Noise exposure and hearing status among call center operators.

INTRODUCTION: The overall objective of the study was to assess noise exposure and audiometric hearing threshold levels (HTLs) in call center operators.
MATERIALS AND METHODS: Standard pure-tone audiometry and extended high-frequency audiometry were performed in 78 participants, aged 19 to 44 years (mean ± standard deviation: 28.1 ± 6.3 years), employed up to 12 years (2.7 ± 2.9 years) at one call center. All participants were also inquired about their communication headset usage habits, hearing-related symptoms, and risk factors for noise-induced hearing loss (NIHL). Noise exposure under headsets was evaluated using the microphone in a real ear technique as specified by ISO 11904-1:2002. The background noise prevailing in offices was also measured according to ISO 9612:2009. RESULTS AND DISCUSSION: A personal daily noise exposure level calculated by combining headset and nonheadset work activities ranged from 68 to 79 dBA (74.7 ± 2.5 dBA). Majority (92.3%) of study participants had normal hearing in both ears (mean HTL in the frequency range of 0.25-8 kHz ≤20 dB HL). However, their HTLs in the frequency range of 0.25 to 8 kHz were worse than the expected median values for equivalent highly screened otologically normal population, whereas above 8 kHz were comparable (9-11.2 kHz) or better (12.5 kHz). High-frequency hearing loss (mean HTLs at 3, 4, and 6 kHz >20 dB HL) and speech-frequency hearing loss (mean HTLs at 0.5, 1, 2, and 4 kHz >20 dB HL) were noted in 8.3% and 6.4% of ears, respectively. High-frequency notches were found in 15.4% of analyzed audiograms. Moreover, some of call center operators reported hearing-related symptoms.
CONCLUSIONS: Further studies are needed before firm conclusions concerning the risk of NIHL in this professional group can be drawn.

Introduction

In past decades, there has been an increase in the usage of wired and wireless headsets in various occupational sectors of industry. One of the most dynamic and fastest growing branches all over the world where communication headsets are necessary to perform basic duties is call centers. According to the European Contact Centre Benchmark Platform, in 2012, 3.4 million workers in Europe were employed by more than 35,000 call centers.[1] In Poland, the estimated number of call center operators is around 23,700.[2]

The most common occupational health problems in this professional group are visual problems due to working with video display units, voice disorders due to continuous talking, and auditory problems due to communication headset usage and acoustic shocks.[34] However, relatively few studies have been published on the risk of hearing impairment in call center operators due to usage of communication headsets. Such a situation has probably been in part due to the difficulties in the measurement set-up and in the evaluation of the exposure itself.

Although several standards describe methods for general noise measurements in occupational settings (e.g., ISO 9612:2009[5], these are not directly applicable to noise assessments under communication headsets. For measurements under occluded ears, specialized methods have been specified by the International Standards Organization (ISO 11904-1:2002[6] and ISO 11904-2:2004[7] such as the microphone in a real ear and manikin techniques. Simpler methods have also been proposed in some national standards such as the use of general purpose artificial ears and ear simulators in conjunction with single number corrections to convert measurements to the equivalent diffuse field (AS/NZS 1269.1:2005, CSA Z107.56-13).[8910]

Nowadays in Poland, noise exposure evaluation from communication headsets, especially in call center operators, is not routinely performed. Only a few studies have, to date, been conducted.[11] Thus, there is no data on the scale of noise exposure and risk of noise-induced hearing loss (NIHL) in this professional group. Therefore, the overall purpose of this study was to evaluate hearing status of call center operators in relation to their noise exposure.

Materials and Methods

A study was conducted in call center operators, including questionnaire surveys, measurements of noise from headphones and background noise, and hearing tests.

The study group comprised 78 workers employed in one call center. They were recruited by advertisement and received financial compensation for their participation in the study. The study design and methods were approved by the Bioethical Commission of the Nofer Institute of Occupational Medicine, Lodz, Poland (resolution no. 03/2015 of 18 February 2015).

Questionnaire surveys

All participants filled in a questionnaire developed to enable identification of occupational and nonoccupational risk factors of NIHL and self-assessment of hearing status. The questionnaire consisted of items on (a) age and gender, (b) education and/or profession, (c) work history, including time of employment/exposure to noise and/or usage of headsets at current and previous workplaces, (d) data concerning current job (details of work pattern and equipment used, preferred volume control setting, type of calls typically handled, etc.), (e) medical history (past middle-ear diseases, and ear surgery, hereditary disorders, cholesterol levels, arterial hypertension, head trauma, etc.), (f) physical features (body weight, height, skin pigmentation), (g) lifestyle (smoking, noisy hobbies, using portable media players, attending disco/bars, rock concerts, etc.), and (h) hearing-related symptoms such as hearing impairment, difficulties in hearing or understanding whisper, normal speech and speech in noisy environment, as well as presence of tinnitus and hyperacusis.

In addition, participants’ hearing ability was assessed using a (modified) Amsterdam Inventory for Auditory Disability and Handicap [(m)AIADH].[12] This questionnaire consists of 30 questions and comprises two control questions not included in the assessment. The questions are divided into five parts (subscales) assessing separately: (a) the ability of discrimination of sounds (subscale I), (b) auditory localization (subscale II), (c) understanding speech in noise (subscale III), (d) intelligibility in quiet (subscale IV), and (e) detection of sounds (subscale V).

The respondents reported how often they were able to hear effectively in the situations specified above. The four answer categories were as follows: almost never, occasionally, frequently, and almost always. Responses to each question were coded on a scale from 0 to 3; the higher the score, the smaller the perceived hearing difficulties. The total score per participant was obtained by adding the scores for 28 questions. Maximum total score of the questionnaire was 84. In addition, the answers for each subscale were summed up (maximum score for subscale I was 24, whereas for the other scales, the total was 15 each).[12]

Noise exposure evaluations

To assess the noise exposure of call center operators, noise levels generated by the headsets and background noise levels were measured and information on typical working pattern was also collected.

The background noise levels were measured using integrating-averaging sound level meter and personal sound exposure meters, that is, a SVANTEK (Poland) sound and vibration analyzer type SVAN 958 (with a SVANTEK type SV12L preamplifier and a SVANTEK type SV22L 1/2-in. microphone) and Brüel & Kjær (Denmark) personal dosimeters type 4443 (with Brüel & Kjær type MM0111 1/2-in. microphones). The measuring instruments were placed consecutively at different locations in the call center to give a representative sample of typical background noise levels. The following noise parameters were determined according to PN-N-01307:1994[13] and ISO 9612:2009[5]: (a) A-weighted equivalent-continuous sound pressure level (SPL), (b) maximum A-weighted SPL with S (slow) time constant, and (c) peak C-weighted SPL.

Noise exposure from communication headsets (during phone calls as well as while awaiting them) was evaluated using a microphone in the real ear (MIRE) technique, as specified in ISO 11904-1:2002.[6] According to this standard, a miniature microphone probe, the SVANTEK type SV25S (connected to the one of two available inputs of the dual-channel acoustic dosimeter type SV102) was placed at the entrance of the open ear canal of call center operators, and the aforesaid noise parameters together with SPLs in 1/3-octave bands (from 20 to 10,000 Hz) were determined. Simultaneously, the second channel of dosimeter (equipped with a SVANTEK standard 1/2-in. microphone type SV25D) was used for assessment of noise exposure outside the other ear.

According to ISO 11904-1:2002,[6] results of the frequency analysis under headphone were then converted into corresponding diffused-field levels to obtain the diffuse-field-related A-weighted SPLs. A job-based measurement strategy according to ISO 9612:2009[5] was applied for exposure evaluation from both the headsets and background noise.

Each single noise sample was measured using dual-channel acoustic dosimeter type SV102 lasted approx. 30 min, whereas samples of background noise lasted on average 62 min.

In general, 241 (2 × 103 + 35) random noise samples of SPLs were collected.

Because participants used single-ear headsets, noise exposure was separately assessed for ear without and with headphone. In the latter case, for each study participant, daily noise exposure level (LEX,8h) was calculated by combining headset and nonheadset work activities and taking into account declared time of the headset usage (per working day), using the following equation:

where LH-C,Aeq,T is the energy mean of diffuse-field-related A-weighted equivalent-continuous SPLs under headset, in dBA; LB-N,Aeq,T is the energy mean of A-weighted equivalent-continuous SPLs of background noise, in dBA; TH is the declared time of the headset usage per working day, in hours; TB-N is the effective duration of nonheadset work activities, in hours; To is the reference duration, To = 8 h.

Similar calculation was made for the nonequipped ear. However, in this case, instead of diffuse-field-related A-weighted equivalent-continuous SPL under headset, the noise level measured outside the aforesaid ear was taken into consideration.

Hearing examinations

The standard pure-tone air conduction audiometry (PTA) and extended high-frequency audiometry (EHFA) were performed in all participants of the study. The auditory rest before the audiological evaluations was 14 h.

Hearing threshold levels (HTLs) for each ear were determined for both standard frequencies from 0.25 to 8 kHz (0.250, 0.5, 0.75, 1, 1.5, 2, 3, 4, 6, and 8 kHz) as well as extended frequencies from 8 to 18 kHz (8, 9, 10, 11.2, 12.5, 14, 16, and 18 kHz) with 5 dB steps. The bracketing method according to ISO 8253-1:2010[14] has been used in case of standard pure-tone audiometry. The similar methodology has been applied for EHFA. But in the latter case, the initial familiarization was performed using a tone of 11.2 kHz. The order of tones was from 11.2 upward to 18 kHz, followed by the lower frequency range, in the descending order (i.e., from 11.2 to 8 kHz). However, HTLs at 18 kHz was not included into analysis due to many missing data.

The PTA was always performed first, followed by the EHFA. In both cases, the right ear was tested first. The hearing examinations were conducted with the VIDEOMED Smart Solution (Poland) clinical audiometer, model AUDIO 4002 with the Holmberg GMBH & CO. KG Electroacoustik (Germany) headphones type HOLMCO P-81 for the PTA, and the Sennheiser Electronic GmbH & Co. KG (Germany) headphones type HDA 200 for EHFA. Prior to the audiological evaluations, otoscopy was performed.

Hearing tests were conducted in a quiet room located in the call center where the A-weighted equivalent-continuous SPL of background noise did not exceed 35 dBA.

Data analysis

Audiometric HTLs in call center operators were compared to age-related reference data from highly screened and unscreened populations according to ISO 7029:2017[15] and ISO 1999:2013.[16] Differences in HTLs between participants’ left and right ears as well as between ears exposed and nonexposed to noise from headsets were also explored.

The prevalence of normal audiograms, high- and speech-frequency hearing loss were analyzed among call center operators. Normal hearing was defined as having mean HTL between 0.25 and 8 kHz lower than or equal to 20-dB HL. In contrast, speech- and high-frequency hearing loss were defined as pure-tone mean of >20 dB HL at 0.5, 1, 2, and 4 kHz, and 3, 4, and 6 kHz, respectively. Percentages of ears with HTLs exceeding 20 dB HL at any of high frequencies (3–6 kHz) and at speech frequencies (0.5, 1, 2, and 4 kHz) were also calculated.

Recommendations of the British Society of Audiology were applied for assessment of severity of hearing loss among call center operators.[17] In contrast, to identify early signs of NIHL, the prevalence of high-frequency notched audiograms was analyzed. According to Cole’s recommendation, a high-frequency notch was defined as a HTL at 3 and/or 4 and/or 6 kHz at least 10 dB HL greater than at 1 or 2 kHz and at 6 or 8 kHz.[18]

Answers to the questionnaire and frequency of some outcomes (e.g., prevalence of hearing-related symptoms) were presented as proportions with 95% confidence intervals (95% CI). Differences between participants’ right and left ears in proportions of some hearing tests results (e.g., percentage of notched audiograms) were analyzed using Fisher’s exact test, whereas the differences in hearing thresholds levels between ears exposed and nonexposed to noise from headsets were analyzed using t test for dependent data or Wilcoxon singed-rank test, where applicable. Similar tests were used for comparison of HTLs in call center operators with reference data from highly screened nonnoise-exposed population as well as when analyzing the differences in audiometric thresholds between participants’ left and right ears.

In contrast, t test for independent data or Mann–Whitney U test, where applicable, was applied for assessment of differences between noise levels under headsets for various volume settings.

Statistical analysis was conducted with an assumed level of significance P = 0.05. However, when comparing noise levels obtained from MIRE technique for various volume settings, to avoid the risk of mass significance, a P value divided by the number (N) of possible comparisons (P = 0.05/N) was set as a limit for statistical significance. The STATISTICA software (StatSoft, Inc. (2010). STATISTICA (data analysis software system), version 9.1. www.statsoft.com.) was used for statistical analysis.

Results

Study group characteristics and questionnaire data

The study group comprised 37 females and 41 males aged 19 to 44 years [mean ± standard deviation (SD): 28.1 ± 6.3 years], employed up to 12 years in call center (mean ± SD: 2.7 ± 2.9 years), including over half (56.0%) less than 2 years. Almost all participants (98.7%) used the Plantronics single-ear headsets with microphone. Every fifth worker put the headphone alternately on both ears, whereas the others put it always on the same preferred right (42.3%) or left (36.6%) ear.

Nearly two-thirds of current call center operators (61.5%, 95% CI: 50.4%–71.5%) were exposed to noise at the previous workplace, of which 70.6% (95% CI: 56.9%–81.3%) to loud noise. Furthermore, nearly half of them declared the frequent (at least a few times per month) attending music clubs, pubs, or loud music concerts (48.7%, 95% CI: 37.8%–59.7%). A somewhat lower percentage (37.5%, 95% CI: 27.2%–49.1%) used portable media players every day (or a few times a week) for at least 1 h a day. Only a few participants (4.0%, 95% CI: 1.0%–11.7%) had noisy hobby (e.g., shooting or motor sport).

Among other risk factors for NIHL (such as smoking, elevated blood pressure, diabetes, white-finger syndrome, light skin pigmentation, and ototoxic antibiotic treatments),[19] the most frequent was smoking. Over half of study participants (59.0%, 95% CI: 47.9%–69.2%) were smokers at present or in the past.

Noise exposure evaluation

Table 1 summarizes measurement results of the background noise and noise from headsets. In particular, it presents both, uncorrected (at real ear) and corrected (diffuse-field-related), A-weighted equivalent-continuous SPLs measured using MIRE technique. Figure 1, in turn, shows 1/3-octave band frequency spectra of noise measured in the real ear under headset.

Table 1

Summary results of noise measurements at workplaces in call centre

Noise parameter	Total	Headset volume setting, V
	Mean±SD [L eq]	V ≤50%	50% <V <100%	V =100%
SPLs in the real ear with headphone
Diffuse-field-related A-weighted equivalent-continuous SPL [dB]	76.1±4.1 [78.3]	72.7±3.7[74.0]a,b	76.0±3.0 [77.1]a	77.2±3.7 [79.1]b
Uncorrected A-weighted equivalent-continuous SPL [dB]	80.0±4.2 [82.2]	76.4±3.6 [77.6]a,b	79.5±2.8 [80.5]a	81.3±3.7 [83.1]b
Uncorrected A-weighted maximum SPL [dB]	96.2±5.7	93.4±5.7b	94.7±3.6	97.6±5.7b
Uncorrected C-weighted peak SPL [dB]	112.9±5.4	111.4±3.1	112.5±5.7	113.8±5.9
SPLs outside the ear without headphone
A-weighted equivalent-continuous SPL [dB]	67.9±3.6 [69.4]	66.8±3.4 [68.1]	67.1±2.8 [67.9]	68.7±3.7 [70.2]
A-weighted maximum SPL [dB]	83.6±8.4	82.4±7.6	82.3±7.5	84.8±9.0
C-weighted peak SPL [dB]	108.6±8.5	106.3±7.7	110.0±8.7	108.9±8.7
SPLs of the background noise
A-weighted equivalent-continuous SPL [dB]	58.1±5.7 [61.5]
A-weighted maximum SPL [dB]	79.7±6.7
C-weighted peak SPL [dB]	98.1±8.0

L eq is the energy average of N samples of measured A-weighted equivalent-continuous sound pressure levels. SD = standard deviation, SPL = sound pressure level. a,b Significant differences between SPLs corresponding to various volume settings of communication headsets (P < 0.05/3).

Figure 1

1/3-octave band spectra of noise measured under headsets using MIRE technique

Generally, headsets generated noise (dominated by frequency content from 250 to 3150 Hz) at diffuse-field-related A-weighted equivalent-continuous SPLs ranging from 63 to 88 dBA, however, with only 3% and 18% of cases exceeding 85 and 80 dBA, respectively [Figure 2]. In contrast, background noise remained within the range of 54 to 74 dBA.

Figure 2

Distributions of diffused-field related A-weighted equivalent-continuous sound pressure levels produced by communication headsets.

It is obvious that higher volume settings of communication headsets were associated with higher levels of produced sounds, but without significant impact on A-weighted equivalent continuous SPLs measured outside the ear without headphone [Table 1].

According to the questionnaire responses, workers spent from 2.0 to 10.0 h per day (5.2 ± 2.0 h, median = 6.0 h) on phone calls. Furthermore, majority of them (77.6%) usually set the volume of communication headset at over 50% of maximum value, including 14.9% at 100% of maximum volume. Subsequently, the personal daily noise exposure level determined on the basis of data from the MIRE technique remained within the range of 68 to (79 dBA (mean±SD: 74.7 ± 2.5 dBA, median= 75.0 dBA), whereas the LEX,8h levels obtained for ear without headset varied from 63 to 70 dBA (mean±SD: 66.6 ± 1.7 dBA, median =67.0 dBA) [Figure 3].

Figure 3

Distributions of daily noise exposure levels (LEX,8h) in call center operators. Data represents the LEX,8h levels determined for ear with (a) and without (b) headphone taking into account both headset and nonheadset work activities

Generally, these values did not exceed the Polish maximum admissible intensity (MAI) values for occupational noise (LEX,8h = 85 dBA),[20] nor the lower exposure action value (LEX,8h = 80 dBA) specified by Directive 2003/10/EC.[21] In addition, A-weighted maximum SPLs and C-weighted peak SPLs (measured during both headset and nonheadset activities) were also lower than Polish MAI values which are equal to 115 dBA and 135 dBC, respectively.[20]

It is worth stressing that noise levels occurring during phone calls were higher than recommended level (LAeq,T = 65 dBA) to ensure proper working conditions at workplaces in observational dispatcher cabins, telephone remote control rooms used in management procedures, on premises for precise works, and so on.[13] Not surprisingly, therefore, that noise prevailing in call centers was assessed as annoying by a number of workers.

Results of audiometric tests

Audiometric HTLs determined in the 78 call center operators (156 ears) together with the expected HTLs for comparable highly screened (otologically normal) and unscreened populations are presented in Table 2 and Figure 4, respectively. Reference data on hearing thresholds (in the frequency range of 0.25–12.5 kHz) for the highly screened otologically normal population were obtained from ISO 7029:2017[15] while for unscreened populations (within 0.5–8 kHz) from ISO 1999:2013.[16]

Table 2

Standard pure-tone audiometry (PTA) and extended high-frequency audiometry (EHFA) hearing thresholds in call centre operators together with the expected median values of hearing thresholds for comparable highly screened otologically normal population according to ISO 7029:2017[15]

	Frequency (Hz)	Hearing threshold level (dB HL)
		Mean±SD
		Left ear	Right ear	Ear without headphone*	Ear with headphone*	Both ears	ISO 7029:2017
PTA	250	14.4±5.5a	11.6±5.1a	13.4±6.0	13.8±5.6	13.0±5.4c	0.2±0.3c
	500	15.2±4.7	14.6±4.6	15.2±4.8	15.4±5.1	14.9±4.6c	0.3±0.4c
	750	16.4±5.1	15.5±4.9	16.8±5.3b	15.4±5.4b	16.0±5.0c	0.3±0.4c
	1000	15.6±5.7a	14.2±4.8a	15.2±5.4	14.6±5.7	14.9±5.3c	0.3±0.5c
	1500	14.7±5.3a	12.9±5.3a	14.6±5.2	14.6±6.2	13.8±5.4c	0.4±0.6c
	2000	13.1±8.0	12.6±8.2	13.1±9.4	13.5±7.7	12.9±8.1c	0.8±1.3c
	3000	11.8±8.8	11.0±10.3	11.4±11.0	11.0±5.9	11.4±9.5c	0.7±1.0c
	4000	10.4±9.1	10.1±12.3	10.4±12.3	9.9±8.6	10.2±10.8c	0.9±1.2c
	6000	6.7±10.6	4.6±10.9	6.8±12.6	4.3±9.8	5.6±10.8c	1.2±1.5c
	8000	14.5±10.3	12.9±10.2	13.5±10.4	13.8±10.3	13.7±10.3c	1.4±1.8c
EHFA	8000	6.4±8.7a	12.9±10.2a	8.0±10.9	13.8±10.3	7.6±9.8c	1.4±1.8c
	9000	0.3±10.1a	8.8±10.8a	3.3±13.4	7.1±9.3	2.9±11.5	2.1±3.1
	10,000	5.1±11.7	5.5±12.3	5.7±13.9	2.1±10.9	5.0±11.7	4.5±4.4
	11,200	7.1±12.3	5.0±11.8	8.6±14.1	5.3±11.2	7.6±12.6	6.4±5.9
	12,500	7.6±15.4	8.1±12.9	8.6±16.0	8.3±12.8	7.6±15.3c	11.1±8.7c
	14,000	4.0±15.0a	7.6±15.3a	6.6±18.1	7.3±16.3	6.5±16.9	–
	16,000	14.2±17.7a	9.0±18.3a	15.4±19.8	7.2±17.1	16.7±19.4	–
Number of cases		78	78	56	56	156	156

SD = standard deviation. aSignificant differences between left and right ear (P < 0.05), bsignificant differences between ear without and with headphone (P < 0.05), csignificant differences between call centre operators and comparable highly screened otologically normal population (P < 0.05).*Data concerns participants who used headphone on one ear only.

Figure 4

Statistical distributions of hearing threshold levels (HTLs) in call center operators and unscreened Swedish (a), Norwegian (b), and United States (c) populations according to ISO 1999:2013.[16] Data are given as 10th/50th/90th percentiles. Rhombus and whiskers represent call center operators, whereas solid lines represent reference populations

The latter standard, among other things, contains examples of databases for unscreened populations of three typical industrialized countries, that is, Sweden (database B2), Norway (B3), and the United States (database B4). However, it does not include formulae applicable for calculation of the statistical distribution of hearing thresholds as function of age and gender. Thus, the distribution of HTLs in comparable unscreened Swedish, Norwegian, and US populations were estimated taking into account mean age and sex compositions in the study participants. In contrast, the expected medians of the HTLs in the reference highly screened population were determined using formulae given in ISO 7029:2017.[15]

As demonstrated, call center operators’ HTLs in the frequency range of 0.25 to 8 kHz were significantly higher than expected median values for comparable (due to age and gender) highly screened nonnoise-exposed population, whereas at higher frequencies, they were close (9–11.2 kHz) or better (at 12.5 kHz) than predictions [Table 2]. Furthermore, in the frequency range below 3000 to 4000 Hz, a tendency to worse hearing thresholds among call center operators compared to unscreened populations was observed when analyzing results of PTA testing in comparison with reference data from databases B2, B3, and B4 from ISO 1999:2013[16] [Figure 4].

There were significant differences in mean hearing thresholds between left and right ear at 0.25, 1, 1.5, 8, 9, 14, and 16 kHz (P < 0.05) [Table 2]. Standard pure-tone audiometry hearing thresholds (below 2 kHz) were worse in left ear compared to right ear, whereas opposite relation was observed for high-frequency audiometry thresholds (at 8, 9, 14, and 16 kHz). However, there were no significant differences in HTLs (in the whole frequency range excluding 750 Hz) between ears without and with headphone, when analysis was limited to participants who reported usage of headsets on one ear only [Table 2].

Generally, 92.3% (95% CI: 83.9%–96.7%) of study population had normal hearing (mean HTL in the frequency range of 0.25–8 kHz ≤20 dB HL for both ears). Furthermore, average HTL at 0.25, 0.5, 1, 2, and 4 kHz did not exceed 20 dB HL in 70 (89.7%, 95% CI: 80.8%–95.0%) participants [147 (94.2%; 95% CI: 89.3%–97.1%) ears]. Mild hearing loss (expressed as mean HTL at 0.25, 0.5, 1, 2, and 4 kHz between 21 and 40 dB HL)[17] was observed in nine (5.8%, 95% CI: 2.9%–10.7%) of analyzed ears, two-thirds of which in case of left ear. Neither moderate (41–70 dB HL) nor severe (71–95 dB HL) hearing losses were found in study participants.

High-frequency hearing loss (mean HTL at 3, 4, and 6 kHz >20 dB HL) and speech-frequency hearing loss (mean HTL at 0.5, 1, 2, and 4 kHz >20 dB HL) were noted in 8.3% and 6.4% of audiograms, respectively [Table 3]. In contrast, high-frequency notched audiograms were found in the 15.4% of analyzed ears. Most of them occurred at 4 kHz. However, neither prevalence of hearing loss nor high-frequency notching differed significantly between left and right ear [Table 3].

Table 3

Prevalence of hearing loss, high-frequency hearing loss, speech-frequency hearing loss, and high-frequency notched audiograms in call center operators

	Total	Left ear	Right ear
	Proportion (95% CI) (%)
Mean hearing threshold level >20 dB HL
Frequency range of 0.25–8 kHz	5.1 (2.5–10.0)	3.8 (0.9–11.3)	6.4 (2.5–14.6)
Frequencies of 0.5, 1, 2, and 4 kHz	6.4 (3.4–11.6)	3.8 (0.9–11.3)	9.0 (4.2–17.7)
Frequency range of 3–6 kHz	8.3 (4.9–13.9)	7.7 (3.3–16.2)	9.0 (4.2–17.7)
Hearing threshold level in at least one frequency >20 dB HL
Frequency range of 0.25–8 kHz	38.5 (31.2–46.3)	33.3 (23.9–44.4)	43.6 (33.2–54.6)
Frequencies of 0.5, 1, 2, and 4 kHz	20.5 (14.9–27.6)	19.2 (12.0–29.5)	21.8 (14.0–32.3)
Frequency range of 3–6 kHz	16.7 (11.6–23.4)	14.1 (7.9–23.8)	19.2 (12.0–29.5)
High-frequency notching
At 3, 4, or 6 kHz	15.4 (10.5–22.0)	17.9 (10.9–28.1)	12.8 (7.0–22.3)
Bilateral notching at 3, 4 or 6 kHz	3.8 (0.9–11.3)

CI = confidence interval.

Self-assessment of hearing capability

Generally, majority of participants (87.2%, 95% CI: 77.7%–93.0%) assessed their hearing as good. However, some of them reported gradually progressing hearing impairment (24.4%, 95% CI: 16.2%–35.1%) and complained of difficulty in hearing whisper (15.4%, 95% CI: 8.9%–25.2%), problems with understanding speech in noisy environment (28.2%, 95% CI: 19.4%–39.1%), hyperacusis (15.4%, 95% CI: 8.9%–25.2%), and having tinnitus after work (6.4%, 95% CI: 2.5%–14.6%).

Call center operators examined using the (m)AIADH obtained the mean total score 85.7 ± 10.1% of the maximum value (84), which was close to normative value and suggested no hearing problems [Table 4]. Only a small percentage of participants (7.8%, 95% CI: 3.4%–16.4%) obtained the total score under 70% of the maximum value. Relatively low scores were more frequent in subscales evaluating auditory localization (subscale II) and intelligibility in noise (subscale III), because 24.4% (95% CI: 16.2%–35.1%) and 16.9% (95% CI: 10.1%–27.0%) of call center scored below 70% of maximum value.

Table 4

Hearing ability in terms of score in the (modified) Amsterdam Inventory for Auditory Disability and Handicap [(m)AIADH] in call centre operators

	(m)AIADH score
	Mean±SD
Total	71.9±8.5
Subscale
I—Distinction of sounds	21.8±2.2
II—Auditory localization	12.3±2.5
III—Intelligibility in noise	12.0±1.5
IV—Intelligibility in quiet	13.0±1.8
V—Detection of sounds	12.9±2.1

SD = standard deviation.

Discussion

The main objective of this study was to analyze the audiometric HTLs of call center operators in relation to their noise exposure. Noise levels under communication headsets were assessed using the MIRE technique as specified in ISO 11904-1:2002,[6] whereas background noise prevailing in offices was measured according to ISO 9612:2009.[5] Determined under headsets the diffuse-field-related A-weighted equivalent-continuous sound pressure level remained within the range of 63 to 88 dBA, however, with only 3% and 18% of cases exceeding 85 and 80 dBA, respectively. In contrast, the noise occurring outside the ear without headphone varied from 61 to 76 dBA and was in majority (78%) cases higher than the recommended in Poland A-weighted equivalent-continuous sound pressure level (65 dBA) to ensure proper working conditions at workplaces in observational dispatcher cabins, telephone remote control rooms used in management procedures, on premises for precise works, etcand so on.[13] Not unsurprisingly, therefore, it was regarded as an annoying by some workers.

Almost all participants (98.7%) used single-ear headsets. Subsequently, the personal daily noise exposure levels determined by combining headset and nonheadset work activities varied from 68 to 79 dBA and 63 to 70 dBA for ears with and without headphone, respectively. Thus, our noise measurement showed that call center operators are unlikely to be exposed to noise exceeding lower and upper exposure action values from Directive 2003/10/EC.[21]

Generally, the aforesaid outcomes are in agreement with the results of some earlier investigations, although different methods were used to assess sound immission from communication headsets.[112223]

For example, Patel and Brougthon[22] visited 15 call centers in the United Kingdom to evaluate if there was a risk to hearing from working in call center. They measured noise exposure in 150 operators and revealed that corrected noise levels generated by headsets fitted on the KEMAR manikin varied from 65 to 88 dBA, whereas background noise levels were between 57 and 66 dBA. Subsequently, taking into account the time spent by workers on phone calls, the estimated daily noise exposure level ranged from 67 to 84 dBA or 87 dBA in case of using for estimation mean or maximum corrected noise levels, respectively. On that basis, Patel and Broughton concluded that daily noise exposure level of call center operators is unlikely to exceed 85 dBA, and therefore, the risk of hearing impairment is extremely low.[22]

Later, Smagowska[11] reported noise levels at 18 workstations in call center in Poland. Measurements were performed with a miniature microphone placed in the entrance of the external ear canal according to ISO 11904-1:2002; however, measured levels were not corrected to obtain free- or diffuse-field-related A-weighted equivalent-continuous sound pressure levels under headsets. Noise levels during phone calls varied from 68 to 91 dBA, while anticipating a phone call remained within the range of 55 to 65 dBA. Subsequently, daily noise exposure levels ranged from 62 to 87 dBA, showing that noise at call center workstations can be an annoying factor contributing to hearing loss in some cases.

More recently, Gerges et al.[23] analyzed the results of 166 noise level measurements in various call centers in Brazil. These measurements were also conducted according to methodology described in ISO 11904-1:2002.[6] However, contrary to our study, every single measurement lasted much longer and included the whole working shift. Therefore, the measuring equipment (with mini-microphone) was installed at the beginning of the participant’s working day, and it was removed at the end. Diffuse-field-related A-weighted sound pressure levels determined on the basis of these measurements remained within the range from 71 to 85 dBA, however, with only 14.4% of cases exceeding 80 dBA.[23]

In contrast, according to the latest study by Venet et al.[24] comprising 39 French call center operators (working with headsets), mean value of diffuse-field related A-weighted equivalent-continuous sound pressure level measured under headset using manikin technique was 69.6 ± 3.7 dBA. Consequently, both maximum and mean daily noise exposure level normalized for an equivalent 8-h exposure duration (equal to 75.5 and 65.7 ± 3.6 dBA, respectively) was well below the lower action level according to Directive 2003/10/EC.[21]

Regarding hearing status, almost all (92.3%) our study participants had mean audiometric hearing thresholds (in the frequency range 0.250–8 kHz) within normal limits (i.e., ≤20 dB HL) in both ears. Neither moderate (41–70 dB HL) nor severe (71–95 dB HL) hearing losses were observed in call center operators. Only mild hearing loss (21–40 dB HL) was found in 5.8% of analyzed ears. Furthermore, both high-frequency and speech-frequency hearing losses were observed in less 10% of analyzed audiograms. Such results of pure-tone audiometry are not surprising.

First, according to the ISO 1999:2013[16] model, the lowest noise exposure level normalized to 8-h working day (or 40-h working week) which may cause any noise-induced permanent hearing threshold shift is equal to 75 dBA. Furthermore, a permanent shift of hearing threshold greater or equal to 25 dB HL in speech frequencies should not take place in males with healthy ears, provided the exposure to noise does not exceed 15 years for 85 dBA level and 6 years for 90 dBA level.[16] However, in this study, call center operators were exposed to relatively low noise levels. The highest individual daily noise exposure level was 79 dBA. Furthermore, our study participants were relatively young (59% were aged 19–30 years) with job seniority up to 12 years (including 56% less than 2 years).

Second, individual susceptibility (or vulnerability) to noise, along with the degree of hearing loss, varies greatly among people. It is believed that NIHL is a complex disease, resulting from interaction between intrinsic and environmental factors. Besides, well-known environmental factors contributing to NIHL, such as exposure to noise and some other factors may also play a role. They include coexposures to ototoxic substances (organic solvents and heavy metals); noise and vibration; ototoxic drugs (aminoglycosides) and heat. Associations have also been observed between several individual factors and NIHL, including smoking, elevated blood pressure, diabetes, cholesterol levels, skin pigmentation, gender and age, and genetic predisposition as suggested by clinical knowledge and guidelines.[19] In our study, nearly half of call center operators declared the frequent presence at music clubs, pubs, or loud music concerts, and somewhat lower percentage used portable media players. Among other risk factor for NIHL, the most frequent was smoking. However, the aforesaid risk factors did not play important role probably due to relatively young age of call center operators.

Third, our findings are in line with the observations from some earlier studies. For instance, Mazlan et al.[25] examined call center operators in Malaysia, among others, to analyze the prevalence of hearing loss in relation to the duration of service. Their study group comprised 136 workers, aged 18 to 35 years, wearing headphones and receiving calls continuously for 7 h. As in our study, the majority (47%) of Malaysian participants has been working between 2 to 3 years, and the longest duration of service was 8 years in three participants. The average noise level from headphones was found to be 58 dBA.

Results of pure-tone audiometry revealed that 78.8% of examined call center operators had normal hearing in both ears, and only 21.2% of them were found to have hearing impairment in either one or both ears. (Normal hearing was defined as having HTL between −10 dB HL and 20 dB HL for all frequencies from 250 to 8000 Hz. Hearing impairment was defined as having HTL of more than 20 dB HL in at least one frequency.) That prevalence was comparable to prevalence of hearing loss in normal participants used as controls in other Malaysian studies. Furthermore, there was no association between hearing loss and duration of employment. Thus, it was concluded that there was no evidence of NIHL among call center operators with prolonged exposure to noise from headphones and the duration of service.[25]

More recently, Ayugi et al.[26] conducted a descriptive cross-sectional study in 1351 call center operators (aged 19–55 years) to study the prevalence of symptoms of acoustic shock syndrome. They noted such symptoms in 384 (13%) of study participants. Blockage or fullness of the ears (27.7%), headache (25.8%), otalgia (24.9%), tinnitus (21.3%), hoarseness of voice (21.8%), and hyperacusis (19.5%) were the most common complaints. However, despite the numerous symptoms of acoustic shock syndrome, only 21 (i.e., 5.5% of 384 and 1.6% of 1351) workers had a form of hearing loss. Twelve females had mild hearing loss, whereas only one man had severe hearing loss.

However, different conclusions were formulated by El-Bastar et al.,[27] who analyzed the prevalence of sensory-neural hearing loss (SNHL) among older 58 telephone operators, including those using headphones (age: 46.3 ± 8.1 years, time of employment: 20.6 ± 9.1 years) in comparison with 30 administration staff workers (age: 47.2 ± 8.1 years, time of employment: 21.7 ± 8.2 years). They found that telephone operators had significantly higher prevalence of acoustic shock symptoms and decreased hearing sensitivity compared to the controls. In particular, they noted 44.8% cases of SNHL among the telephone operators versus no cases among the controls; all of them were bilateral in distribution and concluded that among other analyzed factors, only headset use (odds ratio = 5.2, 95% CI: 1.7–16.1) and age (OR = 1.1, 95% CI: 1.0–1.2) were significant risk factors for developing SNHL among telephone operators.[27]

Because majority of our study participants had hearing thresholds within normal limits, to identify early signs of NIHL, the prevalence of high-frequency notches in audiograms was analyzed. Generally, various definitions of audiometric notches have been proposed. In this investigation according to Cole’s recommendation, a high-frequency notch was defined as a HTL at 3 and/or 4 and/or 6 kHz at least 10 dB HL greater than at 1 or 2 kHz and at 6 or 8 kHz.[18] Such notches, mostly occurring at 4 kHz, were found in the 15.4% of analyzed ears.

Recently, Corroll et al.[28] analyzed the prevalence of audiometric notches among the United States adult population (aged 20–69 years) based on data collected within the 2011 to 2012 National Health and Nutrition Examination Survey. They found that generally nearly one-fourth adults (24%) had bilateral or unilateral audiometric notches. Among participants not reporting work exposure this number was estimated to be 20%. The presence of notches increased with age, ranging from 17.6% among persons aged 20 to 29 years to 18.6% among persons aged 30 to 39 years. That study defined the presence of a high-frequency notch when any threshold at 3, 4, or 6 kHz exceeded the averaged threshold at 0.5 and 1 kHz by ≥15 dB HL and the 8-kHz threshold was at least 5 dB HL lower (better) than maximum threshold at 3, 4, or 6 kHz. Despite the difference in the notch definitions, our findings are comparable with those obtained by Corroll et al.[28] Thus, the prevalence of high-frequency notches in call center operators corresponded to that occurring in not occupationally exposed to noise population.

However, comparison of call center operators to highly screened otologically normal nonnoise-exposed population (according to ISO 7029:2017[16] revealed that their HTLs at extended frequencies were comparable (9–11.2 kHz) or better (12.5 kHz) than expected due to age and gender, whereas at standard frequencies, they were worse than predictions. Moreover, a similar tendency was observed in the lower frequency range (below 3000–4000 Hz) when HTLs of call center operators were compared to unscreened populations (databases B2, B3, and B4) according to ISO 1999:2013.[16] It is worth underlining that databases B2 and B3 represent populations who have not been exposed to occupational noise, whereas participants with occupational noise exposure are included in database B4.[16]

More recently, in the above citied study, Venet et al.[24] also analyzed auditory fatigue among call center dispatchers working with headsets. However, due to much lower noise exposure levels (up to 75.5 dBA, with mean value 65.7 ± 3.6 dBA), no significant temporary changes in hearing was detected with either pure-tone audiometry or the EchoScan test. In conclusion, it was suggested that dispatchers’ fatigue was probably due to duration of the work shift or to the tasks they performed rather than to the noise exposure under a headset.[24]

According to a subjective evaluation, in this study, majority of call center operators had good hearing. Results of the (modified) Amsterdam Inventory for Auditory Disability and Handicap also suggested no substantial hearing problems among study participants. Nevertheless, some of them complained of problems with understanding speech in noisy environment (28.2%), difficulty in hearing whisper (15.4%), hyperacusis (15.4%), and having tinnitus after work (6.4%).

Conclusion

Noise measurements showed that the mean daily personal noise exposure level of call center operators is unlikely to exceed the lower exposure action value (LEX,8h = 80 dBA) from Directive 2003/10/EC.[21]

Almost all of call center operators had normal hearing. However, despite the young age and short time of usage of communication headsets, a number of workers complained of some hearing-related symptoms and had high-frequency notched audiograms typical for NIHL. Moreover, comparison of call center operators to highly screened otologically normal nonnoise-exposed population (as specified in ISO 7029:2017[15] revealed that their HTLs at standard frequencies (0.25–8 kHz) were worse than expected with regard to age and sex, whereas at extended frequencies, they were comparable (9–11.2 kHz) or better (12.5 kHz).

Further studies are needed, comprising a greater number of participants, as well as longer time of employment, before firm conclusions concerning the risk of NIHL in the call center operators can be drawn, especially in the light of alarming information coming from recently published paper presenting a case report of NIHL in 30-year-old call center operator.[29]

Financial support and sponsorship

This study was supported by the Ministry of Science and Higher Education of Poland (Grant IMP 17.3/2015–2016).

Conflicts of interest

There are no conflicts of interest.