Ocular microtremor (OMT) is a high-frequency, low-amplitude physiologic tremor of the eye that originates in brainstem neurons (1) and is present in all individuals, even when the eye appears at rest or in the primary position (1–11). The frequency of OMT is decreased in patients with neurologic disorders involving the brainstem (7,12), including those with head injury, in whom it correlates with level of consciousness and prognosis (5,6). Both IV (11,13) and volatile (10) anesthetics reversibly decrease OMT frequency, suggesting that OMT may be useful as a measure of anesthetic depth. In a previous study involving 30 patients anesthetized with sevoflurane, we reported that OMT predicted the response to verbal command in patients emerging from anesthesia (10).
Movement in response to a painful stimulus is often used as an indicator of inadequate anesthetic depth and is the basis of the minimum alveolar anesthetic concentration (MAC) of volatile anesthetics (14). This movement is governed by subcortical structures (15) and, although an electroencephalograph (EEG)-based technique may predict movement in anesthetized patients (16), most reports found EEG-based techniques to be poorly predictive (17–21). However, the auditory evoked potential index relies on subcortical pathways and may be more accurate (22). Because OMT is related to tonic activity in subcortical neurons (23), it might also predict movement in anesthetized patients.
Previous studies of the effects of general anesthesia on OMT have relied on post hoc computer analysis of 30- to 60-s epochs of OMT wave forms (11) or on real-time measurements based on a 3-s epoch confirmed visually on an oscilloscope (10). Both of these techniques require cumbersome equipment and some operator experience to identify the characteristic OMT signal. The overall aim of this study was to evaluate an automated system of OMT signal analysis in a diverse population of patients undergoing general anesthesia with and without muscle relaxation. In a multicenter trial involving four centers in three countries, we examined the accuracy of OMT to identify the unconscious state and to predict movement in response to stimulation. We also tested the effects of neuromuscular blockade and patient position on OMT.
The study protocol was approved by the local ethics committees of the participating centers in Dublin (Ireland), Bristol (UK), St. Louis, MO, and Memphis, TN. Two-hundred-fourteen adult patients scheduled to undergo surgery other than intracranial or cardiothoracic procedures under general anesthesia gave written, informed consent to participate. Exclusion criteria were age <18 yr and history of ocular trauma or ocular surgery.
OMT was measured by using the closed-eye piezoelectric technique as described previously (24). Briefly, a piezoelectric element was mounted in a Perspex rod, and its tip was covered with silicone rubber (Fig. 1). The tip was placed gently over the closed eyelid and held in position by tape. The signal from the sensor was amplified, high-pass-filtered at 10 Hz, and low-pass-filtered at 150 Hz. OMT is a continuous oscillation with a displacement of the eye surface of approximately 150–2000 nm (24). In a clinical study, OMT frequency and amplitude were determined by visual inspection of an OMT trace displayed on an oscilloscope (11). In this study, calculation of OMT frequency and amplitude was performed by an electronic module. The signal frequency (peaks in oscillations per second) and amplitude were displayed on a light-emitting diode display. Data capture was every 0.5 s, and the module displayed an average of 5 recordings from a 2.5-s epoch.
The anesthesiologist was blinded to the measured OMT. OMT recordings were taken at the following time points: before induction, after the induction of anesthesia (loss of response to verbal command), before and 30 s after laryngeal mask airway (LMA) insertion or laryngoscopy/intubation, before and after surgical incision, during maintenance of anesthesia (unstimulated and during surgical stimulation), and at first return of response to response to verbal command. Limb or head movement response within 30 s of LMA insertion or surgical incision was noted. Any visible limb movement, including flexor or withdrawal responses, coughing, or gagging, was considered movement. The eye probe was removed before changes in patient positioning, e.g., from supine to prone, and then reapplied. In 100 patients (in 1 center: Dublin), OMT readings were additionally recorded continuously (i.e., automated recording of 2.5-s epochs) throughout each case. Notable events were marked on the recording, including the administration of neuromuscular blocking drugs and anticholinesterases, changes in the train-of-four (TOF) ratio to peripheral nerve stimulation, and patient movement.
Anesthetic technique was not specified by the study protocol. Most patients received sedative premedication, either 1 h before surgery or on arrival in the holding area. The baseline OMT signal for these patients was therefore made after the administration of such drugs. Anesthesia was induced with propofol in all patients and was maintained with sevoflurane (n = 124), isoflurane (n = 71), or desflurane (n = 11) in nitrous oxide and oxygen. One-hundred-thirty-one patients received neuromuscular blocking drugs. Neuromuscular blockade was assessed by visual assessment of response to TOF stimulation. Return of four twitches of TOF was confirmed before the first assessment of movement response to a verbal stimulus at the end of anesthesia.
Statistical analysis was performed with JMP (SAS Institute, Cary, NC). The Kolmogorov-Smirnov test was used to test the normality of data distribution. Serial changes in OMT frequency were compared by using analysis of variance with the post hoc Tukey test. The χ2 test and logistic regression were used to evaluate the prediction of movement in response to stimulation (LMA insertion) and to evaluate the prediction of response to command at emergence. Receiver operating characteristic (ROC) curves were constructed. Results are expressed as mean ± SD; P values <0.05 (two tailed) were considered statistically significant. Unless otherwise indicated, results are pooled data from all four study sites. Continuous recordings were made in one site only (Dublin).
OMT frequencies at each of the major time points satisfied tests for normality of distribution. Figure 2 shows OMT frequency at postinduction loss of responsiveness, during maintenance of anesthesia when unstimulated and during surgical stimulation, and at first response to command after termination of the anesthetic for all patients. The mean OMT frequency decreased significantly at loss of responsiveness and remained decreased significantly below the baseline (awake) level at each time point during unconsciousness. At return of consciousness, defined by response to command, OMT frequency was significantly more than any mean value during unconsciousness. Figure 3A shows the ROC curve for OMT frequency to predict return of response to verbal command at emergence. The area under the curve was 0.939. Figure 3B shows the same data in terms of the relations between the OMT frequency and the probability of unresponsiveness. There were no reported cases of explicit awareness. Figure 4 shows a sample trace of continuously recorded OMT frequency in a patient receiving general anesthesia that consisted of induction with fentanyl and propofol and maintenance with sevoflurane and nitrous oxide in oxygen.
There were no differences in OMT frequencies among patients anesthetized with different volatile anesthetics or between male and female patients (data not shown). Only one patient moved in response to the initial surgical incision; this was insufficient to permit analysis of OMT frequency as a predictor of this response. Continuous recording during the period of surgical incision in the presence or absence of neuromuscular blockade showed an increase in both the frequency and the amplitude of OMT 15–30 s after incision. Fifteen patients moved in response to the insertion of an LMA. OMT frequency before LMA insertion was significantly higher in patients who moved compared with those who did not move (Fig. 5A). The ROC curve area to predict movement was 0.783 (not shown). Continuous recording of OMT during the period of airway instrumentation in the presence or absence of neuromuscular blockade showed an increase in OMT frequency and amplitude 15–30 s after airway instrumentation commenced. The extent of this increase and that after surgical incision was highly variable and transiently exceeded baseline (awake) values in some cases.
There was no significant difference in OMT frequency between patients who received neuromuscular blocking drugs and those who did not move at any of the preselected points (Fig. 5B). Continuous recording of OMT in the minutes after the administration of neuromuscular blocking drugs and after the administration of neostigmine/glycopyrrolate confirmed no direct effect on OMT frequency. OMT amplitude appeared to decrease with the onset of neuromuscular blockade, although we were unable to identify a graded relationship between OMT amplitude and TOF. The mean OMT frequency during the maintenance of anesthesia (unstimulated) for patients placed supine was 24 ± 11 Hz and for those placed prone was 27 ± 9 Hz (no significant difference). OMT frequency did not change significantly in those patients turned from the supine to the prone position (paired Student’s t-test).
The main finding of this multicenter study was that OMT frequency, measured with an automated signal analysis module, in patients anesthetized with a variety of drugs, decreased consistently after the induction of anesthesia and increased as consciousness returned. OMT frequency increased after stimuli expected to stimulate the reticular formation (RF), such as LMA insertion or surgical incision, and predicted patient movement in response to LMA insertion. OMT remained measurable by the automated system despite neuromuscular blockade and was also measurable when patients were placed in the lateral or supine position.
OMT is one of three involuntary eye movements that occur even when the eye is apparently still: the others are microsaccades and slow drift (25). OMT has the highest frequency and lowest amplitude of any known physiologic tremor (9). Its physiologic role, if any, is undetermined, and although it was once thought to function in preventing retinal image decay, this is now thought unlikely (26). Spauschus et al. (1) found OMT to be coherent between the right and left eye of individual subjects, irrespective of convergence or visual fixation. This indicates a central neurogenic origin rather than a myogenic one. The oculomotor neurons are embedded in the RF, and it is postulated that OMT reflects a spillover effect of active RF neuron activity impinging on neighboring oculomotor neurons. Because the RF is the main locus that controls the state of arousal of the neocortex, OMT has been proposed as a surrogate measure of the state of arousal (5). In support of this, OMT frequency is completely absent only in brainstem death (4,5). After head injury, OMT was found to be decreased, and the frequency of OMT correlated with the level of consciousness and the eventual outcome (5,6).
In a previous study involving 22 patients not receiving muscle relaxants (11), propofol decreased OMT frequency in a dose-dependent manner between a predicted plasma concentration of propofol of 1 and 2 μg/mL, with no further decreases at 3–5 μg/mL. A subsequent study evaluated OMT during propofol induction and sevoflurane/nitrous oxide maintenance of anesthesia in 30 patients (10). OMT remained depressed in patients receiving sevoflurane/nitrous oxide and increased at emergence in every patient. Increasing sevoflurane from 1 to 2 MAC did not result in further decreases in OMT. Together with the observations on the effects of propofol, this suggests an all-or-none type of response of OMT frequency to state of consciousness in patients receiving general anesthetics, although a graded response may occur before consciousness is lost. In the latter study (10), OMT predicted the return of consciousness, determined by response to command, with a high degree of accuracy. The present study found a similar discrimination for this variable: indeed, the area under the ROC curve was similar to that previously reported, despite the use of a variety of anesthetic regimens in four different centers. Importantly, OMT frequency remained measurable in the subgroup of patients who were given muscle relaxants, although the OMT amplitude was reduced.
The external ocular muscles consist of both skeletal and smooth muscle fibers, and tremor continues during profound neuromuscular blockade, presumably because of smooth muscle activity. Among the purposes of this study was the need to evaluate the direct effects of neuromuscular blockade and to validate the accuracy of our automated system to measure OMT frequency even at low amplitudes. Our results indicate that OMT frequency was measurable even at TOF = 0, and the induction or reversal of neuromuscular blockade had no discernible effect on measured OMT frequency.
In this study, OMT was measured with a piezoelectric strain gauge via a probe placed gently on the closed eyelid. Earlier descriptions of OMT were based on measurement of deflection of a light beam focused on a mirror mounted in the eye (2) or on accelerometry with a sensor mounted on scleral contact lenses (27). The strain gauge technique for measurement of OMT was developed by Bengi and Thomas (28) and is postulated to be particularly accurate because it adds minimally to corneal inertia which, because of the extremely low amplitude of this tremor, may affect measurements made with corneal contact methods (24). In previous studies in anesthetized patients, the piezoelectric probe was placed directly on the sclera (11,13) or on the closed eyelid (10), as used in this study. We have now applied this latter technique to more than 300 patients without report of eye injury. Because of concerns about eye injury after spinal surgery in the prone position (29), we suggest that caution should be exercised in the use of this device in such patients. More information is expected regarding the risk factors that contribute to the development of this complication (30). It is noteworthy, however, that the OMT signal is completely dampened at a pressure much less than that required to cause scleral damage.
An insufficient number of patients moved in response to surgical incision to allow a meaningful analysis of OMT prediction of this variable, although a high OMT frequency before airway instrumentation predicted movement. Movement in response to pain is the basis of MAC, a measure partially or completely independent of cortical activity (15). This may explain why EEG-based measures predict movement inconsistently (17,20), whereas measures that rely partly on brainstem pathways are reported to be predictive (22). For example, the auditory evoked potential index predicted response to LMA insertion during propofol and alfentanil anesthesia (18) and during sevoflurane anesthesia (22). Basal RF activity and, secondarily, OMT frequency reflect tonic inputs from multiple projection pathways derived from cortical and subcortical structures. We hypothesize that OMT may be indicative of an aggregate of anesthetic effects on these regions and thus may predict movement independently of its ability to indicate level of consciousness. Nonetheless, the extent to which a movement response is indicative of cortical activity and, therefore, the potential for awareness, is difficult to estimate, thus making this hypothesis difficult to test, and the accuracy of OMT frequency for predicting movement in response to incision remains to be defined.
Several new technologies for measuring the state of hypnosis have been developed. A previous study reported a comparison of OMT frequency and the bispectral index as predictors of response to verbal command in patients receiving sevoflurane anesthesia (10). OMT values showed less overlap between responsive and unresponsive patients and more accurately identified patients regaining consciousness after anesthesia. Comparisons with other technologies have not been performed.
A limitation to this technique for intraoperative monitoring of anesthetic effect is interference with signal recording because of electrocautery or gross patient movement. This is a consequence of the high sensitivity of the probe required by the extremely low amplitude of the tremor. Nonetheless, the measured amplitude during such events is widely different from that of OMT, and during these periods the light-emitting diode displays an error message.
In summary, these results demonstrate that an automated signal analysis module for real-time measurement of OMT frequency may be used as a measure of anesthetic effect across a broad range of patients anesthetized with a variety of anesthetic regimens. OMT responses to sensory stimulation during anesthesia suggest that this measure reflects the balance between central nervous system depressant effects of anesthetics and RF stimulation by external stimuli. Validation of this technique by prospective titration of anesthesia based on OMT frequency remains to be performed.
The authors thank Linda McEvoy, RN (Beaumont Hospital, Dublin), for valuable assistance.
1. Spauschus A, Marsden J, Halliday DM, et al. The origin of ocular microtremor in man. Exp Brain Res 1999;126:556–62.
2. Alder FH, Fliegelman F. The influence of fixation on the visual acuity. Arch Ophthalmol 1934;12:475–83.
3. Eizenman M, Hallett PE, Frecker RC. Power spectra for ocular drift and tremor. Vision Res 1985;25:1635–40.
4. Coakley D, Thomas JG. The ocular microtremor record as a potential procedure for establishing brain death. J Neurol Sci 1977;31:199–205.
5. Shakhnovich AR, Thomas JG. Microtremor of the eyes of comatose patients. Electroencephalogr Clin Neurophysiol 1977;42:117–9.
6. Coakley D, Thomas JG. The ocular microtremor record and the prognosis of the unconscious patient. Lancet 1977;1:512–5.
7. Bolger C, Bojanic S, Sheahan NF, et al. Ocular microtremor in patients with idiopathic Parkinson’s disease. J Neurol Neurosurg Psychiatry 1999;66:528–31.
8. Bolger C, Bojanic S, Sheahan NF, et al. Ocular microtremor in oculomotor palsy. J Neuroophthalmol 1999;19:42–5.
9. Bolger C, Bojanic S, Sheahan NF, et al. Dominant frequency content of ocular microtremor from normal subjects. Vision Res 1999;39:1911–5.
10. Kevin LG, Cunningham AJ, Bolger C. Comparison of ocular microtremor and bispectral index during sevoflurane anaesthesia. Br J Anaesth 2002;89:551–5.
11. Bojanic S, Simpson T, Bolger C. Ocular microtremor: a tool for measuring depth of anaesthesia? Br J Anaesth 2001;86:519–22.
12. Bolger C, Bojanic S, Sheahan N, et al. Ocular microtremor (OMT): a new neurophysiological approach to multiple sclerosis. J Neurol Neurosurg Psychiatry 2000;68:639–42.
13. Coakley D, Thomas JG, Lunn JN. The effect of anaesthesia on ocular microtremor. Br J Anaesth 1976;48:1121–2.
14. Eger EI II, Saidman LJ, Brandstater B. Minimum alveolar anesthetic concentration: a standard of anesthetic potency. Anesthesiology 1965;26:756–63.
15. Rampil IJ, Mason P, Singh H. Anesthetic potency (MAC) is independent of forebrain structures in the rat. Anesthesiology 1993;78:707–12.
16. Sebel PS, Lang E, Rampil IJ, et al. A multicenter study of bispectral electroencephalogram analysis for monitoring anesthetic effect. Anesth Analg 1997;84:891–9.
17. Rampil IJ, Laster MJ. No correlation between quantitative electroencephalographic measurements and movement response to noxious stimuli during isoflurane anesthesia in rats. Anesthesiology 1992;77:920–5.
18. Doi M, Gajraj RJ, Mantzaridis H, Kenny GN. Prediction of movement at laryngeal mask airway insertion: comparison of auditory evoked potential index, bispectral index, spectral edge frequency and median frequency. Br J Anaesth 1999;82:203–7.
19. Kochs E, Kalkman CJ, Thornton C, et al. Middle latency auditory evoked responses and electroencephalographic derived variables do not predict movement to noxious stimulation during 1 minimum alveolar anesthetic concentration isoflurane/ nitrous oxide anesthesia. Anesth Analg 1999;88:1412–7.
20. Dwyer RC, Rampil IJ, Eger EI II, Bennett HL. The electroencephalogram does not predict depth of isoflurane anesthesia. Anesthesiology 1994;81:403–9.
21. Katoh T, Suzuki A, Ikeda K. Electroencephalographic derivatives as a tool for predicting the depth of sedation and anesthesia induced by sevoflurane. Anesthesiology 1998;88:642–50.
22. 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.
23. Bolger C, Bojanic S, Phillips J, et al. Ocular microtremor in brain stem death. Neurosurgery 1999;44:1201–6.
24. Sheahan NF, Coakley D, Hegarty F, et al. Ocular microtremor measurement system: design and performance. Med Biol Eng Comput 1993;31:205–12.
25. Gaymard B, Pierrot-Deseilligny C. Neurology of saccades and smooth pursuit. Curr Opin Neurol 1999;12:13–9.
26. Steinman RM, Cushman WB, Martins AJ. The precision of gaze: a review. Hum Neurobiol 1982;1:97–109.
27. Brown P, Day BL. Eye acceleration during large horizontal saccades in man. Exp Brain Res 1997;113:153–7.
28. Bengi H, Thomas JG. Three electronic methods for recording ocular tremor. Med Biol Eng 1968;6:171–9.
29. Lee LA, Lam AM. Unilateral blindness after prone lumbar spine surgery. Anesthesiology 2001;95:793–5.
30. Roth S, Barach P. Post-operative visual loss: still no answers—yet. Anesthesiology 2001;95:575–7.