IV sedation is a useful technique in ophthalmic surgery. Sedation helps to keep the patient relaxed and comfortable during both the placement of a local anesthetic and surgery. It may also produce postoperative amnesia. However, during ophthalmic surgery excessively deep levels of sedation may lead to sudden head movements, hampering surgery with potentially disastrous consequences if the eye is damaged (1). Hence, IV sedation in eye surgery has been slow to gain acceptance. Although propofol effectively induces sedation with moderate reductions in intraocular pressure (2–4), and a low incidence of side effects, such as nausea and vomiting (5,6), the desired level of sedation may be difficult to maintain with sedation scores varying from minimal sedation to general anesthesia. A reliable and objective continuous measurement of the level of consciousness might make it easier to more accurately maintain the appropriate level of sedation. Bispectral index (BIS), a derived electroencephalograph (EEG) parameter that has been extensively validated as a monitor for the depth of hypnosis and middle-latency auditory-evoked potentials (AEP), that quantify the level of consciousness by measuring the cortical response to acoustic stimuli, have both been reported to be objective continuous measures of the adequacy of anesthesia and predictors for loss of consciousness (7–9). One commercially available monitor records middle-latency AEP and converts the waveform into an index: Alaris AEP index (AAI). The current study was designed to assess the performance of BIS, AAI, and the combination of BIS and AAI as a guide to propofol administration for sedation in ophthalmic surgery. The aim was to achieve light to moderate levels of sedation with particular emphasis on prevention of head movement. In addition, we paid special attention to how well the index of each device could be maintained within the target range.
One-hundred adult patients, ASA physical status I/II/III, consecutively scheduled for elective eye surgery (cataract, glaucoma, or retina) were enrolled in the study, which was approved by the Institutional Medical Ethics Committee of the Academic Medical Center, University of Amsterdam. Written informed consent was obtained from all participants.
Pregnant patients, patients with neuro-sensory hearing loss, unable to give informed consent, swallowing disorders, patients taking medications that affected the central nervous system, patients with dementia, and previous experience with IV sedation were all excluded from the study. All patients were preoperatively screened according to the ASA guidelines for general anesthesia, were instructed to fast overnight and did not receive premedication with oral benzodiazepines or analgesics. An anesthetic nurse explained the sedation procedure, the local anesthetic block of the eye, the intraoperative procedure, and the postoperative phase in the postanesthesia care unit.
Patients were randomly allocated to one of four groups upon arrival in the operating room (OR): BIS-guided, AAI-guided, BIS/AAI-guided, and clinically guided (n = 25 per group). A sealed envelope was opened which contained a computer generated randomization assignment and research number. None of the investigators had prior knowledge of the randomization scheme.
BIS scores were measured using an Aspect 2000 ™ BIS XP monitor, software version 3.12 (Aspect Medical Systems, Natick, MA). The smoothing time of the BIS was set at 15 s. The EEG was recorded using an XP BIS “Quatro” sensor secured on the prepared skin of the forehead as recommended in the Aspect Medical Systems BIS XP manual. The AEP Monitor/2 uses an algorithm to express the amplitude and the latency of the middle-latency AEP to provide a simple numerical index (AAI) that ranged between 0 and 100. The AEP Monitor/2, A-line ™ ARX index, (Danmeter A/S, Odense, Denmark) (Software version 1.6 and AAI version 4.2) was used to obtain both the AEP and the AAI Index. Earphones connected to the device provided the auditory stimulus: the AEP was elicited with a bilateral click stimulus of 70 dB intensity, 2 ms duration, and repetition rate of 9 HZ. Three silver-silver chloride electrodes (Danmeter, Odense, Denmark) were positioned on the prepared skin of the middle forehead (+), left forehead (ref.), and left mastoid (−) as recommended by the manufacturer. Processing time for the AAI was 30 s for the first signal, with total update delay of 10 s. Electrode impedance for both monitors was kept <5 kΩ. Every patient had both BIS sensors and AEP electrodes on the forehead and the anesthesiologist providing sedation was blinded to the nonallocated device.
The level of consciousness as determined on the 5-point Observer’s Assessment of Alertness Sedation Scale (OAAS) (Table 1) was intermittently monitored and recorded every 3 min.
A 5-lead electrocardiogram with automated ST segment evaluation, heart rate, noninvasive arterial blood pressure, and oxyhemoglobin saturation (Spo2) by pulse oximeter were continuously monitored (Hewlett Packard Component Monitoring System, Böblingen HP, Germany). Data files containing hemodynamic, oxygen saturation, processed EEG parameters, and electromyographic activity (EMG) of both EEG devices were logged automatically on all patients by Rugloop© vII (DEMED, Temse, Belgium) for subsequent analysis.
Once baseline recordings of the hemodynamic variables, Spo2, and BIS/AAI data were obtained, propofol 1% (Astra Zeneca, Netherlands) was administered using an infusion device (Graseby 3400) connected through an RS-232 serial interface to a laptop computer. The algorithm used for propofol infusion was written in GFA Basic for Windows by one of the coauthors. The implemented pharmacokinetic model and pump mechanical performance were identical to those of the Diprifusor© (Astra Zeneca, Graseby 3500, Macclesfield, Cheshire, UK). It uses the pharmacokinetic model described by Marsh et al. (10) to deliver propofol at a targeted blood concentration. The keo implemented into our system was 0.29 min−1 (11). Propofol was infused at a preset effect site target concentration of 0.5 μg · mL−1 in patients older than 70 yr and 1.0 μg · mL−1 in all other patients. The propofol concentration was increased every 3–5 min in steps of 0.1 or 0.2 μg · mL−1, respectively. The effect site target concentration was adjusted until the patient had reached a BIS value of 75 or an AAI of 40 and maintained during surgery between 70–90 (BIS) or 30–60 (AAI). In the combined BIS/AAI-guided group, the indices were kept within these ranges as well. The propofol effect site concentration was decreased when the patient became unresponsive or uncooperative and the procedure temporarily halted. In the non-BIS/AEP-guided group, propofol was administered until the patient had reached Level 3 of the OAAS/S. At this level of sedation, patients typically have lost anxiety and apprehension for the treatment procedure, are cooperative and relaxed, adequately responding to verbal commands or physical stimulus (i.e., gentle hand squeeze).
Supplemental oxygen at 2 L/min was administered via a nasal cannula in all patients. No opioids were given. When the patient was adequately sedated (clinically and BIS values of 75 (range 70–90) or AAI of 40 (range 35–60)), 5 mg S-ketamine 25 (Pfizer, Netherlands) was administered once as an analgesic adjunct, before peri-ocular local anesthetic infiltration which was given by the eye surgeon in the OR. After 3–5 min, 6–8 mL of an equal mixture of bupivacaine 0.5% and lidocaine 4% was injected with hyaluronidase (150 IU), using a 23 G, 1.5 in. Atkinson needle for the retrobulbar block (transcutaneous inferotemporal) or a 19 G, 1 in. subtenon anesthesia cannula (transconjunctival nasal).
All patients were kept normothermic with forced air (Warm Air Hyperthermia System-The Surgical Company, Cincinnati OH).
At the end of surgery, the propofol infusion was discontinued and the patient was allowed to recover in the postanesthesia care unit. One hour later, the patient was interviewed and discharged to the ward or home.
After the operation, the surgeon (who was blinded for the EEG device) assessed operating conditions, according to the following criteria: the amount of movement during penetration of the needle and during surgery, falling asleep with sudden awakening, need for verbal encouragement to lie still, bleeding, pain, patient talking, and cooperation. For each criterion, 5 or more occurrences were scored as 3, (maximum score: 21, worst operating conditions), 3–4 events as 2, 1–2 events as 1, and no event as 0 (ideal operating conditions).
To assess the patient’s experiences regarding the surgical procedure, a questionnaire was administered 1 h after surgery, which included the following items: present experience with the local anesthetic block (apprehension, pain), pain during the surgical procedure, recall of intraoperative events such as remembering the peri-ocular block, pain, conversation, and sound of instruments, satisfaction with the care provided (1 = uncomfortable and dissatisfied, 10 = very comfortable and satisfied), postoperative pain and emesis.
An initial sample size of 23 patients per group was calculated based on a 80% power to detect a difference in means of one standard deviation on a primary outcome measure between each pair of groups, assuming an overall two-sided significance level of 0.05 and correcting for 3 adhoc comparisons. Eventually, 25 patients per group were included to enable nonparametric testing, should this be necessary for reasons of non-normally distributed data.
Group comparisons were made by nonparametric Kruskal–Wallis tests for the total event score and surgical satisfaction, and by one-way analysis of variance for total propofol dose. In addition, differences in the proportions of out-of-range BIS and AAI values during sedation with only BIS-guided or AAI-guided monitoring respectively, were tested for significance with the Mann–Whitney U-test. P values <0.05 were considered significant.
For each patient the association between BIS values and EMG, between surgical satisfaction and event score during surgery were determined with the Pearson’s product-moment correlation coefficient (r). The mean r along with its 95% confidence interval indicated the overall association of BIS with EMG or surgical satisfaction with event score. Statistical analysis was performed with the SPSS package (Version 12; SPSS, Chicago, IL).
Neither orbital nor systemic complications occurred in any of the study groups after injection of the local anesthetic. There was no hyper/hypotension, dysrhythmia, obstructed airway, respiratory depression, or psychomimetic emergence reactions in any of the patients during sedation. Surgery proceeded as planned in every patient.
In four patients (one in the BIS-guided group; one in the BIS/AAI-guided group; two in the clinically guided group), propofol sedation was converted to general anesthesia because of excessive head movement. In these patients, the data collected during the sedation phase were included in the analysis. A man from the BIS-guided group was very fearful of the operation and remained anxious during sedation. He kept moving his head, talked and although he was frequently encouraged to lie still, remained uncooperative at the target BIS range. The total event score in this patient was 12. In the BIS/AAI-guided group, one woman with severe chronic obstructive pulmonary disease began breathing in such a labored manner during sedation that microsurgery was impossible due to the increased movement of the head. The total event score in this patient was 10. In the clinically guided group, two men were uncooperative, alternately asleep or talking, excessively moving their heads, and requiring continuous encouragement to lie still. Although the target effect site propofol concentrations were changed in both directions, the event scores did not decrease, impeding surgery. The OAAS/S was 4–3 and the corresponding total event scores were 16 and 13. Demographic and clinical characteristics for each group are shown in Table 2.
During insertion of the needle, or during administration of the local anesthetic, none of the patients showed unwanted head movement. All patients had either an OAAS/S 4 or 3.
Forty-eight patients recalled the block being administered, (14 in the BIS-guided, 11 in the AAI-guided, 9 in the BIS/AAI-guided, and 14 in the clinically guided group); 10 patients experienced discomfort and pain, but none of these patients reported pain during the surgery.
The AAI showed large variability and was out of the target range 58% of the time vs 7% for BIS. In the combined BIS and AAI-guided group, the AAI showed such a great variability that it was impossible to target the sedation level on the information from the AAI. Neither BIS nor AAI monitoring significantly improved the operating conditions from the surgeon’s standpoint, nor did guided intervention reduce the amount of propofol needed (Tables 3 and 4).
Seventy-five patients (19 in the BIS-guided, 16 in the AAI-guided, 20 in the BIS/AAI-guided, and 20 in the clinically guided group) were entirely pain-free during the operation. Additional local anesthetic was administered whenever needed in the other patients. Fourteen patients (three in the BIS-guided, four in the AAI-guided, five in the BIS/AAI-guided, and two in the clinically guided group) reported that they were completely unaware of the operation. Eighty-one patients (18 in the BIS-guided, 21 in the AAI-guided, 20 in the BIS/AAI-guided, and 22 in the clinically guided group), respectively remembered fragments of conversations. Eighty-nine patients had partial to no amnesia, but did not experience this as a problem (21 in the BIS-guided, 23 in the AAI-guided, 22 in the BIS/AAI-guided, and 23 in the clinically guided group). Ninety-four patients (22 in the BIS-guided group, 24 in the AAI-guided, 24 in the BIS/AAI-guided, and 24 in the clinically guided group) indicated a high preference to undergo future ophthalmic surgery under a local anesthetic with IV sedation. One patient preferred only a local anesthetic, three patients expressed a preference for general anesthesia, and two patients were indifferent about the technique. The auditory clicks did not cause discomfort for any of the patients.
Postoperative nausea, emesis, dizziness, confusion, pain, or agitation was not observed in any patient.
Figure 1 depicts the trend of the AAI, BIS, and EMG variable in patient No. 1. The BIS and EMG trends varied in parallel in all patients. The mean correlation (r) between BIS and EMG during surgery with propofol sedation was 0.46 (95% CI: 0.42–0.50) for all patients.
The mean Pearson correlation coefficient between surgical satisfaction and events score during surgery was −0.84 (95% CI: −0.77 to −0.89) for all patients.
In the present study, we compared BIS or AAI-guided titration of propofol for ophthalmic procedures with titration based on clinical observations only. We were unable to demonstrate a benefit of BIS or AAI guidance in terms of reduced incidence of unintended head movements, surgeon satisfaction with operating conditions, or patient satisfaction.
Propofol sedation for diagnostic and therapeutic procedures is growing in popularity. Careful titration of the sedative drug to the desired clinical end-point for that specific procedure and to the individual patient’s need is required. In ophthalmic surgery, for example, unexpected movement of the head because of excessive level of sedation is a dreaded scenario. Therefore, assessing the level of consciousness during sedation is essential in determining the dose of the sedative drug. We wished to determine whether cerebral monitoring can prevent the occurrence of over-sedation, e.g., to prevent patients from falling asleep with sudden awakening. When the study was designed, two fundamentally different monitoring devices were commercially available, BIS (Aspect Medical) and the AEP Monitor/2 (Danmeter). Both monitors use proprietary algorithms to convert the EEG and provide the observer with a dimensionless numerical index from 0 to 100, where 90–100 represents an awake EEG and 0 represents complete electrical silence (cortical suppression) (7–9).
AEP measurements have been reported to correlate well with consciousness transitions and the modified OAAS/S during anesthesia (8). The AEP index is a reliable guide to the depth of sedation and movement in response to several noxious stimuli (12). However, the manufacturer of the AEP/2 monitor does not recommend using the device for sedation. The AAI was highly variable in our study. It was out of the target range in 58% of cases. Similar high AAI variability was observed by Struys et al. (9) during propofol administration in sedative doses. The auditory-evoked response was contaminated by spontaneous muscle activity from the forehead (frontalis and corrugator muscles) and from muscle activity behind the ear (picked up from the electrode on the mastoid bone). This muscle artifact makes recording the middle latency auditory-evoked response difficult in patients or volunteers who are not paralyzed.
In our study, we used the BIS range guidelines reported by Johansen and Sebel (7) and Struys et al. (9) A cut-off value of 75 indicated a transition from light-to-moderate (light hypnotic state) to deep sedation level. We found that in 7% of cases BIS was under this cut-off value.
The accuracy of BIS monitoring in the OR and intensive care unit settings has been reported.
During light-to-moderate procedural sedation where patients maintain a certain degree of neuromuscular activity, cerebral monitoring has produced conflicting results. In the presence of high EMG activity, a high value of BIS need not correspond with a lighter level of sedation (Aspect BIS-XP manual). The effect of EMG activity on the BIS in different studies is inconclusive due to different software revisions. BIS-XP was designed to provide improved identification and filtering of electro-oculographic and EMG artifacts. EMG values more than the threshold of 42 dB (interpreted as high-frequency, low-amplitude waves) are considered to falsely increase BIS values, and Vivien et al. (13) have shown that high EMG activity was still present with the BIS-XP monitor. The overlap of EEG (0.5–30 Hz) and EMG (30–300 Hz band) signals is likely to remain a limitation to the success of BIS monitoring in the intensive care unit or in the OR in unparalyzed patients for light-to-moderate sedation (7,13,14). Renna et al. (15) administered two different fentanyl doses, each resulting in muscle rigidity, and observed that the correlation between EMG and BIS was much stronger than that between OAAS/S scores and BIS. They concluded that in clinical situations associated with high EMG activity, BIS becomes a highly EMG-dependent variable and, as a consequence, BIS variations may not correspond to variations in depth of sedation. High BIS values should be interpreted concurrently with the amount of EMG activity. The influence of neuromuscular blockade on BIS had been investigated (13,16–18). BIS decreased after the administration of a muscle relaxant in awake volunteers and sedated patients to values observed during deep anesthesia. However, in healthy young volunteers receiving mivacurium for neuromuscular blockade, BIS values and EMG activity were unaltered during propofol anesthesia (17).
In the current study, where patients were lightly to moderately sedated during surgery and breathing spontaneously without an airway device, a correlation of 0.46 between EMG and BIS suggests that BIS values are influenced by EMG to such an extent that they may not be accurate enough to assess the level of consciousness.
Ninety-four percent of patients in the present study indicated a strong preference to undergo future eye operations under a regional block combined with propofol sedation during surgery. However, most studies reported on propofol sedation administered selectively during the placement of the local anesthetic block (19). One reason why ophthalmic surgeons have been reluctant to continue sedation throughout the surgical procedure is fear of unwanted head movement as a result of over-sedation. The data from the present study suggest that continuing sedation throughout the ophthalmic surgical procedure is feasible, provided that the level of sedation is carefully monitored by an experienced sedation provider. There is no evidence that adding cerebral monitoring enhances the maintenance of the desired sedation level.
In 4% of the procedures, sedation was converted to general anesthesia. These patients were all adequately sedated, but started to move their heads as soon as the surgeon initiated surgery. They were uncooperative, although none complained of pain. All were amnesic for the sedation period. They were disappointed when told that they had received a general anesthetic. Janzen et al. (20) reported that two of their 20 patients using patient-controlled sedation with propofol showed significant head movement, probably attributable to disorientation during emergence from sedation.
Subanesthetic doses of ketamine provide analgesia without ventilatory impairment. However, side effects, such as hypertension and psychomimetic emergence reactions, limit its use as the sole sedative drug. The combination of propofol and ketamine has been shown to attenuate hemodynamic responses and to prevent ketamine-induced emergence reactions. Ten patients in our study group experienced pain during the administration of the block, suggesting that the ketamine dose was insufficient to provide adequate analgesia for these patients. Frey et al. used higher doses of ketamine in their study (5).
In the current study, we administered a single bolus dose of ketamine 5 mg before the regional block. Reports of the influence of ketamine on the EEG are contradictory. With a bolus of ketamine (1 mg · kg−1) administered as the single drug, the AAI increased significantly, possibly because of increased muscle tone (21). However, a bolus of ketamine (0.4 mg · kg−1) followed by a continuous infusion (1 mg · kg−1 · h−1), in combination with propofol delivered by a target-controlled infusion, did not cause significant changes of the AAI, whereas BIS increased for a period of 3–8 min after the ketamine bolus (22). Another study reported that BIS did not change during ketamine anesthesia, not even in combination with propofol (14). Our single 5 mg analgesic ketamine dose was 10–15 times smaller than the dose at which the aforementioned EEG and AEP changes were observed. Therefore we consider it unlikely that this single analgesic dose had a substantial effect on the observed BIS and AAI trends.
In conclusion, cerebral monitoring to guide propofol titration to sedative levels did not improve the quality of operating conditions when compared with sedation guided by clinical observation only.
The authors would like to thank the ophthalmic surgeons of the Academic Medical Center, University of Amsterdam, for participating in the study. The authors also thank Renee van Keule and Hetty van den Wall Bake (anesthetic nurses) for their dedicated assistance, and Eric J.Weber for providing the latest software version of the AEP monitor. We also thank the Dutch representatives of Aspect Medical Systems and the Alaris Medical Systems for the loan of their monitors.
1. Hamilton RC, Gimbel HV, Javitt JC. The prevention of complications of regional anesthesia for ophthalmology. Ophthalmol Clin North Am 1990;3:111–25.
2. Ferrari LR, Donlon JV. A comparison of propofol, midazolam and methohexital for sedation during retrobulbar and peribulbar block. J Clin Anesth 1992;4:93–6.
3. Polarz H, Bohrer H, von Tabouillot W, et al. Behavior of intraocular pressure in anesthesia with isoflurane in comparison with propofol/alfentanil. Anasthesiol Intensivmed Notfallmed Schmerzther 1995;30:96–8.
4. Neel S, Deitch R Jr, Moorthy SS, et al. Changes in intraocular pressure during low dose intravenous sedation with propofol before cataract surgery. Br J Ophthalmol 1995;79:1093–7.
5. Frey K, Sukhani R, Pawlowski J, et al. Propofol versus propofol–ketamine sedation for retrobulbar nerve block: comparison of sedation quality, intraocular pressure changes and recovery profiles. Anesth Analg 1999;89:317–21.
6. Weir PM, Munro HM, Reynolds PI. Propofol infusion and the incidence of emesis in pediatric outpatient strabismus surgery. Anesth Analg 1993;76:760–4.
7. Johansen JW, Sebel PS. Development and clinical application of electroencephalographic bispectrum monitoring. Anesthesiology 2000;93:1336–44.
8. Litvan H, Jensen EW, Revuelta M,et al. Comparison of auditory evoked potentials and the A-line ARX Index for monitoring the hypnotic level during sevoflurane and propofol induction. Acta Anaesthesiol Scand 2002;46:245–51.
9. Struys MM, Jensen EW, Smith W, et al. Performance of the ARX-derived auditory evoked potential index as an indicator of anesthetic depth: a comparison with bispectral index and hemodynamic measures during propofol administration. Anesthesiology 2002;96:803–16.
10. Marsh B, White M, Morton N, Kenny GN. Pharmacokinetic model driven infusion of propofol in children. Br J Anaesth 1991;67:41–8.
11. White M, Schenkels MJ, Englbers FH, et al. Effect-site modelling of propofol using auditory evoked potentials. Br J Anaesth 1999;82:333–9.
12. Kurita T, Doi M, Katoh T, et al. Auditory evoked potential index predicts the depth of sedation and movement in response to skin incision during sevoflurane anesthesia. Anesthesiology 2001;95:364–70.
13. Vivien B, Di Maria S, Quattara A, et al. Overestimation of bispectral index in sedated intensive care unit patients revealed by administration of muscle relaxant. Anesthesiology 2003;99:9–17.
14. Sakai T, Singh H, Mi WD, et al. The effect of ketamine on clinical endpoints of hypnosis and EEG variables during propofol infusion. Acta Anaesthesiol Scand 1999;43:212–16.
15. Renna M, Wigmore T, Mofeez A, Gilbe Ch. Biasing effect of the electromyogram on BIS: a controlled study during high-dose fentanyl induction. J Clin Monit 2002;17:377–81.
16. Bruhn J, Bouillon TW, Shafer SL. Electromyographic activity falsely elevates the bispectral index. Anesthesiology 2000;92:1485–7.
17. Greif R, Greenwald S, Schweitzer E, et al. Muscle relaxation does not alter hypnotic level during propofol anesthesia. Anesth Analg 2002;94:604–8.
18. Messner M, Beese U, Romstock J, et al. The bispectral index declines during neuromuscular block in fully awake persons. Anesth Analg 2003;97:488–91.
19. Habib NF, Balmer HG, Hocking G. Efficacy and safety of sedation with propofol in peribulbar anesthesia. Eye 2002;16:60–6.
20. Janzen PRM, Christys A, Vucevic M. Patient controlled sedation using propofol in elderly patients in day-case cataract surgery. Br J Anaesth 1999;82:635–6.
21. Nishiyama T. Changes in the auditory evoked potentials index by induction doses of four different intravenous anesthetics. Acta Anaesthesiol Scand 2005;49:1326–9.
© 2006 International Anesthesia Research Society
22. Vereecke HEM, Struys MMR, Mortier EP. A comparison of bispectral index and ARX-derived auditory evoked potential index in measuring the clinical interaction between ketamine and propofol anaesthesia. Anaesthesia 2003;58:957–61.