Sage Journals: Discover world-class research

Abstract

Background

The prevalence of age-related hearing loss is expected to increase with the ageing population in Singapore resulting in increased demand for pure tone audiometry (PTA). One way of meeting the increased demand for PTA is the Automated Method of Testing Auditory Sensitivity (AMTAS). Nonetheless, before AMTAS may be implemented, it is important to validate its results by comparing them against those of conventional PTA for the diverse Singaporean population.

Methods

100 participants were recruited from the Otorhinolaryngology clinic at a public hospital. PTA was conducted for participants via the conventional method and AMTAS. The hearing thresholds obtained via the two methods were then compared. A feedback questionnaire on the use of AMTAS was also completed by all participants.

Results

Air and bone conduction threshold results from 80 participants were analysed. The absolute mean difference in air conduction thresholds obtained via the two methods ranged from 3.30 to 9.62 dB. The absolute mean difference in bone conduction thresholds obtained via the two methods ranged from 8.16 to 9.38 dB. Most participants indicated that the machine was easy to use. Other feedback collected indicated that the AMTAS testing environment was noisy and that testing was fast.

Conclusion

This study finds that despite differences in hearing thresholds obtained via AMTAS and manual PTA, these differences are within the acceptable 10 dB test-retest variation. Nonetheless, existing issues related to the AMTAS software and noise levels in the testing environment will need to be addressed before it can be used in the clinic.

Keywords

Automated method of testing auditory sensitivity hearing audiometry automated audiometry

Introduction

The hearing of Singapore residents is becoming a national concern. In 2017, 63.7% of 662,480 Singapore residents aged 60 years and above were estimated to have hearing loss.¹ Hearing loss is defined as a 4-frequency average (0.5 kHz, 1 kHz, 2 kHz, and 4 kHz) greater than 25 dB HL. This number is likely to be higher now with the ageing population in Singapore and the higher prevalence of hearing loss with increasing age.^1,2 Recognizing this issue, hearing screening was included in a nationwide functional health screening program. Preliminary data from this showed that at least 16,100 out of 23,100 people screened in 2018, would need further detailed hearing assessment or diagnostic pure tone audiometry (PTA).³ As the screening program expands, the demand for diagnostic PTA is predicted to increase further. Yet, despite this, it is difficult to increase the capacity for conventional diagnostic PTA in Singapore due to manpower, space, and financial constraints.⁴ One way of meeting this demand is via automated audiometry specifically the Automated Method of Testing Auditory Sensitivity (AMTAS) software.

Automated audiometry is diagnostic PTA that is conducted without direct access to trained personnel.⁵ This is done using various equipment and has been validated in numerous countries since 2010.^5–9 AMTAS is a computer software used with a clinical audiometer that allows the full clinical diagnostic PTA, to be conducted without trained personnel.¹⁰ Over the past decade, numerous studies have been conducted on it by the developers and independent groups.^9,11–13 These studies concluded that variations in air and bone conduction thresholds obtained via AMTAS and manual PTA were within the acceptable range of 10 dB.^9,11 Thus, AMTAS is now commercially available.¹⁰ The strength of the evidence obtained overseas for AMTAS is the reason that AMTAS is selected as the automated audiometry solution in this study. There is however limited information about the validity and feasibility of AMTAS especially in a country with multi-ethnicities who speak different languages.

Studies on AMTAS were done mainly in Western countries with pre-dominantly English native speakers and no such studies have been done in Asian countries to our knowledge. More importantly, AMTAS is only available in English with hearing test instructions recorded in English by a native English speaker. Given the large range in English fluency within the Singapore population, this could affect the accuracy of the hearing test results obtained via AMTAS.¹⁴ Nonetheless, with the potential benefits of AMTAS in alleviating the manpower demand for diagnostic PTA in Singapore, it is important to validate the use of it in Singapore’s heterogeneous population.

In this study, we have two aims. The first aim is to determine the validity of AMTAS by determining whether there is any difference in hearing thresholds (air and bone conduction) between AMTAS and conventional manual PTA. The second aim is to determine the feasibility of using AMTAS in a specialist outpatient clinic at a restructured government hospital.

Methods

Study design

To validate AMTAS^®, we designed a prospective single group within-subject repeated measures study. Validity was determined by comparing the pure tone hearing obtained via AMTAS^® with those obtained from an established gold standard assessment method, conventional manual PTA.¹⁵ We conducted both manual PTA and AMTAS^® for study participants and compared the hearing thresholds obtained via the two methods. Feasibility of AMTAS^® was assessed based on whether there was any change in the hearing profile of participants obtained via the two methods and on participant feedback in a short questionnaire completed at the end of testing.

Participants

One hundred study participants were recruited from outpatients attending an Otorhinolaryngology Head and Neck specialist outpatient clinic within a government-restructured hospital in Singapore. These patients were referred for PTA testing by specialist doctors, as part of routine clinical care. Outpatients were recruited if they were above 12 years old and English literate. They were excluded if they had impacted ear cerumen, and/or visible active ear infections. Ethics approval was obtained from Singhealth Centralized Institutional Review Board (CIRB) (Reference number: 2017/3069). Written informed consent was obtained from all participants. All procedures were performed in accordance with the ethical principles of the declaration of Helsinki 1964.

Manual conventional pure tone audiometry

Manual PTA was conducted by a trained audiologist or audiometric technician with at least 1 year of experience from a team of six. Manual PTA was carried out according to the recommended PTA procedure by the British Society of Audiology¹⁶ and a mastoid placement of the bone conductor was used.

Testing was conducted in an American National Standards Institute (ANSI) certified sound booth using a GN Otometrics Madsen Itera (Copenhagen, Denmark) audiometer with TDH39 circum-aural headphones, and mastoid placement of a B71 bone oscillator was used for bone conduction. Threshold seeking was done with the modified Hughson-Westlake method. Hearing thresholds were plotted manually into the audiogram module in NOAH software version 4.4 by the tester.

AMTAS^®

AMTAS^® is a software developed in 2002.¹⁷ It automates the process of obtaining a full diagnostic pure tone audiogram using a touchscreen computer and any audiometer that can be controlled with a computer.¹⁷ In this study, the AMTAS^® Pro version 1.1 was used with the GSI Audiostar Pro audiometer (Grason-Stadler, Eden Prairie, MN, USA) and a touchscreen desktop computer.

With AMTAS^®, the audiometer was hidden, and study participants were presented with a touchscreen monitor to make their responses (Figure 1). Trained personnel placed the forehead calibrated B81 bone conductor in the centre of the participant’s forehead, and it was secured with an AMBAND or a calibrated bone conduction band. Thereafter, circum-aural Sennheiser HDA 200 high frequency headphones were placed over both ears (Figure 2). This concludes the involvement of any personnel in the testing.

Figure 1.

AMTAS setup.

Figure 2.

Final placement of transducers in AMTAS testing.

After transducer placement, study participants viewed an approximately three-minute animated video with English voiceover narration by a native speaker on hearing test instructions. They could adjust the volume to a comfortable level by using a slider at the bottom (centre) of the screen at any time. During testing, a forced choice method was used, and participants indicated whether they heard a tone or not by touching the corresponding icon on the computer screen. Masking noise was also always presented in the non-test ear for both air and bone conduction masking. Six air conduction frequencies (250 Hz, 500 Hz, 1 kHz, 2 kHz, 4 kHz, 8 kHz) and four bone conduction frequencies (500 Hz, 1 kHz, 2 kHz, 4 kHz) were tested and when two adjacent thresholds differed by 20 dB or more, the inter-octave frequencies were also tested. Results were presented graphically in an audiogram (Figure 3).

Figure 3.

Sample Audiogram generated via AMTAS.

Testing procedures

Outpatients referred for PTA were taken through the clinical protocol of history taking, otoscopy, and tympanometry. Eligible patients may be recruited after this or after they have done manual PTA. Depending on when the participants were recruited, they may either undergo AMTAS or manual PTA first. The timing of recruitment was dependent on the number of patients referred for PTA and availability of the AMTAS machine. No break was given between the two tests unless participants requested for one or if they were found to be tired. This was due to time constraints within the clinic.

AMTAS testing was done in a quiet room located in a corner of the clinic. The ambient noise level was measured three times on a typical weekday afternoon using a Class I sound level meter. The average of those readings was taken to be representative of the ambient noise level in the quiet room.

Statistical analysis

Hearing thresholds obtained via AMTAS and manual PTA were entered into a Microsoft Excel data sheet. Statistical analysis was conducted using the Statistical Package for Social Sciences (SPSS) for Windows, version 16.0 (SPSS Inc., Chicago, IL, USA). Hearing threshold data was classified according to air or bone conduction threshold and frequency. Exploratory data analyses were conducted to determine the use of parametric or non-parametric equivalent. This data was then summarized using the mean and standard deviation. The difference between hearing thresholds obtained via manual PTA and AMTAS was obtained by subtracting the hearing thresholds obtained via AMTAS from those obtained via manual PTA (Manual PTA hearing threshold – AMTAS hearing threshold). We analysed the absolute differences in the hearing thresholds.

Results

Study participants

Of the 100 participants recruited, four were excluded from analyses as they did not complete testing via AMTAS for reasons summarized in Table 1. One participant completed air conduction testing for the left ear only and data collected for this ear was included in the analysis. Another 16 participants were excluded from the analysis due to a reliability rating of “Poor” by the AMTAS software (Table 2).¹⁸ Hence, data from 80 participants was eventually analysed.

Table 1.

Reasons for exclusion of participants.

Reason	Number of participants
Patient was tired	1
Patient did not understand instructions and left in the middle of testing before testing for a single ear could be completed	1
AMTAS software hanged and could not proceed	1
Obtained a quality rating of “FAIL” with few responses recorded during AMTAS	1

Table 2.

Quality indicators considered in reliability rating by AMTAS (Taken from Margolis et al. 2007).

Quality indicator	Description
Masker alert rate	Number of thresholds for which there is over or under making divided by the total number of thresholds
Time per trial	Average time across all observation intervals
False alarm rate	The number of times the tester reported hearing a stimulus when no stimulus was presented divided b the total number of trials without a stimulus
Average test-retest difference	The average difference in threshold measures obtained for stimuli that were tested twice
Quality check fail rate	The total number of occurrences of failure to respond to stimuli presented above threshold divided by the number of measured thresholds
Air-bone gap >50 dB	Number of air-bone gaps exceeding 50 dB
Air bone gap < −10 dB	Number of reverse air-bone gaps i.e., bone conduction threshold worse than air conduction threshold exceeding 10 dB
Average air-bone gap	The average difference between air and bone conduction thresholds

Of the 80 participants (159 ears), there were 52 females (65%), median age of 46 years (Range = 21–74 years) and 28 males (35%), median age of 55 years (Range = 21–78 years). There was no significant difference in the median age between gender (p = 0.377 >0.05). Age of study participants was not normally distributed (p-value = 0.00 <0.05, Shapiro-Wilk test).

Among the 159 ears analysed, the majority had sensorineural hearing loss (56.0%), followed by normal hearing (35.2%) as well as conductive and mixed hearing loss respectively (4.4%). Of the 103 ears with hearing loss, most (62.2%) were of a mild to moderate degree (Table 3).

Table 3.

Distribution of hearing loss severity.

Hearing loss severity	Number (%)
Up to mild (≤40 dB HL)	32 (31.1%)
Up to moderate (≤55 dB HL)	32 (31.1%)
Up to moderately-severe (≤70 dB HL)	25 (24.3%)
Up to severe (≤90 dB HL)	12 (11.7%)
Up to profound (>90 dB HL)	2 (1.9%)
Total	103 (100.0%)

Air conduction thresholds

Air conduction thresholds were tested for six frequencies and data was obtained for between 153 and 159 ears for each frequency. The variation in the number of ears analysed for each frequency was due to the exclusion of “no response” data (no threshold determined at audiometer limits), and because one participant only completed testing for one ear. The absolute mean differences in air conduction thresholds obtained from AMTAS and manual audiometry ranged from 3.30 to 9.62 dB and the standard deviations from 3.37 to 8.65 dB (Table 4).

Table 4.

Absolute mean differences in air conduction thresholds at each frequency.

	n	Absolute mean difference (dB)	Standard deviation (SD)	95% confidence intervals (CI)_
250 Hz	154	7.01	5.62	6.12–7.91
500 Hz	158	5.22	4.60	4.50–5.94
1 kHz	159	3.30	3.37	2.77–3.83
2 kHz	159	9.62	8.65	8.27–10.98
4 kHz	159	6.42	5.10	5.62–7.21
8 kHz	153	6.01	5.63	5.11–6.91

Of these differences, more than 80% were within 10 dB for all frequencies except 2 kHz. For most frequencies, the air conduction thresholds obtained via AMTAS were better than those obtained via manual audiometry except for 500 Hz and 8 kHz (Table 5 and Figure 4).

Table 5.

Differences in air conduction thresholds between manual audiometry and AMTAS (Manual – AMTAS).

Frequency	n	Within ± 10 dB	<0	0	>0	Comment
250 Hz	154	126 (81.8%)	54 (35.1%)	30 (19.5%)	70 (45.5%)	Manual thresholds poorer
500 Hz	158	148 (96.1%)	67 (42.4%)	49 (31.0%)	42 (26.6%)	AMTAS thresholds poorer
1 kHz	159	157 (98.7%)	27 (17.0%)	68 (42.8%)	64 (40.3%)	Manual thresholds poorer
2 kHz	159	126 (79.2%)	53 (33.3%)	34 (21.4%)	72 (45.3%)	Manual thresholds poorer
4 kHz	159	139 (87.4%)	17 (10.7%)	36 (22.6%)	106 (66.6%)	Manual thresholds poorer
8 kHz	153	134 (87.5%)	58 (37.9%)	43 (28.1%)	52 (34.0%)	AMTAS thresholds marginally poorer

Figure 4.

Graph showing mean air conduction thresholds obtained via manual audiometry and AMTAS.

Bone conduction thresholds

Bone conduction thresholds were tested for the corresponding frequencies when air conduction thresholds exceeded 25 dB HL at any of the four frequencies (500 Hz, 1 kHz, 2 kHz, and 4 kHz). Data was obtained for between 65 and 71 ears for each frequency. The absolute mean differences in bone conduction thresholds obtained from AMTAS and manual audiometry ranged from 8.16 to 9.38 dB and the standard deviations from 6.70 to 11.40 dB (Table 6).

Table 6.

Absolute mean differences in Bone conduction thresholds at each frequency.

	n	Absolute mean difference	SD	CI
500 Hz	65	9.38	8.17	7.35–11.40
1 kHz	71	8.16	6.16	6.70–9.62
2 kHz	67	8.58	6.14	7.08–10.08
4 kHz	71	8.38	7.05	6.70–10.05

Of these differences, more than 70% were within 10 dB for all frequencies. For all frequencies, the bone conduction thresholds obtained via AMTAS were better than those obtained via manual audiometry (Table 7 and Figure 5).

Table 7.

Differences in bone conduction thresholds between manual audiometry and AMTAS (Manual – AMTAS).

Frequency	n	Within ± 10 dB	<0	0	>0	Comment
500 Hz	65	46 (70.7%)	15 (23.1%)	12 (18.4%)	38 (58.5%)	Manual thresholds poorer
1 kHz	71	52 (73.2%)	19 (26.8%)	13 (18.3%)	39 (54.9%)	Manual thresholds poorer
2 kHz	67	48 (71.6%)	17 (25.4%)	8 (11.9%)	42 (62.7%)	Manual thresholds poorer
4 kHz	71	54 (76%)	25 (38.5%)	17 (23.9%)	29 (44.6%)	Manual thresholds poorer

Figure 5.

Graph showing mean bone conduction thresholds obtained via manual audiometry and AMTAS.

Changes in hearing profile

The hearing profile of participants obtained via manual audiometry and AMTAS is summarized in Table 8. 75.4% of the 159 ears analysed had a similar hearing profile for both methods. There was a decrease in the number of ears with normal hearing and sensorineural hearing loss with many of these ears (12 out of 13) showing a mixed hearing loss when tested via AMTAS.

Table 8.

Hearing Profile of participants when tested via manual audiometry and AMTAS.

Hearing profile	Manual audiometry	AMTAS	Positive predictive value (PPV)	Negative predictive value (NPV)
Normal	56 (35.2%)	50 (31.4%)	89.3%	94.2%
SNHL	89 (56.0%)	82 (51.6%)	92.1%	90%
Conductive	7 (4.4%)	8 (5.0%)	100%	99.3%
Mixed	7 (4.4%)	19 (11.9%)	100%	92.1%
Total	159	159

Feedback questionnaire

All 100 study participants completed the feedback questionnaire on AMTAS regardless of whether they successfully completed testing via AMTAS. More than 85% of the participants indicated that the machine was easy to use, and instructions were easy to follow by indicating 4–5 in a 5-point rating scale (Table 9). Other notable written and verbal feedback given by participants indicated that the AMTAS testing environment was slightly noisy and that testing was a bit fast (Table 10).

Table 9.

Summarized responses to questions from feedback questionnaire.

Question	Responses (n = 100) on a scale of 1–5
Question	Rating 1 to 3	Rating 4 to 5
1) Was this machine easy to use?	13 (13%)	87 (87%)
Answer on a scale of 1 to 5 with 1 being not easy to use and 5 being very easy to use	13 (13%)	87 (87%)
2) were the instructions easy to understand and follow?	10 (10%)	90 (90%)
Answer on a scale of 1 to 5 with 1 being not easy to understand and 5 being very easy to understand	10 (10%)	90 (90%)

Table 10.

Detailed verbal feedback from participants.

Verbal feedback
Participant 003
Reported that he finds it easier to distinguish the tones when there is masking noise. However, he noted that it might be a bit difficult for some people to distinguish the two tones
Likes to close his eyes to concentrate when doing the test, so sometimes this might lead to him pressing the wrong button indicating the wrong response
Reported that he could still hear the background air-con sound with the headphones
Participant 005
Reported that the instructions were a bit too soft
Participant 006
Reported that bone conductor band was a bit tight and painful
Participant 017
Reported that she initially got confused between the masking noise and tone. Did not note the instructions at the beginning to ignore the masking noise
Participant 023
Finds the room noisy and he could hear other sounds
Participant 034
Participant reported that he got confused towards the end
Participant 046
Reported that test was more difficult to do as compared to conventional PTA. Felt that it took a bit longer as well and she got bored during the testing
Participant 048
Reported that he did not know how long to wait before making response
Participant 049
Reported that he found it hard to concentrate with all the external sounds. Felt that the test was too fast for him
Participant 084
Was confused initially and feels that there should be a more obvious indication for him to listen out for the sounds

Discussion

AMTAS is an automated method of conducting PTA that has been studied in various Western countries but never in a multi-ethnic Asian country like Singapore.^9,11,13,17 Therefore, in this study, we aimed to validate the hearing thresholds of Singapore residents obtained via AMTAS by comparing them with those obtained via conventional PTA and to determine the feasibility of using AMTAS in a clinic within a restructured public hospital in Singapore.

This study showed that hearing thresholds obtained via AMTAS were largely similar to those obtained via conventional PTA. There is however greater variation in BC thresholds obtained via the two methods. In terms of feasibility of use in the clinic, patient feedback regarding the ease of use of the machine is encouraging but issues concerning noise levels during testing and differences in hearing profiles need to be addressed before AMTAS may be implemented.

Validation of hearing thresholds

In this validation study, we looked at absolute mean differences between the hearing thresholds obtained via the two methods instead of mean differences. This contrasts with other studies on automated audiometry which reported both.^6,9,11 This is because both positive and negative differences in hearing thresholds obtained via the two methods are expected. These would cancel each other out in the calculation of mean differences which could artificially reduce the differences obtained.⁹ Hence, in this study, we focused only on absolute mean differences, which resolves this issue.

Air conduction thresholds

The absolute mean differences in Air Conduction (AC) thresholds obtained in this study via the two methods are within 10 dB (3.30–9.62 dB) which is the accepted test-retest variability for pure tone audiometry.¹⁹ These results corroborate those obtained by Eikelbloom’s group which found absolute mean differences in hearing thresholds from AMTAS and manual PTA ranging from 4.9 to 7.0 dB.⁹

The greatest absolute mean difference observed in this study was at 2 kHz and it is noted that it is 71.7% larger than that obtained in the study by Eikelbloom’s group (9.62 dB vs 5.6 dB respectively).⁹ This large absolute mean difference persisted even when outliers were excluded (Absolute mean difference = 8.96 dB). Moreover, the percentage of 2 kHz thresholds obtained via the two methods within 10 dB was also the lowest among the frequencies (79.2%). One proposed reason for the discrepancy observed at 2 kHz is the central masking effect. In AMTAS, masking noise is continuously presented in the non-test ear. This is unlike manual PTA whereby masking is done only when required. The masking noise in AMTAS could have contributed to the central masking effect which has been shown to affect the 2 kHz AC threshold more thereby resulting in a greater difference for the 2 kHz threshold obtained via the two methods.²⁰

The larger absolute difference at 2 kHz observed in this study as compared to the study by Eikelbloom’s group may be due to the different headphones used for manual PTA in the two studies. In this study, circum-aural headphones were used whereas in the study by Eikelbloom’s group, supra-aural headphones were used. It is postulated that the increased bulkiness of the circum-aural headphones used in this study could have affected the headphone placement. This may in turn alter the ear canal acoustics in some participants introducing greater variation at the 2 kHz AC threshold which was found to be more susceptible.²¹

In this study, AC thresholds obtained via AMTAS were generally found to be better than those obtained via manual PTA except at 500 Hz. One reason for this may be the learning effect. As most study participants were tested via manual audiometry first, they may have learned to detect the tones better. In addition, the rate of stimuli presentation in AMTAS is determined by the user’s response time.¹⁷ Thus, if participants responded at a consistent rate, the rate of stimuli presentation may become predictable. This could result in false responses and better hearing threshold results. Although this may be detected by AMTAS via “catch trials” or intervals whereby no tone is presented, the possibility of “misses” cannot be excluded.

Bone conduction thresholds

Overall, the absolute mean differences for Bone Conduction (BC) thresholds are greater than those for AC thresholds. This is also seen in Eikelbloom’s study and in a meta-analysis by Margolis’s group.^9,11 One reason for this could be the variability of bone conduction thresholds obtained via mastoid placement of the bone conductor during manual PTA as a result of acoustic radiation which could result in better than actual BC thresholds.^22,23 Another factor influencing the differences found between BC thresholds obtained via the two methods is the forehead placement of the bone conductor in AMTAS as compared to the ipsilateral mastoid placement of the bone conductor in conventional PTA. Various studies have shown that the two placement methods result in differences in the bone conduction thresholds obtained.^24,25

Nonetheless, despite the variability, the absolute mean differences in BC thresholds obtained in this study is within 10 dB (8.16–9.38 dB). The results obtained here also closely corroborate the results obtained in Eikelbloom’s study which found absolute mean differences in BC ranging from 7.2 to 9.4 dB.⁹ This suggests that the variability in BC thresholds obtained via the two methods are still within acceptable limits.

Feasibility of using AMTAS within the clinic

The main reason for conducting this study was to determine whether AMTAS was suitable for use within a quiet room at a clinic within a public restructured hospital in Singapore. This was premised on a prior study which showed that differences in AMTAS thresholds obtained in a quiet room compared to thresholds obtained manually were within the test-retest variation of 10 dB.⁹

The AMTAS testing room had average ambient sound levels of 17.9 dB SPL at 8 kHz and this sound level increased steadily to 43.5 dB SPL as the frequency decreased. Of these readings, the average sound level at 250 and 500 Hz exceeded the MPANLs for the HDA200 headphones used.²⁶ This suggests that hearing thresholds at 250 and 500 Hz obtained via AMTAS may be poorer than expected given that the noise could mask out the tones. Six participants also gave feedback that the room was noisy. It is however unclear whether the noise they referred to was from the environment or from the masking noise presented continuously in the non-test ear. In addition, in this study a larger proportion of AC thresholds obtained at 500 Hz via AMTAS are poorer than those obtained via manual PTA (42.4%). Despite this, 96.1% of the 500 Hz AC thresholds obtained via the two methods were within 10 dB. Most of the bone conduction thresholds obtained via AMTAS are also better than those obtained via manual PTA. This suggests that the impact of ambient noise on hearing thresholds obtained via AMTAS may be limited.

Another consideration is consistency of hearing profiles when participants were tested via manual audiometry and AMTAS. In this study, 75.4% of the ears had unchanged hearing profiles when tested via AMTAS. However, 24.6% of the ears changed from normal/sensorineural hearing loss to conductive/mixed hearing loss. This could be due to better bone conduction thresholds obtained via AMTAS which widened air-bone gaps. Although this occurred in a minority of ears, the possible significant effects of this such as the wrong investigations being ordered, and the wasted resources means that it must be addressed before AMTAS may be used in the clinic.

In terms of ease of use, most study participants found the machine easy to use and the instructions easy to understand. Some issues identified are that the instructions were confusing and that the testing was a bit fast. Device failure also occurred in one instance. A recent update to the AMTAS software, which allowed the test to be paused and introduced written instructions, may resolve some of these issues. This has implications on future testing protocols as for example, we can continuously monitor the background noise levels and pause the test when noise levels exceed thresholds. The written instructions may be used for certain patients such as those with very poor hearing for whom the audio instructions in the video may be difficult for them to hear and understand. The visual written instructions may thus help clarify any doubts on test instructions if patients miss critical information from the video.

Future studies should look into the demographic details and hearing profiles of patients tested to adjust for the effects of covariates such as age, gender, educational levels, cognition, type, and degree of hearing loss. This may help clinicians better understand the profile of patients better suited for AMTAS testing thereby improving the accuracy of hearing test results obtained via AMTAS. These individuals can then be triaged and separated from the manual audiometry patient pool, to reduce wait time for patients and overall burden on PTA testing in audiometric booths.

Conclusion

This study finds that despite differences in hearing thresholds obtained via AMTAS and manual PTA, these differences are within the acceptable 10 dB test-retest variation. The results obtained are also similar to those obtained in other studies. Nonetheless, important existing issues related to maintaining background noise levels below permissible standards during AMTAS testing, ensuring consistency of hearing profiles obtained via AMTAS and manual audiometry, as well as the ease of use of AMTAS will need to be addressed before it is implemented in the clinic to ensure result accuracy. As AMTAS is a software, the ease of use issues may be resolved in future with software updates. Should AMTAS be implemented, clinics can save on labour cost, space, and financial constraints of needing a sound-treated audiometric booth.

Footnotes

Acknowledgements

First, we would like to thank all the audiologists and audiometric technicians who helped in recruiting study participants. Secondly, we would like to thank Mr Eros Zeus Lomboy for providing his technical expertise regarding the AMTAS software and hardware. Thirdly, we would like to thank Associate Prof. Yuen Heng Wai for his guidance in this study. Finally, we would like to thank Prof. William Hal Martin and Ms Esther Seah as well as the rest of the team from the National University of Singapore for their assistance in measuring the noise levels in this study.

Author contributions

SL conceptualised the study and came up with the study design. SK researched literature and directed the study. KC was involved in editing the manuscript, data analysis, and providing critical inputs. HY was involved in researching literature, protocol development, gaining ethical approval, patient recruitment, data analysis, and in writing the first draft of the manuscript. All authors reviewed and edited the manuscript and approved the final version of the manuscript.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Changi General Hospital Innovation Grant (COCHFINN1704).

Ethical approval

Singhealth Centralized Institutional Review Board (CIRB) (Reference number: 2017/3069).

Informed consent

Written informed consent was obtained from all subjects before the study.

Data availability

The datasets generated and/or analysed during the current study are available from Hazel Yeo Kai Hui.

ORCID iDs

Hazel Yeo Kai Hui

Kenneth Chua Wei De

References

Lee Jek Chong

Danker

Wong

, et al. Hearing loss amongst the elderly in a Southeast Asian population - a community based study. Ann Acad Med Singapore 2017; 46: 145–154.

Epidemiology and Disease Control Division . Ministry of health S. Mumbai, India: National Health Survey, 2010, https://www.moh.gov.sg/content/dam/moh_web/Publications/Reports/2011/NHS2010-low_res.pdf

Ministry of Health and Temasek Foundation Cares . More seniors to benefit from functional screening. News Highlights, https://www.moh.gov.sg/news-highlights/details/more-seniors-to-benefit-from-functional-screening (2018, toegang verkry 10 Julie 2020).

Ministry of Health. Speech by Mr Gan Kim Yong. Minister for health, at the ministry of health committee of supply debate 2019. Speeches, https://www.moh.gov.sg/news-highlights/details/peech-by-mr-gan-kim-yong-minister-for-health-at-the-ministry-of-health-committee-of-supply-debate-2019 (2019, toegang verkry 10 Julie 2020).

Shojaeemend

Ayatollahi

. Automated audiometry: a review of the implementation and evaluation methods. Healthc Inform Res 2018; 24: 263.

Brennan-Jones

Eikelboom

Swanepoel

, et al. Clinical validation of automated audiometry with continuous noise-monitoring in a clinically heterogeneous population outside a sound-treated environment. Int J Audiol 2016; 55: 507–513.

Masalski

Grysiński

Kręcicki

. Hearing tests based on biologically calibrated mobile devices: comparison with pure-tone audiometry. JMIR MHealth UHealth 2018; 6: e10.

Mahomed

Swanepoel

Eikelboom

, et al. Validity of automated threshold audiometry: a systematic review and meta-analysis. Ear Hear 2013; 34: 745–752.

Eikelboom

Swanepoel

Motakef

, et al. Clinical validation of the AMTAS automated audiometer. Int J Audiol 2013; 52: 342–349.

10.

Grason-Stadler . AMTAS Pro-Diagnostic automated audiometry, https://www.grason-stadler.com/solutions/software/gsi-amtas (2019, toegang verkry 15 September 2020).

11.

Margolis

Glasberg

Creeke

, et al. AMTAS®: automated method for testing auditory sensitivity: validation studies. Int J Audiol 2010; 49: 185–194.

12.

Margolis

Moore

BCJ

. AMTAS(®): automated method for testing auditory sensitivity: III. sensorineural hearing loss and air-bone gaps. Int J Audiol 2011; 50: 440–447.

13.

Rao

Altus

Schroeder , et al. Comparison of thresholds obtained with the radioear B-71 and B-81 bone-conduction transducers. Ear Hear 2020; 41: 1775–1778.

14.

Low

. English in Singapore and Malaysia-differences and similarities. In: Kirkpatrick (ed) A (red) the routledge handbook of world englishes. 1st ed. London, UK: Routledge, 2010, p. bll234-bll236.

15.

Baiduc

Poling

Hong

, et al. Clinical measures of auditory function: the cochlea and beyond. Dis Mon 2013; 59: 147–156.

16.

British Society of Audiology . Recommended procedure: pure-tone air-conduction and bone-conduction threshold audiometry with and without masking. Seafield, UK: Bristish Society of Audiology, 2017, pp. 1–36.

17.

Margolis

. Adaptive apparatus and method for testing auditory sensitivity. 6496585, United States, 2019, https://patents.google.com/patent/US6496585B1/en

18.

Margolis

Saly

Chap

, et al. Qualind^TM: a method for assessing the accuracy of automated tests. J Am Acad Audiol 2007; 18: 78–89.

19.

Franks

. Hearing measurement. In: Occupational exposure to noise: evaluation, prevention and control. Geneva: World Health Organization, 2001, pp. bll183–bll231.

20.

Zeng

Liu

Gong

, et al. [An analysis of the results of central masking effect on healthy ears]. Lin Chuang Er Bi Yan Hou Ke Za Zhi 2001; 15: 507–509.

21.

Paquier

Koehl

Jantzem

. Influence of headphone position in pure-tone audiometry. In: Proc Acoust 2012 - Acoust 2012 Jt congr (11ème congrès français d’Acoustique - 2012, 2012, http://hal-univ-brest.archives-ouvertes.fr/view_by_stamp.php?&halsid=sebc3a7ve6gfbteugn6eum4v06&label=UNIV-BREST&langue=fr&action_todo=view&id=hal-00742225&version=1

22.

de Matos

Valle Sde

Dias

, et al. Acoustic radiation effects on bone conduction threshold measurement. Braz J Otorhinolaryngol 2010; 76: 654–658.

23.

Coles

Lutman

Robinson

. The limited accuracy of bone-conduction audiometry: its significance in medicolegal assessments. J Laryngol Otol 1991; 105: 518–521.

24.

Reinfeldt

Stenfelt

Håkansson

. Estimation of bone conduction skull transmission by hearing thresholds and ear-canal sound pressure. Hear Res 2013; 299: 19–28.

25.

Small

Hatton

Stapells

. Effects of bone oscillator coupling method, placement location, and occlusion on bone-conduction auditory steady-state responses in infants. Ear Hear 2007; 28: 83–98.

26.

Margolis

Madsen

. The acoustic test environment for hearing testing. J Am Acad Audiol 2015; 26: 784–791.

A pilot study to validate AMTAS in a specialist outpatient clinic at a public restructured hospital in Singapore

Abstract

Background

Methods

Results

Conclusion

Keywords

Introduction

Methods

Study design

Participants

Manual conventional pure tone audiometry

AMTAS®

Testing procedures

Statistical analysis

Results

Study participants

Air conduction thresholds

Bone conduction thresholds

Changes in hearing profile

Feedback questionnaire

Discussion

Validation of hearing thresholds

Air conduction thresholds

Bone conduction thresholds

Feasibility of using AMTAS within the clinic

Conclusion

Footnotes

Acknowledgements

Author contributions

Declaration of conflicting interests

Funding

Ethical approval

Informed consent

Data availability

ORCID iDs

References

AMTAS^®