Bone anchored hearing systems (BAHS) are an established method for the treatment of conductive and mixed hearing loss [
The audiological benefits of BAHS are well documented and the devices are widely used [
Especially in young children, but also in a growing population of older children, adolescents, and even adults, there is a demand for solutions in which users can benefit from the advantages of a BAHS without having to undergo surgery.
Until recently, there have been mainly two different nonsurgical solutions to use BAHS sound processors: headbands and softbands [
There are two main disadvantages which are associated with the use of these nonimplantable wearing options. One is the sound attenuation by the skin, which increases with high frequencies and reaches, e.g., approximately 15 dB at 3000 Hz [
Recently, a new wearing method for BAHS sound processors without the need for surgery has become available. The Baha SoundArc (Cochlear Inc, Mölnlycke, Sweden) is shown in Figure
Left: dummy head with a medium power sound processor (Cochlear Baha 5) mounted on a Baha SoundArc. Middle: view of the SoundArc and the sound processor separately. Right: sound processor mounted on a softband.
From a clinical point of view, an important question in this context is whether the audiologic performance with the BAHS sound processor is not compromised by choosing the new wearing option. To our knowledge, this has not been investigated to date. With this study, we would like to start to fill this gap.
The primary aim of this investigation is therefore to test the audiological benefit of a current BAHS sound processor worn on a SoundArc in terms of soundfield hearing thresholds, speech understanding in quiet, and speech understanding in noise in persons with a purely conductive hearing loss, as this is the population which is expected to benefit most likely from the new option. The second aim of this study is to compare the audiological benefit with the softband, as this is the most frequently used solution, which is already available today. The third aim is to investigate and to quantify the binaural benefit when using two sound processors on a single SoundArc instead of just one.
This prospective study was approved by the local ethical committee of Bern (KEK-BE 2017-00642). All tests were performed at the University of Bern in accordance with the declaration of Helsinki.
15 young and normal hearing volunteers (aged 18 to 34 years, 6 women, 9 men) participated in this study. Their air conduction (AC) and bone conduction (BC) thresholds were better than 20 dB HL at all audiometer frequencies between 250 Hz and 6000 Hz. A bilateral conductive hearing loss was simulated for the duration of the tests by blocking both ears with a combination of ear plugs (E-A-Rsoft™, 3M, Berkshire, UK) and an additional filling of the remaining volume within the pinna with silicon mould material (Otoform Ak®, Dreve Otoplastik GmbH, Unna, Germany).
The rationale for the choice of normal hearing subjects and a simulated conductive hearing loss rather than real patients is as follows. All currently available nonsurgical wearing options for BAHS sound processors are known to introduce a considerable high frequency skin attenuation at higher frequencies [
A combination of earplugs and silicon mould was used to simulate conductive hearing loss, as either of these blocking methods by itself did not yield a sufficiently high sound attenuation in preliminary tests and circumaural hearing protectors cannot be used together with either the softband or a SoundArc wearing option, because of their size and spatial interference.
After otoscopy and pure tone audiometry, both ears of each subject were blocked as described above. Then all participants underwent a total of 5 series of measurements, one in each of the following 5 conditions: (i) unaided, (ii) aided with one BAHS sound processor mounted on a SoundArc, (iii) aided with 2 BAHS sound processors on a single SoundArc, (iv) aided with one BAHS sound processor on a softband, and (v) aided with 2 BAHS sound processors on a single softband. The order of these 5 conditions was varied systematically between subjects to minimize the effects of learning or fatigue.
The following measurements were performed in each of these conditions: the sound field thresholds were measured using narrow-band noise at 0.25, 0.5, 1, 2, 3, 4, and 6 kHz. The speech reception threshold (SRT) in quiet, i.e., the presentation level required for 50% word understanding, was measured using German two-digit numbers from the Swiss Version of the Freiburger Test [
The SRT in noise, i.e., signal-to-noise ratio required for 50% speech understanding in noise, was measured using the adaptive OLSA test [
Schematic representation of the spatial settings used for the measurements of the speech reception threshold (SRT) in noise. Uncorrelated noise was emitted by 4 loudspeakers (grey speakers) to create an approximated diffuse noise field. Sentences (white speaker) were presented either from the front, from the side ipsilateral to the sound processor, or contralateral to it.
Sound localization was measured using 12 loudspeakers spaced 30° apart in a circle with 1 m diameter and centred around the head of the subject. 36 bursts of white noise with a duration of 200 ms were presented in a randomized order and 3 bursts from every loudspeaker at level of 60, 65, and 70 dB SPL, respectively. Subjects indicated from which of the 12 loudspeakers they thought the sound had been presented. Sound localization was measured in all aided conditions.
All of the above measurements were carried out in a sound-attenuated room (6 x 2 x 4 m3) with an almost frequency independent reverberation time of approximately 0.14s. JBL Professional Control® 1 PRO loudspeaker (JBL Professional, Northridge, California, USA) was used in all tests.
Baha 5 sound processors (Cochlear Inc., Sweden) were used in all aided conditions. They were programmed for each subject and for each of the test conditions separately, following the manufacturer’s recommended fitting procedure for conductive hearing loss, using the most recent fitting software (Cochlear BAHA Fitting Software v4.45, Cochlear Inc.) and BC-direct measurements. The settings “position compensation”, “automatic sound classifier”, and “adaptive microphone directionality” were selected.
Baha SoundArcs and softbands were fitted individually for each subject according to the instructions provided by the manufacturer with the products. The force with which the disc with the attached sound processor was pressed to the head was measured 3 times in each subject and for each mounting method using a spring balance (Pesola type Medio-Line 40003, Schindellegi, Switzerland) and the results were then averaged.
To assess the treatment effects, linear mixed-effects models were implemented for each outcome measure. The treatment condition was included as fixed effect (i.e., unaided, SoundArc unilateral, SoundArc bilateral, softband unilateral, and softband bilateral). For SRTs in noise, the test situations (speech from the front, speech from the side of the first sound processor, and speech from the contralateral side) were additionally considered as fixed effects. To account for multiple measures, the subject IDs were included as random effects. Post hoc comparisons between the tested wearing options (i.e., SoundArc versus softband) and number of devices used (unilateral versus bilateral) were performed with general linear hypothesis testing using two-tailed tests and Holm correction [
Figure
Aided and unaided sound field threshold measured with narrow-band noise. Mean values and standard deviations are shown. See Table
Comparison between the unaided and the different unaided soundfield thresholds. All data shown represents average values over the 4 frequencies 0.5, 1, 2, and 4 kHz.
| ||||||
---|---|---|---|---|---|---|
Softband | SoundArc | |||||
Unilateral | Bilateral | Unilateral | Bilateral | |||
Unaided | vs. | aided | | | | |
(p <.001) | (p <.001) | (p <.001) | (p <.001) | |||
Unilateral | vs. | bilateral | - | 0.3 | - | -0.8 |
(p = 1) | (p = 1) | |||||
Softband | vs. | SoundArc | -1.2 | -1.1 | ||
(p = 1) |
Aided soundfield thresholds are significantly better (p<.001) than unaided by 23.6 to 25.1 dB, when averaged over the same 4 frequencies between 0.5 and 4 kHz, as above. Statistically, results do not differ between any of the 4 wearing options (softband or SoundArc, unilateral or bilateral) significantly, as shown in detail in Table
Figure
Aided and unaided speech recognition scores for monosyllabic words. Mean values and standard deviations are shown. Symbols and error bars are shifted horizontally for a better visual discrimination between the measurements.
Figure
Comparison between the unaided and unaided speech reception thresholds (SRT) in quiet.
| ||||||
---|---|---|---|---|---|---|
Softband | SoundArc | |||||
Unilateral | Bilateral | Unilateral | Bilateral | |||
Unaided | vs. | aided | | | | |
(p <.001) | (p <.001) | (p <.001) | (p <.001) | |||
Unilateral | vs. | bilateral | - | 0.2 | - | -0.8 |
(p = .98) | (p = .98) | |||||
Softband | vs. | SoundArc | - | - | -2.2 | -1.2 |
(p = .26) | (p = .98) |
Speech reception thresholds (SRT) in quiet, i.e., presentation levels required for 50% understanding of 2-digit numbers. Individual results (symbols) and mean values (horizontal lines) are shown. Details of the statistical analyses are given in Table
Figure
Speech reception thresholds (SRT) in noise. More negative values, i.e., closer to the top of the graph, represent better speech recognition in noise. Mean values across all subjects and standard deviations are shown.
If speech is presented from the front (left hand panel in Figure
If speech is presented at the side of the Baha (middle panel in Figure
If a single Baha is used unilaterally and speech is presented from the side of the device (right hand panel in Figure
Figure
Results of the sound localization measurement with 12 loudspeakers spaced 30° apart. Individual results (symbols) and mean values (lines) are shown. A mean error of 90° corresponds to the chance level expected for guessing.
The average force needed to lift the holding disc from the skin of the subjects was 1.69 ±0.21 N for the softband and somewhat higher (1.80 ±0.16 N) for the SoundArc. The difference is not statistically significant (p=.178).
We believe that these results can be adequately summarized into 3 main findings: (1) there is a clear improvement with a Baha mounted on either a SoundArc or a souftband, when compared to the unaided condition, (2) audiologically, no statistically significant difference between the use of a softband and a SoundArc was found, and (3) there is an added audiologic benefit from an additional,
The benefit of the aided versus the unaided situation can be seen in the soundfield threshold measurements (Figure
No statistically significant difference was found between sound processors worn on either a SoundArc or a softband in any of the measurements. This includes sound field thresholds, speech understanding in quiet and in noise, and sound localization. These results seem reasonable, as the same sound processors were used in all tests and the placement as well as the force measured is similar for both mounting methods.
As bone conduction has a much lower sound attenuation across the head than air conduction [
As a side effect, the results of the tests in noise shown in Figure
There are two potential issues with the relevance of the presented work: first, the practical importance of the SoundArc and similar devices as an alternative to the already existing softband and second, the applicability of our results to real patients.
Regarding the first of these potential issues, we believe that a new wearing option can be of substantial importance for current and future patients, if it is perceived as attractive by the users. It is known that a significant number of patients chooses not to use or to discontinue the use of bone conduction devices because either a surgical intervention is needed or the nonsurgical wearing option is not aesthetically appealing [
All of our tests were performed with normal hearing subjects with bilaterally blocked ears rather than with patients with bilateral conductive hearing loss and with medium-output-power sound processors rather than with power or super-power sound processors.
A number of advantages, but also some limits are associated with this choice. Using young, adult, normal hearing subjects allowed us to study the impact of the devices in subjects with bilaterally normal inner ears and a pure conductive hearing loss. We believe that, in this way, the hearing of children with a purely conductive hearing loss and normal inner ears is adequately simulated. Unlike children, these young adults can easily perform all the rather taxing and time consuming tests required for this study. We expect children and adolescents with a purely conductive hearing loss to become one of the important group of potential users of the SoundArc, as this is an important group for BAHS already today at our centre and at other centres worldwide [
As subjects with an additional cochlear hearing loss were not included, we cannot directly draw conclusions from our results on the efficacy of the SoudArc or the softband wearing options for BAHS sound processors in mixed hearing loss or single sided deafness. Nevertheless, we feel that the most important target group for these wearing options is reasonably represented. Future studies may shed more light on the benefit in these groups of users.
As a by-product of this study, we have presented a method which can reliably, safely, and reversibly simulate a considerable conductive hearing loss in the order of magnitude of 45 dB without having to resort to circumaural hearing protectors, which preclude the use of several types of bone conduction hearing devices.
The results of our study suggest that in subjects with a bilateral conductive hearing loss and a medium power BAHS sound processor mounted on a SoundArc device can benefit from significant improvements in terms of sound field hearing thresholds, speech understanding in quiet, and speech understanding in noise, when compared to the unaided condition. We have not found any significant difference in the audiological performance when compared to the mounting on a softband. The use of an additional
The data used to support the findings of this study are available from the corresponding author upon request.
The work of Tom Gawliczek was supported by a research grant from Cochlear Bone Anchored Hearing Solutions, Mölnlycke, Sweden, and by Cochlear Europe.
We would like to thank Cochlear Inc. for their support, Gianni Pauciello from the Department of ENT, Head and Neck Surgery, Inselspital, for the photographs, and Dr. Lukas Bütikofer, senior statistician at the CTU of the University of Bern for his invaluable help with the statistical analysis.