59 Diagnostic Reference Data for the Monaural Brain-stem Auditory Evoked Response (BAER) Cyril D Govender Μ (Audiology) UDW Department of Speech and Hearing Therapy University of Durban - Westville ABSTRACT The objective of the investigation was to establish diagnostic reference data for the normal BAER. BAERs were elicited from the target (R) ear using clicks presented at 70dBnHL. Relevant latency and amplitude data were obtained from 60 selected normal hearing Indian undergraduate females (N=30; Τ age = 20.33gears) and male (N=30; Τ age = 21.33 gears) students aged between 18 and 25 gears (X age =20.73gears). Diagnostic reference data were established for the absolute latencies ofpeaks I to VI; relative latencies of peaks I-III;III-V and I- V; absolute amplitudes ofpeaksland Vand the relative amplitude ratio of peaks V:I. These results are discussed in terms of the literature and implications for clinical application and further research. OPSOMMING Die doel van hierdie ondersoek was om diagnostiese verwgsingsdata vir normale ouditief ontlokte breinstamresponse daar te stel. Hierdie response is van die regteroor ontlok deur middel van klikgeluide wat bg 70dBnGPaangebied is. Toepaslike data vir latentheid en amplitude is vir 60normaalhorendeIndier voorgraadsestudente verkrg. Vroulike (N=30;Τouderdom = 20.33jaar) en manlike (N=30;Τouderdom = 21.33 jaar) studente, tussen die ouderdom 18 tot 25jaar (X ouderdom = 20.73jaar) is vir die doel van hierdie studie geselekteer. Diagnostiese verwgsingsdata is verkrg vir die absolute latenthede van pieke I tot VI; relatiewe latenthede van pieke I-IU; ΙΙΙ-Ven I-V; absolute amplitude van pieke I en V en die relatiewe amplitude-verhouding van pieke V:I. Hierdie resultate is bespreek met verwgsing na die literatuur. Die implikasies vir kliniese toepassing en verdere navorsing is ook aangedui. INTRODUCTION During the past decade and a half, there has been a formidable increase in the use of specialised audiological test procedures in otoneurological diagnosis. Since the initial description of a procedure from the human scalp (Jewett & Williston 1971), the measurement of the brain-stem auditory evoked response (BAER) has become the most recent electrophysiological pro- cedure to be integrated into testing protocols. This procedure subsequently had a major impact on the disciplines of audio- logy, otology and neurology, (Schwartz & Berry, 1985). The development of the BAER was focused on two principal areas of application: I i. the evaluation and diagnosis of peripheral auditory pro- blems and related pathology, and ii. the assessment of the neural integrity of the acoustic nerve and caudal levels of the brain-stem afferent audi- tory pathway (Hecox & Jacobson, 1984). However, despite the reported robustness and stability of the BAER as a reliable assessment tool, it is critical to the effective use of this measurement to have diagnostic reference data that are collected within the individual laboratory or clinic. Several investigators have reported on the myriad of variables that can potentially alter one or more of the important parameters of the BAER, and hence lead to misinterpretation. It is, therefore, appropriately suggested by Schwartz & Berry (1985), that it is not advisable for any clinician to depend on published diag- nostic reference data for interpreting BAERs. This emerges from the opinion that there is a lack of uniformity in BAER Die Suid-Afrikaanse Tydskrif vir Kommunikasieafykins, Vol. 37 1990 measurement variables among investigators in various clinics and laboratories around the world. Furthermore, although there is available information on diag- nostic reference data on the normal BAER (see table 1, adapted from Schwartz & Berry, 1985), such information is lacking in South Africa for any age or population group. Due to the absence of standard to specify recording parameters and methods used to measure the BAER, (this is clearly seen in table 1 whereby the various clinics and or laboratories differ in their test protocols), it is imperative that any clinic, includ- ing the Audiology Clinic at the University of Durban-Westville establishes its own diagnostic reference data based on its own test equipment'and protocol. BAER interpretation maybe con- founded by the influence of various factors, viz. differences m: a. electrical and electromagnetic field variation between clinical/laboratory sites b. the use of different stimuli c. recording and analysis parameters d. electrode placement e. transducer type and f. transducer placement. The above differences may lead to small but significant changes in peak latency, amplitude and morphology. In addi- tion a number of investigators (Beagly & Sheldrake, 1978; Jerger & Hall, 1980; and Stockard et al. 1978) reported that there is a ftiarked sex effect on the normal BAER. Generally, females show shorter latencies and larger amplitudes than SASHA 1990 R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 2) 60 Cyril D Govender males. The difference is greater for peak V than for the earlier peaks, and also produces shorter IPLs within females. The clinical implication of this differential effect is that it is advis- able to establish separate reference data for males and females to avoid misinterpretations (Hall, 1984). Therefore a study designed to elicit and to examine the BAER in a group of nor- mal male and female subjects under controlled conditions would provide relevant information, leading to the establish- ment of suitable diagnostic reference data for the interpreta- tion of BAER in the Audiology Clinic at the University of Durban-Westville. METHOD Aim The aim of the study was to examine certain response charac- teristics of the BAER obtained in a selected group of normal hearing Indian male and female subjects. This would allow for the generation of diagnostic reference data for the interpreta- tion of BAER at the Audiology Clinic at the University of Durban-Westville. The response characteristics of the normal BAER to be examined were as follows: a. absolute latencies of peaks I to VI b. relative (interpeak) latency (IPL) of peaks ITII, III-V and peaks I-V. c. peak to trough amplitudes of peaks I and V d. amplitude ratio obtained by comparing the absolute amplitudes of peak I and V. SUBJECTS A total of 60 randomly selected subjects (X age 20.7 years) comprising of 30 males (X age 21.3 years) and 30 females (X age 20.3 years) contributed relevant data fc*· the purpose of this study. All subjects had normal hearing thresholds of 0- 26dB for AC and BC for the test frequencies 250Hz to 8000Hz, SRTs which were within ±5dB of the PTAs, speech dis- crimination scores of 92% to 100% at 35dBsl, normal type A tympanograms, static compliance measures ranging between 0.28 and 2.5cc, and contralateral acoustic reflex thresholds of between 70dBsl and 90dBsl. Subjects also had a negative his- tory of neurological abnormalities, consuming known oto- toxic drugs and of excessive noise exposure as evidenced in the pre-test case history questionnaire (Govender, 1989). Fur- thermore, all subjects were right-handed as determined by a standardised handedness questionnaire (Lazarus, 1989), so as to select the right ear as the test ear. This was to ensure consis- tency in testing the same ear for all subjects. PROCEDURE All subjects were assessed in the supine position on a standard patient couch with appropriate head propping by using a pillow to minimize postural muscle activity around the head and neck regions (Chiappa et al. 1979). Subjects were en- .couraged to relax and to fall asleep during the recording ses- sion as this would reduce myogenic activity from influencing the responses. Therefore actual testing commenced only if subjects appeared to be relaxed or asleep. The electrode sites were cleaned of all debris with omni-prep skin preparing paste and were slightly abraded to assist in reducing skin resistance. Self-adhesive silver-silver chloride electrodes were then arranged so that the electrical potential difference was measured between a pair of electrodes as sug- gested by Schwartz & Berry (1985). The positive electrode was placed on the high forehead just below the hairline, the nega- tive electrode was placed on the ipsilateral mastoid, while the contralateral mastoid was used as the site for the ground elec- trode. Prior to fixing the electrodes to the skin, standard EEG paste (colloidon) was applied between the surface of the elec- trode and the skin. The electrodes were then fixed firmly to the skin. The recording electrodes were connected to a high gain (104) differential, low noise, biological preamplifier (the Cadwell Quantum 84 preamplifier). The inter-electrode impedance was always below 3000Ω before testing commenced. Elec- trode activity was differentially filtered with a filter bandpass setting of 100Hz to 3000Hz, as suggested by the Cadwell pro- gramme for BAER testing. Alternating click stimuli with a duration of lOOpsec were pre- sented at a rate of 11.29 per sec at 70dBHL to the target right ear via a TDH-39 earphone, housed in an Amplivox freefield audiocup. The headphone position was held constant for all subjects. The no-test left ear was masked with 60dBHL of broadband noise so as to prevent its participation in obtaining the monaural BAER in the right ear. The sweep time was set at 1 so that each division on the moni- tor screen represented 1 msec. Therefore, the BAER was ob- served over a time frame of 10msec post-stimulus. A total of 2048 clicks was presented to ensure waveform build-up and clarity. The artifact rejection facility of the evoked potential system was switched on continuously so as to allow for auto- matic rejection of artifacts. All responses were recorded using the built in Alps printer of the evoked potential system. All BAERs were elicited using the Cadwell Quantum 84 com- puter based, software run, auditory evoked response audio- meter. Testing was conducted in a electro-magnetically screened anechoic sound treated chamber meeting noise level requirements set by ANSI (1977). The actual test run was initiated as soon as all test parameters ι were set. Prior to any stimulus presentation, a control run was j done to allow for comparing and identifying of true responses. : At least two trials were done to ensure waveform repeatability ! and consistency. NB. See appendix 1 for a summary of the j BAER test protocol used in this study. I I MEASUREMENTS MADE ! ι 1. Absolute latencies. , The absolute latencies of peaks I to VI in msec were made from stimulus onset to the positive peak of each component wave of the BAER. 2. Relative or Interpeak IPL latencies. The relative latencies in msec, of peaks ITII, III-V and I-V were recorded from the target (R) ear. These were automatically calculated by the computer of the evoked response system. 3. Absolute amplitudes. Peak to trough amplitudes of peaks I and V inmicrovolts were measured. These were measured from the-positive peak to the following negative trough of each of the named component waves of the BAER. The South African Journal of Communication Disorders. Vol. 37, 1990 R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 2) Diagnostic Reference Data for the Monaural Brain-stem Auditory Evoked Response (BAER) 61 4. Amplitude ratio. This ratio was obtained by comparing the amplitude of peak V to that of peak I. The ratio was obtained by dividing the amplitude of peak I by that of peak V, as suggested by Musiek & Gollegly, (1985). ANALYSIS OF DATA The data were analysed by using conventional statistical pro- cedures which allowed for the generation of means, ranges and standard deviations. RESULTS Table 2 summarises the statistical analyses made on the va- rious latency and amplitude measurements obtained from the combined group (N=60). Table 3 reflects a summary of the statistical analysis conduc- ted for the various latency and amplitude measurements ob- tained among 30 female and 30 male subjects. This table was drawn up to specifically reflect the data for males and females separately. Tables 2 and 3 therefore represent the diagnostic reference data related to the aim of the study. DISCUSSION Table 2 provides a summary of the latency and amplitude measures made on the monaural BAER elicited in 60 normal hearing young students. The diagnostic reference data reflec- ted in table 2 is to be discussed in relation to previously pub- lished data under the following headings: a) Absolute latencies Table 2 reflects the overall group absolute latency values for peaks I to VI. The obtained range extends from 2.08 to 7.49 msec in response to clicks presented at 70dBnHL. The latency range for peaks I to V was 2.8 to 6.00 msec. Beagley & Shel- drake (1978), in reporting their "normative data" obtained from 5 male and 5 female subjects (age range 21-30 years) for peaks I to V show a similar latency range of 2.1 to 6.1 msec in response to clicks presented at 70dBsl. However, in comparing the findings of the above studies to those obtained in other clinics and laboratories (see table 1), it is evident that there are small variations between and among the values reported. These variations may be attributed to dif- ferences in intensity, polarity of clicks, repetition rates, filter setting and other aspects of test protocols used. The fact that these discrepancies between and among clinics and laborato- ries exist, highlights the need for each facility to generate its own diagnostic reference data. An examination of the individual studies reported in' table 1 reveals two major differences: Laboratory/Clinic 1 2 3 Filter settings (Hz) Wave latency (ms) Laboratory/Clinic 1 2 3 Filter settings (Hz) I II III IV V I-III III-V I-V Jewett'and Williston, 60-75 ? ? 10-10,000 1.5 2.6 3.5 4.3 5.1 - - - / Ί ' 9 7 0 ) Lev and Sohmer, 65 ? 2 250-5000 1.5 2.5 3.5 - 5.0 - - - (1972) Picton et al. 60 ? 10 10-3000 1.5 2.6 3.8 5.0 5.8 - - - (1974) Starr and Achor, 65 alt. 10 100-3000 1.6 2.8 3.8 4.8 5.5 - - - (1975) Stockard and Rossiter, 60 ^ r a r . 10 100-3000 1.9 3.0 4.1 5.2 5.9 2.1 1.9 4.0 (1977) Rosenhamer et al. 60 ? 16.6 180-4500 1.7 2.9 3.9 5.2 5.9 2.26 2.0 4.27 (1978) Row, 60 ? 10 100-3000 1.9 2.9 3.8 5.1 5.8 1.97 1.97 3.94 (1978) Gilroy and Lynn, 75 ? 11 150-3000 1.55 2.67 3.60 4.69 5.40 2.05 1.9 3.83 (1978) Beagley and Sheldrake, 70 ? 10 250-3200 2.1 3.3 4.3 5.3 6.1 2.2 1.8 4.0 (1978) Chiappa et al. 60 alt. 10 100-3000 1.7 2.8 3.9 5.1 5.7 2.1 1.9 4.0 (1979) Schwartz and Berry, 75 dB rar. 11.1 75-1500 1.65 2.85 3.8 4.99 5.66 2.05 1.85 4.00 (1985) (nHL) 1 = Stimulus intensity level in dB 2 = Stimulus polarity | 3 = Repetition rate (cps) Table 1: Normative ABR latency data across 11 laboratories. Adapted from Schwartz & Berry, ( 1 9 8 5 ) . Die Suid-Afrikaanse Tydskrif vir Kommunikasieafivykinijs, Vol. 37. 1990 R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 2) 6 2 Cyril D Govender i. Some researchers did not always clearly define the refe- rence intensity level for the clicks used, ie. whether the clicks were presented in dBsl, dBnHL, dBspl or dBHL. This aspect is considered important as it reflects on exactly how loud a click is presented. Furthermore, it is well documented that there is a direct relationship be- tween intensity level and the latency of the peaks ob- tained in the BAER (Rowe 1978, Stockard et al. 1978), ie. as the intensity of the click increases, the latencies of the peaks decreases and vice-versa (Moore 1983). It is, therefore, suggested that investigators clearly define the reference intensity levels used. This suggestion is sup- ported by the American Electroencephalic Society guidelines for clinical evoked potential studies (1984). Adherence to these guidelines would permit the uni- form use of reference intensity levels and would allow for the comparison of reference data between and among clinics and laboratories. ii. Investigators did not always clearly define the polarity of clicks used. BAERs are affected by the acoustic phase of clicks (Gastone et al. 1987). Rarefaction clicks evoke shorter peak I and V latencies (Coats & Martin 1977); although these vary considerably among subjects. Con- densation clicks, however, appear to delay peak I and V latencies (Schwartz & Berry 1985), while alternating clicks do not seriously compromise BAERs, but serve to enhance the clarity of the response (Schwartz & Berry, 1985). It is therefore recommended that investigators clearly define the acoustic phase of clicks used in the generation of "normal" reference data. This would faci- litate the development of uniform procedures and hence allow for inter-facility comparison. It is conceivable that among the variables mentioned earlier, both variation in intensity reference levels and the polarity of clicks used, may principally account for the differences in nor- mal values reported in table 1 and in this investigation. In considering the appearance of individual peaks, all six peaks except for peak IV were consistently elicited in this investigation. Six subjects (2 males and 4 females) did not pro- duce clear and measurable peak IV latencies. According to Beagley & Sheldrake (1978), peak IV tends to be a more labile peak while Rowe (1978), states that it may, in some normal subjects be absent. This may be attributed to the fact that peak IV sometimes tends to fuse with peak V thereby making it indistinct (Chiappa et al. 1979). Peak I was consistently elicited in a latency range of 1.87 to 2.29 msec. This finding is consistent with Picton's (1986), recommended range of 1.4 to 2.5 msec. The mean of 2.08 msec, however, appears to be slightly delayed when compared to the means reflected in table 1. Among the other variables men- tioned earlier, this may be attributed to variations in testing and measurement protocols used in the different laboratories and clinics, particularly to differences in intensity levels of click phases used. Generally, the absolute latencies of peaks II, III, IV and V show close approximation with those presented by Beagley & Shel- drake (1978). However, in comparison with the other studies reflected in table 1, these latencies appear to be slightly delay- ed, but the inter-facility standard deviations show close agree- ment. This provides support for the claim that the BAER is a stable and reliable measure at moderate to high intensity levels. The inspection of the peak V latency, which, according to Schwartz &Berry (1985) and Stockard et al. (1978), should occur within 4.00 msec after peak I, reveals that the finding of this study is consistent with the above i.e. peak V (see table 1) had a mean latency period of 3.90 msec after peak I. According to several researchers, e.g. Beagley and Sheldrake (19 78), Stoc- kard et al. (1978) and Picton (1986), peak V is the most consis- tent and prominent of the BAERs, and is probably most useful diagnostically. Its appearance at a mean latency period of 6.00 msec in this study conforms well with findings reported by Stockard & Rossiter (1977), at 5.9 msec, Rosenhamer et al. (1978), at 5.9 msec and of Rowe, (1978) at 5.8 msec. Further- more, the standard deviation of the peak V latency as reported by all these researchers did not differ significantly from that of this study, i.e. a standard deviation of 0.23. Clearly, peak V latency appears to be robust in character. It is reliable and sta- ble even under varying measurement conditions. This conten- tion is supported by the evidence presented in table 1. It is, therefore, not surprising to note that peak V (a rostral compo- nent of the BAER) has received widespread clinical attention in differential diagnosis of otoneurologic disorders, as well as for the estimation of hearing sensitivity (Schwartz & Berry, 1985). The finding that there were some overall variations be- tween this study and of those summarised in table 1, illustrates and highlights again the need for each clinic or test facility to establish its own diagnostic reference data. b) Relative or inter-peak latencies Table 2 also reflects the mean relative latency values obtained in the group of 60 normal hearing students. These include the values for the relative latencies of peaks ITII; III-V and peaks I- V. Crucial to the differential diagnosis of space occupying lesions, either intrinsic or extrinsic to the brain-stem, is the time difference between peaks. These time differences are reflected by the time intervals between the following: i. peaks ITII as representing peripheral transmission time from stimulus onset to the ponto-medullary junction in the lower pons (Stockard et al. 1978) ii. peaks III-V as reflecting central transmission time from caudal pons to the midbrain (Schwartz & Berry, 1985) and | iii. peaks I-V as representing both peripheral and central transmission time from stimulus onset to the midbrain (Schwartz & Berry, 1985). , i Peripheral transmission time is determined by middle-earj function, cochlea mechanics, cochlea transduction, synaptic; and cochlea nerve conduction velocity while central transmis-1 sion time is associated with fibre conduction velocity and syn-. aptic transmission of brain-stem tracts and nuclei (Cornacchia et al. 1983). The mean relative latency values obtained in this investiga- tion were as follows: peaks ITII = 2.01 msec peaks III-V = 1.88 msec ' . ·" peaks I-V = 3.90 msec These values coincide well with previously published data as reflected in table 1. Furthermore, these values fit in well with the suggested values presented by,Schwartz & Berry (1985); these being ± 2 msec for peaks I-III, and III-V and ± 4 msec, for peaks I-V in normal hearing subjects. According to Rowe The South African Journal of Communication Disorders, Vol. 37 1990 R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 2) Diagnostic Reference Data for the Monaural Brain-stem Auditory Evoked Response (BAER) 6 3 (1978), these relative latency values should not vary between and among laboratories and clinics using the same rate of stimulus presentation. In summary, this investigation established relative latency values which are similar to those found in other clinics and laboratories having used a click stimulation rate of 10-12 clicks per sec (see table 1). The lack of variability in these measures between and among facilities using the same click rate, makes them robust measures of peripheral and central transmission time within the auditory system. Therefore, the interpeak latency measures are suitable for assessing patholo- gies which may affect the transmission of auditory impulses in the peripheral and brain-stem part of the auditory system. c) Absolute amplitudes In referring to table 2, two absolute amplitude measures were considered in this investigation. These were the absolute amp- litudes of peaks I and V. The obtained values in μ ν were used for the computation of the more clinically acceptable ampli- tude measure, i.e. the amplitude ratio of peak V compared to peak I. Table 2: Summary statistics for absolute and relative latencies in milli-seconds and absolute and rela- tive amplitude measurements in micro-volts of the monaural BAER for the combined group. (N=60) BAER Measures Statistical Measures Absolute Latencies (msec) X Range SD Peak I 2.08 1.87-2.29 0.08 Peak II 3.02 2.17-3.37 0,18 Peak III 4.12 3.71-4.52 0.17 Peak IV (N=54) 5.13 4.42-5.83 0.30 Peak V 6.00 5.50-6.58 0.23 Peak VI >'7.49 6.54-8.21 0.32 Relative Latencies (msec) ^Peak I-III 2.01 1.69-2.41 0.15 Peak III-V 1.88 1.48-2.41 0.18 Peak I-V 3.90 3.45-4.37 0.22 Absolute Amplitudes (μν) Peak I J 0.17 0.08-0.36 0.05 Peak V |0.24, 0.08-0.50 0.08 Relative Amplitude (μν) Peak V:I 1.50 0.55-3.85 0.67 N.B. For Peak IV, Combined group N=54 Table 2 above reflects the diagnostic reference data for the various latency and amplitude measurements obtained from the combined group (N=60). 1 There is consensus among researchers, viz. Schwartz & Berry, (1985), Rowe, (1978), and Chiappa et al.(l979), that absolute amplitude measures are not normally distributed; are highly susceptible to myogenic activity and noise levels; are difficult to replicate, and are easily influenced by minor alterations in recording techniques. Consequently, the measurement of ab- solute amplitudes do not enjoy the stability and reliability of their latency counterparts (Schwartz & Berry, 1985). In this study, the mean peak I amplitude value was ο. 17 μν and that of peak V was 0.24 μν. Chiappa et al. (1979), presented a mean peak amplitude value of 0,28 μν and a mean peak V value of 0.47μν. Stockard et al. (1978), published a mean value of 0,23 μ ν for peak I and 0,35 μν for peak V. It is not clear that there are no close approximations between and among repor- ted measures. These reported variations in amplitude measures between and among normal hearers may be attributed to the present system of signal averaging and use of artifact rejection (Fer- nandes, 1989). Theoretically, a wanted evoked potential is extracted from ongoing EEG by signal averaging and the use of artifact rejection. That is, by increasing the signal-to-noise ratio. Waveform and amplitude build up is, therefore, a pro- duct of time-locked averaging together with the rejection of other contaminating artifacts, e.g. myogenic and other cere- bral activity. It has been found that consensus among researchers on how much of averaging and/or artifact rej ection is required before a response is judged as acceptable or not, is lacking. According to Hyde (1985), the choice of the number of clicks presented for averaging is often "based on popular consensus rather than on quantitative rationale". Due consideration has not been given to the influence of differences in "internal noise levels" among normal hearers when reference data are established. That is, some normal subjects may have higher internal noise levels, requiring longer periods of averaging with greater number of averages within a trial before eliciting an appropriate response than subjects who have lower internal noise levels (Hyde, 1985). Therefore a choice of either 1048,200 or 2048 clicks to elicit a suitable averaged response may not be appropriate for all normal hearers. Furthermore, since the amplitude of a res- ponse is partly dependent on the number of averages that occur in a trial, it is reasonable to assume that response ampli- tudes will differ between and among individuals. This there- fore may account for the variability in amplitude measure- ments that are reported in the literature. Similarly, the use and control of artifact rejection to eliminate unwanted noise is not consistent in studies that have reported on normal amplitude values. It is therefore not surprising to find variations in the reported amplitude values between and among studies. The consistent and approriate application of signal averaging and management of artifact rejection needs to be given careful attention in future research. Attention needs to be focused on decisions pertaining to the: i. Actual number of averages required in a trail (i.e. 1048, 2000 or 2048 clicks) before a response is regarded as representative of a "true neurogenic" response. ii. Use and control of artifact rejection so that the final res- ponse is truely representative of the BAER without being contaminated by other artifacts. A reasonable course of action, is to set the artifact rejection limits so that little of the "well behaved" (low variance) acti- vity is rejected, while all of the high variance (bursts of elec- tromyogenic noise) activity is. This may be done by "tuning" the rejection level while observing the displayed activity, so that only about 5-10% of the good activity is rejected. Perhaps, the manufacturers of evoked potential systems need to incor- /» Suid-Afrikaanse Tydskrif vir Kommimikasirafityikhi/is. Vol. .17, 1990 R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 2) 6 4 Cyril D Govender porate additional desirable features that will allow for the dis- play of the input EEG during averaging, rejection of trails in which large voltage artifacts occur and an assessment of ampli- tude variability within an averaging run. The above may assist in establishing appropriate reference data for amplitudes which may be used routinely in BAER interpretation. Thereafter, such reference data should be applied widely to assess how otoneurologic pathologies in- fluence the measures, and to document the obtained patterns for ongoing comparisons. d) Relative amplitude - the peak V:I amplitude ratio Table 2 reveals, that the mean amplitude ratio obtained in this investigation was 1.50. This is consistent with the findings of Chiappa et al. (1979); Rowe (1978), and that of Starr & Achor (1975), who have all reported that a value greater than 1.00 be considered as normal. In order to detect abnormality, Musiek et al. (1984), state that amplitude ratio should be less than 1.00. Stockard et al. (1978), however, state that a complete absence of peak V in the presence of peak I is an indication of relative amplitude abnormality. Differing in this opinion, Starr & Achor (1975), state that a peak V:I amplitude ratio ofless than 0,5 at 55 dBsl is abnormal. Later in 1978, Stockard et al. suggested that the peak V abso- lute amplitude value which is reduced by more than 3 sd from the normal mean, together with a peak I amplitude that is larger than peak V, and an inter-trial variation ofless than 10% are all necessary for the peak V:I amplitude ratio to be defined as abnormal. Chiappa et al. (1979), agree with Starr & Achor (1975),in the lOoftheir 104 normal subjects displayed apeak I amplitude which was larger than peak V. The findings of this investigation are in part agreement with Starr & Achor (19 75), and with Chiappa et al. (1979), since 12 subjects (5 females and 7 males) displayed peak I amplitudes which were larger than peak V, although the overall mean was 1.50. The obser- ved differences in amplitude ratios appear to be due to normal variations that occur within and among normal individuals. This contention is in keeping with Stockard et al.'s (1977), statement that "alterations of BAER morphology in the ab- sence of quantifiable latency or absolute amplitude abnor- mality are not considered abnormal per se, because of the variability of BAER waveforms within and among normal individuals." However, Schwartz & Berry (1985), are of the opinion that there is a dearth of well documented literature concerning the use of the V:I amplitude ratio in a large pathologic population. They suggest that considerable research is needed on the con- founding effects of such variables as stimulus polarity, repeti- tion rates, filter characteristics, electrode sites etc., prior to the general use of this measure in clinical practice. The inves- tigator concurs with the above recommendation. Due con- sideration should also be given to inter and intra individual variations when examining amplitude data. Furthermore, and improvement in signal averaging and artifact control may aid in resolving the issue of obtaining variable amplitude meas- ures in normal hearers. SEPARATE DIAGNOSTIC REFERENCE DATA FOR MALES AND FEMALES In response to the suggestion made by several researchers, viz. Stockard et al. (19 78), (19 79); Jerger & Hall (1980), and Jerger & Johnson (1988), that diagnostic reference data be establish- ed separately for males and females, the raw data was further treated to reflect this separation. Table 3, reflects the means, ranges and standard deviations for the various BAER measurements as obtained from 30 females and 30 males. On inspection and comparison of the mean absolute latency values obtained for the two groups, it is evident that for all six peaks, females tended to show shorter latency values than males. This is also evident for the peak I-III and peak I-V rela- tive latency values. The absolute and relative amplitude measures show no such differences, implying that there are no observable differences between sexes for these measures in this investigation. How- ever, further research focusing on the appropriate use of signal averaging and artifact rejection may produce realistic ampli- tude measures in normal hearers. Once this has been achieved, it is suggested that the effect of sex difference on amplitude measurements be reassessed. The question of whether there is a statistically significant sex difference effect on the normal BAER, needs to be research- ed further. ι In the interim, the fact that there are observed latency differen-j ces between the sexes as seen in table 3, is supportive of the suggestion that separate diagnostic reference data be esta-i blished for the two sexes. The establishment of such data, would prevent the clinician from applying inappropriate sex| related reference data to interpret the BAER. | / The South African Journal of Communication Disorders, Vol. 37, 1990 R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 2) Diagnostic Reference Data for the Monaural Brain-stem Auditory Evoked Response (BAER) 6 5 Table 3: Summary statistics for absolute and relative latencies in milli-seconds and absolute and relative amplitude measurements i n micro-volts of the monaural evoked BAER in females (N=30) and males (N=30) BAER Measures (P=Peak) Statistical Measures BAER Measures (P=Peak) Males (N=30) BAER Measures (P=Peak) Females (N=30) Males (N=30) Absolute Latencies X Range SD X Range SD Ρ I 2.06 1.87-2.19 0.07 2.11 1.92-2.29 0.10 Ρ II 3.02 2.71-3.29 0.14 3.4 2.79-3.37 0.14 Ρ III 4.08 3.71-4.52 0.18 4.16 3.87-4.50 0.16 Ρ IV 5.10 4.42-5.83 0.37 5.17 4.58-5.52 0.21 Ρ V 5.98 5.54-6.58 0.24 6.03 5.50-6.42 0.22 Ρ VI 7.44 6.75-7.96 0.30 7.55 6.54-8.21 0.33 Relative Latencies Ρ I-III 2.00 1.74-2.41 0.16 2.03 1.69-2.37 0.15 Ρ III-V 1.87 1.48-2.41 0.18 1.87 1.56-2.29 0.17 Ρ I-V 3.84 3.49-4.37 0.22 3.91 3.45-4.33 0.23 Absolute Amplitude Ρ I 0.17 0.08-0.36 0.05 0.16 0.09-0.29 0.05 Ρ V 0.25 0.09-0.44 0.08 0.22 0.08-0.50 0.08 Relative Amplitude Ρ V:1 1.50 0.74-2.50 0.51 1.50 0.55-3.85 0.81 N.B. For peak IV: Female no. = 26 Male no. = 28 Table 3 represents the means, ranges and standard deviations for the various BAER measurements as obtained for the females and males respectively. CONCLUSION Diagnostic reference data were established for both the com- bined group (N=60) and separately for females and males. Similarities an differences between this study and of those reported in the literature were noted and discussed. The simi- larities in absolute latency measures were attributed to close approximations between testing protocols used, whilst varia- tions were primarily related Jto, among other variables, dif- ferences in reference intensity levels and the polarity of clicks. Despite the difference between this study and of those reflec- ted in table 1, the absolute latency of peak V remained resistant to variations in stimulus, recording and "normal subject" variables. Therefore this measure,appears to be robust and maybe reliably used in otoneurological diagnosis and for es- timating hearing sensitivity. The relative latency values generated are consistent with those reported in the literature (see table 1), and this is attri- buted to the fact that click presentation rates used are similar, i.e. 10-12 clicks per sec., in each of the studies. It is therefore suggested that clinicians may confidently use these measures to assess otoneurological pathologies that may upset the con- duction of impulses in the auditory periphery (e.g. multiple sclerosis), provided that the click rate used is 10 to 12 per sec. Differences in amplitude measures between this study and among other studies were noted. These variations were, among other factors, attributed to the manner in which signal averaging and artifact rejection have been manipulated in obtaining the average BAER. Further research in this respect has been suggested. However, the relative amplitude value of 1.50 obtained in this study is consistent with those reported in the literature. This implies that the RA measure is less variable in normals and therefore, may be used as a more sensitive measure of brain-stem auditory function than absolute ampli- tude measures. In view of the demonstrated differences in reference data be- tween and among clinics and laboratories, the writer is of the opinion that clinicians should exercise caution in using refer- ence data established elsewhere, especially if reported testing protocols differ in stimulation, recording and normal subject variables, e.g. sex. The observation that there were differences between the sexes is strongly supportive of the suggestion that separate reference data be established for the sexes (Stockard et al. 1978; Jerger & Hall, 19.80). This would allow for the accurate clinical interpretation of the BAER obtained in the two sex groups. Therefore it is recommended that each clinic generates its own reference data commensurate with its needs. Furthermore, noting that this study fell short of giving due consideration to age-related data across the continuum, inter- aural latency differences, use of different repetition rates, stimulus intensity reference levels and click polarity, future research considering the above, needs to be conducted to ex- tend the present reference data base. The need for consensus to be reached among researchers and clinicians with respect to test protocols used in BEAR testing cannot be overemphasized. Perhaps, an international con- ference involving the various disciplines that use this test pro- R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 2) 6 6 Cyril D Govender cedure should be held, in order to formulate a standard guide- line or protocol for the use of interpretation of the BAER. This would facilitate inter-clinic and/or laboratory comparisons, and perhaps aid in resolving some of the controversies that exist in BAER testing and interpretation. In the interim, it is important the researchers and clinicians clearly define the parameters of their test protocols in estab- lishing reference data. In addition, such data should be ap- plied within populations having known otoneurological pa- thologies to assess the extent to which the reference data is able to differentiate normal from pathological ears. The latter is also applicable to the reference data generated in this study. REFERENCES American Electroencepalographic Society. Guidelines for clinical evoked potential studies. Journal of Clinical Neurophysiology, 1, 3-53, 1984. American National Standards Institute·.-Standard criteria for per- missibility of ambient noise during audiometric testing. ANSI 53:1977 New York: American National Standards Institute, 1977. Beagley, H.A., and Sheldrake, J.B. Differences in brain-stem response latency with age and sex. British Journal of Audioloyy, 12, 69-77, 1978. Chiappa, K.H., Gladstone, K.J., and Young, R.R. Brain-stem auditory evoked responses. Studies of waveform variations in 50 normal human subjects. Archives of Neurology, 36, 81-87, 1979. Coats, A.C. and Martin, J.L. Human auditory nerve action potentials and brain-stem evoked responses. Archives of Otolaryngology, 103, 605-622, 1977. Cornacchia, L., Martini, Α., and Morra, B. Air and bone conduction brain-stem responses in adults and infants. Audiology, 22,430- 437, 1983. Fernandes, C.M.C. ENT Department, Medical School, University of Natal, Personal Communication, 1989. Gastone, G.C., and Grigg, M.M. Auditory evoked potentials. In E. Niedermeyer and F. Lopes da Silva (eds.), Elect roencephalo- yraphy: Basic principles, Clinical applications and Related fields. Baltimore: Urban and Schwarzenberg. 1987. Govender, C.D. An investigation into the effects of sex difference and contralateral masking on the monaural auditory evoked res- ponse (BAER) obtained in a group of normal hearing Indian undergraduate university students. Unpublished masters disser- tation, University of Durban-Westville, 1989. Hall, J.W. Auditory brain-stem audiometry. In J. Jerger {ed.), Hearing Disorders in Adults. San Diego: College-Hill Press, Inc, 1984. Hecox, K., and Jacobson, J.T. Auditory evoked potentials. In J.L. Northern (ed.), Hearing Disorders. Boston: Little-Brown, 1984. Hyde. M.L. Instrumentation and Signal Processing. In J.T. Jacobson (ed.), The Auditory Brain-stem Response, San Diego: College-Hill Press, Inc, 1985. Jerger, J., and Hall, J. Effects of age and sex on auditory brain-stem res- ponse. Archives of Otolarynyoloyy 106, 387-391, 1980. Jerger, J., and Johnson, K. Interactions of age, gender and sen- sorineural hearing loss on ABR latency. Ear and Hearing, 9. 190-197, 1988. Jewett, D.L., and Williston, J.S. Auditory evoked responses for fields averaged from the scalp of humans. Brain, 94, 681-696, 1971. Lazarus, T. Annets Handedness Questionnaire. Modified and adap- ted for S.A. use. Unpublished, available from author, Depart- ment of Psychology, University of Durban-Westville, 1989. Moore, E.J. Effects of stimulus parameters. In E.J. Moore (ed.), Bases of auditory brain-stem evoked responses. New York: Grune and Stratton, 1983. Musiek, F.E., and Gollegly, K.M. ABR in Eighth nerve and low brain- stem lesions. In J.T. Jacobson(ed.), The auditory brain-stem res- ponse. San Diego: College-Hill Press, 1985. Musiek, F.E., Kibbe, K., Rackliffe, L., and Weider, D.J. The auditory brain-stem response I-V amplitude ratio in normal, cochlear and retrocochlear ears. Ear and Hearing, 5, 52-55, 1984. Picton, T.W. Abnormal brain-stem auditory evoked potentials: A ten- tative classification. In R.Q. Cracco and I. Bodis-Wollner (Eds.), Frontiers of clinical neuroscience. Evoked Potentials, Vol. 3. New York: Alan R. Liss Inc, 1986. Rosenhamer, H., Lindstrom, B., and Lundborg, J. On the use of click evoked electric brain-stem responses in audiological diagnosis, i. The variability of the normal response. Scandinavian Audiol- ogy, 7, 197-206, 1978. Rowe, M.J. Normal variability of the brain-stem auditory evoked res- ponse in young and old subjects. Electroencephalography and Clinical Neurophysiology, 44, 459-470, 1978. Schwartz, D.M., and Berry, G.A. Normative aspects of the ABR. In J.T. Jacobson (ed.), The Auditory Brain-stem Response. San Diego: College-Hill Press, 1985. Starr,Α., and Achor, J. Auditory brain-stem responses in neurological disease. Archives of Neurology, 32, 761-768, 1975. Stockard, J.J., and Rossiter, V.S. Clinical and pathologic correlates of brain-stem auditory response abnormalities. Neurology, 27, 316-325, 1977. , Stockard, J.J., Stockard, J.E., and Sharbrough, F.W. Non-pathologicl factors influencing brain-stem auditory evoked potentials.] American Journal of Electroencephalographic Technology, 18,: 171-209,1978. ' ; R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 2) 6 7 APPENDIX I BEAR - TEST PROTOCOL TECHNICAL AND PROCEDURAL CONSIDERATIONS STIMULUS : TRANSDUCER ELECTRODES EVOKED RESPONSE AUDIOMETER ELECTRODE SITES POLARITY REPETITION RATE FILTER PASS BAND SWEEP TIME TIME FRAME NO. OF CLICKS PER TRIAL NO. OF TRIALS CONTRALATERAL MASKING LEVEL OF TEST EAR STIMULUS ARTIFACT REJECTION I RECORDING OF RESPONSES / / TEST ENVIRONMENT PATIENT STATE clicks - ΙΟΟμ sec. duration electrodynamic TDH-39P earphones housed in free field audio-cups. self-adhesive silver-silver chloride. Cadwell Quantum 84 positive - Fz high forehead negative - ipsilateral mastoid ground - contralateral mastoid alternating 11,29 per sec 100Hz - 3000Hz 1 division = 1 msec 10 msec post-stimulus 2048 Minimum-Two to ensure waveform repeatability. 60dBHL kept constant at 70 dBnHL switched on by built in ALPS Printer ANECHOIC Chamber - electromagnetically screened low noise levels ANSI (1979). appeared to be relaxed or asleep lying in a supine position on a standard patient couch. N.B.: A control run prior to stimulation was done to allow for comparing and identifying true responses. R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 2) < */ * ν U v -sex m i<§ r' JSpvs. •s T a l k i n g to P r o f e s s i o n a l s The Needier WestdeneOrganisation P.O. Box 28975 Sandringham 2131 Telephone (011) 485-1302/3/4/5' Wm R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 2)