Pilot study: Comparison of pure tone audiometry and otoacoustic emis- sion based hearing assessment for classical music students S. Dance 1 School of the Built Environment and Architecture London South Bank University London, UK G. Zepidou 2

AECOM Ltd Sunley House, CR0 2AP London, UK

ABSTRACT Since the enforcement of the Control of Noise at Work Regulations research has been undertaken in collaboration with the Royal Academy of Music investigating the hearing acuity of more than 5000 students between 2007 and 2021. Standard pure tone audiometric screening methods were employed for both entry and exit testing of undergraduate and postgraduate students. The results of these in- vestigations have informed a pilot study, run in September 2021, comparing two hearing health sur- veillance methodologies: pure tone audiometry and otoacoustics emissions involving 256 classical music students. This comparison was necessary as music students had been found to be able to game the Bekesy test because of their acute listening ability. Results showed that otoacoustic emissions were able to identify, at an early stage, hearing damage in an additional 3.8% of students compared to audiometry. The test was found to be quicker, more convenient and offered greater objectivity. This provides reassurance that otoacoustic emissions can be an excellent tool for assessing the hear- ing health of classical music students and identifying hearing damage at an early stage.

1. INTRODUCTION The Control of Noise at Work Regulations 2005 (CoNAWR2005) [1] was introduced to protect em- ployees from health risks associated with noise. The Regulations were enforced in 2006 for all sectors other than the entertainment sector who were allowed a 2-year transitional period. With regards to sound exposure, it is known that excessive exposure results in Noise Induced Hearing Loss (NIHL) typically at 4 kHz [2]. Approximately half of NIHL occurs within the first 3 years of excessive sound exposure with the remaining NIHL occurring over the following 42 years [3,4]. Therefore, it is criti- cally important to protect new employees at the earliest opportunity. The CoNAWR were developed on research on the effects of occupational noise exposure on hearing health which was based on industrial environments and did not include effects of entertainment sound or music exposure due to the absence of relevant data. However, recently there has been some research investigating this field of study [5-7]. In addition, enforcing the Regulations on the entertainment sector is particularly dif- ficult due to the nature of their work and sound being a deliberate product.

Since 2007, through a long-term collaboration with the Royal Academy of Music (RAM), ap- proximately 330 students each year have been assessed using the standard Bekesy pure tone audio- metric (PTA) test procedure to determine hearing acuity [8]. Information collected over 15 years resulted to numerous findings from this longitudinal study. Early data from the first 1300 students tested indicated a consistent reduction in hearing acuity at 6 kHz (high frequency dip) [9]. The study also found that, based on the Health and Safety Executive (HSE) categorisation scheme [10], 94% of

worm 2022

students had “Normal” hearing (acceptable hearing ability), 4.5% presented “Warning” levels of hear- ing loss (mild hearing impairment) and 1.5% “Referral” levels of hearing (poor hearing) [11]. Note that according to the HSE categorisation scheme, typically 75% of population for each age band and gender would have hearing within normal limits, 20% would normally present a mild hearing impair- ment and 5% would need to be referred for further investigation. Test data from the first 2500 students indicated that 50% of the students tested, despite their already accumulated sound exposure, had on average negative hearing loss [12], where the sum of hearing levels at 1, 2, 4, and 6 kHz is below zero, when compared to the general population allowing for age and gender [3]. Study results from 229 students repeat audiometric testing before the end of their studies (3.5 year interval), indicated an improvement in musicians hearing but the improvement lessened for students regularly exposed to higher levels of sound, e.g. Brass and Jazz students [13]. A reason behind the excellent hearing results and improved hearing during student exit testing could be fact that with their well-trained ears and developed sensitivity to sound/changes in pitch, music students could simply be better at detect- ing pure tones than general population of same age. Hence, the authors concluded that another audi- ometric testing methodology to assess hearing acuity of classical music students was required. 2. HEARING RESEARCH Otoacoustic Emissions (OAE) based hearing assessments have been traditionally carried out to test the hearing acuity of new-born babies. It has been shown that a reduced OAE amplitude is a precursor to NIHL and hence may act as a biomarker for susceptibility to noise induced hearing loss [14]. In 2008, a comparative study was carried out using 154 young adults (aged 18-25) regularly exposed to occupational noise and 99 non-exposed (of same age group) and their hearing was assessed over a 3- year period using pure tone audiometry and Transient-Evoked OAEs (TEOAE). The young adults of both groups would have also been exposed to social noise, personal music players, live events, and nightclubs. Results showed that the hearing threshold values, an average across the frequencies of 3, 4, 6 kHz, of the exposed group were 1.7 dB worse compared to that of the non-exposed group whereas, according to the ISO 1999 methodology [3], a 6 dB reduction was being predicted based on cumulative amount of occupational and social noise exposure [16]. The difference between the meas- ured and predicted threshold shift, 270%, could not be due to a misremembering of the number of social events attended, or the length of time listening to personal music players. Hence, it can be reasonably inferred that social noise exposure and music has a different effect on hearing than occu- pational noise of equivalent noise energy. This agrees with the results of the RAM studies focused on the effect of the sound of performance on the hearing of classical musicians [8-13]. The 2008 report by Lutman, Davis and Ferguson concluded, “OAE may be beneficial to supplement PTA, as audiometry has limited sensitivity to detect Noise Induced Hearing Loss, 15 dBHL is required before confidence can be attached to the change in individuals. Future research may show that PTA and OAE have complimentary roles.” [16].

In November 2019, Distortion Product OAE (DPOAE) based hearing screening was trailed by audiologists at the Royal College of Music (RCM). In total 80 classical music students were assessed over a two-week period. The newly introduced British Association of Performance Arts Medicine (BAPAM) Best Practice Guidance reflected this approach. It was decided that the same OAE testing methodology would be used to further assess the heating acuity complementing the PTA audiometry already undertaken.

worm 2022

3. PILOT STUDY A study of 310 classical music students was undertaken over a three-week period in September 2021. The hearing conservation programme commenced with an one-hour seminar on risks to hearing as- sociated with the sound of music performance. The hearing assessment itself consisted of a short questionnaire, pre-completed by each student, which included information on instrument studied and years of music exposure plus additional information on leisure activities, part-time work, and music listening habits. Each student was allocated a 10-minute slot for the hearing assessment comprising OAE and PTA. OAE testing was undertaken in an anechoic chamber, ambient noise level LAeq, 30 minutes 18 dBA, using a Path Medical OAE Senti instrument and dual ear probes with eight sizes of ear tips to ensure a good ear canal seal [18]. The instrument was used in simultaneous mode to speed up the testing procedure, so that an assessment could be undertaken in under three minutes. Results were provided at twelve octave band centre frequencies 1000-6500 Hz at 500 Hz intervals for each ear. Due to Covid-19 related restrictions, an otoscope examination was not possible which led to difficulty in establishing the correct size of ear tip to provide a seal of the ear canal. As such the students themselves checked the fit of the ear probes. In addition, the instrument automatically cali- brated, or failed to calibrate which allowed the ear tip to be swapped to the next size to ensure the quality of the assessment. The ear tips were washed and/or disinfected prior to each use.

Assessment categorisation of overall estimated outer hair cell (OHC) damage was based on the Hearing Coach software recommendations (OAE Expert package) [19] using the following criteria along with the corresponding young adult audiometry-based hearing categorisation [10], see Table 1.

Table 1: Categories of Hearing Damage for 18-25 year olds Hearing Coach Estimated Outer

Threshold of Hearing Levels: Sum of 1, 2, 3, 4, 6 kHz (dBHL)

Haircell Damage

Normal Hearing 0-20% <46 “Normal”

Mild Hearing Loss 21-40% 46-95 “Warning” Moderate Hearing Loss 41-60% Severe Hearing Loss 60-80% >95 “Referral” Profound Hearing Loss >=81%

Pure tone audiometry was carried out in a sound-proof booth compliant with ISO 8253-1 [21] using Amplivox 850 Mark 4s and TDP49 audio cups; the system being calibrated annually in accord- ance with ISO 389 [20]. The Bekesy test used started with a 1kHz familiarisation in the left ear fol- lowed by testing of the left ear at 7 octave band centre frequencies, 500-8000 Hz, then repeated in the right ear. This allowed air conduction pure tone hearing threshold levels to be determined. 4. RESULTS AND DISCUSSION The pilot study results presented within this section relate to students who were successfully assessed using both PTA and OAE test methodologies (n=256). In total 310 students were assessed; however, 54 students (17.4%) could not take the OAE test due to minor amounts of ear wax preventing the necessary autocalibration . The population consisted of 142 females and 114 males with the students separated into six music departments of ten or more students (n>10). Students had an average age of 21.8 years with a standard deviation of 2.8 years. An analysis of the estimated Outer Hair Cell (OHC) hearing damage of all students based on five Hearing Coach categories is given in Table 2.

worm 2022

Table 2: Number of students in each Hearing Category based on Estimated OHC Damage

Hearing Loss Number by Cat- egory based on Estimated OHC Damage (Left Ear, n=250*)

Number by Cat- egory based on Estimated OHC Damage (Right Ear, n=256)

Percentage of Students (Left Ear)

Percentage of Students (Right Ear)

Normal 175 181 70.0 70.7

Mild 57 52 22.8 20.3

Moderate 13 14 5.2 5.47

Severe 4 7 1.6 2.73

Profound 1 2 0.4 0.78

* Temporary issues with equipment resulted in a smaller number of left ear assessments compared to those for the right ear

The percentages presented hereafter are an average of both ears. Table 2 indicates that 70.4% of the students had “Normal” hearing, 26.9% with ‘Mild or Moderate” hearing damage, and 2.8% with “Severe or Profound” hearing damage very similar to the percentage breakdown in the Health and Safety Executive (HSE) guidance notes based on pure tone audiometry [10].

A comparative analysis has been undertaken, Table 3, based on categorisation of the audiograms resulting from the Bekesy tests for the identical student population taken at the same time. These are based on the sum of 1000, 2000, 3000, 4000, 6000 Hz threshold of hearing levels (dBHL) for each ear, categorised in accordance with HSE guidance [10]. An additional category has also been in- cluded, “Exceptional”, defined as hearing gain. Note: zero hearing loss is a statistical averaged thresh- old of hearing level for a population of 18-25 year olds.

Table 3: Hearing Category based on Audiometric results for Classical Music Students, n=256

Number of Students (Left Ear)

Number of Stu- dents (Right Ear)

Percentage of Students (Left Ear)

Percentage of Students (Right Ear)

Exceptional 129 168 50.4 65.6

Normal 75% of Pop. 242 248 94.5 96.9

Warning 25% of Pop. 14 8 5.5 3.1

Referral 5% of Pop. 0 0 0.0 0.0

The 2021 comparative study produced results in-line with our earlier research based on the first 1300 musicians tested. Given the high number of students with “Normal” hearing, the proposed new category “Exceptional” proved useful as a differentiator between normal population and classical music students more than half of which were having negative hearing loss, again in-line with our earlier finds when 50% of 2536 students had hearing gain [12]. Results from pure tone audiometric testing indicated that musicians quickly familiarise themselves with the testing procedure and, taking advantage of their music listening skills, exhibit excellent results and game the Bekesy test. This is

worm 2022

confirmed by the consistently better scores for the right ear, the second ear under test and is in line with Lawton’s findings on learning bias [22].

Comparing the two hearing assessment methods, as postulated by Lutman, Davis and Ferguson [16], OAEs did show up potential hearing damage earlier than pure tone audiometry as 2.0% of left ears and 3.5% right ears were found as having “Severe” or “Profound” cochlear damage compared to 0% of students with “Referrals”; see Tables 2 and 3. Thus, 3.8% of the students (9) were made aware of potential hearing impairment that would have gone undiagnosed if only PTA was carried out.

An assessment of the level of learning for each music department assessed has been carried out by comparing left/right ear results for averaged estimated OHC damage % (objective method) and summed thresholds of hearing levels dBHL (subjective method); see Table 4.

Table 4: Hearing Assessment of Classical Music organised in Six Music Departments, n=256 Left Ear Averaged OHC Dam-

Left Ear Averaged

Right Ear

Right Ear

Difference

Difference

Averaged

Averaged

in Left/Right

in Left/Right

Summed Threshold

OHC Damage

Summed Threshold

age (%)

Ear (dBHL)

Ear (%)

of HLs (dBHL)

(%)

of HLs (dBHL)

All 15.7 ± 16.30 0.58 ± 16.63 16.3 ± 17.50 -7.09 ± 23.19 7.67 -0.6

Strings 14.8 ± 15.22 2.38 ± 27.11 15.5 ± 20.98 -9.79 ± 22.58 12.17 -0.7

Woowind 15.2 ± 16.47 -1.04 ± 24.52 17.2 ± 17.94 -9.65 ± 19.36 8.61 -2.0

Voice 17.4 ± 16.48 9.46 ± 22.55 22.6 ± 11.01 1.81 ± 19.35 7.65 -5.2

Brass 21.7 ± 13.04 1.65 ± 30.62 17.0 ± 18.89 -5.7 ± 25.98 7.35 4.7

Music Th. 16.3 ± 27.32 0.00 ± 35.44 14.0 ± 20.92 -3.83 ± 29.74 3.83 2.3

Piano 14.9 ± 23.56 0.82 ± 27.67 13.8 ± 15.12 0.27 ± 23.24 -0.55 1.1

Results presented in Table 4 indicated that the objective testing method (OAE) provided signifi- cantly more symmetric results between the two ears compared to the subjective testing method (PTA). Overall, PTA results presented a right ear bias of the order of 7.7 dB across the five summed frequen- cies (range: 12.17 to -0.55 dBHL), where the OAE test results difference between the ears was just - 0.6% (range: 4.7 to -5.2 %). It can also be seen that the right ear had the greater standard deviation for both OAE and PTA based tests indicating a greater variation in the results.

Analysing the results for each department, it was found that musical theatre students had the larg- est standard deviation in both hearing assessment methodologies suggesting greater variation amongst these older individuals. Studying the differences between the ears, audiometry showed the left ear had consistently worst results, for all departments, except for pianists who exhibited similar levels of hearing loss at both ears, see Table 4. This agrees with Vinay et al [23] who found that during PTA testing of 32 young adults (non-musicians) with good hearing, < 15 dBHL at any fre- quency, the summed threshold of hearing increased during right ear testing on average by 11.4 dB.

In contrast to PTA, increased hearing loss symmetry was found during the OAE assessment with the averaged estimated OHC hearing damage difference between the two ears being within maximum of 5%. This confirmed a preliminary result from our earlier feasibility study [15]. It is believed that

worm 2022

more data would need to be collected to determine the influence of specific instruments on the hearing of the players, particularly for asymmetric instrument players, such as the string players.

Further information on the population distribution of each of the six departments based on otoa- coustic emission tests using the Hearing Coach software categorisation [19] and pure tone audiometry using the HSE categorisation scheme [10] including the new ‘Exceptional’ hearing category intro- duced, is presented in Table 5 and Table 6 respectively.

Table 5: Hearing Category Percentage Breakdown by Department using OHC (%)Damage Category Voice String Musical Th. Piano Woodwind Brass

Left Righ

Left Righ

Left Righ

Left Righ

Left Righ

Left Righ

t

t

t

t

t

t

Normal 73.8 69.0 68.4 72.2 57.7 50.0 70.6 60.0 69.4 80.0 63.6 72.7

Mild 19.0 31.0 25.3 16.5 34.6 30.8 17.6 30.0 22.2 11.4 27.3 18.2

Moderate 4.8 0.0 5.1 6.3 7.7 11.5 0.0 10.0 8.3 5.7 9.1 0.0

Severe 2.4 0.0 1.3 3.8 0.0 7.7 5.9 0.0 0.0 0.0 0.0 9.1

Profound 0.0 0.0 0.0 1.3 0.0 0.0 5.9 0.0 0.0 2.9 0.0 0.0

Table 6: Hearing Category Population Percentage Breakdown by Department based on PTA Voice String Musical Th. Piano Woodwind Brass

Left Right Left Right Left Right Left Right Left Right Left Right

Exceptional 45.2 66.7 29.1 36.7 42.3 53.8 50.0 60.0 47.2 69.4 36.4 45.5

Normal 97.6 100 100 100 80.8 88.5 90.0 100 94.4 91.7 81.8 90.9

Warning 2.4 0.0 0.0 0.0 19.2 11.5 10.0 0.0 5.6 8.3 18.2 9.1

Referral 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

Results in Table 5 indicate that Musical Theatre students had consistently the highest OHC hear- ing damage when compared to the other departments assessed. This was particularly obvious for the “Normal”, “Mild” and “Moderate” categories. Results agreed with those from the PTA testing where Musical Theatre had the highest incidence of warning levels compared to other students.

Nevertheless, results from Tables 5 and 6 confirm the finding that RAM students assessed, in- cluding those from the Musical Theatre department, had better hearing than that expected of the gen- eral population according to the HSE categorisation scheme. None of the RAM students assessed had “Referral” levels of hearing loss during the 2021 testing where during previous years testing approx- imately 1.5% of the students tested was found to have poor hearing [8,9,11,12]. It should be noted that COVID restrictions did curtail many activities which are known to be influence hearing acuity for these students, for example live events and nightclubs at the time of the study.

Although an analysis of hearing thresholds across the testing frequencies is not required by the CoNAWR2005, it was decided at an early stage of the RAM study that such an analysis should be carried out to investigate further the hearing of musicians, any patterns in hearing thresholds and more

worm 2022

importantly, any signs of early hearing loss; a warning sign for both the musicians and the Academy that is not revealed by the HSE assessment procedure. During the 2021 study details from the OAE testing frequency content were also analysed.

Figure 1 presents the estimated OHC damage (%) (1000 to 6500 Hz) averaged for 250 students as measured by the OAE instrumentation and the PTA hearing threshold values, n=256.

35100015002000250030003500400045005000550060006500

Estimated Outer Haircell Damage

8

Threshold of Hearing Levels (dBHL)

30

6

25

4

20

2

(%)

15

0

10

-2

5

-4

1000 2000 3000 4000 6000 8000

0

Frequency (Hz)

Left

Left Ear Right Ear

Right

Figure 1. Averaged OHC Damage (%), n=250 and Averaged Hearing Thresholds, n=256

The shape of the hearing damage presented in Figure 1 closely matches that found in the earlier feasibility study with three central frequencies, 4000, 4500 and 5000 Hz showing “Mild”, >20%, OHC damage [15]. This is in broad agreement with air conduction PTA results which indicate NIHL first occurring at 4 kHz [2]. It should be noted that automated screening audiometry does not normally measure the 5 kHz frequency. By way of comparison Figure 1 also shows the PTA results for the same cohort of classical music students, n=256, for six frequencies: 1000, 2000, 3000, 4000, 6000 and 8000 Hz. It can be clearly seen that pure tone audiometry shows a steep rise in the thresholds of hearing at 6 kHz for music students. A 6 kHz notch rather than the expected NIHL 4 kHz notch [2] was found for the left ear, not seen in the right ear. This result agrees with the findings of Backus and Williamon who undertook a similar audiometric study of 162 students at the Royal College of Music [7] that revealed a notch at 6 kHz in the left ear only. It should be noted that in their study the right ear was tested first. This is a sign of noise induced hearing loss linked with musicians’ noise exposure but also considered the earliest audiometric indicator of impending NIHL for musicians.

OAE offered a clear advantage in terms of frequency granularity, 12 vs 7 frequencies tested. This granularity exposed a new result that of a consistent 5 kHz loss for musicians and that all music departments had “Mild” damage at 4000, 4500 and 5000 Hz. This agreed with the feasibility study result [15]. In terms of PTA music induced hearing loss occurred primarily at 6 kHz (left ear) and 8 kHz (right ear), agreeing with results [7-9, 11-13].

5. CONCLUSIONS By analysing the results of 256 RAM students, OAEs detected 41 out of 506 ears assessed as having “Moderate”, “Severe”, or “Profound” hearing damage (27, 11 and 3 ears respectively). This equates to 8.1% of student ears with indications of moderate or significant hearing damage. The PTA testing

worm 2022

found, 22 out of 512 ears assessed as having a mild hearing impairment (Warning category) with no students found as having poor hearing (Referral category); this equates to 4.3% of student ears with mild hearing damage. As such, it has been shown that the OAE method detected nearly twice the incidence of hearing damage compared to that of PTA amongst music students.

Results of OAE testing showed good ear symmetry, on average a -0.6% difference, whereas the average audiogram showed a 7.7 dB bias towards the right ear for the music students, the second ear tested. This result confirmed that music students appear to be able to game the test due to their acute musical ability and their focus on listening, something not possible with the OAE based test. This preliminary research has shown the value of otoacoustic emission testing. The system was found to have multiple advantages compared to audiometry: mobility, speed, simplicity, objectivity with no learning bias, and results which indicate at the earliest possible stage risk of permanent hearing dam- age. This study showed an additional 3.8% of students identified as having hearing damage.

The hearing categorisation for pure tone audiometry showed the need for a new category for classical music students. It was proposed that ‘Exceptional’ be introduced meaning hearing gain based on the ISO 1999 standard necessary to provide greater differentiation for music students. It is also strongly suggested that 5 kHz should be included in automated health surveillance, as this was consistently found to be the frequency where OAEs determined the OHCs were most damaged. 6. REFERENCES 1. The Control of Noise at Work Regulations, Health and Safety Executive, (2005). 2. Alberti P. W. Noise and the ear . In: Scott-Brown’s Otolaryngology , Volume 2, Butterworth-Heinemann, (1997). 3. BS EN ISO 1999: 2013 Estimation of noise-induced hearing loss, Geneva, Switzerland . 4. BS EN ISO 7029: 2017. Statistical distribution of hearing thresholds as a function of age, Geneva, Switzerland. 5. Royster J., Royster L. , Killion M. Sound exposures and hearing thresholds of musicians , JASA , 89 , (1991) 2793. 6. Fearn F.W. Hearing loss in musicians, Journal of Sound and Vibration , 163 , (1993) 372. 7. Backus B., Williamon A. Noise-induced hearing loss in orchestral musicians, Int. Sym. Perf. Science , (2009) 8. Dance S., Zepidou G., Shearer D. Face the Music, Proc Int. Congress on Acoustics , Aachen, (2019) 9. Zepidou G., Dance S. Hearing loss amongst classical music students , ICBEN, London (2011) 68-73. 10. Health and Safety Executive. The control of noise at work regulations: Guidance on Regs L108 3 rd Ed (2021). 11. Dance S., Morant S., Zepidou G. Hearing loss amongst classical music students, Proc. Inst. of Acoustics (2008). 12. Dance S., Dymock B. Sound exposure and the hearing of musicians, Proc. Euronoise 2015 , Maastricht, (2015) 13. Dance S., Shearer D. Facing the Music, Int. Congress on Sound and Vibration , London, (2017) 14. Bamiou D., Lutman M. Interaction of noise induced hearing loss and ageing: epidemiological aspects. In: Luxon L,

Prasher D (eds). Noise and its effects. Chichester: John Wiley, (2007) 64-84. 15. Dance S., Ballestero E. Feasibility Study for OAE Hearing Assessment, Internoise 2022 , Glasgow, (2022). 16. Lutman M., Davis A., Ferguson A. Epidemiological evidence for the effectiveness of the noise at work regulations

RR669, Health and Safety Executive, (2008). 17. Shepard R., Ryan F., Checkley P., Cordeaux C. Hearing Conservation for Performers: Best Practice Guidance,

British Association for Performance Arts Medicine, (2020). 18. www.pathme.de 19. www.hearingcoachsoftware.com 20. ISO 389-1: 2017. Acoustics. Reference zero for the calibration of audiometric equipment - Part 1: Reference equiv-

alent threshold sound pressure levels for pure tones and supra-aural earphones, Geneva, Switzerland, (2017). 21. ISO 8253-1: 2010 Acoustics Audiometric test methods Part 1: Pure-tone air and bone conduction audiometry, Switzerland (2010). 22. Lawton B., Variability of the threshold of hearing: It’s important in the case of Noise Induced Hearing Loss, ISVR

Tech Report No. 336 ( 2015). 23. Vinay S., Svenson U., Kvaløy O. A comparison of test–retest variability and time efficiency of auditory thresholds

measured with pure tone audiometry and new early warning test, Applied Acoustics , 90 , (2015) 153–159.

worm 2022