading EEG features
A
Awake
B0
Sleepiness, light level of anesthesia
Alpha waves and/or typical artifact constellations Beta and theta waves
B1 B2 C0 C1 C2 D0 D1 D2 E0
E1 F0
Flat to moderate level of anesthesia
Increasing amount of theta waves
Moderate to deep level of anesthesia
Increasing amount of delta waves
Deep level of anesthesia Continuous high delta activity and very slow delta waves, transition to burst suppression EEG Very deep level of anesthesia
Burst suppression activity, continuous EEG suppression
F1 Examples of the respective EEG recordings from stages A-F are illustrated in Fig. 1.
utes) were continuously monitored throughout the procedure. The time elapsing between the first injection of propofol and final withdrawal of the endoscope was recorded (procedure duration). Patient co-operation was rated (on a visual analog scale ranging from 0 [poor] to 10 [excellent]) by the endoscopist, who was blinded to the propofol titration technique used. Tolerability of the procedure was rated by the patients 4 hours after ERCP by using a visual analog scale (0-10 points). After the procedure, the time until the patient was fully alert (conversant and awake) was recorded and the degree of residual sedation was assessed by using a postanesthesia recovery score (PARS) 30 minutes after completion of the ERCP.23 This was done by a well-trained staff nurse not otherwise involved in the study. For PARS, patients were assigned points of 0, 1, or 2 for each of the following categories: (1) activity (inability to move limbs, ability to move 2 or 4 limbs with or without command); (2) respiration (evidence of apnea, labored breathing or normal breathing pattern); (3) circulation (blood pressure compared with baseline before sedation: ± 50% with reference to baseline, ± 20-50% to baseline, ± 20% to baseline); (4) consciousness (hypnotic, arousable or fully awake); and, (5) skin color (cyanotic, pink, or normal). Complete recovery is indicated by the maximum PARS of 10 points.23 Tolerability of the procedure was noted by the patient 4 hours after ERCP using a visual analog scale (0-10 points). Sedation regimens After a loading dose of 40 mg (<70 kg body weight) or 60 mg (≥70 kg body weight) of propofol (Disoprivan 1%, Zeneca, Plankstadt, Germany) had been injected in both groups, repeated doses of 20 mg of propofol were given intravenously to maintain an adequate level of sedation VOLUME 56, NO. 6, 2002
EEG monitoring for sedation during ERCP
(with no limit on total dose). No other medication was used for induction of sedation. An assisting physician (who was not involved in the endoscopic procedure and who had special training in intensive care medicine) always performed the sedation procedure. For patients in group A (conventional sedation technique) the assisting physician ensured that the patient tolerated the procedure well and that vital functions were unimpaired (heart rate ≥50/minute, oxygen saturation ≥90%, systolic pressure ≥90 mm Hg) by titrating repeated propofol boluses cautiously and increasing the oxygen supply to 4 to 6 L per minute if oxygen saturation decreased to less than 90%. A small preamplifier and a unit for analysis and display of EEG data was used (Narcotrend, MT MonitorTechnik, Bad Bramstedt, Germany). This device was developed by one of the investigators (AS) and coworkers at Hanover Medical School. The EEG was recorded by means of 3 self-adhesive ECG electrodes placed on the forehead of the patient. The system performed automated analysis of EEG segments of 20 seconds’ duration (20-second epochs). After extensive artifact analysis, the EEG epochs are automatically classified by multivariate statistical procedures by using a scale from A (awake) to F (very deep sleep) with 14 substages (adopted from a visual EEG classification system proposed by Kugler22,24). The algorithm includes adaptations for agerelated EEG changes from childhood to old age.22 The main features of stages A to F are illustrated in Table 2 and Figure 1. Validation studies have shown that the Narcotrend provides an accurate assessment of hypnotic depth compared with conventional EEG analysis during propofol sedation.25,26 Before the present study, this device was used during 85 interventional endoscopic procedures and the sedation levels D0 to D2 were found to be optimal with regard to patient tolerance and avoidance of cardiorespiratory side effects.25 Therefore, EEG stages D0 to D2 were considered the target level. EEG registration was performed in group A but not accessible to the physician who administered the propofol; the EEG data were stored for later analysis. In group B, EEG analysis was displayed online on a personal computer monitor screen (Fig. 2) and the physician responsible for the sedation titrated the repeated propofol doses to achieve and maintain the target sedation level of D0 to D2. The endoscopist as well as the patient were blinded to the respective propofol titration technique used. Statistics All data are given as mean (standard deviation). For statistical comparison the Wilcoxon signed rank test, the Mann-Whitney rank sum test, and the Fisher exact test were used whenever appropriate, and a p value of < 0.05 was regarded as significant. Data analysis was performed on a personal computer with statistical software (Instat, Graph Pad, San Diego, Calif).
RESULTS There were no significant differences between the groups with regard to special patient characteristics VOLUME 56, NO. 6, 2002
T Wehrmann, J Grotkamp, N Stergiou, et al.
Figure 1. Characteristics of EEG stages A through F according to Kugler.24
that might have impacted co-operation under sedation or the performance of associated or interventional procedures (Tables 3 and 4). Only 6 of the ERCP procedures in group A (15%) and 8 in group B (20%) were performed for diagnosis. The mean duration of the endoscopic procedures was similar in both groups (group A, median 35 minutes, range 1784 minutes; group B, median 37 minutes, range 1891 minutes). Propofol dosage The mean propofol dose administered was 374 (166) mg (range 90 to 920 mg; median 360 mg) in group A and 290 (158) mg (range 60 to 740 mg; median, 300 mg) in group B (p = 0.02). Similarly, the mean propofol dose per kilogram of body weight GASTROINTESTINAL ENDOSCOPY
819
T Wehrmann, J Grotkamp, N Stergiou, et al.
EEG monitoring for sedation during ERCP
Figure 2. Photograph of PC monitor screen of Narcotrend device displaying raw EEG curve (top), EEG analysis (bottom right and left) and respective EEG stage (bottom middle), in this case C0.
Table 3. Patient characteristics with respect to potential impact on quantity and quality of sedation Group A (n = 40) Alcohol abuse (n) Regular use of sedatives or psychotropic drugs (n) Regular use of narcotics (n) Regular use of non-narcotic analgesics (n) Prior difficulty with sedation during endoscopic examination (n)
Group B (n = 40)
10 9
9 7
2 14
3 12
3
2
Differences between both groups are not significant. Patients in group A underwent ERCP under sedation with propofol based on clinical monitoring alone and patients in group B with the aid of additional EEG monitoring.
related to the duration of the procedure was significantly different between the groups (0.17 [0.10] mg propofol/kg/min in group A vs. 0.13 [0.08] mg/kg/min in group B; p < 0.01). As comedication, 20 mg of Nbutylscopolamine was required in 2 patients of group A and in 3 patients of group B (not significant [ns]). Sedation efficacy Patient co-operation, as rated by the endoscopist, was similar for both groups (ns; Table 5). Also, patient assessment of procedure tolerability was similar with either sedation regime (Table 5). 820
GASTROINTESTINAL ENDOSCOPY
Table 4. Endoscopic procedures performed in the 2 study groups Cholangiography Pancreatography Biliary sphincterotomy Pancreatic sphincterotomy Sphincter Oddi manometry Stone extraction Balloon dilation Biliary endoprosthesis Pancreatic endoprosthesis Intraductal US Miscellaneous
Group A (n = 40)
Group B (n = 40)
37 11 19 4 6 12 3 12 3 6 6
36 12 16 5 6 11 2 13 4 5 5
Vital signs and adverse effects The mean increase in heart rate (initial heart rate vs. maximum heart rate during procedure, as measured by continuous ECG recording) was not significantly different between group A and group B (Table 5). A temporary decrease in heart rate below 50 per minute occurred in 1 patient in each group. The mean decline in oxygen saturation (initial oxygen saturation vs. lowest oxygen saturation during procedure) was not significantly different between the groups (Table 5) and a potentially harmful decrease in oxygen to less than 90% was observed with similar frequency in both groups (5/40 patients in group A vs. 4/40 patients in group B; ns). A brief decrease in oxygen saturation to less than 85% was recorded for 2 patients in each group VOLUME 56, NO. 6, 2002
EEG monitoring for sedation during ERCP
T Wehrmann, J Grotkamp, N Stergiou, et al.
Table 5. Parameters of sedation efficacy and alterations of vital signs in 2 study groups Group A (n = 40) Patient co-operation 8.6 (1.5) (7-10) (points) Patient tolerance (points) 8 (1) (6-10) Increase in heart rate (%) 5 (3) (1-8) Decrease in mean arterial 14 (7) (8-22) blood pressure (mm Hg) Decline in oxygen saturation 6 (4) (1-11) (%)
Group B (n = 40) 8.8 (1.6) (7-10) 9 (1) (6-10) 4 (3) (1-8) 11 (6) (5-19)* 5 (3) (1-10)
All data given as mean (SD) (ranges). Differences between both groups are not significant except for *p < 0.05. Patients in group A underwent ERCP under sedation with propofol based on clinical monitoring alone and patients in group B with aid of additional EEG monitoring.
(ns). All of the latter patients responded rapidly to an increase of supplemental oxygen. Assisted bag ventilation because of prolonged apnea, as has been reported with propofol sedation,6-8 was not required for any patient in either group. Mean decrease in mean arterial blood pressure was 14 (7) mm Hg in group A and 11 (6) mm Hg in group B (p < 0.05; Table 5). A decrease in systolic blood pressure below 90 mm Hg, however, was observed in only 1 patient in group A versus none in group B (ns). Post-ERCP pancreatitis, defined according to criteria proposed by Cotton et al.,27 was recorded in 2 of 40 patients (5%) in group A (both cases mild) and 3 of 40 patients (8%) in group B (2 mild, 1 moderate; ns vs. group A). There were no other immediate complications of the endoscopic procedures (e.g., perforation or sepsis). Recovery The mean time until patients were fully alert after the endoscopic procedure was 20 (8) minutes (range 6-36 minutes) in the patients who had conventional propofol-titration (group A) but 16 (7) minutes (range 5-27 minutes) in those with EEGassisted propofol titration (group B)(p = 0.02; Fig. 3). Also, the postanesthesia recovery score (30 minutes after completion of procedure) tended to be better in group B (8.4 [1.0]; range 6-10) than in group A (8.0 [0.9]; range 6-10), but the difference failed to reach statistical significance (p = 0.06; Fig. 3). EEG analysis The respective proportion of time during which patients in both groups remained in the different EEG stages is displayed in Table 6. The predefined target level of sedation (D0 to D2) was maintained during 75% of the procedure time in group B but in VOLUME 56, NO. 6, 2002
Figure 3. Box-Whisker plot of time until patients are fully alert after endoscopic procedure (recovery time, top) and postanesthesia recovery score (PARS; obtained 30 minutes after completion of ERCP) in 80 patients randomly assigned for titration of propofol sedation with (group B) or without EEG-monitoring (group A). *p < 0.05 vs. group A.
only 58% of time in group A (p < 0.05). Deeper sedation levels were reached significantly more often in the patients of group A versus those in group B (stages E-F, 25% in group A vs. 11% in group B; p < 0.05). Individual examples of recordings from each group are shown in Figure 4. DISCUSSION The duration of ERCP, now largely a therapeutic procedure, has a wide span, but is sometimes more than 1 hour, and long duration can lead to respiratory and cardiovascular changes. Therefore, appropriate levels of sedation must be achieved while averting the potentially adverse effects of sedative drugs. The need for proper sedation during ERCP is GASTROINTESTINAL ENDOSCOPY
821
T Wehrmann, J Grotkamp, N Stergiou, et al.
EEG monitoring for sedation during ERCP
Table 6. Respective proportion of time during which patients remained in different EEG stages EEG stage
Group A (n = 40)
Group B (n = 40)
A B C D E F
4% 1% 12% 58% 23% 2%
5% 1% 8% 75%* 10%* 1%
*p < 0.05 vs. group A.
Figure 4. EEG stages (i.e., level of sedation) over time during ERCP in a patient from group A (propofol titration based on clinical data) (top), compared with EEG stages for a patient from group B (propofol administration based on additional EEG monitoring) (bottom). Note that the target level (D) of sedation was achieved more constantly in the patient from group B.
particularly obvious when technically demanding interventional procedures are performed. Three randomized controlled studies have shown that sedation efficacy is greater and patient recovery shorter when ERCP is performed with the patient under sedation with propofol compared with the traditional midazolam.7,8,10 In addition, a significantly higher success rate was found for performance of endoscopic sphincter of Oddi manometry under propofol sedation when compared with a historical control group sedated with midazolam.6 Propofol, however, causes more severe cardiorespiratory side effects and repeated administration is necessary because of its short half-life. Therefore, comprehensive patient monitoring (e.g., pulse oxymetry, ECG, and automated blood pressure recording) is required and propofol administration has to be titrated carefully to avoid oversedation. In previous studies, which were based on subjective parameters 822
GASTROINTESTINAL ENDOSCOPY
of the desired level of sedation, there were large variations in the initial and repeated dosages of propofol given for sedation during endoscopic procedures.1-9 It is known that there is a great individual variability in the drug levels needed to achieve certain desired (or undesired) effects. Usually the effect of the sedation is assessed by observing patient awareness, compliance with the procedure, pain reactions, and reflex status. However, these parameters are not easy to monitor in a darkened procedure room by a physician who is usually not positioned near the patient’s head (a position routinely occupied by the endoscopist and assisting nurse). With use of a drug like propofol, primarily a hypnotic agent, in these circumstances, the level of sedation is especially difficult to assess by clinical observation and cardiorespiratory parameters alone.19 EEG-guided sedation is being used increasingly by anesthesiologists to achieve exact titration of hypnotic agents.14-21 For general anesthesia, it has been shown that EEG guidance avoids both underdosing and overdosing and is valuable for detecting potentially harmful situations for the brain (e.g., hypoxic episodes). Visual EEG classifications, as proposed originally by Kugler,24 have been used in several studies with different hypnotic drugs.17,28 For practical purposes in GI endoscopy, automated analysis rather than visual interpretation of the EEG recording (which is time consuming and requires special knowledge) is desirable. Numerous quantitative parameters have been studied for their ability to describe and characterize the EEG. Spectral parameters such as the median or the spectral edge frequency have been used most often.10,18,29-31 However, in many cases the complex EEG signal cannot be adequately characterized by a single parameter.26,32 Thus, for the classification algorithms in the Narcotrend device used in the present investigation, a multiparametric approach with statistical discriminant analysis including several parameters derived from the EEG is applied. A validation study demonstrated that the Narcotrend device provides a reliable description of the EEG.25 The present study demonstrates that EEGguided propofol sedation during ERCP is associated VOLUME 56, NO. 6, 2002
EEG monitoring for sedation during ERCP
with a significant lowering of total propofol dose and, accordingly, more rapid patient recovery. Further, the decline in mean arterial blood pressure was less pronounced in the patients in whom propofol administration was EEG-guided. This is due to more constant achievement of a proper level of sedation (Fig. 4, Table 6) with the aid of EEG recording. This is also reflected by the fact that, although the total dose of propofol administered was significantly lower in the patient group in which EEG monitoring was applied, sedation efficacy was not lowered compared with that for patients in whom propofol administration was based on clinical monitoring alone. Struys et al.33 documented that a closed-loop controlled administration of propofol with an EEG parameter as the control variable is clinically acceptable for general anesthesia and associated with better control of arterial blood pressure than conventional propofol dosing. EEG monitoring in the present study did not, in general, affect patient safety or sedation efficacy (with the exception of the influence on the blood pressure). With respect to cost-effectiveness, the additional costs of EEG monitoring (in Germany the Narcotrend device is available for about 10,000 U.S. dollars) must be balanced against the reduced need for propofol and faster patient recovery. In the present study, there was a saving of $3.00 per procedure (corresponding to 60 mg of propofol). Therefore, the addition of EEG monitoring seems to increase costs. However, avoidance of oversedation because of propofol administration (e.g., in patients with hepatic encephalopathy or patients taking neuroleptic drugs in whom propofol dosing must be adjusted) is a worthwhile goal irrespective of cost considerations. The number of patients in the present study is too small to allow evaluation of differences in the occurrence rates of rare life-threatening complications, for example clinically relevant hypoxemia necessitating assisted ventilation, as have been reported.68 Furthermore, for study purposes patients who required emergency ERCP, the subgroup at greatest risk for complications, were excluded. In conclusion, the results of the present study demonstrated that additional EEG monitoring results in more effective and economic titration of propofol administration for sedation during interventional endoscopic procedures. DISCLOSURE During the study period a prototype EEG monitor Narcotrend was given as a grant from the Department of Anesthesiology, Medical University of Hanover. A. Schultz is a coinventor and holds the patent, “Method and device for evaluating an EEG carried out in the context of anesVOLUME 56, NO. 6, 2002
T Wehrmann, J Grotkamp, N Stergiou, et al.
thesia and intensive care” but he has not received any license fees. None of the investigators have commercial associations with the Narcotrend device, produced by MT Monitor Technik, Bad Bramstedt, Germany (i.e., equity ownership or interest, consultancy or institutional and corporate associations). REFERENCES 1. Gepts E, Claeys MA, Camu F, Smekans L. Infusion of propofol as sedative technique for colonoscopies. Postgrad Med J 1985;61:120-6. 2. Dubois A, Balatoni E, Peters JP, Baudoux M. Use of propofol for sedation during gastrointestinal endoscopies. Anesthesia 1988;43:75-80. 3. Pohlmann S, Herden HN, Hagenmueller F. Propofol-Narkose fuer die Endoskopie-gefaehrlicher als Midazolam? [in German with English abstract]. Bildgebung 1993;60(suppl):61-3. 4. Kostash MA, Johnston R, Bailey RJ, Konopad EM, Guthriern LP. Sedation for colonoscopy: a double-blind comparsion of diazepam/meperidine, midazolam/fentanyl and propofol/fentanyl combinations. Canad J Gastroenterol 1994;8:27-31. 5. Hofmann C, Kiesslich R, Brackertz A, Jung M. Propofol for premedication in gastroscopies—a randomized comparison to midazolam. Z Gastroenterol 1999;37:589-95. 6. Schmitt Th, Seifert H, Dietrich CF, Caspary WF, Wehrmann T. Intravenous sedation with propofol during endoscopic sphincter of Oddi manometry. Z Gastroenterol 1999;37:219-27. 7. Wehrmann T, Kokabpick S, Lembcke B, Caspary WF, Seifert H. Efficacy and safety of intravenous propofol sedation during routine ERCP: a prospective, controlled study. Gastrointest Endosc 1999;49:677-83. 8. Jung M, Hofmann C, Kiesslich R, Brackertz A. Improved sedation in diagnostic and therapeutic ERCP: propofol is an alternative to midazolam. Endoscopy 2000;32:233-8. 9. Koshy G, Nair S, Norkus EP, Hertan HI, Pitchumoni CS. Propofol vs. midazolam and meperidine for conscious sedation in GI endoscopy. Am J Gastroenterol 2000;95:1476-9. 10. Krugliak P, Ziff B, Rusabrov Y, Rosenthal A, Fich A, Gurman GM. Propofol vs. midazolam for conscious sedation guided by processed EEG during endoscopic retrograde cholangiopancreatography: a prospective, randomized double-blind study. Endoscopy 2000;32:677-82. 11. Reimann FM, Samson U, Derad M, Fuchs M, Schiefer B, Stange EF. Synergistic sedation with low dose midazolam and propofol for colonoscopies. Endoscopy 2000;32:239-44. 12. Sebel PS. Propofol: a new intravenous anaesthetic. Anaesthesiology 1989;71:260-77. 13. McCollum JSC, Dundee JW, Halliday NJ, Clarke RSJ. Dose response studies with propofol in unpremedicated patients. Postgrad Med J 1985;61:85-7. 14. Cockshott ID, Briggs LP, Douglas EJ, White M. Pharmacokinetics of propofol in female patients: studies using single bolus injection. Br J Anaesth 1987;59:1103-10. 15. Searle NR, Sahab P. Propofol in patients with cardiac disease. Can J Anaesth 1993;40:816-8. 16. Doenicke AW, Kugler J, Schellenberger A, Gürtner T. The use of electroencephalography to measure recovery time after intravenous anaesthesia. Br J Anaesth 1966;38:580-90. 17. Doenicke AW, Löffler B, Kugler J, Suttmann H, Grote B. Plasma concentration and EEG after various regimens of etomidate. Br J Anaesth 1982;54:393-400. 18. Schwilden H, Stoeckel H. Quantitative EEG analysis during anaesthesia with isoflurane in nitrous oxide at 1.3 and 1.5 MAC. Br J Anaesth 1987;59:738-45. GASTROINTESTINAL ENDOSCOPY
823
T Wehrmann, J Grotkamp, N Stergiou, et al.
19. Heier T, Steen PA. Assessment of anaesthesia depth. Acta Anaesthesiol Scand 1996;40:1087-100. 20. Smith WD, Dutton RC, Smith NT. Measuring the performance of anesthetic depth indicators. Anesthesiology 1996; 84:38-51. 21. Guérit J-M. Neuromonitoring in the operating room: why when, and how to monitor? Electroencephalogr Clin Neurophysiol 1998;106:1-21. 22. Schultz B, Schultz A, Grouven U. Sleeping stage based systems (Narcotrend). In: Bruch HP, Köckerling F, Bouchard R, Schug-Paâ C, editors. New aspects of high technology in medicine. Bologna: Monduzzi Editore; 2000. p. 285-91. 23. Kankaria A, Lewis JH, Ginsberg G, Gallagher J, Al-Kawas F, Nguyen CC, et al. Flumazenil reversal of psychomotor impairment due to midazolam or diazepam for conscious sedation for upper endoscopy. Gastrointest Endosc 1996;44: 416-21. 24. Kugler J. Elektroenzephalographie in Klinik und Praxis. Stuttgart: Thieme; 1981. p. 120-46. 25. Schultz B, Grouven U, Schultz A. Automatic classification algorithms of the EEG-monitor Narcotrend for routinely recorded EEG-data from general anaesthesia: a validation study. Biomed Eng 2002;47:9-13. 26. Schultz A, Grouven U, Schultz B. SEF90, SEF95 and the median do not discriminate deep EEG stages during propofol/ remifentanil anaesthesia [abstract]. Eur J Anaesthesiol 2000;17(Suppl 10):22.
EEG monitoring for sedation during ERCP
27. Cotton PB, Lehman GA, Vennes J, Geenen JE, Kozarek RA, DiMagno EP. Endoscopic sphincterotomy complications and their management. An attempt at consensus. Gastrointest Endosc 1991;37:383-93. 28. Doenicke AW, Roizen MF, Rau J, O’Connor M, Kugler J, Klotz U, et al. Pharmacodynamics of propofol in a new solvent. Anesth Analg 1997;85:1399-403. 29. Drummond JC, Brann CA, Perkins DE, Wolfe DE. A comparison of median frequency spectral edge frequency a frequency band power ratio, total power and dominance shift in the determination of depth of anaesthesia. Acta Anaesthesiol Scand 1991;35:693-9. 30. Schwender D, Daunderer M, Mulzer S, Klasing S, Finsterer U, Peter K. Spectral edge frequency of the electroencephalogram to monitor “depth” of anaesthesia with isoflurane or propofol. Br J Anaesth 1996;77:179-84. 31. Gurman GM, Fajer S, Porat A, Rusabrov Y. Use of EEG spectral edge as an index of equipotency in a comparison of propofol and isoflurane for maintenance of general anesthesia. Eur J Anaesthesiol 1994;11:443-8. 32. Levy WJ. Intraoperative EEG patterns: implications for EEG monitoring. Anesthesiology 1984;60:430-40. 33. Struys MM, De Smet T, Versichlen LF, Van de Velde S, Van den Brocke R, Mortier EP. Comparison of closed-loop controlled administration of propofol using bispectral index as the controlled variable versus “standard practice” controlled administration. Anesthesiology 2001;95:6-17.
Availability of Journal back issues As a service to our subscribers, copies of back issues of Gastrointestinal Endoscopy for the preceding 5 years are maintained and are available for purchase from Mosby until inventory is depleted. Please write to Mosby, Subscription Customer Service, 6277 Sea Harbor Dr., Orlando, FL 32887 or call 800-654-2452 or 407-345-4000 for information on availability of particular issues and prices.
824
GASTROINTESTINAL ENDOSCOPY
VOLUME 56, NO. 6, 2002
Background: Endoscopy with the patient under sedation with propofol requires careful monitoring of patient consciousness and vital signs to achieve the desired hypnotic effect without overdosage. Because level of consciousness is difficult to judge by clinical observation alone, electroencephalogram monitoring has been used to guide general anesthesia. Methods: Eighty consecutive patients (mean [SD] age 62 [14] years) undergoing ERCP for various indications were randomly allocated to 2 groups. In group A (n = 40), propofol sedation was guided by conventional monitoring (heart rate, blood pressure, oxygen saturation, electrocardiogram), whereas electroencephalogram monitoring was performed but not displayed to the physician who administered the drug. In group B (n = 40), electroencephalogram monitoring was displayed online and used to guide propofol administration to maintain a preselected sedation level. Procedurerelated parameters, recovery time, and quality (postanesthesia recovery score), as well as patient cooperation and tolerance for the procedure (visual analog scale) were prospectively assessed. Results: The groups were comparable with regard to demographic, clinical, and procedure-related parameters. Mean propofol dose was significantly lower in group B than in group A (respectively, 290 [158] mg vs. 374 [166] mg; p = 0.02). The mean decrease in arterial blood pressure was significantly lower in group B than in group A (respectively, 11 [6] mm Hg vs. 14 [7] mm Hg; p < 0.05). Clinically relevant changes in vital signs to below critical values were observed in both groups, albeit infrequently. Efficacy of sedation was also rated similar in both groups, whereas mean recovery time was significantly faster in group B than in group A (respectively, 16 [7] minutes vs. 20 [8] minutes; p = 0.02). Accordingly, the recovery score tended to be better in group B compared with group A (respectively, 8.4 [1.0] points vs. 8.0 [0.9] points; p = 0.06). The predefined target level of sedation was maintained during 75% of the procedure time in group B but in only 58% of the time in group A (p < 0.05), and deeper sedation levels were achieved significantly more often in group B patients compared with group A patients (p < 0.05). Conclusion: Electroencephalogram monitoring enables more effective titration of propofol dosage for sedation during endoscopy and is, therefore, associated with faster patient recovery. (Gastrointest Endosc 2002;56:817-24.)
There is growing interest in sedation by intravenous administration of propofol during endoscopic procedures.1-11 Propofol, a substituted phenol derivative, is an established anesthetic agent in clinical practice.12 It has a rapid onset of action and short half-life that results in rapid recovery from anesthesia.13,14 However, it has pronounced cardiorespiratory depressant effects.12,13 Therefore, Received November 8, 2001. For revision February 20, 2002. Accepted June 19, 2002. Current affiliations: Department of Internal Medicine I, Academic Hospital Hanover-Siloah, Hanover, Germany, Department of Internal Medicine, St. Barbara Hospital, Gladbeck, Germany, and Department of Anesthesiology IV, Medical University of Hanover, Hanover, Germany. Reprint requests: Prof. Dr. Till Wehrmann, Department of Internal Medicine I, Academic Hospital Hanover-Siloah, Roesebeckstr. 15, 30449 Hanover, Germany. Copyright © 2002 by the American Society for Gastrointestinal Endoscopy 0016-5107/2002/$35.00 + 0 37/1/129603 doi:10.1067/mge.2002.129603 VOLUME 56, NO. 6, 2002
propofol administration has to be titrated to maintain the desired hypnotic effect, usually by continuous intravenous infusion or intermittent bolus injection. Because propofol has a narrow therapeutic range and patients’ transition from conscious to deep sedation is easily obtained, the intermittent bolus technique is preferred. Administration of propofol doses is guided by observation of level of consciousness and vital signs (e.g., automated registration of blood pressure, electrocardiogram [ECG], oxygen saturation). The main goal of titration is to obtain that level of sedation during the entire endoscopic procedure that allows maximum patient tolerance without inducing cardiorespiratory side effects. However, level of consciousness cannot be reliably judged by somatic or hemodynamic responses alone, especially when the drug is used intravenously for anesthesia. In this context, it is helpful that alterations of cerebral function caused by anesthetic agents are reflected in the electroencephalogram GASTROINTESTINAL ENDOSCOPY
817
T Wehrmann, J Grotkamp, N Stergiou, et al.
EEG monitoring for sedation during ERCP
Table 1. Demographic and clinical data Age, y (SD) (range) Gender (male:female) ASA class I (n) II (n) III (n) IV (n)
Group A (n = 40)
Group B (n = 40)
61 (15) (26-91) 15:25
62 (14) (24-92) 14:26
3 8 20 9
3 8 21 8
Study covers 80 patients who underwent ERCP under sedation with propofol based on clinical monitoring with (group A) or without the aid of additional EEG monitoring (group B). Differences between both groups are not significant. ASA, American Society of Anesthesiology.
(EEG) rhythm.16-21 Increasing hypnotic effect is associated with a slowing of the EEG. Thus, evaluation of the EEG provides important and objective information about the hypnotic state of the patient. A novel EEG monitor for guiding sedation and anesthesia, developed by a research group at Hanover Medical School, incorporates algorithms for an automatic assessment of the EEG.22 Therefore, the efficacy of EEG monitoring with this device during propofol sedation for ERCP was evaluated in a prospective, randomized, and controlled study. PATIENTS AND METHODS A total of 90 adult patients undergoing ERCP under sedation with propofol were randomly assigned by computer-generated list to have conventional monitoring without (group A) or with additional EEG monitoring (group B). Exclusion criteria were age less than 18 years, pregnancy, inability to obtain informed consent because of medical condition (e.g., septic cholangitis, acute biliary pancreatitis), partial or total gastrectomy, or critical illness (ASA-class V). Each patient participated only once, that is, re-examinations were not included. After randomization, 4 patients had to be excluded because they refused participation. In 4 patients cannulation was unsuccessful, and in 2 patients malignant gastric outlet obstruction hindered positioning of the duodenoscope. The latter 6 patients were excluded because the short duration of the procedure (median 6 minutes) did not allow meaningful comparisons. Clinical data for the remaining 80 patients are shown in Table 1. Written informed consent was obtained from all patients, and the study protocol was approved by the ethics committee of our institution. Parameters assessed All endoscopic examinations were performed by a single experienced investigator (T.W.) who used standard techniques and duodenoscopes with the patient always positioned on the left side. Heart rate (3-lead ECG), oxygen saturation (pulse oxymetry with continuous oxygen supplementation with 2 L per minute through nasal prongs) and blood pressure (automated measurement every 5 min818
GASTROINTESTINAL ENDOSCOPY
Table 2. EEG stages A-F according to Kugler24 Stage
State
Leading EEG features
A
Awake
B0
Sleepiness, light level of anesthesia
Alpha waves and/or typical artifact constellations Beta and theta waves
B1 B2 C0 C1 C2 D0 D1 D2 E0
E1 F0
Flat to moderate level of anesthesia
Increasing amount of theta waves
Moderate to deep level of anesthesia
Increasing amount of delta waves
Deep level of anesthesia Continuous high delta activity and very slow delta waves, transition to burst suppression EEG Very deep level of anesthesia
Burst suppression activity, continuous EEG suppression
F1 Examples of the respective EEG recordings from stages A-F are illustrated in Fig. 1.
utes) were continuously monitored throughout the procedure. The time elapsing between the first injection of propofol and final withdrawal of the endoscope was recorded (procedure duration). Patient co-operation was rated (on a visual analog scale ranging from 0 [poor] to 10 [excellent]) by the endoscopist, who was blinded to the propofol titration technique used. Tolerability of the procedure was rated by the patients 4 hours after ERCP by using a visual analog scale (0-10 points). After the procedure, the time until the patient was fully alert (conversant and awake) was recorded and the degree of residual sedation was assessed by using a postanesthesia recovery score (PARS) 30 minutes after completion of the ERCP.23 This was done by a well-trained staff nurse not otherwise involved in the study. For PARS, patients were assigned points of 0, 1, or 2 for each of the following categories: (1) activity (inability to move limbs, ability to move 2 or 4 limbs with or without command); (2) respiration (evidence of apnea, labored breathing or normal breathing pattern); (3) circulation (blood pressure compared with baseline before sedation: ± 50% with reference to baseline, ± 20-50% to baseline, ± 20% to baseline); (4) consciousness (hypnotic, arousable or fully awake); and, (5) skin color (cyanotic, pink, or normal). Complete recovery is indicated by the maximum PARS of 10 points.23 Tolerability of the procedure was noted by the patient 4 hours after ERCP using a visual analog scale (0-10 points). Sedation regimens After a loading dose of 40 mg (<70 kg body weight) or 60 mg (≥70 kg body weight) of propofol (Disoprivan 1%, Zeneca, Plankstadt, Germany) had been injected in both groups, repeated doses of 20 mg of propofol were given intravenously to maintain an adequate level of sedation VOLUME 56, NO. 6, 2002
EEG monitoring for sedation during ERCP
(with no limit on total dose). No other medication was used for induction of sedation. An assisting physician (who was not involved in the endoscopic procedure and who had special training in intensive care medicine) always performed the sedation procedure. For patients in group A (conventional sedation technique) the assisting physician ensured that the patient tolerated the procedure well and that vital functions were unimpaired (heart rate ≥50/minute, oxygen saturation ≥90%, systolic pressure ≥90 mm Hg) by titrating repeated propofol boluses cautiously and increasing the oxygen supply to 4 to 6 L per minute if oxygen saturation decreased to less than 90%. A small preamplifier and a unit for analysis and display of EEG data was used (Narcotrend, MT MonitorTechnik, Bad Bramstedt, Germany). This device was developed by one of the investigators (AS) and coworkers at Hanover Medical School. The EEG was recorded by means of 3 self-adhesive ECG electrodes placed on the forehead of the patient. The system performed automated analysis of EEG segments of 20 seconds’ duration (20-second epochs). After extensive artifact analysis, the EEG epochs are automatically classified by multivariate statistical procedures by using a scale from A (awake) to F (very deep sleep) with 14 substages (adopted from a visual EEG classification system proposed by Kugler22,24). The algorithm includes adaptations for agerelated EEG changes from childhood to old age.22 The main features of stages A to F are illustrated in Table 2 and Figure 1. Validation studies have shown that the Narcotrend provides an accurate assessment of hypnotic depth compared with conventional EEG analysis during propofol sedation.25,26 Before the present study, this device was used during 85 interventional endoscopic procedures and the sedation levels D0 to D2 were found to be optimal with regard to patient tolerance and avoidance of cardiorespiratory side effects.25 Therefore, EEG stages D0 to D2 were considered the target level. EEG registration was performed in group A but not accessible to the physician who administered the propofol; the EEG data were stored for later analysis. In group B, EEG analysis was displayed online on a personal computer monitor screen (Fig. 2) and the physician responsible for the sedation titrated the repeated propofol doses to achieve and maintain the target sedation level of D0 to D2. The endoscopist as well as the patient were blinded to the respective propofol titration technique used. Statistics All data are given as mean (standard deviation). For statistical comparison the Wilcoxon signed rank test, the Mann-Whitney rank sum test, and the Fisher exact test were used whenever appropriate, and a p value of < 0.05 was regarded as significant. Data analysis was performed on a personal computer with statistical software (Instat, Graph Pad, San Diego, Calif).
RESULTS There were no significant differences between the groups with regard to special patient characteristics VOLUME 56, NO. 6, 2002
T Wehrmann, J Grotkamp, N Stergiou, et al.
Figure 1. Characteristics of EEG stages A through F according to Kugler.24
that might have impacted co-operation under sedation or the performance of associated or interventional procedures (Tables 3 and 4). Only 6 of the ERCP procedures in group A (15%) and 8 in group B (20%) were performed for diagnosis. The mean duration of the endoscopic procedures was similar in both groups (group A, median 35 minutes, range 1784 minutes; group B, median 37 minutes, range 1891 minutes). Propofol dosage The mean propofol dose administered was 374 (166) mg (range 90 to 920 mg; median 360 mg) in group A and 290 (158) mg (range 60 to 740 mg; median, 300 mg) in group B (p = 0.02). Similarly, the mean propofol dose per kilogram of body weight GASTROINTESTINAL ENDOSCOPY
819
T Wehrmann, J Grotkamp, N Stergiou, et al.
EEG monitoring for sedation during ERCP
Figure 2. Photograph of PC monitor screen of Narcotrend device displaying raw EEG curve (top), EEG analysis (bottom right and left) and respective EEG stage (bottom middle), in this case C0.
Table 3. Patient characteristics with respect to potential impact on quantity and quality of sedation Group A (n = 40) Alcohol abuse (n) Regular use of sedatives or psychotropic drugs (n) Regular use of narcotics (n) Regular use of non-narcotic analgesics (n) Prior difficulty with sedation during endoscopic examination (n)
Group B (n = 40)
10 9
9 7
2 14
3 12
3
2
Differences between both groups are not significant. Patients in group A underwent ERCP under sedation with propofol based on clinical monitoring alone and patients in group B with the aid of additional EEG monitoring.
related to the duration of the procedure was significantly different between the groups (0.17 [0.10] mg propofol/kg/min in group A vs. 0.13 [0.08] mg/kg/min in group B; p < 0.01). As comedication, 20 mg of Nbutylscopolamine was required in 2 patients of group A and in 3 patients of group B (not significant [ns]). Sedation efficacy Patient co-operation, as rated by the endoscopist, was similar for both groups (ns; Table 5). Also, patient assessment of procedure tolerability was similar with either sedation regime (Table 5). 820
GASTROINTESTINAL ENDOSCOPY
Table 4. Endoscopic procedures performed in the 2 study groups Cholangiography Pancreatography Biliary sphincterotomy Pancreatic sphincterotomy Sphincter Oddi manometry Stone extraction Balloon dilation Biliary endoprosthesis Pancreatic endoprosthesis Intraductal US Miscellaneous
Group A (n = 40)
Group B (n = 40)
37 11 19 4 6 12 3 12 3 6 6
36 12 16 5 6 11 2 13 4 5 5
Vital signs and adverse effects The mean increase in heart rate (initial heart rate vs. maximum heart rate during procedure, as measured by continuous ECG recording) was not significantly different between group A and group B (Table 5). A temporary decrease in heart rate below 50 per minute occurred in 1 patient in each group. The mean decline in oxygen saturation (initial oxygen saturation vs. lowest oxygen saturation during procedure) was not significantly different between the groups (Table 5) and a potentially harmful decrease in oxygen to less than 90% was observed with similar frequency in both groups (5/40 patients in group A vs. 4/40 patients in group B; ns). A brief decrease in oxygen saturation to less than 85% was recorded for 2 patients in each group VOLUME 56, NO. 6, 2002
EEG monitoring for sedation during ERCP
T Wehrmann, J Grotkamp, N Stergiou, et al.
Table 5. Parameters of sedation efficacy and alterations of vital signs in 2 study groups Group A (n = 40) Patient co-operation 8.6 (1.5) (7-10) (points) Patient tolerance (points) 8 (1) (6-10) Increase in heart rate (%) 5 (3) (1-8) Decrease in mean arterial 14 (7) (8-22) blood pressure (mm Hg) Decline in oxygen saturation 6 (4) (1-11) (%)
Group B (n = 40) 8.8 (1.6) (7-10) 9 (1) (6-10) 4 (3) (1-8) 11 (6) (5-19)* 5 (3) (1-10)
All data given as mean (SD) (ranges). Differences between both groups are not significant except for *p < 0.05. Patients in group A underwent ERCP under sedation with propofol based on clinical monitoring alone and patients in group B with aid of additional EEG monitoring.
(ns). All of the latter patients responded rapidly to an increase of supplemental oxygen. Assisted bag ventilation because of prolonged apnea, as has been reported with propofol sedation,6-8 was not required for any patient in either group. Mean decrease in mean arterial blood pressure was 14 (7) mm Hg in group A and 11 (6) mm Hg in group B (p < 0.05; Table 5). A decrease in systolic blood pressure below 90 mm Hg, however, was observed in only 1 patient in group A versus none in group B (ns). Post-ERCP pancreatitis, defined according to criteria proposed by Cotton et al.,27 was recorded in 2 of 40 patients (5%) in group A (both cases mild) and 3 of 40 patients (8%) in group B (2 mild, 1 moderate; ns vs. group A). There were no other immediate complications of the endoscopic procedures (e.g., perforation or sepsis). Recovery The mean time until patients were fully alert after the endoscopic procedure was 20 (8) minutes (range 6-36 minutes) in the patients who had conventional propofol-titration (group A) but 16 (7) minutes (range 5-27 minutes) in those with EEGassisted propofol titration (group B)(p = 0.02; Fig. 3). Also, the postanesthesia recovery score (30 minutes after completion of procedure) tended to be better in group B (8.4 [1.0]; range 6-10) than in group A (8.0 [0.9]; range 6-10), but the difference failed to reach statistical significance (p = 0.06; Fig. 3). EEG analysis The respective proportion of time during which patients in both groups remained in the different EEG stages is displayed in Table 6. The predefined target level of sedation (D0 to D2) was maintained during 75% of the procedure time in group B but in VOLUME 56, NO. 6, 2002
Figure 3. Box-Whisker plot of time until patients are fully alert after endoscopic procedure (recovery time, top) and postanesthesia recovery score (PARS; obtained 30 minutes after completion of ERCP) in 80 patients randomly assigned for titration of propofol sedation with (group B) or without EEG-monitoring (group A). *p < 0.05 vs. group A.
only 58% of time in group A (p < 0.05). Deeper sedation levels were reached significantly more often in the patients of group A versus those in group B (stages E-F, 25% in group A vs. 11% in group B; p < 0.05). Individual examples of recordings from each group are shown in Figure 4. DISCUSSION The duration of ERCP, now largely a therapeutic procedure, has a wide span, but is sometimes more than 1 hour, and long duration can lead to respiratory and cardiovascular changes. Therefore, appropriate levels of sedation must be achieved while averting the potentially adverse effects of sedative drugs. The need for proper sedation during ERCP is GASTROINTESTINAL ENDOSCOPY
821
T Wehrmann, J Grotkamp, N Stergiou, et al.
EEG monitoring for sedation during ERCP
Table 6. Respective proportion of time during which patients remained in different EEG stages EEG stage
Group A (n = 40)
Group B (n = 40)
A B C D E F
4% 1% 12% 58% 23% 2%
5% 1% 8% 75%* 10%* 1%
*p < 0.05 vs. group A.
Figure 4. EEG stages (i.e., level of sedation) over time during ERCP in a patient from group A (propofol titration based on clinical data) (top), compared with EEG stages for a patient from group B (propofol administration based on additional EEG monitoring) (bottom). Note that the target level (D) of sedation was achieved more constantly in the patient from group B.
particularly obvious when technically demanding interventional procedures are performed. Three randomized controlled studies have shown that sedation efficacy is greater and patient recovery shorter when ERCP is performed with the patient under sedation with propofol compared with the traditional midazolam.7,8,10 In addition, a significantly higher success rate was found for performance of endoscopic sphincter of Oddi manometry under propofol sedation when compared with a historical control group sedated with midazolam.6 Propofol, however, causes more severe cardiorespiratory side effects and repeated administration is necessary because of its short half-life. Therefore, comprehensive patient monitoring (e.g., pulse oxymetry, ECG, and automated blood pressure recording) is required and propofol administration has to be titrated carefully to avoid oversedation. In previous studies, which were based on subjective parameters 822
GASTROINTESTINAL ENDOSCOPY
of the desired level of sedation, there were large variations in the initial and repeated dosages of propofol given for sedation during endoscopic procedures.1-9 It is known that there is a great individual variability in the drug levels needed to achieve certain desired (or undesired) effects. Usually the effect of the sedation is assessed by observing patient awareness, compliance with the procedure, pain reactions, and reflex status. However, these parameters are not easy to monitor in a darkened procedure room by a physician who is usually not positioned near the patient’s head (a position routinely occupied by the endoscopist and assisting nurse). With use of a drug like propofol, primarily a hypnotic agent, in these circumstances, the level of sedation is especially difficult to assess by clinical observation and cardiorespiratory parameters alone.19 EEG-guided sedation is being used increasingly by anesthesiologists to achieve exact titration of hypnotic agents.14-21 For general anesthesia, it has been shown that EEG guidance avoids both underdosing and overdosing and is valuable for detecting potentially harmful situations for the brain (e.g., hypoxic episodes). Visual EEG classifications, as proposed originally by Kugler,24 have been used in several studies with different hypnotic drugs.17,28 For practical purposes in GI endoscopy, automated analysis rather than visual interpretation of the EEG recording (which is time consuming and requires special knowledge) is desirable. Numerous quantitative parameters have been studied for their ability to describe and characterize the EEG. Spectral parameters such as the median or the spectral edge frequency have been used most often.10,18,29-31 However, in many cases the complex EEG signal cannot be adequately characterized by a single parameter.26,32 Thus, for the classification algorithms in the Narcotrend device used in the present investigation, a multiparametric approach with statistical discriminant analysis including several parameters derived from the EEG is applied. A validation study demonstrated that the Narcotrend device provides a reliable description of the EEG.25 The present study demonstrates that EEGguided propofol sedation during ERCP is associated VOLUME 56, NO. 6, 2002
EEG monitoring for sedation during ERCP
with a significant lowering of total propofol dose and, accordingly, more rapid patient recovery. Further, the decline in mean arterial blood pressure was less pronounced in the patients in whom propofol administration was EEG-guided. This is due to more constant achievement of a proper level of sedation (Fig. 4, Table 6) with the aid of EEG recording. This is also reflected by the fact that, although the total dose of propofol administered was significantly lower in the patient group in which EEG monitoring was applied, sedation efficacy was not lowered compared with that for patients in whom propofol administration was based on clinical monitoring alone. Struys et al.33 documented that a closed-loop controlled administration of propofol with an EEG parameter as the control variable is clinically acceptable for general anesthesia and associated with better control of arterial blood pressure than conventional propofol dosing. EEG monitoring in the present study did not, in general, affect patient safety or sedation efficacy (with the exception of the influence on the blood pressure). With respect to cost-effectiveness, the additional costs of EEG monitoring (in Germany the Narcotrend device is available for about 10,000 U.S. dollars) must be balanced against the reduced need for propofol and faster patient recovery. In the present study, there was a saving of $3.00 per procedure (corresponding to 60 mg of propofol). Therefore, the addition of EEG monitoring seems to increase costs. However, avoidance of oversedation because of propofol administration (e.g., in patients with hepatic encephalopathy or patients taking neuroleptic drugs in whom propofol dosing must be adjusted) is a worthwhile goal irrespective of cost considerations. The number of patients in the present study is too small to allow evaluation of differences in the occurrence rates of rare life-threatening complications, for example clinically relevant hypoxemia necessitating assisted ventilation, as have been reported.68 Furthermore, for study purposes patients who required emergency ERCP, the subgroup at greatest risk for complications, were excluded. In conclusion, the results of the present study demonstrated that additional EEG monitoring results in more effective and economic titration of propofol administration for sedation during interventional endoscopic procedures. DISCLOSURE During the study period a prototype EEG monitor Narcotrend was given as a grant from the Department of Anesthesiology, Medical University of Hanover. A. Schultz is a coinventor and holds the patent, “Method and device for evaluating an EEG carried out in the context of anesVOLUME 56, NO. 6, 2002
T Wehrmann, J Grotkamp, N Stergiou, et al.
thesia and intensive care” but he has not received any license fees. None of the investigators have commercial associations with the Narcotrend device, produced by MT Monitor Technik, Bad Bramstedt, Germany (i.e., equity ownership or interest, consultancy or institutional and corporate associations). REFERENCES 1. Gepts E, Claeys MA, Camu F, Smekans L. Infusion of propofol as sedative technique for colonoscopies. Postgrad Med J 1985;61:120-6. 2. Dubois A, Balatoni E, Peters JP, Baudoux M. Use of propofol for sedation during gastrointestinal endoscopies. Anesthesia 1988;43:75-80. 3. Pohlmann S, Herden HN, Hagenmueller F. Propofol-Narkose fuer die Endoskopie-gefaehrlicher als Midazolam? [in German with English abstract]. Bildgebung 1993;60(suppl):61-3. 4. Kostash MA, Johnston R, Bailey RJ, Konopad EM, Guthriern LP. Sedation for colonoscopy: a double-blind comparsion of diazepam/meperidine, midazolam/fentanyl and propofol/fentanyl combinations. Canad J Gastroenterol 1994;8:27-31. 5. Hofmann C, Kiesslich R, Brackertz A, Jung M. Propofol for premedication in gastroscopies—a randomized comparison to midazolam. Z Gastroenterol 1999;37:589-95. 6. Schmitt Th, Seifert H, Dietrich CF, Caspary WF, Wehrmann T. Intravenous sedation with propofol during endoscopic sphincter of Oddi manometry. Z Gastroenterol 1999;37:219-27. 7. Wehrmann T, Kokabpick S, Lembcke B, Caspary WF, Seifert H. Efficacy and safety of intravenous propofol sedation during routine ERCP: a prospective, controlled study. Gastrointest Endosc 1999;49:677-83. 8. Jung M, Hofmann C, Kiesslich R, Brackertz A. Improved sedation in diagnostic and therapeutic ERCP: propofol is an alternative to midazolam. Endoscopy 2000;32:233-8. 9. Koshy G, Nair S, Norkus EP, Hertan HI, Pitchumoni CS. Propofol vs. midazolam and meperidine for conscious sedation in GI endoscopy. Am J Gastroenterol 2000;95:1476-9. 10. Krugliak P, Ziff B, Rusabrov Y, Rosenthal A, Fich A, Gurman GM. Propofol vs. midazolam for conscious sedation guided by processed EEG during endoscopic retrograde cholangiopancreatography: a prospective, randomized double-blind study. Endoscopy 2000;32:677-82. 11. Reimann FM, Samson U, Derad M, Fuchs M, Schiefer B, Stange EF. Synergistic sedation with low dose midazolam and propofol for colonoscopies. Endoscopy 2000;32:239-44. 12. Sebel PS. Propofol: a new intravenous anaesthetic. Anaesthesiology 1989;71:260-77. 13. McCollum JSC, Dundee JW, Halliday NJ, Clarke RSJ. Dose response studies with propofol in unpremedicated patients. Postgrad Med J 1985;61:85-7. 14. Cockshott ID, Briggs LP, Douglas EJ, White M. Pharmacokinetics of propofol in female patients: studies using single bolus injection. Br J Anaesth 1987;59:1103-10. 15. Searle NR, Sahab P. Propofol in patients with cardiac disease. Can J Anaesth 1993;40:816-8. 16. Doenicke AW, Kugler J, Schellenberger A, Gürtner T. The use of electroencephalography to measure recovery time after intravenous anaesthesia. Br J Anaesth 1966;38:580-90. 17. Doenicke AW, Löffler B, Kugler J, Suttmann H, Grote B. Plasma concentration and EEG after various regimens of etomidate. Br J Anaesth 1982;54:393-400. 18. Schwilden H, Stoeckel H. Quantitative EEG analysis during anaesthesia with isoflurane in nitrous oxide at 1.3 and 1.5 MAC. Br J Anaesth 1987;59:738-45. GASTROINTESTINAL ENDOSCOPY
823
T Wehrmann, J Grotkamp, N Stergiou, et al.
19. Heier T, Steen PA. Assessment of anaesthesia depth. Acta Anaesthesiol Scand 1996;40:1087-100. 20. Smith WD, Dutton RC, Smith NT. Measuring the performance of anesthetic depth indicators. Anesthesiology 1996; 84:38-51. 21. Guérit J-M. Neuromonitoring in the operating room: why when, and how to monitor? Electroencephalogr Clin Neurophysiol 1998;106:1-21. 22. Schultz B, Schultz A, Grouven U. Sleeping stage based systems (Narcotrend). In: Bruch HP, Köckerling F, Bouchard R, Schug-Paâ C, editors. New aspects of high technology in medicine. Bologna: Monduzzi Editore; 2000. p. 285-91. 23. Kankaria A, Lewis JH, Ginsberg G, Gallagher J, Al-Kawas F, Nguyen CC, et al. Flumazenil reversal of psychomotor impairment due to midazolam or diazepam for conscious sedation for upper endoscopy. Gastrointest Endosc 1996;44: 416-21. 24. Kugler J. Elektroenzephalographie in Klinik und Praxis. Stuttgart: Thieme; 1981. p. 120-46. 25. Schultz B, Grouven U, Schultz A. Automatic classification algorithms of the EEG-monitor Narcotrend for routinely recorded EEG-data from general anaesthesia: a validation study. Biomed Eng 2002;47:9-13. 26. Schultz A, Grouven U, Schultz B. SEF90, SEF95 and the median do not discriminate deep EEG stages during propofol/ remifentanil anaesthesia [abstract]. Eur J Anaesthesiol 2000;17(Suppl 10):22.
EEG monitoring for sedation during ERCP
27. Cotton PB, Lehman GA, Vennes J, Geenen JE, Kozarek RA, DiMagno EP. Endoscopic sphincterotomy complications and their management. An attempt at consensus. Gastrointest Endosc 1991;37:383-93. 28. Doenicke AW, Roizen MF, Rau J, O’Connor M, Kugler J, Klotz U, et al. Pharmacodynamics of propofol in a new solvent. Anesth Analg 1997;85:1399-403. 29. Drummond JC, Brann CA, Perkins DE, Wolfe DE. A comparison of median frequency spectral edge frequency a frequency band power ratio, total power and dominance shift in the determination of depth of anaesthesia. Acta Anaesthesiol Scand 1991;35:693-9. 30. Schwender D, Daunderer M, Mulzer S, Klasing S, Finsterer U, Peter K. Spectral edge frequency of the electroencephalogram to monitor “depth” of anaesthesia with isoflurane or propofol. Br J Anaesth 1996;77:179-84. 31. Gurman GM, Fajer S, Porat A, Rusabrov Y. Use of EEG spectral edge as an index of equipotency in a comparison of propofol and isoflurane for maintenance of general anesthesia. Eur J Anaesthesiol 1994;11:443-8. 32. Levy WJ. Intraoperative EEG patterns: implications for EEG monitoring. Anesthesiology 1984;60:430-40. 33. Struys MM, De Smet T, Versichlen LF, Van de Velde S, Van den Brocke R, Mortier EP. Comparison of closed-loop controlled administration of propofol using bispectral index as the controlled variable versus “standard practice” controlled administration. Anesthesiology 2001;95:6-17.
Availability of Journal back issues As a service to our subscribers, copies of back issues of Gastrointestinal Endoscopy for the preceding 5 years are maintained and are available for purchase from Mosby until inventory is depleted. Please write to Mosby, Subscription Customer Service, 6277 Sea Harbor Dr., Orlando, FL 32887 or call 800-654-2452 or 407-345-4000 for information on availability of particular issues and prices.
824
GASTROINTESTINAL ENDOSCOPY
VOLUME 56, NO. 6, 2002