Copyright @ 200 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited. 8 Reliability of Vestibular Evoked Myogenic Potentials in Healthy Subjects *Suwicha Isaradisaikul, ���Darcy A. Strong, ���Jamie M. Moushey, ���Sandra A. Gabbard, ���Steven R. Ackley, and ���Herman A. Jenkins *Chiang Mai University, Chiang Mai, Thailand ��Marion Downs Hearing Center, University of Colorado Health Sciences Center and ��University of Colorado, Boulder, Colorado, U.S.A. Objective: To analyze test-retest reliability of vestibular evoked myogenic potential (VEMP) responses with and with- out the use of electromyography (EMG) monitoring in people with normal audiovestibular function. Patients: Twenty adult volunteers with no history of ear dis- ease, normal otoscopic examination, normal pure-tone audio- metry thresholds, and normal tympanograms. Interventions: Prospective evaluation of VEMP responses with and without the use of EMG monitoring in 2 separate sessions 1 to 4 weeks apart. Main Outcome Measures: Threshold repeatability, p13 and n23 latency, p13-n23 interlatency, and interamplitude and inter- aural amplitude difference from the first and the second ses- sions were assessed via the intraclass correlation coefficient. Results: Test-retest reliability of p13-n23 interamplitude was found to be excellent, and the reliability of threshold and latency was found to be fair to good (with the exception of poor reliability for p13 latency in the EMG monitoring condition). Conclusion: Overall, VEMP response parameters were found to have fair to good test-retest reliability. The intraclass correlation coefficient value for amplitude was found to be more reliable than latency, with the latency of n23 more reliable than the latency of p13. Clinicians should consider these findings when interpreting VEMP responses. Mainte- nance of symmetric head rotation with and without EMG monitoring produced reliably reproducible results, the VEMP amplitude being the best criteria. Key Words: Test-retest reliabilityVVestibular evoked myogenic potentialsVVestibular function test. Otol Neurotol 29:542Y544, 2008. Vestibular evoked myogenic potential (VEMP) testing assesses the otolith organ (saccule) and inferior vestibu- lar nerve function. During VEMP testing, a high-level acoustic stimulus is introduced into the ear, and a short latency electromyogram (EMG) response is recorded from the tonically contracted sternocleidomastoid (SCM) muscle. The presence of VEMP responses in people with normal vestibular function has been well documented. Recorded responses in healthy individuals show a biphasic waveform with a positive peak (p13) and a negative peak (n23) (1,2). Among VEMP studies in normal subjects, a variety of protocol parameters have been used. Stimuli that have been used to evoke VEMP responses include air- and bone-conducted tone bursts, air-conducted clicks, fore- head taps, and galvanic stimulation, with air-conducted tone bursts and clicks as the most widely used stimuli. The test position to activate and maintain SCM muscle tonicity has been variable. Reported positions include sitting, supine, recumbent, and prone (3Y5). The location of the surface electrodes on the SCM muscles and recom- mended EMG target levels have also varied in previous studies. The optimal placement for the electrode over the SCM muscle seems to be at the midpoint between the mastoid tip and the base of the SCM muscle (6). Vestib- ular evoked myogenic potential waveform response amplitude is influenced by the EMG level of the SCM muscle. The higher the EMG level of the SCM muscle, the larger the amplitude of the response waveform (2,7). Akin et al. (2) suggested that EMG target levels range from 30 to 50 KV. Young (5) kept active SCM muscle contraction levels between 50 and 200 KV. All of these parameters affect test results. Thus, VEMP waveforms should be reproducible under controlled circumstances and stimulus parameters within the same subject. Address correspondence and reprint requests to Suwicha Isaradisaikul, Department of Otolaryngology, Faculty of Medicine, Chiang Mai Uni- versity, 110 Intawaroros Road, Sriphum Mueang, Chiang Mai 50200, Thailand E-mail: sisaradi@mail.med.cmu.ac.th This study was presented at the Association for Research in Otolar- yngology 30th Midwinter Meeting, Denver, CO, February 11, 2007. Otology & Neurotology 29:542Y544 �� 2008, Otology & Neurotology, Inc. 542

Copyright @ 200 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited. 8 The purpose of this study was to analyze test-retest reliability of VEMP responses with and without the use of EMG monitoring in people with normal audiovestib- ular function. MATERIALS AND METHODS Subjects enrolled in the study had no history of ear, vestib- ular, or neurologic disease. All subjects had normal hearing sensitivity and normal tympanograms on the days of testing. Furthermore, they had the ability to maintain contraction of the SCM muscle for the duration of the test. The study was con- ducted at the University of Colorado Hospital, and the protocol was approved by the Colorado Multiple Institutional Review Board. Before the first session, each subject had a hearing screening. Before each session, an otoscopic examination and tympano- metry were performed. Air-conducted 500-Hz alternating tone-burst stimuli at a rate of 5 per second (1-ms rise/fall, 2-ms plateau) were presented monaurally through ER3A insert earphones (Etymotic Research, Elk Grove Village, IL, USA). The EMG signal was amplified and band passYfiltered from 10 to 1,500 Hz with a Blackman gating function. Vestibular evoked myogenic poten- tial responses were recorded using a Biologic Explorer Evoked Potential System (Bio-Logic Systems Corp., Mundelein, IL, USA). Disposable, pregelled, foam-backed, side-snap surface electrodes (Viasys Healthcare, Inc., Madison, WI, USA) were used. To ensure consistent placement of the SCM electrode for both sessions, a measurement of the distance from the mastoid tip to the base of the SCM muscle was recorded. This measure- ment was then divided in half to analyze the exact midpoint of the SCM for reference electrode placement. The active elec- trode was placed on the sternum with the ground electrode on the forehead. Subjects were in a supine position at an approxi- mately 30-degree angle from the floor. To ensure reflex sym- metry, EMG monitoring was performed. Subjects wee instructed to lift and turn their heads as far as possible toward the shoulder contralateral to the stimulus presentation to uni- laterally contract the SCM muscle. Thus, 2 different methods were used to ensure symmetric activation: 1) formal EMG measurement and 2) maintenance of similar degree of head rotation/elevation. Responses to 100 stimuli were averaged for each recorded response, and a minimum of 2 responses were obtained for each stimulus intensity. Electromyogram monitoring was performed using a Pathway TR-10 (Prometheus Group, Dover, NH, USA), which provides visual monitoring of EMG levels. The EMG stand-alone mon- itoring electrode was placed above the ipsilateral SCM elec- trode. Subjects were instructed to maintain a constant EMG target level of 40 KV or greater during the test and to relax their necks between each test run to reduce muscle fatigue. Vestibular evoked myogenic potential testing was performed first without the use of EMG monitoring then with the use of EMG monitoring. Subjects participated in 2 sessions with each session 1 to 4 weeks apart. Electromyogram levels were main- tained between 40 and 150 KV during each test session. Response thresholds were analyzed using a down 10-, up 5-dB step procedure. Absolute latencies and peak-to-peak amplitudes were calculated from the average of 2 responses at 110 dB HL (115.5 dBpeakSPL). Interaural amplitude differ- ence (IAD) ratio was calculated by dividing the interaural peak- to-peak amplitude difference by the sum of the amplitudes of both ears (5,7). Results were interpreted by the same rater. The data were analyzed using SPSS version 11.0 (SPSS, Inc., Chicago, IL, USA). Repeatability of threshold, p13 latency, n23 latency, p13-n23 interlatency, and interamplitude of 40 ears, and the IAD ratio of the 20 subjects were assessed by the intraclass correlation coefficient of reliability (ICC). Intraclass correlation coefficient of reliability values 0.8 or greater are considered to represent excellent reliability values less than 0.4 may be taken to represent poor reliability. Values between 0.4 and 0.75 are considered as fair to good reliability. RESULTS There were 23 subjects enrolled in the study. Three subjects were excluded: 2 subjects did not return for the second session, and 1 subject showed absence of a VEMP response in 1 ear in the second session. Vestibu- lar evoked myogenic potential responses in 20 subjects TABLE 1. Means and standard deviations of VEMP parameters Parameter First session Second session With EMG monitoring Without EMG monitoring With EMG monitoring Without EMG monitoring Threshold, dBHL 104.6 (T5.2) 103.2 (T5.9) 105.2 (T3.7) 104.4 (T4.7) p13 latency, ms 14.44 (T1.92) 14.08 (T1.81) 14.81 (T2.28) 14.30 (T1.67) n23 latency, ms 21.16 (T2.11) 20.72 (T2.17) 21.53 (T2.58) 21.38 (T2.36) p13-n23 interlatency, ms 6.72 (T1.64) 6.64 (T1.55) 6.72 (T2.23) 7.08 (T2.36) p13-n23 interamplitude, KV 160.71 (T101.11) 171.51 (T120.83) 158.09 (T89.47) 162.82 (T91.50) IAD ratio, % 18.8 (T16.5) 21.0 (T16.5) 21.3 (T21.7) 19.6 (T20.4) EMG indicates electromyogram IAD, interaural amplitude difference VEMP, vestibular evoked myogenic potential. TABLE 2. The ICC values and their 95% CIs for repeatability between the first and the second session with and without the use of EMG monitoring Parameter ICC value (95% CI) With EMG monitoring Without EMG monitoring Threshold 0.62 (0.39Y0.78) 0.53 (0.27Y0.72) p13 latency 0.37 (0.07Y0.60) 0.64 (0.42Y0.79) n23 latency 0.65 (0.44Y0.80) 0.70 (0.50Y0.83) p13-n23 interlatency 0.43 (0.14Y0.65) 0.46 (0.18Y0.67) p13-n23 interamplitude 0.86 (0.75Y0.92) 0.81 (0.67Y0.89) IAD ratio 0.67 (0.74Y0.92) 0.61 (0.25Y0.82) CI indicates confidence interval EMG, electromyogram IAD, inter- aural amplitude difference ICC, intraclass correlation coefficient of reliability. 543 RELIABILITY OF VEMPS IN HEALTHY SUBJECTS Otology & Neurotology, Vol. 29, No. 4, 2008