Keywords
aural atresia - bone conduction hearing aids - conductive hearing loss
Introduction
Atresia of the external auditory canal affects 1 in every 10 thousand to 20 thousand
live births. It is mostly present unilaterally (only 30% of the patients are affected
bilaterally), in the right ear of male patients,[1]
[2] and it can be associated with microtia.[3] Reports from Latin America indicate that it is more prevalent there, affecting 5[4] to 21 out of every 10 thousands newborns.[5] A large percentage of these cases occur for unknown reasons, while some types are
genetic and associated with craniofacial disorders. Most of the patients present with
an air conduction pure-tone average (PTA) of 60dB to 70 dB, with a bone conduction
(BC) PTA in the normal range.[6]
The treatment of patients with atresia and microtia involves esthetic and functional
aspects. Reconstructive esthetic surgery with autologous cartilage has had successful
results.[7] In the functional approach to hearing loss, BC hearing aids (spectacle frames as
well as rigid and softband systems) can be initially used. The disadvantages of these
prostheses include limited functional gain, visibility, cosmetic unattractiveness,
and pain due to pressure on the skin.[3]
The adhesive BC device (aBCD) called ADHEAR (MED-EL, Innsbruck, Austria), released
in 2017,[8] is another nonsurgical option.[9] The system consists of an adhesive adapter that attaches to the hairless skin behind
the pinna in the mastoid area and is connected to an audio processor (AP). The adhesive
adapter can only be used on healthy skin, is water-resistant, and can stay on the
skin for several (3 to 7) days.[10] The AP receives the sound waves and turns them into vibrations. Clinical studies
have shown that the audiological performance of the ADHEAR in quiet and noise is comparable
to that of traditional BC hearing aids. However, these studies found the adhesive
device to have several advantages, namely, superiority in wearing comfort, wearing
time, and subjective satisfaction.[11]
[12]
[13]
Bone conduction implants (BCIs) are another alternative to BC hearing aids.[14] The active transcutaneous BCI BONEBRIDGE (MED-EL) is one of the systems available.
It consists of an implantable coil and transducer that convert the delivered signals
into vibrations that are subsequently transmitted to the inner ear via the skull.
Transcutaneous direct stimulation of the bone minimizes the risk of skin irritation
and achieves good sound transmission.[15] As the BCI lies completely under the skin, it is not visible, and the complication
rates are very low.[16]
[17] The BONEBRIDGE device has been approved for sale in the European market in 2012
for use in adults and, in 2014, for children over 5 years of age.[16]
The present research is of particular interest for Latin American countries, which
present the highest prevalence of outer ear malformations, greater than the average
reported worldwide.[18]
[19] The high costs of implants are an access barrier for the most disadvantaged segments
of the population.
Therefore, the aim of the present prospective and multicentric study was to evaluate
the audiological benefits and subjective satisfaction with a BCI hearing system in
patients with conductive hearing impairment over a period of one year, and to compare
these results with the benefits obtained with an aBCD in the same group of users.
Materials and Methods
Subjects – The study cohort comprised 10 children < 18 years of age who underwent BONEBRIDGE
BCI602 implantation. The average age was of 10 (range: 5 to 14) years. All subjects
had moderate conductive hearing loss with BC thresholds ≤ 25 dB HL and an average
air-bone gap (ABG; 4-frequency PTA, 4PTA) > 20 dB. The patients were diagnosed with
congenital microtia (6 unilateral and 4 bilateral) associated with atresia of the
external auditory canal ([Table 1]).
Table 1
Data analyzed in the sample of the present study
|
Subject number
|
Age at implantation (in years)
|
Sex
|
Side of the hearing loss
|
Tested ear
|
Ipsilateral AC 4PTA (in dB HL)
|
Ipsilateral BC 4PTA (in dB HL)
|
|
1
|
10.6
|
Female
|
Bilateral
|
Left
|
64
|
6
|
|
2
|
11.9
|
Male
|
Unilateral
|
Right
|
94
|
23
|
|
3
|
14.4
|
Male
|
Unilateral
|
Right
|
78
|
0
|
|
4
|
9.0
|
Male
|
Bilateral
|
Right
|
69
|
14
|
|
5
|
8.4
|
Female
|
Bilateral
|
Right
|
71
|
9
|
|
6
|
12.6
|
Female
|
Unilateral
|
Left
|
71
|
19
|
|
7
|
8.9
|
Male
|
Unilateral
|
Right
|
63
|
3
|
|
8
|
12.3
|
Female
|
Bilateral
|
Right
|
64
|
15
|
|
9
|
10.9
|
Male
|
Unilateral
|
Right
|
60
|
6
|
|
10
|
5.8
|
male
|
Unilateral
|
Left
|
70
|
3
|
Abbreviations: 4PTA, four-frequency pure-tone average; AC, air conduction; BC, bone conduction.
Note: The 4PTA was calculated from the frequencies of 0.5 kHz, 1 kHz, 2 kHz, and 4 kHz;
the results presented are the means of the 4 frequencies for each subject.
Procedure – The present study was approved by the Ethics Committees of both clinics (Comités
Institucionales de Ética en Investigación, CIEIs, no. 291/2020) and was performed
according to the Declaration of Helsinki. Informed consent was obtained from the patients
prior to study inclusion.
On the day of the activation of the implant system, BCI users were enrolled in the
study, and tests were performed in the unaided condition. The subjects were then tested
at 1, 6, and 12 months after device activation in the BCI-aided condition. After completion
of this stage of the protocol, the same users were asked to stop using the BCI audio
processor and instead use the aBCD for 4 weeks. At the end of this period, the measurements
were repeated in the aided condition with the aBCD.
Audiological tests – The audiological assessment consisted of basic audiometric tests and sound field
measurements in the unaided and aided conditions with the BCI and aBCD hearing systems,
in an audiometric sound-attenuated room. Calibrated loudspeakers were set up at a
distance of 1 m from the center of the subject's head, at ear level. For the audiological
tests, the aBCD was used in program one (automatic) and with the volume at the preferred
level of the patients. The SAMBA audio processor (MED-EL) was tested with the personalized
fitting of the patient in the universal program. Both devices were operating with
automatic beamformer, directional microphones focusing to the front in the S0 and
S0N0 test setup. All devices were supplied with a new battery prior to testing. For
all sound field measurements, the contralateral ear was plugged with a foam earplug
and covered with an earmuff.
The auditory tests (thresholds and warble tone stimuli) were performed in a soundproof
booth using a SENTIERO ADVANCED (Path Medical, Germering, Germany) audiometer in 1
center and an AC40 (Interacoustics, Middelfart, Denmark) in the other. The 4PTA was
calculated from the frequencies of 0.5 kHz, 1 kHz, 2 kHz, and 4 kHz.
The word recognition score (WRS) was measured in the sound field in quiet with the
speaker at 0° azimuth (S0). The percentage of words correctly recognized by the patient
was assessed. Each list comprised 25 phonetically-balanced disyllabic words.[20]
[21]
To measure speech intelligibility in noise, the speech signal (65 dB SPL) as well
as the noise signal (60 dB SPL or 65 dB SPL) were provided from the front (S0N0).
Questionnaires – Subjective satisfaction was assessed by means of the hearing-specific Parents' Evaluation
of Aural/Oral Performance of Children (PEACH) rating scale.[22] Satisfaction with the device itself was evaluated using the Audio Processor Satisfaction
Questionnaire (APSQ) and a BCI/aBCD comparison questionnaire. The PEACH questionnaire, developed by Ching and Hill,[23] comprises 13 questions about the child's behavior in everyday life in relation to
a range of hearing and communication scenarios. There are 5 possible answers, ranging
from never (0%) to always (75% to 100%). The APSQ questionnaire[24] consists of 21 items that refer to wearing comfort, social life, usability, and
device conveniences. The responses are on a Likert scale, with 5 options ranging from
never (0%) to always (100%). The custom-made BCI/aBCD comparison questionnaire has 13 questions about the preferences of the user regarding
topics like device use, wearing comfort, and sound quality. All questionnaires were
thoroughly explained by the study personnel and filled out by a proxy. The proxy for
a particular subject was always the same person (the child's mother, for example)
throughout the study.
Statistics – The statistical analysis was performed using GraphPad Prism (GraphPad Software, San
Diego, CA, United States) software, version 7.04. The Shapiro-Wilk test was used to
test for normal distribution. The Wilcoxon signed-rank test with Bonferroni correction
was applied to compare results between conditions on the following tests: 4PTA sound
field thresholds, speech in quiet, speech in noise with 7 comparisons per test, resulting
in a corrected p-value of 0.0071, and wearing time results of the APSQ questionnaire with 6 comparisons,
resulting in a corrected p-value of 0.0083. The remaining results of the APSQ and PEACH questionnaires were
analyzed by two-way repeated measures analysis of variance (ANOVA) with the Bonferroni
multiple comparison test (APSQ: F
3,135 = 5.37, p = 0.0016; PEACH: F
1,27 = 55.15; p < 0.0001).
Results
Hearing thresholds – The mean hearing threshold for the frequencies of 0.5 kHz, 1 kHz, 2 kHz, and 4 kHz
(4PTA) in the unaided condition was of 65 ± 4.3 dB, which improved significantly,
to 23 ± 8.1 dB, after using the BCI for 1 month (p = 0.0020). Compared with the unaided results, the performance of the subjects further
improved significantly after using the BCI for 6 months, with a mean 4PTA of 22 ± 8.7 dB
(p = 0.0020), and 12 months, with a mean 4PTA of 20 ± 7.0 dB (p = 0.0020). Using the aBCD, a mean 4PTA of 33 ± 5.3 dB was measured, which was significantly
higher compared with the mean 4PTA hearing threshold after using the BCI for 12 months
(p = 0.0020; [Fig. 1]).
Fig. 1 PTA4 sound field thresholds, an average of the frequencies 0.5, 1, 2, and 4 kHz,
in dB HL. Bone conduction implant (BCI) at 1, 6 and 12 months after device activation
(T1 – T3). Adhesive bone conduction device (aBCD). Min to max (whiskers), mean (cross)
and median (line).
Speech recognition in quiet – The speech recognition in quiet in the unaided condition presented a mean WRS of
33 ± 11%. After using the BCI or aBCD, the speech recognition improved significantly
compared with the unaided condition (all; p = 0.0020). After using the BCI for 1, 6, and 12 months, the mean WRS values were
of 97 ± 4.8%, 99 ± 2.5%, and 99 ± 2.5% respectively. With the aBCD, a mean WRS of
91 ± 7.4% could be achieved. No significant difference was found between the BCI and
aBCD results ([Fig. 2]).
Fig. 2 Speech in quiet. Word recognition score (WRS) in %. Bone conduction implant (BCI)
1, 6 and 12 months after device activation (T1 – T3). Adhesive bone conduction device
(aBCD). Min to max (whiskers), mean (cross) and median (line).
Speech recognition in noise – When speech recognition in noise was measured at a signal-to-noise ratio (SNR)
of +5 dB ([Fig. 3A]), the average WRS in the unaided situation was of 24 ± 11%. Compared with the unaided
condition, the speech in noise results improved significantly after using the BCI
for 1, 6, and 12 months (all; p = 0.0020), to mean WRS values of 87 ± 9.1%, 89 ± 5.5%, and 93 ± 5.9% respectively.
The mean WRS using the aBCD was of 81 ± 8.3%, which was also a significant improvement
compared with the unaided condition (p = 0.0020). The speech in noise result at +5 dB SNR after 12 months (p = 0.0039) using the BCI was significantly better compared with the result with the
aBCD. The result after 6 months using the BCI was close to statistical significance
when compared with the result with the aBCD (p = 0.0078).
Fig. 3 Speech in noise. Word recognition score (WRS) in % at a signal to noise ratio (SNR)
of A) 5 dB SNR and B) 0 dB SNR. Bone conduction implant (BCI) at 1, 6 and 12 months
after device activation (T1 – T3). Adhesive bone conduction device (aBCD). Min to
max (whiskers), mean (cross) and median (line).
At 0 dB SNR, the unaided speech in noise result presented a mean WRS of 17 ± 13%,
which improved significantly with the BCI after 1, 6, and 12 months of use to mean
scores of 78 ± 16%, 80 ± 13%, and 85 ± 10% respectively. With the aBCD, a mean WRS
of 75 ± 10% was measured, which was also statistically different compared with the
unaided condition (all; p = 0.0020). The speech in noise results using the BCI at 0 dB SNR were not statistically
different from the results with the aBCD at any time point (1 month: p = 0.4063; 6 months: p = 0.2871; 12 months: p = 0.0195).
Subjective satisfaction and adherence to use – The three questionnaires used in the present study were filled out for all ten children.
The hearing-specific PEACH questionnaire was applied to evaluate the performance of
the users in relation to a range of hearing and communication scenarios. Significant
differences were observed between the BCI and aBCD. The mean overall score was of
91 ± 12 points for the BCI, and of 78 ± 13 points for the aBCD (p = 0.0002). The mean score on the quiet dimension was of 88 ± 11 points for the BCI, and of 78 ± 11 points for the aBCD (p = 0.0013). The mean score on the noise dimension was of 79 ± 12 points for the BCI, and of 68 ± 13 points for the aBCD (p = 0.0011). In each case, the differences favored the use of the BCI over the aBCD.
The mean daily wearing time was of 11 ± 3.0 hours per day for the BCI, and of 9 ± 2.5 hours
per day for the aBCD ([Fig. 4]).
Fig. 4 Parents' Evaluation of Aural/Oral Performance of Children (PEACH) questionnaire results,
comparing the bone conduction implant (BCI) and adhesive bone conduction device (aBCD).
Questions regarding communication scenarios in quiet and noise were evaluated. Standard
deviation (whiskers).
Regarding the audio processor-specific APSQ questionnaire, no statistical difference
was found in any of the domains when comparing the BCI with the aBCD. The following
topics were covered by the wearing comfort domain: Comfort when wearing the AP, use of a phone at the processor side, physically
active lifestyle with the AP, wearing glasses or head gear (cap, hat, or helmet) and
the AP at the same time, and general satisfaction. In the wearing comfort domain, the BCI users reported a mean satisfaction rate of 79 ± 17%, and the aBCD
users, 67 ± 23%. The social life domain consists of items regarding AP-related improvement of confidence, independence,
group communication, and ease/enjoyment of social or cultural activities with the
help of the device. The mean score on the social life domain was of 92 ± 14% for BCI users, and of 77 ± 19% for aBCD users. In the usability domain, the mean score of the BCI users was of 93 ± 7% and that of the aBCD users
was of 80 ± 16%. The usability domain evaluated the following topics: AP positioning, sound location, exchanging
batteries, turning the AP on and off, and maintenance. The device conveniences domain analyzed skin health, sweating or pressure at the AP position, and the AP
falling off or malfunctioning. Using the BCI, the users reported a mean satisfaction
rate of 79 ± 15%, and the aBCD users, 82 ± 16%.
Regarding the BCI/aBCD comparison questionnaire, favorable results for the BCI were observed in most
items (Q1, 2, 3, 4, 6, 7, 9, 10, and 13). Greater dispersion in the responses was
found for items Q8 (“What device was better to use during sports?”), Q11 (“With what
device do you hear less feedback/whistling?”), and Q12 (“With what device was it more
comfortable to wear headwear (such as hat, helmet) and the processor at the same time?”)
([Fig. 5]).
Fig. 5 BCI/aBCD comparison questionnaire. Results of thirteen questions regarding the subject's
preference, comparing the bone conduction implant BONEBRIDGE (BB) to the adhesive
bone conduction device ADHEAR.
Discussion
The present study assessed audiological performance and subjective satisfaction in
a cohort of children with moderate to severe conductive hearing loss provided with
an active BCI system and compared these outcomes with those of the use if a nonsurgical
aBCD in the same subjects. The study aimed to answer the question of whether or not
surgical treatment is necessary by comparing treatment with a nonsurgical approach
in the same patient group.
Both devices provided substantial, clinically relevant hearing rehabilitation for
the patients, which is consistent with previously published outcomes.[25]
[26]
The present study had the added value of being able to compare the hearing performance
with both devices in the same group of users, reducing the possibility of bias. The
audiological results between the BCI and aBCD during the first month after implant
activation are comparable, although with a trend toward better performance with the
BCI. After 1 year of implant use, superior results for the BCI versus the aBCD could
be found in sound field thresholds (mean 4PTA: 20 dB versus 33 dB respectively; p = 0.0020) and in speech in noise at the SNR of +5 dB, (93 versus 81% respectively;
p = 0.0039). A study[27] in which the audiological performance of ADHEAR was tested after 1, 6, and 12 months
of use showed no improvement in the outcomes over time. Even comparing an acute test
with the results after 2 months of use revealed stable audiological performance with
the ADHEAR.[13] However, several studies[28]
[29]
[30] have shown that the BCI requires an acclimatization time of several months to reach
its performance plateau. The reasons for the differences we have observed between
both devices could lie in the different transducer design and placement, higher output
with the active BCI, and transmission loss through the skin with the passive aBCD.
The transducer of the aBCD is located outside and on top of the skin, transferring
vibrations passively through the skin with associated signal dampening.[31]
[32]
[33] The transducer of the active implant sits in the skull bone and stimulates the bone
directly. Besides the differences in output parameters of these devices, the active
design of the implant enables signals to be transferred to the cochlea with minimal
transfer loss. Gavilan et al.[33] (2019) compared the audiological performance of the aBCD with a passive BCI (as
opposed to the active BCI used in the present study). Both systems transfer vibrations
passively through the skin, and comparable results between the aBCD and the passive
transcutaneous BCI were reported.
Good, aided speech perception is particularly important for children, especially in
noisy environments such as school. Research[34]
[35]
[36] has shown that even untreated unilateral hearing loss negatively affects language
development, communication skills, academic progress, and social/emotional development.
Although more costly and invasive, an active BCI appears to be the best device for
hearing rehabilitation in these patients.
Percutaneous bone-anchored hearing aids (BAHAs) also directly stimulate the bone.
However, device-associated skin complications and related inconveniences have been
regularly reported.[37]
[38]
We have seen a ceiling effect in WRS results in quiet with the BCI, as well as in
some aBCD cases. Therefore, additional test setups like speech in noise were needed
to evaluate performance differences. A speech in quiet test at a lower presentation
level would have been a valuable addition to the standard 65 dB SPL. The relatively
small sample size and the restrictive inclusion criteria are further limitations of
the present study. The indication of the BCI is for up to 45 dB HL BC hearing loss,
and both devices could be used for patients with unilateral deafness if hearing of
the contralateral ear would fall within the indication criteria. However, the population
of the present study was chosen to facilitate optimal comparability of both systems
within the same subjects.
The questionnaire results showed high subjective satisfaction with the aBCD, which
is consistent with the published results of a comparable patient group.[27] Although most results with the APSQ audio processor-specific questionnaire were
slightly better for the BCI compared with the aBCD, there was no statistically significant
difference. However, the BCI/aBCD comparison questionnaire provided a clearer picture,
as users had to choose between the BCI or aBCD or report equal performance. On the
individual level, the BCI was mostly chosen as the preferred solution due to its superior
output. One user preferred the aBCD as no surgery is required. In addition, the subjects
preferred the BCI to the aBCD in terms of sound quality, cosmetic appeal, and ease
of use. The BCI was also mostly chosen as the better option for those who wear glasses
and the AP at the same time. However, several patients preferred the aBCD during sports.
It is possible that users perceive the BCI processor to be more fragile and costly
to repair than the aBCD.
Regarding hearing-specific subjective satisfaction, the results of the PEACH questionnaire
revealed a clear superiority of the BCI over the aBCD, and they are in line with the
audiological results comparing both systems. As another measure of overall satisfaction,
the wearing time results support the subjective satisfaction findings, as the aBCD
was used sufficiently (9 hours per day); however, the BCI audio processor was used
2 hours longer on average. Although the active BCI was superior to the passive aBCD
in most objective and subjective results, the overall good results and high adherence
to use reinforce how useful the aBCD can be for this group of patients. Lastly, there
were no adverse events reported with either of the devices: both were well tolerated
by the patients, and no problems were reported during the course of the present study.
Conclusion
In the sample of the present study, hearing performance with the passive transcutaneous
device was clinically sufficient and, regarding certain results, comparable to the
active BCI. However, superiority of the implant was shown in terms of quality of life
and after device acclimatization in speech in noise. Thus, the aBCD should be considered
an alternative in cases in which surgery is not desired or not possible.