Introduction
A bone-anchored hearing aid (BAHA) is an implantable bone-conduction system for hearing rehabilitation. When the device was first introduced, the main indication for implantation was a conductive or mixed hearing loss, and patient outcomes were reported to be very successful.Reference Cooper, Burrell, Powell, Proops and Bickerton1 It was subsequently noted that BAHAs may also be beneficial in cases of profound unilateral deafness.Reference Hol, Snik, Mylanus and Cremers2
The implantable bone-conduction system consists of three parts: the implant, the abutment and the sound processor. The implant is a small titanium fixture or screw, which is surgically implanted into the skull behind the ear. This becomes firmly anchored in the bone by a process of osseointegration. The abutment is a socket attached by an internal screw to the implant. This protrudes through the skin of the scalp and is shaped to hold the snap-fit coupling of the sound processor, allowing the fitting and removal of the sound processor. The sound processor is a removable hearing aid, which converts and processes acoustical signals, and transmits them, via bone conduction, from the fixture to the better-hearing cochlea.Reference Tjellström, Håkansson and Granström3
The introduction of programmable in-the-ear and behind-the-ear digital hearing aids in more recent years has had a major impact within audiology. It is estimated that 90 per cent of all behind-the-ear hearing aid fittings use digital processing.Reference Kochkin4 However, as noted by Olsen et al. prior to 2009 the sound processors used in BAHAs had changed little since their first commercial introduction in the 1980s.Reference Olsen, Henrik and Nielsen5 For example, the Divino® BAHA sound processor, launched in 2005, had some digital technology but it was not programmable. Therefore, if an audiologist wanted to adjust the gain to correspond with a patient's audiogram, they needed to use the trimmer potentiometers on the sound processor, in the same way as on an analogue air-conduction instrument. Bone-anchored hearing aid recipients were therefore unable to benefit fully from the same digital processing technology offered by air-conduction hearing aids.
In 2009, there were two new developments for BAHAs. Firstly, the company Cochlear Ltd (Cochlear Bone Anchored Solutions AB, Mölnlycke, Sweden) launched a new product called the BP100, which incorporated full digital technology for the first time. Secondly, Oticon Medical (Copenhagen, Denmark) (a sister company of Oticon, a multinational hearing aid manufacturer) entered the BAHA market and introduced the Ponto Pro device, which also has full digital processing. The introduction of fully digital BAHAs meant that recipients could benefit from features such as adaptive multiband directionality and automatic noise reduction algorithms.Reference Westerkull6, Reference Flynn, Sadeghi and Halvarsson7 The sound processors produced by both companies may be attached to the original Branemark titanium abutment. Therefore, existing BAHA users are now able to use both of the new devices on their existing abutment (although the companies differ in their attitude as to whether or not this is recommended). This allows direct comparison of both new sound processors, as well as comparison with the existing, older models of BAHA sound processors.
Both the BP100 and the Ponto Pro devices have been shown to provide hearing aid users with better speech processing than previous generations of BAHA sound processors, in a range of user groups.Reference Rosenbom8, Reference Pfiffner, Caversaccio and Kompis9 For example, Pfiffner et al. compared the auditory and subjective benefits of the Divino sound processor with the BP100 device in 20 experienced BAHA users with a conductive or mixed hearing loss.Reference Pfiffner, Caversaccio and Kompis9 Participants wore the BP100 for three months before testing. The results showed no significant difference between the two devices in speech understanding in quiet conditions. However, they did demonstrate improved speech understanding in noise with the BP100, and subjective ratings on the questionnaires showed significantly better scores for the BP100 in noisy and reverberant environments. These improvements were not unexpected, as the new sound processors have fully digital sound processing.
Only one published study has directly compared the two new devices with each other. Olsen et al. conducted a cross-over study, in which 12 new BAHA users with a conductive or mixed hearing loss trialled both a BP100 and a Ponto Pro sound processor.Reference Olsen, Henrik and Nielsen5 Differences in speech perception, sound quality, wind noise, auditory feedback suppression and overall participant preference were assessed using the Glasgow Hearing Aid Benefit Profile questionnaire and the Nordic Co-operation on Disability hearing aid questionnaire. Speech-in-noise testing was performed using the Dantale II test, and mean aided thresholds for sound-field testing were also recorded (the better-hearing ear was blocked with an earmuff).
In that study, the Glasgow Hearing Aid Benefit Profile questionnaire showed similar ratings for both devices, and no significant differences were reported.Reference Olsen, Henrik and Nielsen5 The Nordic Co-operation on Disability questionnaire revealed significantly better median ratings for the Ponto Pro than the BP100 in terms of: speech perception, having a conversation with one person in a car, listening to the television or radio, changing the battery, adjusting the control, and the appearance of the device. There were no statistically significant differences between the two devices for the speech-in-noise test results. The median aided thresholds were statistically significantly better with the Ponto Pro than with the BP100 device (p < 0.001). At the end of the study, 67 per cent of the participants reported a preference for the Ponto Pro over the BP100, but this was not a statistically significant result.
The present study aimed to compare experienced BAHA users' ratings of satisfaction, and perceived benefits and performance of two new BAHA devices (Cochlear Ltd BP100 and Oticon Medical Ponto Pro).
Materials and methods
Ethical Ltd approval for the study was given by the National Research Ethics Service (North West 1 Research Ethics Committee, National Health Service (NHS) Health Research Authority).
Independence of study
Cochlear Ltd and Oticon Medical both provided a number of sound processors for use in this study. Other costs were covered by the Department of Audiology at Manchester Royal Infirmary and a grant from the Hospital Research Department.
Participants
Existing BAHA users were identified and invited by letter to participate in the study. The selected individuals were patients in the clinic who met our inclusion criteria. These criteria included: having a conductive or mixed hearing loss (average unmasked bone conduction in the better-hearing ear of ≤ 45 dB HL, between 500 and 3000 Hz), BAHA experience of at least two years prior to the start of the study and English as a first language.
The sample size was determined by the available funding. Sixteen BP100 devices and 16 Ponto Pro devices were used; hence, the first 16 positive respondents to our invitation letter were enrolled in the study.
Speech testing
The Bench–Kowal–Bamford sentence test is an open-set speech perception test that is widely used for children and adults with hearing impairments.Reference Bench, Kowal and Bamford10 The sentences are spoken either in quiet, or in the presence of multi-talker babble noise, via the Predicting Outcomes of Cochlear Implantation in Adults (POCIA) platform.Reference Fielden, Cooper and Craddock11
We recorded the percentage of correct key words repeated by participants in quiet conditions, at the following signal-to-noise ratios: +5, 0, −5 and −10 dB. In each condition, the speech was presented at 70 dB HL. The stimuli were presented through a speaker positioned 1 m from the participant (at head height), at 0° azimuth of the patient's head. The word lists were presented in a randomised order for each participant for the different signal-to-noise ratios. Participants always started the test by listening to the sentences with no additional noise. They then progressed through the conditions in order of increasing difficulty (i.e. +5, 0, −5 and −10 dB signal-to-noise ratios).
Questionnaires
Two questionnaires were used in this study. The first was the Glasgow Hearing Aid Difference Profile questionnaire,12 which was used to compare the two new digital hearing aids with each other. It was administered using an automated format on Auditbase software (following NHS guidelines).12 This questionnaire is typically administered to existing users of hearing aids for whom a change in prescription is being implemented and evaluated. It has been used previously in BAHA studies.Reference McDermott, Dutt, Tziambazis, Reid and Proops13
The Glasgow Hearing Aid Difference Profile comprises questions (with five-point Likert response scales) related to four standard listening situations (watching television, one-to-one conversation, conversation in a busy street or shop, and a group conversation). There is also the option to add user-defined situations, but this was not utilised here. A low score in this questionnaire indicates that the new hearing aid is perceived as worse by participants, and a high score indicates the new hearing aid is perceived as better.
For that questionnaire, participants were asked to compare the two devices directly in terms of the ‘benefit’ and ‘satisfaction’ for each of the four situations. The median scores for each question were determined for each participant.Reference McDermott, Dutt, Tziambazis, Reid and Proops13 The participants were divided into two groups depending on which device they trialled first. This was necessary as the questions are worded as comparisons. In these comparisons, the first BAHA trialled was considered to be the ‘old’ device and the second BAHA trialled was the ‘new’ device.
The second questionnaire was developed specifically for this study in order to examine the participants' subjective responses to the different BAHAs (Appendix 1). This questionnaire was completed at the start of the study to assess the original hearing aid device. It was then re-administered at the end of each four-week trial period. This questionnaire focuses on the aspects of sound processing that have been shown to be of particular importance to hearing aid users.Reference Meister, Lausberg, Kiessling, Walger and von Wedel14–Reference Bridges, Lataille, Buttorff, White and Niparko16 It comprises questions on speech perception in noise and quiet environments, sound quality, aesthetic appearance of the device, usage, physical comfort, auditory feedback, and handling (rated on five-point Likert scales) (see Appendix 1).
Study design
This study was designed as a blind, cross-over study. The hearing aid users were randomly allocated to two groups, trialling either the BP100 or the Ponto Pro device first. The study involved three clinic visits for each participant.
At the first clinic appointment, consent regarding participation in the study was obtained. A pure tone audiogram was also performed, wherein unmasked bone conduction was measured for both ears at 0.5, 1, 2, 3 and 4 kHz. The first half of this clinic session focused on participants' evaluations of their existing BAHA device and their listening performance. This involved completing the bespoke hearing aid questionnaire and undergoing speech-in-noise testing.
Participants were then fitted with either a BP100 or Ponto Pro device. All devices were fitted according to manufacturers' programming guidelines, using the ‘BC direct’ (direct bone-conduction threshold) function of the BP100 fitting software. (At the time of the study, the ‘BC in situ’ function was not available on the Ponto Pro software, so this was not utilised.) All participants were fitted with the versions of software and devices that were available from the start of October 2010. For the BP100 device, the fitting software ‘1.0 (SR2 1.0.8.5355) Driver version 1.6.8.4407’ was used, and for the Ponto Pro device the fitting software version ‘2009.1’ was used. All product and manufacturer information was blanked out (including the logos on the devices themselves), and we also asked participants to refrain from using the internet to view the different shapes and styles of BAHAs available, although we were unable to enforce this.
After the fitting, all participants were given the relevant instruction manuals for the device, and the different controls were explained. All devices had volume control activated. When any auditory feedback occurred, it was managed following the manufacturers' guidelines. Participants were asked to trial the device for a four-week period at home.
At the second clinic appointment, participants were asked to complete the bespoke hearing aid questionnaire and the first section of the Glasgow Hearing Aid Difference Profile questionnaire to evaluate their experience with the first-trialled device. Speech-in-noise testing was again performed with this device. During the latter half of this appointment, participants were fitted with the second new BAHA device, following the same procedure as for the first-trialled device. Participants trialled this second device at home over another four-week period.
At the final appointment, participants were asked to complete the bespoke hearing aid questionnaire and the second section of the Glasgow Hearing Aid Difference Profile questionnaire. Speech-in-noise testing was performed with this second new BAHA device.
Data analysis
Speech testing
Data analysis focused on: the percentage of correct results at a single, fixed signal-to-noise ratio of −5 dB, and the signal-to-noise ratio needed for 50 per cent correct responses on the speech testing material. Performance measures at other noise levels showed strong floor or ceiling effects. The 50 per cent correct response signal-to-noise ratio data points were obtained by plotting the speech test results (percentage of correct responses) for the different signal-to-noise ratios measured. These were then fitted to a probit curve and the 50 per cent correct response point was interpolated from this. The statistical significance of differences was determined using paired t-tests and Wilcoxon signed rank tests.
Questionnaires
For the Glasgow Hearing Aid Difference Profile questionnaire, median scores and interquartile ranges were calculated, as in the study by McDermott et al.Reference McDermott, Dutt, Tziambazis, Reid and Proops13 For the bespoke hearing aid questionnaire, mean scores and standard deviations were calculated for each device. Wilcoxon signed rank tests were used to examine differences between the devices.
Results
Participant data
A summary of participants' demographical and audiological data is shown in Table I. Two participants did not complete the study; none of their data were used in the subsequent analysis, and the results reported are for those 14 participants who completed the study.
Fig. 1 Boxplots showing: (a) the signal-to-noise ratio at which 50 per cent of responses were correct, and (b) the percentage of correct responses at a fixed signal-to-noise ratio of −5 dB. (Boxes represent the interquartile range, solid horizontal lines represent the median, and ‘whiskers’ show the range (excluding outliers; outliers are identified as individual points).) SNR = signal-to-noise ratio
Table I Demographical and audiological data
*Patients 1 and 12 did not complete the study and so their data are not reported here. †At the start of the study. ‡Average unmasked bone-conduction values (measured at 500–3000 Hz). Pt no = patient number; y = years; BAHA = bone-anchored hearing aid; BC = bone conduction; F = female; M = male
Trial duration
The mean duration of both trials was 28 days (standard error of the mean (SEM) for the first device trialled was = 0.5; SEM for the second device trialled was 4.6). The SEM was greater for the second trial because participant number 8 kept their second device (a BP100) for 91 days.
Speech-in-noise data
The results for both outcome measures (the percentage of correct responses at the fixed signal-to-noise ratio of −5 dB and the signal-to-noise ratio needed for 50 per cent correct responses) indicated that both new devices performed better than the old one (Table II). There were notable improvements (higher mean scores and lower signal-to-noise ratios) for both new devices compared with the original device (see Figure 1). However, there were no statistically significant differences between the two new devices, for either measure (Table III).
Table II Speech-in-noise data
Data represent means (standard deviation). SNR = signal-to-noise ratio
Table III Device comparison statistics
CI = confidence interval; SNR = signal-to-noise ratio
Questionnaire data
Glasgow Hearing Aid Difference Profile
It was not possible for one participant to complete the second half of this questionnaire; however, their data from the first part of the questionnaire have been included in the analysis. The median questionnaire scores were almost identical for the two digital devices (see Table IV). In the questions that asked for a direct comparison between the two devices in terms of ‘benefit’ and ‘satisfaction’, the most common response was ‘3’ on the Likert scale, indicating equal performance for both devices. This finding is in line with the results for the speech-in-noise data, for which no significant differences were observed between the two devices in a range of listening environments.
Table IV Glasgow hearing aid difference profile data
Data represent median raw scores (and interquartile ranges). TV = television
Hearing aid questionnaire
Analysis of the bespoke hearing aid questionnaire data focused on comparisons between: the BP100 and the original device, the Ponto Pro and the original device, and the BP100 and the Ponto Pro.
Comparison of the BP100 with the participants' original BAHA device revealed a significant difference for question 4 only, which related to the handling of the device. Specifically, the rating for battery changing, volume control operation and so on was significantly lower (worse) for the BP100 compared with the old device (p = 0.007).
Comparison of the Ponto Pro with the participants' original device revealed significant differences for three of the questionnaire items (questions 3, 4 and 5). Specifically, the appearance of the Ponto Pro was rated significantly higher (better) than the old device (p = 0.01), but the Ponto Pro was rated significantly lower (worse) than the old device in terms of handling (i.e. battery changing, volume control operation and so on) (p = 0.04) and connection (and disconnection) of the device to the abutment (p = 0.016).
Direct comparison of the BP100 with the Ponto Pro revealed significant differences for questions 5, 12 and 16. Ratings for connecting the BAHA to the abutment were significantly higher (better) for the BP100 compared with the Ponto Pro (p = 0.034). Experiences of auditory feedback were significantly better for the Ponto Pro compared with the BP100 (p = 0.022). (It was not simply that one device was rated better than the other; scores for this question were high for the Ponto Pro and low for the BP100). Finally, the overall sound quality was rated as better for the Ponto Pro than the BP100 (p = 0.048).
At the end of the study, nine participants opted to wear the Ponto Pro and five opted to wear the BP100; this difference was not significant (p = 0.30). We examined the possibility of order effects to see if participants' preferences were biased towards the device trialled first. There were seven participants in each group. Of the seven participants who trialled the Ponto Pro first, all seven decided to keep the Ponto Pro at the end of the study. Of the seven participants who were initially fitted with the BP100, five decided to keep the BP100. The other two participants decided to keep their second device, the Ponto Pro, following their experiences of auditory feedback and ‘beeping’ from the BP100. This suggests that on the whole, participants tended to favour the first device they received (see Table I).
Discussion
Our findings showed that both the Ponto Pro and the BP100 were associated with similar significant improvements in listening in speech-in-noise testing compared with the previously used BAHA devices. The results reflect the noise reduction algorithms and other improvements present in these new fully digital devices. These findings are in agreement with other studies that have compared either the BP100 or the Ponto Pro with older generations of BAHAs. For example, Pfiffner et al. showed that the BP100 was better than older models of BAHA sound processors for listening to speech in noise in experienced users with a conductive hearing loss.Reference Pfiffner, Caversaccio and Kompis9
Better speech understanding in noise was rated the most desired improvement by American hearing aid users.Reference Kochkin17 For every 1 dB improvement in signal-to-noise ratio, it is estimated that there is an improvement of the order of 8–10 per cent in speech audibility.Reference Ricketts18, Reference May, Larsen and Warland19 Our data are in broad agreement with these findings. For example, in the present study, comparison of the Ponto Pro with the older devices revealed an improvement in the 50 per cent signal-to-noise ratio (the ratio at which 50 per cent of responses were correct) of approximately 2 dB. Furthermore, at the fixed signal-to-noise ratio of −5 dB, there was an average improvement of around 18 per cent in speech-in-noise test scores. Hence, even seemingly small improvements in signal-to-noise ratio can have potentially important consequences for hearing in noisy environments.
It is hard to compare our results directly with other studies because of the different languages and assessment techniques used for the speech tests. However, the results reported in other studies for the 50 per cent correct response signal-to-noise ratio, for conditions in which speech was presented from the front with the microphones in omnidirectional mode, are generally similar. For example, Pfiffner et al. showed a 50 per cent correct response signal-to-noise ratio improvement of 1.2 dB for the BP100 in this condition (when compared with the Divino), but they did not test the Ponto Pro.Reference Pfiffner, Caversaccio and Kompis9 Together with our data, there is growing evidence to suggest that, compared to older BAHA models, the Ponto Pro and BP100 improve speech understanding in challenging listening conditions.
Our data can also be examined alongside studies that have used questionnaires to compare either the BP100 or the Ponto Pro with previous BAHA devices. The study by Pfiffner et al.Reference Pfiffner, Caversaccio and Kompis9 is particularly interesting, as their participant group consisted of existing BAHA users with a mixed or conductive hearing loss (similar to the current study), allowing comparisons to be drawn. Using the Abbreviated Profile of Hearing Aid Benefit questionnaire, the authors reported trends on every dimension, indicating that the BP100 was better than the previous devices used. However, these effects only reached significance for two (of the four) dimensions: background noise and reverberation.
In many BAHA studies (including this one), relatively small sample sizes are used. This is often because of the cost of the devices and the relative scarcity of participants. This, combined with the inevitable coarseness of the rating scales, will impede the ability of researchers to demonstrate modest differences in patient perceptions.
Comparisons with other studies
Few studies have directly compared the two new digital BAHAs with each other. Our results show some agreement with those reported by Olsen et al. (in a study of new BAHA users).Reference Olsen, Henrik and Nielsen5 For example, in both studies roughly 65 per cent of participants opted to use the Ponto Pro at the end of the study (67 per cent of participants in the Olsen et al. study and 64 per cent in the present study). However, there are also some differences between the findings of these studies.
In our study, all participants had worn a previous generation BAHA for at least two years, and the differences highlighted between this study and that of Olsen et al. (with regard to the questionnaire data) may reflect differences between new and experienced BAHA users. For example, in the study by Olsen et al., no significant difference was reported between the two devices for ‘putting the device on’. However, in our study, participants preferred the BP100 (over the Ponto Pro) for its connection to the abutment. The previous experience of participants is relevant as the BP100 has a similar shape and attachment mechanism to the previous generations of BAHA devices, and this familiarity may have led to a subjective preference for this device in the present study (compared with the Ponto Pro, which is visually different and has a different attachment mechanism).
The questionnaire data for the BAHA users' existing devices revealed that the majority of participants felt their devices were already providing them with optimal listening capabilities (despite being associated with poorer speech discrimination scores). This will have led to ceiling effects when the questionnaires were completed for the new digital BAHAs. This effect might be especially marked in BAHA users, rather than users of other types of hearing aids, as they may have been told that a BAHA was their only hearing aid option, and therefore may have reduced their expectations and increased subjective satisfaction of their original device.
Auditory feedback
In our bespoke hearing aid questionnaire, experiences of auditory feedback were rated as significantly better for the Ponto Pro compared with the BP100 (p = 0.022). Earlier models of the BP100 have been associated with auditory feedback difficulties. For example, using a first generation model of the BP100, Bosman et al. (unpublished data) reported excessive auditory feedback, leading to under-amplification for participants or poor sound quality. The authors went on to note that this issue was subsequently improved.
In this study, one participant provided us with the following written comment: ‘some sounds cause the device to resonate and produce a screaming sound; some of the sources for this [are the] vacuum cleaner, scanner, printer, TV’. No such comments were made regarding the Ponto Pro. These findings are in broad agreement with Olsen et al.; the authors stated that four participants claimed the BP100 gave frequent auditory feedback, with four reports of artefacts related to the use of the auditory feedback canceller.Reference Olsen, Henrik and Nielsen5 Two participants reported no auditory feedback with the BP100. For the Ponto Pro, two participants in that study reported the rare occurrence of auditory feedback, and two reported frequent auditory feedback.Reference Olsen, Henrik and Nielsen5
These findings suggest that although both fully digital devices give superior performance to previous BAHA devices, further developments would be of value.
Order effects
Our findings showed a trend in which participants tended to favour the first device they trialled. The first device trialled may have made a large impression on participants, marking the change from analogue to digital. As noted by Bille et al., participant bias favouring the new and modern is difficult to address and can influence findings.Reference Bille, Jensen, Kjærbøl, Vesterager, Sibelle and Nielson20 There may have been an unconscious bias toward the initial device received. For the second device trialled, some degree of acclimatisation may have occurred. This may have resulted in diminished appreciation of the change relative to their original device.
• The first two fully digital bone-anchored hearing aids (BAHAs) were launched in 2009
• These devices were compared to each other and to earlier BAHA models
• Both new digital BAHAs performed similarly in speech-in-noise tests
• The new BAHAs showed improvement compared with older models that were not fully digital
Our findings are in contrast to the study by Olsen et al., in which 8 of the 12 participants preferred their most recently trialled BAHA.Reference Olsen, Henrik and Nielsen5 As naive BAHA users, their experience with the first BAHA may have included some general BAHA teething problems, which were overcome by the time they switched to the second device. In our study of experienced BAHA users, this was less likely to be an issue. Of the seven participants who were initially fitted with the BP100, two participants decided to keep their second device (the Ponto Pro). Both participants reported auditory feedback and beeping from their BP100, which may have had a stronger influence on preference than any order effects.
Conclusion
The use of both the Cochlear Ltd BP100 and the Oticon Medical Ponto Pro resulted in superior speech processing compared with the previous generation of BAHA devices. However, despite the differences between these new digital devices in terms of signal processing, there does not appear to be a substantial difference for users, as indicated by the results of objective and subjective tests.
Acknowledgements
We would like to thank all our participants for their involvement in our study. We would also like to thank Oticon Medical and Cochlear Ltd for donating the sound processors. Central Manchester Foundation Trust generously contributed to the funding of this study via a ‘Research for Patient Benefit’ grant. This research was facilitated by the Greater Manchester Comprehensive Local Research Network.
Appendix 1. Hearing aid questionnaire
1 Over the last four weeks, on average how many hours did you use your device(s):
2 Battery life time:
3 The appearance of the device (i.e. shape/ size/ colour):
4 The handling (changing batteries, operating volume control etc.):
5 Connecting and disconnecting the device to my abutment:
6 Understanding speech in quiet:
7 Understanding speech in noise (e.g. at a party):
8 Understanding speech in noise (e.g. at a party), when the person speaking is on the same side as my BAHA:
9 Understanding speech in noise (e.g. at a party), when the person speaking is on the opposite side to my BAHA:
10 Getting used to the sound of my new BAHA device was:
11 Listening to my own voice sounds:
12 Experience of feedback (whistling/ beeping):
13 Listening to music is:
14 The ability to easily ignore other sounds when trying to listen to somebody is:
15 The effort it takes to hear words and follow conversation is:
16 Overall sound quality:
17 Overall speech understanding:
18 How satisfied are you with this BAHA?
