Electroneurographic facial muscle pattern in Bell’s palsy☆☆☆★★★

Electroneurographic facial muscle pattern in Bell’s palsy MATS ENGSTRÖM, MD, PhD, LARS JONSSON, MD, PhD, MARGARETA GRINDLUND, and ERIK STÅLBERG, MD, PhD, Uppsala, Sweden

To study the electroneurographic facial muscle pattern in Bell’s palsy over time, electroneurographic recordings in the frontalis, orbicularis oculi, nasalis, and mentalis muscle regions were performed early (mean, day 11) and 1 and 3 months after the onset of the condition in 30 consecutive patients. The correlation between facial muscle electroneurographic recordings over time was also calculated. An additional aim was to assess whether further prognostic information could be obtained by electroneurographic recordings in more than one facial region. The recovery pattern was similar in all 4 facial regions. Initially, the correlation between the facial recordings was weak (r = 0.20-0.27), but it was improved at follow-up examinations (r = 0.33-0.65). Favorable outcome in 23 of 24 patients (96%) could have been predicted by the initial nasalis and/or mentalis recordings. The gap between patients with favorable outcome and patients with unfavorable outcome increased when the average electroneurography values were calculated from 1, 2, and 4 muscle recordings (4%, 8%, and 15%, respectively). Our results indicate that in Bell’s palsy, electroneurographic examination of more than one facial muscle region may add prognostic information and that the degree of degeneration is initially different in the nerve branches. (Otolaryngol Head Neck Surg 2000;122:290-7.)

From the Departments of Oto-Rhino-Laryngology and Head & Neck Surgery (Drs Engström and Jonsson) and Clinical Neurophysiology (Dr Stålberg and Mrs Grindlund), Uppsala University, Akademiska sjukhuset. Supported by grant 135 (E.S.) from the Swedish Medical Research Council, Stockholm, the Swedish Society of Medicine, Stockholm, and the Swedish Society of Otolaryngology. We thank Gina Eriksson for help with language revision of the manuscript. This article has previously been published in Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine 750. Uppsala University, 1998. Reprint requests: Mats Engström, MD, PhD, Department of Otorhinolaryngology-HNS, Uppsala University, Akademiska sjukhuset, SE751 85 Uppsala, Sweden. Copyright © 2000 by the American Academy of Otolaryngology– Head and Neck Surgery Foundation, Inc. 0194-5998/2000/$12.00 + 0 23/77/98731 290

Bell’s palsy is unilateral weakness or paralysis of the face due to acute peripheral facial nerve dysfunction, with no readily identifiable cause and with some recovery of function within 6 months.1 The pathophysiologic mechanism of impaired nerve function in the disease is still being disputed. Neurophysiologic,2 histologic,3 and magnetic resonance imaging4 studies indicate that the dysfunction is caused by compression, secondary to edema and/or ischemia, in the meatal/labyrinthine portion of the nerve. Circumstantial evidence and deductive reasoning support the hypothesis that herpes simplex virus is involved in the pathophysiology of the disease.5 Electroneurography (ENoG) is the most accurate electric test for providing an objective, quantitative assessment of facial nerve degeneration and prognosis of facial recovery in Bell’s palsy.6-13 The nasolabial fold, used for positioning of the recording electrode in most studies on ENoG and Bell’s palsy, has been reported to be the most convenient and reliable recording site.6-7,10,14 Other facial regions, mainly the orbicularis oris or mentalis muscles, have also been used for ENoG recordings.6,14-16 Despite a large number of ENoG studies, information is limited on the results of positioning the electrodes in other facial regions to establish the facial muscle ENoG pattern in Bell’s palsy over time.6,10,14,16-19 The purpose of the present investigation was to establish the facial muscle ENoG pattern over time in Bell’s palsy. ENoG recordings obtained in the frontalis, orbicularis oculi, nasalis, and mentalis regions of patients in the early stage of and during recovery from Bell’s palsy were analyzed and compared. An additional aim was to determine whether further prognostic information could be obtained by measurements with electrode placements other than in the nasolabial fold or in more than one facial region. Furthermore, to acquire additional knowledge on the degeneration process in the disease, the correlation between the ENoG results over time in the different facial muscles was calculated. We also studied the reliability of the method in the different facial recording sites, determining ENoG test-retest20,21 variability in both healthy volunteers and patients with Bell’s palsy. PATIENTS AND METHODS Patients The investigation included 30 consecutive patients, 13 women and 17 men, with acute idiopathic facial palsy (Bell’s

Otolaryngology– Head and Neck Surgery Volume 122 Number 2

ENGSTRÖM et al

291

Table 1. Correlation coefficients between ENoG values (n = 30) obtained in frontalis, orbicularis, nasalis, and mentalis recordings at initial, first, and second follow-up examinations

Initial examination Frontalis Orbicularis oculi Nasalis Mentalis First follow-up Frontalis Orbicularis oculi Nasalis Mentalis Second follow-up Frontalis Orbicualris oculi Nasalis Mentalis

Frontalis

Orbicularis oculi

Nasalis

Mentalis

– 0.1977 0.2704 0.2401

0.1977 – 0.1964 0.0867

0.2704 0.1964 – 0.2558

0.2401 0.0867 0.2558 –

– 0.3843 0.6414 0.5150

0.3843 – 0.4730 0.4113

0.6414 0.4730 – 0.6486

0.5150 0.4113 0.6486 –

– 0.4930 0.5690 0.5996

0.4930 – 0.3698 0.3282

0.5690 0.3698 – 0.5787

0.5996 0.3282 0.5787 –

palsy). Their ages ranged from 21 to 64 years (median, 36 years). Each of 12 patients had a left-sided palsy and each of 18 a right-sided palsy; none had bilateral palsy or a previous history of facial palsy. Only patients between 18 and 65 years of age were included in the study. Pregnant women and patients with malignant disease were excluded. The investigation was carried out from May 1994 to December 1997 and approved by the Ethics Committee, Faculty of Medicine, Uppsala University. Informed consent was obtained from each patient before his or her enrollment. Clinical Investigation The initial clinical examination was carried out on mean day 11 (range, days 7-17) and the 2 follow-up examinations at 1 month (mean, day 36) and 3 months (mean, day 99) after the onset of palsy. The Yanagihara grading system was used at each examination.22 In this system, each of 10 different aspects of muscle or facial function is graded on a 5-point scale from 0 (no function) to 4 (full function). The functional score is calculated by adding the 10 grades. Complete recovery is a score of 40; in our study, favorable outcome was defined as a functional score of at least 36 at the 3-month follow-up. Neurophysiologic Investigations ENoG was performed on the same days as the clinical examinations. The facial nerve was stimulated with a surface electrode (Dantec 3L523) centered over the stylomastoid foramen. A constant voltage stimulator was used to provide single square wave pulses 0.1 msec in duration at a frequency of less than 1 per second. To secure the highest possible

Fig 1. Placement of ENoG recording surface electrodes (standardized) in frontalis, orbicularis oculi, nasalis, and mentalis muscle regions. Positions for stimulation electrode over stylomastoid foramen and for reference electrode on tip of nose are also shown.

reproducibility, we decided, in accord with the results of previous experiments,7 to have one electrode active and one as inactive as possible but still on the face (to remove artifacts and external interference). The tip of the nose was chosen for reference electrode placement. The recording surface electrode positions were standardized10,14,23 and placed with slight modifications from those previously described,7,14 as shown in Fig 1. The forehead electrode (referred to as frontalis in the text) was placed 2 cm above the eyebrow and 2 cm from the midline. The eye electrode (referred to as orbicularis oculi) was placed on the lateral third on the lower eyelid. The nasal electrode (referred to as nasalis) was secured lateral to the midpoint of the nose (nasalis and paranasal muscles); this is slightly different from the electrode placement used by Thomander and Stålberg.7 The perioral electrode (referred to as mentalis) was placed 2 cm lateral to the midline and 2 cm below the lower lip (the mentalis, depressor labii inferioris, orbicularis oris, and depressor anguli oris muscles). Supramaximal voltage (the voltage required to produce a maximal amplitude + 10%) was used. The optimal stimulating electrode placement was determined from the shape of the obtained signal; minor adjustments were sometimes necessary to obtain a maximal muscle response from the defined muscle region. The recording was accepted when the waveform was maintained and reproducible during at least two repeated stimulations. The muscle compound action potential, measured as a peak-to-peak value, was recorded with a Keypoint EMG instrument (Dantec). The percent amplitude reduction was calculated by comparing the affected side with the healthy side. To confirm that the measured electrical activity (ENoG)

292

ENGSTRÖM et al

Otolaryngology– Head and Neck Surgery February 2000

Fig 2. Scatter plots of ENoG correlation between nasalis, mentalis, and frontalis muscles at initial (mean, day 11 after onset), first (1 month), and second (3 months) follow-up examinations.

was related to and specific for the respective functional facial muscle group, electromyographic recordings (with surface electrodes positioned as for the ENoG recordings) were performed in two healthy volunteers. Activation of the frontalis, orbicularis oculi, nasalis, and mentalis muscles induced electromyographic activity in these muscle groups only. Test-retest Variability Three healthy volunteers—2 women aged 28 and 48 years and 1 man aged 38 years (mean, 38 years)—and 3 patients with Bell’s palsy—3 men aged 21, 37, and 64 years (mean, 41 years)—were included in the ENoG test-retest study. The 3 patients with Bell’s palsy were examined on mean day 14 after the onset of palsy. The electrode placements used were the same as those used for the group of patients with Bell’s palsy. The test-retest was repeated 10 times in each muscle group. For each of the 10 tests, the electrodes were removed and new electrodes were applied. The 10 test-retests were carried out over a period of 2 to 4 hours. In one volunteer, however, the test-retest period was 5 days. To express the test-retest variability, the intratest difference expressed as percent was first calculated from the mean of each of the 10 test-retest sessions; the absolute values of these differences were then taken and averaged. Statistical Analyses The Student paired t test was used to compare the ENoG values. Correlation coefficients were calculated to determine

the relationship between the ENoG values in the different facial recordings at the initial, first, and second follow-up examinations in the 30 patients with Bell’s palsy. RESULTS Electroneurographic Evaluation

Table 1 shows the correlation coefficients for the ENoG values obtained in the 30 patients with Bell’s palsy in the 4 different facial regions at the initial, first, and second follow-up examinations. The correlation values between the muscles were weak at the initial examination (r = 0.20-0.27). At the first and second follow-up examinations, the correlation between the frontalis, mentalis, and nasalis muscles improved (first follow-up, r = 0.52-0.65; second follow-up, r = 0.580.60). A stronger but less pronounced correlation between the orbicularis oculi and the other 3 facial muscles was also evident (first follow-up, r = 0.38-0.47; second follow-up, r = 0.33-0.49). The scatter plots in Fig 2 show the correlation between the nasalis, mentalis, and frontalis recordings at the initial, first, and second follow-up examinations. At the initial examination, it appeared that patients with severe degeneration (high ENoG values) demonstrated a higher degree of correlation between the individual facial muscles (represented by the shaded areas in Fig 2). Fig 3 shows box and whisker plots of the different facial muscle recordings at the initial, first, and second

Otolaryngology– Head and Neck Surgery Volume 122 Number 2

ENGSTRÖM et al

293

Fig 3. Box and whisker plots of frontalis, orbicularis oculi, nasalis, and mentalis ENoG recordings at initial, first, and second follow-up examinations (indicated by 1st, 2nd, and 3rd, respectively). Median, 25th and 75th percentiles, and smallest and largest observed values are given.

follow-up examinations. The ENoG pattern is similar in all 4 facial regions. The highest degree of degeneration, except for the frontalis recording, was demonstrated at the initial examination. In the follow-up period, the degeneration levels decreased. When the ENoG values were compared in the various muscle groups and at different times, a significant difference was observed only when the mentalis recordings were compared with the nasalis recordings at the second follow-up (P = 0.008). Electroneurographic vs Clinical Evaluation and Outcome

The results of the initial ENoG recordings in the different facial muscles in relation to the total clinical outcome at 3 months, as assessed through use of the Yanagihara grading system, are shown in Fig 4. According to our results, the favorable outcome in 23 of the 24 patients (Yanagihara score ≥ 36 at 3 months) could have been predicted by the nasalis ENoG recording if the level had been set at ≤75%. One of the 7 patients with ENoG levels greater than 75% had recovered at 3 months. With the mentalis recording, favorable outcome in the same number of patients and in the same patients could have been predicted with an ENoG level set at ≤78%. The patterns of the frontalis and orbicularis oculi recordings were similar to those of the nasalis and

mentalis recordings. At these levels, however, favorable outcome was falsely predicted in a large number of patients when the ENoG values from the frontalis and orbicularis oris were used (Fig 4). Fig 5 shows the averaged ENoG data obtained from more than one facial region. The high predictive value for favorable outcome was the same for the nasalis and mentalis recordings only (Fig 4) and for the average of these two recordings (Fig 5A). When the ENoG values in all 4 different facial regions were averaged and the same levels were set as for the nasalis and/or mentalis recordings, favorable outcome was correctly predicted in all 24 of the patients (Fig 5B). However, when the averaged ENoG of all 4 regions was used, favorable outcome was falsely predicted in one patient. Figs 4 and 5 show that the gap (represented by the shaded areas) between patients with favorable outcome and patients with unfavorable outcome increases with an increasing number of analyzed (averaged ENoG values) facial muscles. This demarcation zone or gap (shaded areas) was 4% for the individual nasalis and mentalis recordings and 8% for the average ENoG of these muscles. When the ENoG values in all 4 different facial regions were averaged, the gap was 15%. The stimulation strength required to produce supramaximal voltage (the voltage required to produce a

294

Otolaryngology– Head and Neck Surgery February 2000

ENGSTRÖM et al

Fig 4. Initial (mean, day 11) ENoG values in frontalis, orbicularis oculi, nasalis, and mentalis recordings in relation to total clinical outcome at 3 months (mean, day 99) assessed by Yanagihara grading system. Shaded areas in nasalis and mentalis recordings indicate gap between patients with favorable outcome (Yanagihara score ≥ 36) and patients with unfavorable outcome (Yanagihara score < 36) at 3 months.

peak amplitude + 10%) was 40-60 mA in the nasal registration. The same stimulation strength was needed to produce maximal amplitudes in the other facial muscle groups. The waveform morphology and amplitude varied in the different muscles. Examples of the typical waveforms in the different facial regions are shown in Fig 6, with positive deflection downwards. The nasalis recordings demonstrated a biphasic-shaped wave with a high amplitude and an initial negative deflection. A similar wave shape, but with lower amplitudes, was demonstrated in the frontalis and mentalis muscles. The orbicularis oculi muscle recordings demonstrated a wave with positive double spikes. Test-retest Variability

The percent ENoG amplitude reductions (mean, day 14) in the 3 patients with Bell’s palsy were 39%, 67%, and 85%, respectively, in the nasalis recordings. The

intrasubject side-to-side amplitude difference (averaged intratest absolute difference) in the 3 healthy volunteers and the 3 patients is shown in Table 2. The lowest variability was demonstrated in the nasalis recordings of both volunteers (mean ± SD, 6% ± 4%) and patients (3% ± 2%). As shown in Table 2, the mean variability values were lower in all 4 recording sites in the patients than in the volunteers. DISCUSSION

The ENoG pattern over time was similar in all 4 facial muscles studied in the patients with Bell’s palsy except for a delayed recovery in the mentalis muscle at the second follow-up. To our knowledge, this is the first report on serial follow-up ENoG recordings from different facial muscles in Bell’s palsy. There are, however, electromyographic studies with integrated surface electromyographic and needle registrations.18,24 In a study by Laskawi and Damenz,18 a reduced innervation pat-

Otolaryngology– Head and Neck Surgery Volume 122 Number 2

ENGSTRÖM et al

A

295

B Fig 5. Averaged initial (mean, day 11) ENoG value for nasalis and mentalis recordings (A) and for frontalis, nasalis, orbicularis oculi, and mentalis recordings (B) in relation to total clinical outcome at 3 months (mean, day 99) assessed by Yanagihara grading system. Shaded areas indicate gap between patients with favorable outcome (Yanagihara score ≥ 36) and patients with unfavorable outcome (Yanagihara score < 36) at 3 months.

Table 2. Test-retest variability in frontalis, orbicularis oculi, nasalis, and mentalis muscle regions in 3 volunteers and 3 patients with Bell’s palsy (mean, day 14 after onset)

Patient 1 Patient 2 Patient 3 Mean for patients Volunteer 1 Volunteer 2 Volunteer 3 Mean for volunteers

Frontalis

Orbicularis oculi

Nasalis

Mentalis

4±3 4±3 7±3 5±3 13 ± 8 17 ± 7 5±4 11 ± 10

10 ± 8 6±3 3±2 6±6 11 ± 2 13 ± 7 9±8 11 ± 8

3±2 4±2 3±2 3±2 6±5 6±2 6±3 6±4

8±5 9 ± 10 2±2 6±7 8±6 11 ± 4 10 ± 6 10 ± 5

Variability is expressed as averaged intratest absolute difference from mean (percent ± SD) for each of 10 test sessions.

tern in 5 standard facial muscles did not show a preferred lesion of the facial nerve in any of these muscles, suggesting that the lesion in Bell’s palsy is located in a portion of the facial nerve with no somatotopic organization. Our findings of a similar ENoG pattern over time in the different muscles are in agreement with these results.18 Initially, there was a weak correlation among the ENoG values in the nasalis (midface), frontalis (upper face), and mentalis muscles (lower face). This may be explained by a difference in the delay and/or degree of degeneration, which might relate to the thicknesses and lengths of the respective nerve branches. Another explanation may be that the initial facial nerve involvement is focal within the nerve and the localization is different for the individual patient. If the nerve axons are somatotopically organized, individual nerve fibers may be involved unevenly before the lesion later spreads to the

entire nerve. In the early stages of palsy, minor differences between these factors may be critical in determining when distal degeneration will occur. The stronger correlation at the follow-up examinations may be related to a more homogenous degeneration of the facial nerve with time. The nasalis and/or mentalis initial ENoGs were the best single recordings in predicting patients with favorable outcome in our study. The gap in the individual nasalis and mentalis initial ENoG recordings between patients with favorable outcome and those with unfavorable outcome was 4%. When the mean ENoG values for both of these muscles were calculated, the gap increased to 8%; when all 4 muscles were included, the gap increased further, to 15%. At the initial examination, patients with severe degeneration (high ENoG values) appeared to demonstrate a higher degree of correlation between the individual facial muscles. This, in

296

Otolaryngology– Head and Neck Surgery February 2000

ENGSTRÖM et al

Fig 6. Typical ENoG waveform morphology with recording electrode placed over frontalis, orbicularis oculi, nasalis, and mentalis muscle regions. Positive deflection downwards.

turn, may more clearly separate the group with favorable outcome from the group with unfavorable outcome. The reliability is increased when the results of more than one test session are analyzed.9,14,25 This refers not only to repeated test sessions in one muscle but also to averaged measures in more than one facial region. In the present study, the number of patients with a severe palsy was too small and the statistical power insufficient to assess the advantage and predictive effect of ENoG in more than one facial region.24 The variation in the waveform morphology and amplitude of the compound action potentials in different facial muscle recordings in the 3 patients and the 3 volunteers is in agreement with previous findings.6,7,14,19 High stimulus magnitude may lead to masseter artifact or cause a degree of pain that precludes testing.10,26 In contrast, low levels of stimulation produce inconsistent responses of low amplitude. To avoid these shortcomings, percutaneous stimulation was used with stimulation electrodes optimally placed by experienced personnel (M.G.).7,10,14,20,25 Supramaximal strength was used to ensure that all nerve fibers were stimulated.10,21 The recording was accepted only when the shape of the obtained signal was consistent at two consecutive stimulations to ensure relative response stability.21 Using this technique, we were able to obtain measurable and reproducible compound action potentials in all 4 facial muscle recordings.

The overall variability was approximately 9% in the healthy volunteers and 7% in the 3 test-retest patients. The intratest variability is mainly due to methodologic factors. It should, however, be noted that variability cannot be measured in patients with ENoG values close to 100% (ie, with severe degeneration), because compound action potentials cannot be elicited. In healthy volunteers, individually optimized lead placement has been shown to reduce intrasubject side-to-side differences.6,10,14,19 As discussed by Kelleher et al,19 however, a recording lead montage site that appears to produce an optimal response in normal muscles may not necessarily be optimal for assessment of facial pathosis. CONCLUSION

The ENoG pattern over time was similar in all 4 facial regions except for a delayed recovery in the lower face (mentalis) at the second follow-up. At the initial examination, the correlation between the different facial muscle ENoG recordings was weak, but at the followup recordings the correlation improved. This may indicate a more homogenous degeneration in the different nerve branches with time. The nasalis and/or mentalis initial ENoG were the best single recordings in predicting patients with favorable outcome. Additional muscle registrations may improve the prognostic accuracy in Bell’s palsy. Further research, including more extensive test-retest validation,26 is required on patients with

Otolaryngology– Head and Neck Surgery Volume 122 Number 2

facial nerve dysfunction before recommendations can be made regarding the clinical advantage of ENoG recordings in more than one facial muscle region. REFERENCES 1. May M, Hughes GB. Facial nerve disorders: Update 1987. Am J Otol 1987;8:167-80. 2. Fisch U, Esslen E. Total intratemporal exposure of the facial nerve. Arch Otolaryngol 1972;95:335-41. 3. O’Donoghue GM, Michaels L. Histopathological aspects of Bell’s palsy. In: Portmann M, editor. Proceedings of the Fifth International Symposium on the Facial Nerve. New York: Masson Publishing USA Inc; 1985. p. 248-52. 4. Engström M, Abdsaleh S, Ahlström H, et al. Serial gadoliniumenhanced magnetic resonance imaging and assesment of facial nerve function in Bell’s palsy. Otolaryngol Head Neck Surg 1997;117:559-66. 5. Adour KK, Ruboyianes JM, von Doersten PG, et al. Bell’s palsy treatment with acyclovir and prednisone compared with prednisone alone: a double-blind, randomized, controlled trial. Ann Otol Rhinol Laryngol 1996;105:371-8. 6. Esslen E. Investigations on the localization and pathogenesis of meatolabyrinthine facial palsies. In: Esslen E, editor. The acute facial palsies. Berlin, Germany: Springer-Verlag; 1977. p. 1-123. 7. Thomander L, Stålberg E. Electroneurography in the prognostication of Bell’s palsy. Acta Otolaryngol (Stockh) 1981;92:221-37. 8. May M, Blumenthal F, Klein SR. Acute Bell’s palsy: Prognostic value of evoked electromyography, maximal stimulation, and other electrical tests. Am J Otol 1983;5:1-7. 9. Fisch U. Prognostic value of electrical tests in acute facial paralysis. Am J Otol 1984;5:494-8. 10. Kartush JM, Lilly DJ, Keminik JL. Facial electroneurography: Clinical and experimental investigations. Otolaryngol Head Neck Surg 1985;93:516-23. 11. Sillman JS, Niparko JK, Lee SS, et al. Prognostic value of evoked and standard electromyography in acute facial paralysis. Otolaryngol Head Neck Surg 1992;107:377-81. 12. Smith IM, Maynard C, Mountain RE, et al. The prognostic value of facial electroneurography in Bell’s palsy. Clin Otolaryngol 1994;19:201-3. 13. Tojima H, Aoyagi M, Inamura H, et al. Clinical advantages of electroneurography in patients with Bell’s palsy within two

ENGSTRÖM et al

14. 15.

16. 17.

18. 19. 20. 21. 22.

23. 24. 25. 26.

297

weeks after onset. Acta Otolaryngol (Stockh) 1994;511(Suppl): 147-9. Coker NJ. Facial electroneurography: Analysis of techniques and correlation with degenerating motoneurons. Laryngoscope 1992; 102:747-59. Podvinec M. Electroneuronography in prognosis of intratemporal facial palsy, as compared to measurements of rheobase and chronaxy. ORL J Otorhinolaryngol Relat Spec 1977;39: 185-94. Mamoli B. Zur Prognoseerstellung peripherer Fazialisparesen unter besonderer Berucksichtigung der Elektroneurographie. Wien Klin Wochenschr 1976;53(Suppl):3-28. Glocker FX, Magistris MR, Rosler KM, et al. Magnetic transcranial and electrical stylomastoidal stimulation of the facial motor pathways in Bell’s palsy: Time course and relevance of electrophysiological parameters. Electroencephalogr Clin Neurophysiol 1994;93:113-20. Laskawi R, Damenz W. Elektrophysologische Untersuchungen zum Schädigungs-muster der mimischen Muskulatur bei der Bellschen Lähmung. Laryngo-Rhino-Otol 1993;72:197-203. Kelleher MJ, Gutnick HN, Prass RL. Waveform morphology and amplitude variability in facial-nerve electroneurography. Laryngoscope 1990;100:570-2. Raslan WF, Wiet R, Zealear DL. A statistical study of the ENoG test error. Laryngoscope 1988;98:891-3. Hughes GB, Josey AF, Glasscock ME, et al. Clinical electroneurography: statistical analysis of controlled measurements in twenty-two normal subjects. 1981;91:1834-46. Yanagihara N. Grading of facial palsy. In: Fisch U, editor. Facial nerve surgery. Proceedings of the Third International Symposium on Facial Nerve Surgery. Amstelveen, The Netherlands: Kugler Medical Publications; 1977. p. 533-6. Smith IM, Murray JAM, Prescott RJ, et al. Facial electroneurography. Arch Otolaryngol Head Neck Surg 1988;114:322-5. Canter RJ, Nedzelski JM, McLean JA. Evoked electromyography in Bell’s palsy: a clinically useful test? J Otolaryngol 1986;15: 344-7. Hughes GB, Nodar RH, Williams GW. Analysis of test-retest variability in facial electroneurography. Otolaryngol Head Neck Surg 1983;91:290-3. Adour KK, Sheldon MI, Kahn ZM. Maximal nerve excitability testing versus neuromyography: prognostic value in patients with facial paralysis. Laryngoscope 1980;90:1540-7.