Noise Exposure and Hearing Loss among Student Employees Working in University Entertainment Venues

Transcription

1 Ann. occup. Hyg., Vol. 46, No. 5, pp , British Occupational Hygiene Society Published by Oxford University Press DOI: /annhyg/mef051 Noise Exposure and Hearing Loss among Student Employees Working in University Entertainment Venues S. SADHRA 1 *, C. A. JACKSON 1, T. RYDER 1 and M. J. BROWN 2 1 Institute of Occupational Health, University of Birmingham, University Road West, Edgbaston, Birmingham B15 2TT; 2 Victoria Hospital, Morecambe, UK Received 7 September 2001; in final form 31 January 2002 Objectives: Most studies to date on sound levels in entertainment establishments have concentrated on exposure levels for the attending public, rather than employees who may be at greater risk of hearing loss. Of particular concern are young employees. The aim of this pilot study was to (i) estimate typical sound levels in different areas where amplified music was played, (ii) measure temporary threshold shift (TTS) and (iii) estimate the dependence of hearing threshold shifts on measured noise levels. Methods: This study focused on students working part-time (up to 16 h/week) in music bars and discotheques in a university entertainment venue. All 28 staff were invited to participate in the study. Pre- and post-exposure audiometry was used to determine hearing threshold at both high and low frequencies. Personal dosemeters and static measurements were made to assess noise levels and frequency characteristics. A questionnaire was used to determine patterns of noise exposure and attitudes to noise levels and hearing loss. Results: Of the 28 student employees working in the three areas, 14 (50%) agreed to take part in the study, giving 21 pre- and post-shift audiograms. The mean personal exposure levels for security staff were higher than those of bar staff, with both groups exceeding 90 db(a). The maximum peak pressure reading for security staff was 124 db. Although TTS values were moderate, they were found to be highly significant at both low and high frequencies and for both ears. Twenty-nine per cent of subjects showed permanent hearing loss of more than 30 db at either low or high frequencies. The correlation between TTS and personal exposure was higher at 4 khz than the low and high frequencies. Conclusions: Contemporary music may be an important yet little considered contributor to total personal noise exposure, especially amongst young employees. Employees need to be better informed of risks of hearing loss and the need to report changes in hearing acuity. Suggestions are made on strategies for improving the assessment of noise exposure in entertainment venues. Keywords: hearing loss; temporary threshold shift; entertainment venues INTRODUCTION Exposure to loud music, especially among young people, is an important source of concern. It has been estimated that some young music lovers will have sustained significant permanent hearing loss by their mid-twenties (Carter et al., 1982). Discotheques and modern day Fun Pubs have had a long-standing association with playing pre-recorded and amplified music for entertainment. The risk of hearing loss from amplified music is dependent on exposure duration, *Author to whom correspondence should be addressed. Tel: sound intensity and the genetic vulnerability of individuals. Over the past 20 yr the power of amplification that is affordable has increased steadily, hence the increase in the potential for hearing damage. Most studies to date on sound levels in entertainment establishments have concentrated on the public attending various music functions or the musicians (Axelsson and Lindgren, 1978; McBride et al., 1992). A few studies have also investigated hearing loss from the use of personal cassette players, or PCPs (Rice et al., 1987). Two excellent reviews have been published on noise exposure from leisure activities (Clark, 1991) and the hearing of classical musicians 455

2 456 S. Sadhra et al. (Palin, 1994). More recently, an epidemiological study on the evaluation of hearing damage from amplified music showed a gradation of audiometric damage from discotheques to PCP to rock concerts (Bisch, 1996). This study also reported a significant difference between the hearing thresholds of people who listen to PCPs for >7 h/week and those who frequently attended concerts when matched with control groups. Noise levels in nightclubs have been reported to frequently exceed 90 db(a), with peaks as high as 109 db(a) (Bickerdike and Gregory, 1980). In previous unpublished studies peak values from music in excess of 140 db(a) have been recorded. The Institute of Hearing Research, in their review of hearing damage from leisure activities (Institute of Hearing Research, 1986), gave a mean 8 h L eq level of db(a) for discotheques. Whilst some data is available for discotheque attendees, very little attention has been given to noise exposure and the risk of hearing loss of employees in discotheques and other leisure or entertainment functions. Clearly the employees in such establishments are exposed more routinely and thus are at a greater risk of hearing damage. A detailed study of discotheques in Hong Kong showed that employees can be exposed to loud noises for up to 8.6 h/day, 6 days/week. In contrast, members of the public (18 30 age group) who attend discotheques are exposed, on average, to high noise levels for only 3.1 h on 1.5 occasions/week (Tan et al., 1990). Within entertainment venues there is a common core of employee types, namely management, bar staff, security and disc jockeys (DJs), who may be routinely exposed to high noise levels. The majority of these employees tend to be young adults working on a part-time basis and may show signs of subjective hearing loss, which may not be identified in its early stages. Furthermore, hearing loss patterns may change in young adults during the initial years of noise exposure. Studies on young adults exposed to music have reported shifts in maximum hearing loss with duration of exposure. Korpert and Winker (1994) measured hearing thresholds of people aged yr from 1971 to 1991 and reported maximum hearing loss at 4 khz at the start of the observation period, while at the end it had shifted to 6 khz. Listening to music via stereo headphones was stated to be the major cause for the threshold shift. Modern nightclubs and discotheques provide a variety of styles of music, therefore it is possible that substantially different noise levels can be encountered on each occasion. These components have made it difficult for many researchers to calculate meaningful and accurate noise exposure levels, resulting in inconsistent study designs (Institute of Hearing Research, 1986). Furthermore, a number of studies have been limited by using audiometric data alone, without noise exposure data or noise exposure data without audiometric data. The infrequent high sound exposure to music may produce temporary threshold shifts (TTS), transient hearing impairment in which there is an increase in the hearing threshold. The rate of TTS recovery varies in individuals from several minutes to several days (Clark, 1991). Repeated TTS over the course of a few weeks to a few years may lead to accumulated cellular damage, causing a permanent threshold shift (PTS). Although TTS cannot predict the extent of PTS, it is a good early indicator of permanent damage (Luz et al., 1973). A few studies have investigated TTS in pop musicians (Axelsson and Lindgren, 1978) and those attending rock concerts (Ulrich and Pinheiro, 1974). These studies showed moderate TTS (up to 30 db at 4 KHz) with recovery within a few days of exposure. The extent of hearing loss amongst adults, especially younger adults, is becoming of increasing concern to both the government and prospective employers. Early signs of PTS may go unrecognized until individuals start their employment with companies who may routinely conduct pre-employment audiometry. However, pre-employment audiometry is not considered important or relevant in a number of work environments where individuals with hearing loss may not be identified nor made aware of risks factors. This pilot study focused on students who worked part-time (up to 16 h) in music bars and discotheques in a university entertainment venue. The aim of this study was to estimate typical sound levels and frequency characteristics of music played in clubs and to assess whether the student staff experienced TTS after noise exposure. The dependence of TTS on measured noise levels was also investigated, as well as knowledge and attitudes of noise levels and hearing loss amongst bar and security staff. METHODOLOGY Music premises This study was conducted in a university Students Guild, which comprised three areas used for musical entertainment, i.e. bars and discotheques. All three areas had hard wooden floors and no soft furnishings. Area 1: bar. A bar room (10 30 m) open 7 nights/week. For 3 nights a coin-operated juke box provides music; on the other 4 nights music is provided by a mobile disco situated in the bar area. The number of customers determines the time the bar remains open, the earliest closing time being 23:30 and the latest 02:00. The music is furnished via a system of 10 speakers suspended from the ceiling, located in corners of the room.

3 Noise exposure and hearing damage at university entertainment venues 457 Area 2: discotheque and bar. A disco (13 20 m) with a bar open 3 nights/week from 21:00 until 02:00. Music is provided via a fixed system with two main speakers, both 2 m in height on a stage 6 m high, located 15 m from the bar. Area 3: discotheque. A disco (13 30 m) without a bar, which is used only when the number of attendees is expected to be particularly high. It is also used for occasional live music performances. The hall has a high ceiling ( 14 m), with a 2 m high speaker stack system in each of the four corners of the room. Selection of subjects for personal dosimetery and audiometry The Students Guild employed a total of 124 staff comprising 112 part-time and 12 full-time staff. The majority of the part-time staff were students who worked as either security or bar staff. All 124 staff were invited to complete a general noise survey questionnaire (see below). All 28 staff working in the three music areas over three sampling days were invited to participate in the study. This involved pre- and post-shift audiometry and wearing of a personal dosemeter for the duration of the workshift. All staff volunteers were given written information explaining the purpose of the study. Those who agreed and completed the consent form were then asked to complete a questionnaire to obtain information on exclusion conditions applied to the study group. The exclusion factors included one or more of the following: history of chronic ear disease, head injury or concussion, significant exposure to noise other than loud music (blasts or explosions at a close distance, shooting, motor sports, etc.), significant exposure to loud music within the past 24 h (which precluded participants from taking part on two successive days/shifts), prolonged or regular use of ototoxic drugs (antibiotics and anti-rheumatics) and perforated or damaged tympanic membranes. The staff who met the selection criteria underwent a visual examination of the external auditory meatus by the occupational health nurse. Examination of the aural canals was undertaken using a Keeler ophthalmoscope fitted with an auroscope attachment. If one or both outer ears were found to have an obstruction, subjects were excluded from the study. Audiometry Guild staff who met the basic criteria for selection underwent audiometry testing to establish TTS for each workshift. Pure tone audiometric testing was carried out before exposure at the beginning of each shift. Audiometry was conducted following the HSE guidelines (Health & Safety Executive, 1995) as closely as possible. Otoscopic examinations were undertaken immediately before testing. All equipment was maintained and calibrated according to the manufacturers instructions. Background noise levels in the audiometry test room were measured during the course of each pre- and post-test session using a hand-held grade 1 Rion SLM. The audiometry was conducted using a Kamplex Computer Audiometer model BA 20, fitted with a noise excluding headset. The pre- and post-shift audiograms enabled a comparison of hearing thresholds across the frequency range khz for both the right and left ears. Postexposure audiometry was conducted 10 min after finishing a shift and leaving the music areas. Noise exposure monitoring: personal and static The noise levels were measured in all three bars and discotheques (areas 1 3) over three consecutive days in a week in June Noise monitoring was conducted from 18:00 to 00:45 in area 1 and 21:00 to 02:15 in areas 2 and 3. The music functions planned for the week of the study were deemed to be typical of a mid-term week. A static monitoring point was established within each of the three music areas where employees (bar and security staff) of the Guild spent the majority of their time. A sound level meter (SLM) was placed behind the bars in areas 1 and 2 and close to where security staff spent most of their time in area 3. In each case the SLM microphone (fitted with a windshield) was attached to an extension cable and suspended from the ceiling 2 m above the floor and 0.5 m from any obstruction such as posts and light fittings. The static SLMs were set to record the A weighted sound pressure levels and octave band analysis (OBA) data for the duration of the music being played. The OBA data was measured over the frequency range Hz. Following the pre-shift audiogram, bar employees were fitted with a personal noise dosemeter. Dosemeter microphones were attached to the shirt collar of the subjects. Due to the requirement for some of the staff to use radio communications with an earpiece, the dosemeter microphone was fitted to the collar on the opposite side to where the subject normally wore their earpiece. The battery powered dosemeter was clipped to the subjects waist belt opposite to the radio, to avoid the possibility of radio interference. The dosemeters were worn throughout the subjects shifts and removed prior to the postshift audiogram for that evening. SLM types and data handling Static noise levels and OBA data were collected using two portable grade 1 SLMs, a Rion NA29E and a Svan 912A. Both instruments were calibrated at 94 db(a) before use and the calibration re-checked at the end of each sampling period. Personal monitoring was conducted with Brüel & Kjær 4428 and CEL 460 personal noise dosemeters. All dosemeters were class 2 with an accuracy of ±1.5 db and all were cali-

4 458 S. Sadhra et al. Table 1. Static noise measurements and calculated L EP,d Area 1 (juke box) Area 1 (mobile disco) Area 2 Area 3 Sampling duration (min) Mean L eq [db(a)] L eq range [db(a)] Estimated L EP,d [db(a)] brated pre- and post-survey using calibrators specified by the dosemeter manufacturers. The stored data from the CEL 460 dosemeters were downloaded using CEL Sound Track v.1.2 and were transferred directly to a Microsoft Excel file for cleaning. The data recorded with Brüel & Kjær 4428 dosemeters were transferred manually. The following information was recorded from the dosemeters; sampling duration, noise dose (L Aeq ) and peak exposure level expressed in db(a). The L EP,d values were calculated using L Aeq values obtained from both the dosemeters and the static SLMs. General questionnaire A questionnaire was despatched using to all 124 Guild staff. The questionnaire was simple to complete and could be returned at no expense via . The questionnaire was divided into four sections detailing: length of employment, work shift patterns and exposure to amplified music at work; non-occupational exposure to music including PCPs, playing an instrument, attending live concerts, shooting, motor sports, etc.; use of hearing protection; knowledge and attitudes to noise levels and hearing loss. Data analysis Once cleaned, the translated files from the sound monitoring equipment were imported into SPSS for analysis. One-way analysis of variance (ANOVA) and within subject t-tests were carried out to make univariate comparisons of the monitored staff and between the pre- and post-shift audiograms at the recorded frequencies. Using SPSS, the hearing thresholds for both ears were assessed and the results from the pre- and postexposure audiograms were correlated in order to establish whether there were any differences in hearing threshold post-exposure at different frequencies. Similarly, analysis was made to examine differences in TTS for both ears at different noise exposure levels and different audiometric frequencies. RESULTS Questionnaire respondents and survey subjects The university Students Guild employed 112 parttime and 12 full-time staff, which included 60 security staff (48%), 46 bar staff (37%), seven bar managers (6%), five security managers (4%) and six technicians/djs (5%). Part-time staff (students) were permitted to work 16 h/week and, depending on job type, may work 2, 3 or 4 nights/week. Full-time staff worked a maximum of 45 h/week over 4 nights with occasional overtime. Over the duration of the study both the security and bar staff worked 4 nights of 10 h/shift. Of the 28 staff working in the three areas, 14 (50%) agreed to take part in the study, with some of the employees participating on more than one occasion. In total 21 pre- and post-shift audiograms were obtained. Four of the 14 subjects worked fulltime and four were females. The mean age of the student employees was 22 yr (range 20 25) and for full-time staff 33 yr (range 22 40). Noise measurements Table 1 shows the noise level ranges measured at the static monitoring points where most employees spent the majority of their time over the shift. The mean static L eq noise levels for the three entertainment areas ranged from 89 to 98 db(a). The bars were in operation all the time whilst the music was being played, therefore the employees were exposed throughout the sampling periods shown in Table 1. Figure 1 shows the variation in the L eq values recorded at the static sampling points from 20:00 to 02:00. The L EP,d values for employees (bar staff) were calculated using the L eq static measurements and found to be in the range db(a) (Table 1). Although these values do not take account of short breaks that staff take during the evening they do provide an estimate of an 8 h equivalent noise dose for comparison with noise action levels as defined in the British Noise at Work Regulations 1989 (NAWR) (Health & Safety Executive, 1989). Octave band analysis was performed to determine the frequency spectrums for the music type played in the three areas. Figure 2 shows the OBA data for

5 Noise exposure and hearing damage at university entertainment venues 459 Fig min L eq for the three music areas over 6 h. Fig. 2. Octave analysis for two music areas, over 4 h band. areas 1 and 2 over the duration of the sampling period. Figures 1 and 2 show that the noise intensity increased gradually from 21:30 to 00:30. The highest 15 min static L eq were recorded at 23:15 for areas 1 and 2 and 00:30 for area 3. The most prominent frequencies for area 1 were 500 and 1000 Hz and for area 2 were 1000 and 2000 Hz. OBA data also showed that over the duration of the function the lower frequencies (250 and 500 Hz) became more prominent. Table 2 shows the results of the personal dosemetry for the bar and security staff. The mean

6 460 S. Sadhra et al. Table 2. Staff personal exposure levels n Mean sampling time (min) Mean L eq [db(a)] SPL (max) [db(a)] Mean L EP,d [db(a)] Bar staff Security staff Table 3. Mean threshold shifts on post-exposure audiograms for 13 of the 14 subjects (21 audiograms) Frequency (khz) Right P Left P < Low frequency TTS mean [db (range)] 8.1 (0 25) < (5 45) <0.01 High frequency TTS mean [db (range)] 20.0 (5 50) < (5 35) <0.01 personal L eq for the security staff [94 db(a)] was higher than the mean exposure level for the bar staff [90 db(a)]. The calculated L EP,d levels for the bar staff exceeded the first noise action level of 85 db(a), whilst the second action level of 90 db(a) was exceeded by the security staff. High peak sound pressure levels were recorded for both groups. The maximum peak pressure readings recorded over the 3 days were 113 and 124 db(a) for the bar and security staff, respectively. Audiometry Hearing threshold shifts were recorded for both ears over the standard frequency range khz. Threshold shifts were calculated for individual frequencies as well as for the low (0.5, 1 and 2 khz) and high (3, 4 and 6 khz) frequency bands. The preshift audiometric data showed that seven of the 14 subjects each had more than 20 db hearing loss for either the low or high frequencies. Four subjects had hearing loss >30 db. One of the four subjects aged 22 had unilateral hearing loss, i.e. the difference in the sums of hearing levels between the two ears exceeded 45 db for the low frequencies. No significant differences were found between pre-shift low and high frequencies (both ears) for the same individuals tested on two consecutive days. The post-shift audiometry data for the subject with unilateral hearing loss was not included in the main analysis. Reliability of the measures of pre-shift audiometric data was examined when audiometry was performed on the same individuals (n = 6) on more than one occasion, and a reliability α coefficient of 0.72 was achieved. Background noise levels measured in the audiometry room at both pre- and post-shift testing were <50 db. Table 3 shows the mean changes in hearing threshold for 13 of the 14 subjects calculated from 21 pre- and post-shift audiograms. The mean TTS at the low and high frequencies were 8 and 20 db for the right ear and 16 and 21 db for the left ear, respectively. However, at 4 khz the mean TTS for the right and left ears were 8 and 5 db, respectively. The Table 4. Personal exposure levels and TTS L eq [db(a)] n Right post pre (db) > P Left post pre (db) recorded TTS values varied between 0 and 50 db. The TTS were highly significant (P < 0.01) for both the low and high frequency band for both ears, with the greater threshold shift at the higher frequencies. Hearing threshold shifts at the individual frequencies were significant for all frequencies except 1.0 and 8.0 khz. At 4 khz the mean threshold shift was more significant (P < 0.01) in the right ear when compared with the left ear (P = 0.02). This may be a consequence of some security staff wearing earpieces, but further investigation would be warranted. The dependence of TTS on personal noise exposure levels was explored by evaluating the correlation between personal L Aeq values and TTS in both ears at high frequency, low frequency and 4 khz. The strongest correlations were obtained at 4 khz (r = 0.42 for the right ear, r = 0.42 for the left ear, n = 22) and although these correlations were moderate, they were consistent for both ears. Table 4 compares the measured personal L eq levels for all subjects and their corresponding threshold shifts for both ears. The subjects were grouped into those with L eq levels either less than or greater than the second action level of 90 db(a), as defined under the NAWR. No significant difference was found in mean threshold shifts at the two noise levels for both ears (P = 0.15 and 0.08 for the right and left ears, respectively). General noise questionnaire In total 65 (52%) employees from the Guild completed the questionnaire, which included the 14 subjects who took part in audiometry and personal dosemetry. Of the employees who completed the

7 Noise exposure and hearing damage at university entertainment venues 461 questionnaire 45 (69%) did not perceive exposure to loud noise as a risk to their hearing. Furthermore, only 13 (20%) of the staff completing questionnaires claimed to have received information on risks of hearing damage from exposure to noise. Fifty-two (75%) claimed that they had not been issued hearing protection and those who were provided with ear defenders did not use them, whilst 62 (95%) were not aware of the NAWR or its basic requirements. DISCUSSION Playing or listening to amplified music has the potential to cause hearing loss in young people, with the risk increasing with higher sound levels and longer exposure durations (Clark, 1991; Bisch, 1996). The Medical Research Council (Institute of Hearing Research, 1986) has estimated that > adolescents suffer from hearing loss due to listening to amplified music in Great Britain. It could also be argued that the risk of hearing loss amongst young individuals has increased over the years due to the greater power of sound amplification, the availability of good quality personal stereos at relatively lower prices and the many establishments playing preamplified music. Although there is an abundance of literature on sound levels in discos, there are only a few studies that have investigated the occupational exposure to loud music for staff employed in such leisure establishments. The other gap in the literature is that studies on leisure music tend to concentrate mainly on the estimation of noise exposure levels and not the associated hearing loss. In this pilot study we have attempted to assess both personal noise exposure level for student bar and security staff and changes in hearing threshold. All staff were employed at a university Guild and the majority worked up to 16 h/week, working typically 2 or 3 nights/week. The majority of staff who completed and returned the questionnaire did not perceive exposure to loud noise as a risk to their hearing and only a few had received information on risks of hearing damage from listening to amplified music. The small numbers that were issued hearing protection did not use them, possibly due to difficulties in communication, discomfort or simply not accepting the need to use them. Given the subject age group and the work environment these findings are not surprising. However, the pre-shift audiometric data showed that four of the 13 staff had >30 db hearing loss for either the low or the high frequency band. The audiometric data was not corrected for age and gender. However, the corrections would appear to be insignificant given that the mean age group for the student employees was only 22 yr. Although all 13 volunteers for the audiometry met the inclusion criteria, one subject showed signs of unilateral hearing loss. Unilateral hearing loss is not normally associated with industrial noise exposure but may result from rifle fire and/or medical factors. None of the 14 subjects in this study regularly (>4 h/week) used a personal stereo. The static sound measurements were found to be better predictors of personal noise exposure levels for bar staff than the mobile security staff. The mean L EP,d levels calculated using dosemetry data for these groups showed that the first action level was exceeded for the bar staff and second action level exceeded for the security staff, who were thus at greater risk of developing hearing damage. The very high peak exposure levels of up to 124 db(a), which may result from shouting or use of personal communication systems (the volume settings of which were not known), are also of particular concern. The noise levels measured in this study are similar to those reported by other workers. Many authors (Rupp et al., 1974; Bickerdike and Gregory, 1980; Gunderson et al., 1997) have shown that mean sound levels exceed 85 db(a) and frequently exceed 90 db(a). However, it is not clear whether the data reported in some studies were based on personal or static exposure measurements. In this study the static samples placed behind both bars gave readings that were on average 7 db higher than the corresponding bar staff personal dosemeters. This difference is probably due to the movement of bar staff, work practices and possible noise shielding due to bar design, although the possibility that dosemeters may have slightly underestimated the personal exposure levels due to body reflections of noise may be possible. As reported in other studies (Bickerdike and Gregory, 1980), as the evening progressed sound levels were found to rise steadily and tended to level off half way through the night. Furthermore, different areas where amplified music was played showed slightly different frequency patterns, but in all three areas the lower frequencies became more prominent, especially after midnight. Regardless of the tolerance of high sound levels, discos should be designed to limit noise exposure of employees to <90 db(a)/shift. This could be achieved by careful consideration of the location of speakers, bars and the use of acoustic walls. A remote microphone could also be connected to the amplifier allowing both the remote measurement of sound levels and control of the output from the unit. Student staff in such establishments should be made more aware of noise levels and the likelihood of hearing damage from amplified music, preferably through development and implementation of hearing conservation programmes. The main aim of this study was to assess changes in hearing threshold amongst the student employees who worked in one of three areas where amplified music was played. Although there may have been variations in the fitting of headsets, the pre-shift

8 462 S. Sadhra et al. audiograms measured for each subject over two consecutive days were consistent with no significant difference (P < 0.01). Audiometric soundproof booths would have been preferable, however, the consistent results of pre-shift audiograms for subjects suggest that the audiometric measurement conditions and background levels were acceptable. Although the postshift audiograms were recorded within a short time of shift completion, it is possible that the true TTS may have been slightly underestimated. This study is further complicated by the fact that post-exposure audiometry was conducted between 02:00 and 03:00, thus human performance was unlikely to be optimum. Analysis of the audiograms showed that TTS was associated with noise exposure at both low and high audiometric frequencies for both ears. Moderate degrees of TTS of 20 db in either ear were recorded for nine (high frequency bands) and six audiograms (low frequency bands). The greatest threshold shifts were observed at the higher frequencies. At 4 khz the mean threshold shift was more significant (P < 0.01) in the right ear when compared with the left ear (P = 0.02). Although these TTS values cannot be used to predict the extent of PTS directly, they do indicate that the subjects are at risk of developing PTS if they are routinely exposed to the measured noise levels. Furthermore, the pre-shift audiometry data do suggest that at least 30% of the subjects have early signs of hearing loss, which may be due to amplified music, as none of the subjects played in bands or had hobbies involving exposure to high noise levels. Clearly, further studies are needed to assess the clinical significance of TTS values reported in this study. Future studies on hearing loss from amplified music should be designed to compare hearing loss amongst age-matched exposed and non-exposed groups, taking account of confounders such as the use of personal stereos, recreational activities that are associated with high noise exposure, and the use of ear protectors. Furthermore, more sensitive techniques, such as otoacoustic emissions (OEA), could be used to provide evidence of cochlear damage due to loud music. There are reports in the literature on the use of OEA in assessing cochlear function from industrial noise exposure (Hotz et al., 1994), however, until recently its use in the assessment of amplified music has been limited (Mansfield et al., 1999). The correlations between TTS and personal L Aeq values at both low and high frequencies were lower and inconsistent between ears, suggesting that measurement at 4 khz is a more reliable indicator of the relationship between TTS and L Aeq than the standard low and high frequency ranges. It is important to emphasize that although utilizing measurements from 14 individuals (50% of the obtainable student staff population), this study is concerned more with the internal validity of the findings, rather than generalizing such findings to the student population at large, as it is never appropriate to generalize an invalid finding (Mant et al., 1996). This pilot study raises a number of issues which need to be examined further. CONCLUSIONS This pilot study has shown that students employed as part-time bar and security staff in a university Guild may be exposed to amplified music for up to 16 h/week. The noise levels depend on job type, but invariably personal exposure levels exceed the first action level of 85 db(a). The mean personal noise exposures for security staff were found to be greater than for bar staff. More importantly, at the recorded personal exposure level hearing threshold shifts were found to be highly significant in both the low and high frequency bands and for both ears. Furthermore, four of the 13 subjects assessed showed a permanent hearing loss of >30 db. It is important that further studies examine both personal exposure levels and hearing loss in larger groups of young people exposed to amplified music. The studies should take account of both variation in music types and duration of exposure for disco employees, as well as personal stereo use and other sources of amplified music, such as attending live concerts and playing musical instruments. Further study designs should also compare hearing loss in exposed and non-exposed groups matched for age and gender and also address the issue of audiometric variability when using test retest methods. Future studies could assess any impact of worker fatigue upon audiometry testing, by using controls with long working hours that have no noise exposure, and the use of regression models to relate threshold changes to exposure (subject numbers allowing) would provide more definitive results. REFERENCES Axelsson A, Lindgren F. (1978) Hearing in pop musicians. Acta Otolaryngol; 85: Bickerdike J, Gregory A. (1980) An evaluation of hearing damage risk to attendees at discotheques. Department of the Environment contract no. DGR 481/99. Leeds Polytechnic. Bisch CM. (1996) Epidemiological evaluation of hearing damage related to strong amplified music (personal cassette players, discotheques, rock concerts). High definition audiometric survey on 1364 subjects. Audiology; 35: Carter NL, Waugh RL, Keen K, Murray N, Bulteau GVG. (1982) Amplified music and young people s hearing. Med J; 2: Clark W. (1991) Noise exposure from leisure activities: a review. J Acoust Soc Am; 90: Gunderson E, Maline J, Catalano R. (1997) Risk of developing noise-induced hearing loss in employees of urban music clubs. Am J Ind Med; 31: Health & Safety Executive. (1989) Noise at work regulations London: HMSO. Health & Safety Executive. (1995) A guide to audiometric testing programmes, guidance note MS 26. London: HMSO.

9 Noise exposure and hearing damage at university entertainment venues 463 Hotz MA, Harris FP, Probst R. (1994) Otoacoustic emissions: an approach for monitoring aminoglycoside-induced ototoxicity. Laryngoscope; 104: Institute of Hearing Research. (1986) Damage to hearing arising from leisure noise. Br J Audiol; 20: Korpert K, Winker N. (1994) Change of hearing threshold of young workers. In Proceedings of the international symposium on new epidemics in occupational health, Helsinki, Finland. Helsinki: Finnish Institute of Occupational Health. pp Luz GA, Fletcher JL, Fravel WJ, Mosko JD. (1973) The relation between temporary threshold shift and permanent threshold shift in rhesus monkeys exposed to impulse noise. Acta Otolaryngol; 312 (suppl.): Mansfield JD, Baghurst PA, Newton VE. (1999) Otoacoustic emissions in 28 young adults exposed to amplified music. Br J Audiol; 33: Mant J, Dawes M, Graham-Jones S. (1996) Internal validity of trials is more important than generalizability. Br Med J; 312: 779. McBride D, Gill F, Proops D, Harrington M, Gardiner K, Attwell C. (1992) Noise and the classical musician. Br Med J; 305: Palin SL. (1994) Does classical music damage the hearing of musicians? A review of the literature. Occup Med; 44: Rice CG, Rossi G, Olina M. (1987) Damage risk from personal cassette players. Br J Audiol; 21: Rupp RR, Banachowski SB, Kiselwich AS. (1974) Hard rock music and hearing damage risk. Sound Vibration; 8: Tan TC, Tsang MC, Wang TL. (1990) Noise surveys in discotheques in Hong Kong. Ind Health; 28: Ulrich RF, Pinheiro ML. (1974) Temporary hearing losses in teenagers attending repeated rock and roll sessions. Acta Otolaryngol; 77: 51 5.