The use of cortical evoked response audiometry in the assessment of noise-induced hearing loss

The use of cortical evoked response audiometry in the assessment of noise-induced hearing loss STEPHEN W. HONE,

FRCSI(ORL),

GARRY NORMAN,

MSC,

IVAN KEOGH,

OBJECTIVES: The study aims were to determine the incidence of exaggerated hearing thresholds in individuals complaining of noise-induced hearing loss (NIHL) as a result of impulse noise using cortical evoked response audiometry (CERA) and to identify any associated audiometric features. SETTING: We conducted an office-based study. STUDY DESIGN: In this prospective case series, 1154 males complaining of NIHL were assessed with pure tone audiometry; 673 had CERA. Pure tone averages (PTA) and hearing disability were calculated using the Irish and American Medical Association systems. A PTA of >10 dB worse than the CERA average was considered evidence of exaggerated thresholds. RESULTS: The mean PTA was 33 dB. Seventy-two percent had a hearing disability of an average of 26% when assessed by the Irish system. Fifty-four percent had a hearing disability of an average of 30% when assessed by the American Medical Association system. Twenty-six percent of subjects had exaggerated thresholds based on CERA. A binaural hearing threshold of >25 dB at 500 Hz had a sensitivity of 94% and a specificity of 59% for the detection of exaggerated thresholds. CONCLUSION: Exaggerated hearing thresholds are common. A hearing threshold of >25dB at 500 Hz should be considered an indication for CERA testing. (Otolaryngol Head Neck Surg 2003;128: 257-62.)

T here has been an escalating rate of legal claims for noise-induced hearing loss (NIHL) in Ireland in recent times. Because there is considerable fiFrom the Department of Otolaryngology–Head and Neck Surgery, Beaumont Hospital. Presented at the Annual Meeting of the American Academy of Otolaryngology–Head and Neck Surgery, Denver, CO, September 9-12, 2001. Reprint requests: Stephen Hone, 4 Ailesbury Gardens, Sydney Parade, Dublin 4, Ireland; e-mail, [email protected]. Copyright © 2003 by the American Academy of Otolaryngology–Head and Neck Surgery Foundation, Inc. 0194-5998/2003/$30.00 ⫹ 0 doi:10.1067/mhn.2003.79

FRCSI,

and VIVIAN KELLY,

MCH, FRCSI,

Dublin, Ireland

nancial gain to be had from exaggeration, accurate assessment of true hearing threshold levels is critical. Initial assessment is by standard behavioral pure tone audiometry. The detection of nonorganic hearing loss or exaggerated hearing loss requires considerable experience and may be considered an art form. Even the most experienced audiologist may be fooled from time to time. A variety of behavioral indices have been described to help identify these individuals,1-3 including poor correlation between pure tone audiogram and speech reception threshold or the acoustic reflex threshold, variable audiometric response, a flat audiometric configuration, hearing ability better than the suggested loss, a significant low-frequency loss, and physiologic inconsistencies. Objective tests of hearing threshold have been used to obtain accurate hearing thresholds in those individuals suspected of exaggerating their hearing loss. Cortical evoked response audiometry (CERA) is the most valuable of these objective tests in adults for the following reasons.4 First, it has good frequency specificity over the speech frequency range (500 to 4000 Hz). Second, it is noninvasive and requires only passive cooperation. Third, it is recorded from a higher auditory level than electrocochleography or brain stem electrical responses and therefore is less subject to organic neurologic disorders. Fourth, it has been demonstrated to have a closer correlation with behavioral audiometry thresholds than brain stem testing.5 This study examines the behavioral and objective audiometric data of individuals complaining of NIHL as a result of impulse noise. Objectives of the study were to determine the incidence of exaggerated thresholds using CERA and to identify any associated audiometric features. SUBJECTS AND METHODS Data for this study were collected prospectively between May 1996 and October 1999 and entered into a database (Microsoft, Excel) on a personal computer. All individuals had a history of expo257

Otolaryngology– Head and Neck Surgery February 2003

258 HONE et al

Table 1. Irish Hearing Disability Assessment System versus the American Medical Association System

Frequencies Disability commences (low fence) Percent disability 100% disability Weighting for binaural loss Correction for age-related hearing loss Allowance for tinnitus

Irish system

AMA system

500 Hz, 1 kHz, 2 kHz, 4 kHz ⬎20 dB 1.25% per dB over 20 dB 100 dB 4:1 By table deduction, ⬎69 y in men, ⬎77 y in women Yes

500 Hz, 1 kHz, 2 kHz, 3 kHz ⬎25 dB 1.5% per dB over 25 dB 91 dB 5:1 None

sure to impulse noise (gunfire) without ear protection and had initiated a medicolegal claim for NIHL. Individuals underwent pure tone audiometry and subsequent cortical evoked response audiometry performed by a single audiologic scientist (G.N.) experienced in the assessment of. NIHL. Early in the series, CERA testing was performed at 1, 2, and 3 kHz in both ears. Subsequent to the development of an Irish system for the assessment of hearing disability, levels of 500 Hz and 4 kHz were also tested.1 Additional testing at 6 and 8 kHz was performed initially when requested by the referring solicitor and later in the series on all subjects. Cortical response audiometry was performed using a Nicolet Pathfinder II unit. Electrodes were placed on the vertex and both mastoids with a ground on the forehead. Patients were seated with eyes open and reading a magazine. Amplifier frequency bandwidth was 1 to 12 Hz. Tone bursts were presented to either ear in a quasi-random fashion at test frequencies using 300-␻ insert earphones. Contralateral masking was applied when necessary. Thresholds were estimated using 10-dB steps. Direct determination of the presence or absence of a response at various intensities was obtained until the response was just detectable in the averaged record. Each threshold judgment was estimated to the nearest 5 dB to allow comparison with the 5-dB increments of pure tone audiometry. For the purposes of this report, hearing disability is defined as the inability to hear everyday sounds, in either quiet or noisy backgrounds, in a manner that is considered to be normal for humans. This is similar to the American Medical Association (AMA) definition of hearing handicap.

None

We assessed 1154 men (mean age, 41 years; age range, 20 to 85 years); 673 underwent cortical evoked response audiometry and are included in this study. All had diagnostic pure tone audiometry performed in a soundproof booth. Pure tone averages (PTAs) were calculated using 500 Hz, 1 kHz, 2 kHz, and 4 kHz. A PTA of ⬎20 dB was considered significant for hearing disability (Irish hearing disability assessment system)1 (Table 1). For comparison purposes, hearing disability was also calculated using the system of the AMA (Table 1).6 In this case, PTA was calculated using 500 Hz, 1 kHz, 2 kHz, and 3 kHz, with a PTA of ⬎25 dB significant for hearing disability. Exaggerated hearing thresholds were considered to be present when the average threshold results obtained by CERA were ⬎10 dB better than the PTA over 500 Hz, 1 kHz, 2 kHz, and 4 kHz.1 RESULTS Pure Tone Audiometry Mean PTA (500 Hz, 1 kHz, 2 kHz, and 4 kHz) was 33 dB (range, 3 to ⬎110 dB) in the right ear and 34 dB (range, 3 to ⬎110 dB) in the left ear. Of 673 subjects, 485 (72%) had a binaural PTA of ⬎20 dB (mean, 41 dB; range, 21 to ⬎110 dB). This equates to an average hearing disability of 26% (range, 1% to 100%), using the Irish system of hearing disability assessment.1 Mean PTA for the frequencies 500 Hz, 1 kHz, 2 kHz, and 3 kHz was 32 dB (range, 0 to ⬎110 dB) for the right ear and the left ear. Of 673, 362 (54%) had a binaural PTA of ⬎25 dB (mean, 45 dB; range, 26 to ⬎110 dB). This equates to an average hearing disability of 30% (range, 1% to 100%), using the AMA system for hearing disabil-

Otolaryngology– Head and Neck Surgery Volume 128 Number 2

HONE et al

259

Fig 1. A comparison of percentage hearing disability as calculated by the Irish and AMA systems. The Y axis is the percentage of the total number with hearing disability as measured by that system.

Table 2. Average binaural pure-tone thresholds for exaggerators and nonexaggerators

Nonexaggerator (n ⫽ 248) Exaggerator (n ⫽ 88)

500 Hz

1 kHz

23.5

23.6

61.8

62.7

2 kHz

3 kHz

4 kHz

6 kHz

32

46

54.8

59.7

57

64.8

70

73.8

77.1

74.6

ity assessment.6 Figure 1 compares those with hearing disability using the Irish and AMA systems. Three hundred thirty-six individuals had CERA performed at 500 Hz, 1 kHz, 2 kHz, and 4 kHz; of the 336, 88 (26%) had an average hearing threshold 10 dB better than their PTA and were considered to have exaggerated hearing loss. Table 2 and Figure 2 present the audiometric data from the nonexaggerators and from the exaggerators. The exaggerators had considerably worse thresholds than the nonexaggerators. This was most marked in the lower frequencies. The audiogram was noted to be “flatter” for the exaggerators and lacked the typical high tone dip that is typical of NIHL. Maximum hearing loss in both groups occurred at 6 kHz. Table 3 presents data on the mean binaural hearing threshold at 500 Hz. A mean threshold of ⬎25 dB HL at 500 Hz had a sensitivity of 94% and a specificity of 59% for the detection of exaggerated hearing thresholds. The positive predictive value, the probability of the individual exaggerating their hearing loss when their threshold at 500 Hz was ⬎25 dB HL, was 46%.

8 kHz

Table 4 presents the audiometric and CERA data for different frequencies averaged from both ears. The highest discrepancy occurred at 1 kHz with an average difference of 41.9 dB. The difference over the averaged frequencies of 500, 1000, 2000, and 4000 was 35.8 dB. The average hearing disability (Irish system) suggested by PTA for exaggerators was 57.5% compared with actual hearing disability (judged by CERA) of only 12.5%. DISCUSSION The results of this study demonstrate that approximately one fourth of our subject population had exaggerated hearing threshold levels. This figure is somewhat higher than data reported from other studies. Rates for exaggerated hearing thresholds or nonorganic hearing loss reported from Canada are 16.7%, and from Australia, 17.7%.2,7 Reasons for this difference may include subject population. Subjects in these 2 studies were claiming industrial noise exposure, whereas all of our subjects had been exposed to highintensity impulse noise.

Otolaryngology– Head and Neck Surgery February 2003

260 HONE et al

Fig 2. Mean hearing thresholds for each frequency on PTA. n Indicates nonexaggerator; m, exaggerator.

Table 3. Number of exaggerators and nonexaggerators with mean binaural hearing threshold level (HTL) greater than 25 dB HL HTL at 500 Hz

Exaggerator

Nonexaggerator

Total

85 3 88

101 147 248

186 150 336

⬎25 db HL ⱕ25 dB HL Total

Table 4. Comparison of audiometric and CERA data for exaggerators

Audio CERA Difference

500 Hz

1 kHz

2 kHz

3 kHz

4 kHz

6 kHz

8 kHz

PTA

61.8 23.4 38.4

62.7 20.8 41.9

64.8 29 35.8

70 39.7 30.3

73.8 46.7 27.1

77.1 57.1 20

74.6 55.4 19.2

65.8 30 35.8

Different studies have used different criteria to define exaggerators on the basis of CERA.2,7 Alberti and colleagues considered a behavioral hearing threshold of ⬎15 dB worse than the CERA threshold to be evidence of nonorganic hearing loss or exaggeration.2 It is unclear from their report whether this applied to one frequency or to

averaged frequencies. Coles and Mason4 examined this area in some depth. The results of their study indicated that if ⱖ3 frequencies in 1 ear have been tested by CERA and this average was ⬎7.5 dB better than pure tone audiometry, some degree of nonorganic hearing loss is likely to be present. For a single frequency, a discrepancy of

Otolaryngology– Head and Neck Surgery Volume 128 Number 2

⬎15 dB is required. The UK hearing disability assessment system published in 1992 recommends repeating CERA if there is a discrepancy of 11 and 15 dB for averaged thresholds.3 If the discrepancy is greater than 15 dB, spurious thresholds are highly likely, and in the absence of convincing evidence to the contrary, the CERA thresholds should be used for hearing disability assessment. We used a discrepancy of ⬎10 dB for averaged frequencies 500 Hz, 1 kHz, 2 kHz, and 4 kHz as recommended by the Irish system for assessment of disability.1 Seventy-two percent of our population had some level of hearing disability when assessed using the Irish system but only 54% when assessed by the AMA system. There are 2 main reasons for this. One is the use of 20 dB as the low fence or threshold for hearing disability in the Irish system compared with 25 dB in the AMA system. The other reason is the incorporation of 4 kHz instead of 3 kHz in the Irish system. This means that the Irish system of hearing disability assessment is more lenient at lower levels of hearing loss than the AMA system. However, the AMA system tends to be more lenient when hearing loss is more severe. This is due to the use of 1.5 as a multiplier for each decibel of hearing loss, over 25 dB, compared with 1.25 for the Irish system. Despite these differences, it can be seen from Figure 1 that the proportion of individuals with different degrees of hearing disability is broadly similar for both systems. The assessment of hearing disability using any system has its problems. There is a continuum of hearing disability that is measurable in most individuals at a hearing threshold of between 15 and 30 dB.1 There is no specific point at which hearing disability suddenly starts. Hearing disability progresses with increasing hearing loss in a sigmoidal fashion.8,9 It progresses slowly initially and then rapidly accelerates and slows again. Therefore the selection of a point at which hearing disability starts has obvious inherent problems. A number of different methods may be used for the selection of such a low fence.1 However, it is important to note that the speech frequencies selected for the assessment system are extremely important in setting a low fence. If more high frequencies are included, the low fence should be

HONE et al

261

higher, and if more low frequencies are chosen, it should be lower. Regardless of the level at which the low fence is set, there is evidence to suggest that individuals with hearing threshold levels of 12 dB at 1 kHz, 2 kHz, and 3 kHz or 15 dB at 1 kHz, 2 kHz, and 4 kHz perform as well as normal hearing controls even in noisy conditions.10 As discussed earlier, CERA is currently the best objective test in adults to verify behavioral hearing threshold levels. However, it is not without drawbacks. It depends on the subjective reading of the presence or absence of certain waveforms. It requires experienced personnel both to perform it and to interpret the results. A subject who continuously moves or the presence of high-amplitude background electroencephalographic waveforms can interfere with the test. Further, it is likely that a small percentage of cases, probably ⬍1%, will have CERA responses that occur only at levels significantly greater than their true hearing threshold.4 In common with other investigators, we noted that subjects with exaggerated hearing thresholds tended to have “flatter” audiograms than did individuals with verified thresholds.3,4 It is likely that these individuals exaggerate their hearing thresholds using the concept of equal loudness. They select a constant level of loudness at which they respond to pure tone audiometry for each frequency. Because of recruitment at frequencies with true hearing loss (the higher frequencies in NIHL), the audiogram tends to become progressively flatter with greater exaggeration. We noted that maximum exaggeration tended to occur at 1 kHz with the least exaggeration at the higher frequencies. This is in keeping with the above explanation and has been noted by some previous investigators.4 Others have noted the greatest difference to occur at 500 Hz, presumably for similar reasons.2 Age-related hearing loss and NIHL tend to affect the higher speech frequencies initially. The lower frequencies are generally not affected to a significant degree. ISO 7029-1990 gives the 10th percentile hearing threshold at 500 Hz for a 70year-old male as 23 dB.11 As significant sensorineural hearing loss at 500 Hz is uncommon, we looked at all those individuals with hearing loss of ⬎25 dB at 500 Hz. This criterion has a high

262 HONE et al

sensitivity but moderate specificity for the identification of exaggerators. Only 3 of 88 exaggerators did not have a hearing loss of ⬎25 dB on pure tone audiometry. Alberti et al2 looked at a number of predictors for exaggerated hearing thresholds based on the hearing threshold at 500 Hz. They found that the hearing threshold at 500 Hz was the best predictor. They reported that a hearing threshold of 50 dB at 500 Hz had a false-positive rate of 11% and a false-negative rate of 50%, whereas a threshold of 30 dB gave a false-positive rate of 36% and a false-negative rate of 6%. In conclusion, we found a high incidence of exaggerated hearing loss in our population of individuals claiming NIHL as a result of impulse noise. CERA is critical to identify these individuals and to obtain accurate thresholds. We found the presence of a flat audiogram and a hearing threshold of ⬎25 dB at 500 Hz to be helpful in identifying exaggerators. We recommend CERA testing in all such cases. REFERENCES

1. Report of the Expert Hearing Group. Hearing disability assessment. 1st ed. Dublin (Ireland): Department of Health and Children; 1998.

Otolaryngology– Head and Neck Surgery February 2003

2. Alberti PW, Hyde ML, Riko K. Exaggerated hearing loss in compensation claimants. J Otolaryngol 1987;16:362-6. 3. King PF, Coles RRA, Lutman ME, et al. Assessment of hearing disability. Guidelines for medico-legal practice. London (UK): Whurr Publishers; 1992. 4. Coles RRA, Mason SM. The results of cortical electrical response audiometry in medico-legal investigations. Br J Audiol 1984;18:71-8. 5. Jones LA, Harding GFA, Smith PA. A comparison of auditory cortical evoked potentials, brainstem evoked potentials and post auricular myogenic potentials in normals and patients with known auditory defects. In: Barber C, editor. Evoked potentials. Lancaster: MTP Press; 1980. 6. Guide for the evaluation of hearing handicap. JAMA 1979;241:2055-9. 7. Rickards FW, De Vidi S. Exaggerated hearing loss in noise induced hearing loss compensation claims in Victoria. Med J Aust 1995;163:360-3. 8. Habib RG, Hinchcliffe R. Subjective magnitude of auditory impairment. A pilot study. Audiology 1978;17:6876. 9. Lutman ME, Robinson DW. Quantification of hearing disability for medico-legal purposes based on self-rating. Br J Audiol 1992;26:297-306. 10. Acton WI. Speech intelligibility in a background of noise and noise-induced hearing loss. Ergonomics 1970;13: 546-54. 11. ISO 7029: 1990 Acoustics: threshold of hearing by air conduction as a function of age and sex for otologically normal persons. Geneva (Switzerland): International Organisation for Standardisation.