Comparing the Results of Sound Field Audiometry With Traditional Headphone-Based Audiometry: Implications for Assessing Hearing Aid Benefits Using Functional Gain
Article information
Abstract
Background and Objectives
This study aimed to compare sound field audiometry (SFA) and headphone-based audiometry (HPA) during the hearing aid fitting process. Factors associated with the differences between the two methods were also investigated.
Subjects and Methods
We retrospectively reviewed 42 patients who underwent both unaided SFA and HPA on the same day.
Results
Hearing thresholds obtained in SFA were lower than those measured with HPA at 250 Hz (p=0.016). No significant differences were observed at other frequencies. When comparing the clinical characteristics of the groups with lower and higher thresholds of SFA at 250 Hz, patients with mixed hearing loss (MHL) were more prevalent in the group with lower hearing thresholds in SFA, whereas those with sensorineural hearing loss (SNHL) were more common in the group with equal or higher thresholds in SFA (p=0.011). Moreover, in subjects with SNHL, no significant difference was observed between the average of the thresholds at the four frequencies (4-FA) of SFA and that of HPA (p=0.156), while a significant difference was observed in subjects with MHL (p=0.028). When compared across the frequency, the thresholds of SFA at 250 Hz and 1,000 Hz were significantly lower than those of HPA (p=0.004 and p=0.009, respectively) in MHL.
Conclusions
Improvement in aided hearing in SFA compared with unaided HPA must be interpreted with caution at lower frequencies in subjects with MHL. Meanwhile, when validating the efficacy in HA users with SNHL, the process can be simplified using unaided HPA instead of SFA; however, unaided SFA should be employed in those with MHL.
Introduction
Hearing loss is a major public health issue, increasing as the elderly population rapidly grows. Approximately one-third of people over the age of 65 experience hearing loss [1]. In the general population, the prevalence of bilateral hearing loss, defined as average pure tone thresholds greater than 25 dB at 500, 1,000, 2,000, and 4,000 Hz, was 5.9%, and that of unilateral hearing loss was 8.2% according to the data from the Korean National Health and Nutrition Examination Survey [2]. Hearing aids (HAs) are a primary rehabilitation tool, and their effectiveness for health-related quality of life has been proven in a systematic review [3]. Use of HAs can lead to various benefits including reductions in psychosocial handicaps associated with hearing loss. In addition, HAs can reduce depression and improve self-confidence and interpersonal relationships [4,5]. In the fitting process of HAs, verification and validation of benefit are an essential part to provide the highest level of care.
There are two primary methods for HA evaluation to document benefit after fitting: focusing on subjective outcomes and focusing on objective outcomes. In an effort to quantify patients’ subjective perceptions of their HA benefits, self-assessment inventories have been developed. Among these inventories, the Hearing Handicap Inventory for the Elderly (HHIE) [6] has been used in numerous studies to demonstrate and quantify the beneficial effects of HA fitting on the emotional and social aspects [7,8]. We have previously demonstrated a significant association of aided word recognition scores (WRS) with subjective outcomes of HA use, assessed with questionnaires. A significant relationship between aided WRS and HHIE was identified not only at short-term follow-up after fitting, but also at long-term follow-up [9].
Objective measurements of HA outcomes can be obtained by real-ear measurement (REM) or functional gain. REM was developed to verify that the prescribed real-ear acoustic response of the HAs met the prescribed target. Although REM and aided speech-in-noise testing provide the most reliable information [10,11], many HA providers are not conducting these tests routinely due to lack of time or constraints of equipment and space [12]. Therefore, commonly performed tests such as aided audiometric threshold and aided WRS can be considered useful evaluation tools for assessing HA outcomes in audiological clinics. Meanwhile, functional gain utilizes frequency-specific signals that can be calculated by subtracting the aided from unaided thresholds in a sound field. The functional gain can detect the degree of audibility of soft sounds provided by the HAs and whether the prescriptive target meets the prescribed value for soft input sounds [13]. The advantage of the functional gain test is that it does not require additional equipment like REM and can be verified using a simple protocol with familiar hearing test equipment.
When a person with hearing loss presents to an audiological clinic, the initial step typically involves audiometry using headphones. If the patient needs HA based on the audiometric test results, counselling and fitting procedures are initiated. To obtain functional gain, it is ideal to perform unaided and aided audiometry in the same sound field. However, this can be a hassle in a busy clinic, and it would be more efficient if aided audiometry could be compared to initial unaided audiometry using headphones on the same day or close to that time point.
A limited number of studies have assessed the difference between thresholds in sound field audiometry (SFA) and headphone-based audiometry (HPA), and detailed features of the difference between the two tests. If the difference is negligible, it may suggest the usefulness of unaided HPA in assessing the functional gain of HA. The main objective of this study was to compare SFA and HPA, which are used indistinctly to evaluate the effectiveness of HAs in the fitting process. We also investigated possible factors associated with the difference between the two tests.
Subjects and Methods
This study was conducted on HA users who visited an audiological clinic from December 2017 to November 2018 for HA adoption and fitting. The inclusion criteria were patients who underwent unaided SFA and unaided HPA on the same day. Patients who use bilateral HAs were excluded from this study.
Forty-two HA users with a mean age of 68.21±20.80 years were enrolled. Fifty percent of the subjects were men, and 42.85% of them used HA on the right side. Twelve patients (28.57%) had complaints of tinnitus. On otoendoscopic examination, 30 patients (71.43%) exhibited normal otoscopic findings, 11 patients (26.19%) showed retraction of the tympanic membrane (TM), and one patient (2.38%) had a TM perforation.
The hearing threshold was based on the average of the responses in dB assessed at the frequencies of 500, 1,000, 2,000, and 4,000 Hz. Thirty-five patients (83.33%) had moderate hearing loss, five (11.90%) had severe hearing loss, and two (4.76%) had profound hearing loss according to the International Bureau of Audiophonology definitions [14].
Regarding the type of hearing loss, 30 (71.42%) patients had sensorineural hearing loss (SNHL), while 12 (28.57%) had mixed hearing loss (MHL) with an air-bone gap larger than 10 dB.
HPA was performed at frequencies of 250, 500, 1,000, 2,000, 3,000, 4,000, and 8,000 Hz using GSI Audio Star Pro (Grason-Stadler) and Telephonics TDH-39P headphones (Telephonics). The contralateral ear was masked by narrow-band noise. SFA was performed using a loudspeaker placed at an angle of 45° azimuth and a 1-m distance from the patient’s test ear. The contralateral ear was masked by narrow-band noise using headphones. The sequence of the two tests was randomized, and the subsequent test commenced immediately after the completion of the preceding one.
The study protocol conformed to the guidelines of the Declaration of Helsinki and Korean Good Clinical Practice. This study was reviewed and approved by the Institutional Review Board of the Samsung Medical Center (IRB No. 2023-11-042-001).
Statistical analysis
Data analysis was undertaken using IBM SPSS Statistics for Windows (Version 27.0; IBM Corp.). A paired t-test was used to compare the unaided thresholds of HPA and SFA at each frequency. The correlation between the two tests was investigated using Pearson correlation analysis. The association of each potentially confounding variable with the difference between the two tests was assessed by univariate analysis. All primary analyses were performed at a 0.05 significance level.
Results
The mean and standard deviations of SFA and HPA thresholds for each of the five frequencies and the average of the four frequencies, including 500, 1,000, 2,000, and 4,000 Hz (4-FA), were analyzed (Fig. 1). Hearing thresholds obtained in SFA were lower than the thresholds measured with HPA at 250 Hz (p=0.016), whereas no significant difference was observed at other frequencies (p=0.285 at 500 Hz, p=0.279 at 1,000 Hz, p=0.543 at 2,000 Hz, p=0.264 at 4,000 Hz, and p= 0.125 at 4-FA).
The average hearing threshold values in sound field audiometry and audiometry using headphones at 0.25, 0.5, 1, 2, 4 kHz, and 4-frequency average (4-FA). *p<0.05.
To investigate the differences at 250 Hz, we compared the clinical characteristics of groups with lower or higher thresholds of SFA at 250 Hz compared to HPA (Table 1). Lower thresholds of SFA at 250 Hz were observed in 22 patients, while equal or higher thresholds were observed in 20 patients. There was no significant difference between the two groups regarding sex, age, side of affected ear, duration from diagnosis to HA fitting, TM findings, presence of tinnitus, causes of hearing loss, severity of hearing loss, and pre-fitting HHIE scores. However, subjects with MHL were more common in the lower threshold group in SFA, whereas those with SNHL were more frequent in the equal or higher hearing thresholds group (p=0.011).
Clinical characteristics of the group with hearing thresholds at 250 Hz in sound field audiometry lower than audiometry using headphones and the group with equal or higher threshold
In Table 2, we compared the clinical characteristics between the SNHL and MHL groups. No statistically significant differences were observed between the two groups regarding sex, age, side of affected ear, duration from diagnosis to HA fitting, presence of tinnitus, and pre-fitting HHIE scores. However, there was a significant difference in severity of hearing loss, indicating the hearing threshold of MHL group was higher (p= 0.001). Abnormal tympanic membrane findings were more frequent in the MHL group, and the causes of hearing loss also differed significantly between the SNHL and MHL groups (p<0.001).
Comparison of clinical characteristics between the sensorineural hearing loss group and the mixed hearing loss group
We investigated the thresholds at each frequency and 4-FA of SFA and HPA according to the type of hearing loss (Table 3 and Fig. 2). In SNHL group, the 4-FA of SFA was 57.25±9.56, and that of HPA was 56.04±7.64, with no significant difference between them (p=0.156). However, in MHL group, a significant difference was observed between the 4-FA of SFA and HPA (68.12±19.30 and 74.27±14.31, respectively; p=0.028). When SFA and HPA thresholds were further compared by frequency in patients with MHL, the thresholds of SFA at 250 Hz and 1,000 Hz were significantly lower than that of HPA (p=0.004 and p=0.009, respectively). Meanwhile, in patient with SNHL, the thresholds of SFA and HPA showed no difference at all frequencies. In addition, we compared the results of the SFA-HPA between the SNHL and MHL groups. Significant differences were observed at low frequencies (250, 500, and 1,000 Hz, p=0.005, p=0.023, and p=0.005, respectively), whereas no significant differences were found at higher frequencies (2,000 Hz and 4,000 Hz) (Table 3).
Hearing thresholds (dB HL) of sound field audiometry and audiometry using headphones according to the frequency
Hearing thresholds of SNHL group and MHL group at 0.25, 0.5, 1, 2, 4 kHz, and 4-FA average. SFA, sound field audiometry; HPA, headphone-based audiometry; SNHL, sensorineural hearing loss; MHL, mixed hearing loss; 4-FA, 4-frequency average.
Discussion
Functional gain is an important tool for verifying HA benefit and is often used along with REM. It is obtained by difference between unaided and aided audiometry at each frequency in the sound field. During this process, unaided SFA and HPA are sometimes used interchangeably in the clinics. In this study, we investigated the clinical usefulness of unaided SFA and HPA in the evaluation of functional gain and the factors associated with differences between the two tests.
Our study shows that hearing thresholds obtained in SFA differed from HPA only at 250 Hz, and SFA and HPA thresholds did not differ at 500, 1,000, 2,000, 4,000 Hz, and in 4-FA. Similarly, in a previous study, the mean sound field threshold was approximately 7.5 dB lower than that measured using earphones, which was attributed to the difference in head diffraction and closed-ear effects [15].
Among patients who demonstrated better hearing in SFA at 250 Hz compared to HPA, the proportion of those with MHL was greater than that of those with SNHL. Notably, patients with MHL hear approximately 6 dB better in the 4-FA hearing thresholds in the sound field audiometry than when wearing headphones, likely due to the contribution of bone conduction. Bone conduction is the way sound energy is transmitted by the skull bones to the cochlea [16]. At lower frequencies and higher stimulation levels, bone conduction sound induces vibrotactile excitation, leading to a multimodal perception [17]. In SFA, this vibrotactile stimulation can be received from multiple angles. On the contrary, in HPA, sound is delivered to a localized area in the ear canal, limiting the potential influence of bone conduction. These findings suggest that patients with MHL may benefit from enhanced hearing in the sound field due to the additional contribution of bone conduction at lower frequencies.
Since there were differences in low frequencies (250 and 1,000 Hz) in the MHL group and no differences in the SNHL group, it can be inferred that the difference at 250 Hz in all participants was influenced by lower threshold of the MHL group. Therefore, when we test functional gain during HA fitting process, it is permissible to use both unaided SFA and HPA in subjects with SNHL. However, in subjects with MHL, as there were significant differences at low frequencies, unaided SFA, instead of aided HPA, should be conducted to obtain an accurate functional gain.
According to the previous study, ambient noise levels in non-sound-treated environments exhibited significantly greater fluctuations than those in sound-treated rooms, with most standard violations occurring in the low to mid-low frequency ranges [18]. In our study, significant discrepancies between SFA and HPA outcomes were observed particularly at low frequencies, underscoring the importance of controlling ambient noise in this frequency range to ensure accurate assessment. Thus, the use of a well–sound-treated anechoic booth and strict adherence to a standardized test environment that takes into account the room acoustics for SFA are crucial for obtaining accurate and consistent test results.
To the best of our knowledge, this is the first study to analyze the difference between SFA and HPA with respect to the process of functional gain assessment and to demonstrate differences across types of hearing loss. However, this study has several limitations. First, the cohort of the study only included moderate, severe, or profound hearing loss and did not include patients with mild hearing loss, because subjects with mild hearing loss usually do not use hearing aids. Therefore, it is uncertain whether the findings of our study are generally applicable regardless of the severity of hearing loss. Second, the small sample size of patients with MHL may have contributed to the observed differences in hearing thresholds at 250 Hz between SFA and HPA.
Functional gain assessments using SFA are important tools for hearing aid verification. In subjects with MHL, hearing thresholds obtained in the sound field were significantly lower than those measured with headphones at lower frequencies. Therefore, improvements in these frequencies using HA compared to unaided HPA should be interpreted with caution, and unaided SFA should be employed instead of HPA to obtain an accurate functional gain in subjects with MHL. Meanwhile, in most HA users who have SNHL, the verification process can be simplified by comparing unaided HPA and aided SFA.
Notes
Conflicts of Interest
The authors have no financial conflicts of interest.
Author Contributions
Conceptualization: Yang-Sun Cho. Data curation: Sung-Min Park, Yang-Sun Cho. Formal analysis: Sung-Min Park. Funding acquisition: Yang-Sun Cho. Investigation: all authors. Methodology: Yang-Sun Cho. Project administration: Yang-Sun Cho. Resources: Miso Lee. Software: Sung-Min Park. Supervision: Yang-Sun Cho. Validation: Sung-Min Park, Yang-Sun Cho. Visualization: Miso Lee, Sung-Min Park. Writing—original draft: Sung-Min Park. Writing—review & editing: Yang-Sun Cho. Approval of final manuscript: all authors.
Funding Statement
None
Acknowledgments
None