Introduction
Percutaneous bone-conduction hearing devices provide an effective hearing rehabilitation option for adult and paediatric patients with conductive or mixed hearing loss.Hearing loss is one of the most prevalent disabilities amongst the UK population,1 and the demand for hearing aids has been rising. Bone-conduction hearing devices were first used in 1977, and since then have undertaken many redevelopments to produce the current bone-conduction hearing device models used today.Reference Tjellström and Granström2
Bone-conduction hearing devices are implanted surgically, under local or general anaesthetic, and involve drilling a titanium implant screw into the mastoid segment of the temporal bone. This implant will, over time, osseointegrate with the surrounding bone to achieve stability. Sound is transmitted via this implant, through the bony skull to the inner ear and stimulates the cochlea thus bypassing or overcoming any conductive hearing losses.
Variations in bone-conduction hearing devices exist.Reference Grant3 Percutaneous bone-conduction hearing devices form a direct connection between the external hearing aid to the implant within the bone via an abutment. Transcutaneous devices transmit signal from the external sound processor to the implant under the skin, with the skin itself intact.
In 2013, the NHS commissioning board outlined the criteria for bone-conduction hearing device surgeries within the NHS.4 However, complications may arise following bone-conduction hearing device surgeries, which may ultimately lead to removal of the device or revision surgeries. These commonly include skin issues such as skin overgrowth, bone-conduction hearing device site infections and granulation tissue formation. In addition, bone-conduction hearing device implants can extrude or fail to achieve osseointegration.
Much has been written about skin complications following percutaneous bone-conduction hearing device surgery.Reference Kiringoda and Lustig5–Reference Shirazi, Marzo and Leonetti8 Transcutaneous surgery has reduced skin complicationsReference Robinette, Sims, Pang and Babu6, Reference Brinkman, Hill, Hone and Kieran9, Reference Shapiro, Ramadan and Cassis10 yet is not suitable for many patients. From our personal experience, percutaneous implant surgery is quicker to perform than transcutaneous surgery. Percutaneous implant surgeries are also MRI compatible.11 The shadow effect from the implant is small and the ipsilateral ear and brain can still be assessed. First-generation transcutaneous implants were not MRI compatible. The newer models are MRI compatible but still retain a significant shadow over the ipsilateral side of the head. Patients that need MRI scans for possible cholesteatoma, or central pathology on the ipsilateral side would not be suitable for a transcutaneous implant. The maximum image artefact according to the manufacturer is 2.7 cm with the Cochlear Osia system (Sydney, Australia) whenthe implant magnet is removed,12 compared to 1.2 cm for the Cochlear Connect system when the abutment is removed.13 Finally, there is evidence that post-implantation costs are higher in passive transcutaneous bone-conduction devices compared to percutaneous devices, due to more frequent explantations.Reference Aukema, Teunissen, Janssen, Hol and Mylanus14 This may change with newer devices and improved technology.
Our unit has historically used Cochlear Connect bone-conduction hearing devices and we were keen to identify our complication rate and any factors that may have contributed to this. With the advent of newer devices such as those manufactured by Oticon Medical (Vallauris, France) using a different implant, abutment design and boasting lower skin complication rates,Reference Lagerkvist, Carvalho, Holmberg, Petersson, Cremers and Hultcrantz15 we wanted to explore our rate of complications. These data will also contribute towards the body of evidence reading skin reaction and complications following specifically Cochlear Connect bone-conduction hearing devices.
The specific aim of this study was to assess the complication rate and potential contributing factors leading to revision procedures for adult patients who underwent bone-anchored hearing aids (BAHA) Connect operations between 2016 and 2021 at the Queens Medical Centre, Nottingham University Hospitals. Previous studies have shown complication rates of about 13 per cent for skin complications and 12 per cent for revision procedures.Reference Hobson, Roper, Andrew, Rothera, Hill and Green7 We also aimed to conduct a literature review to compare our data to the existing published literature.
Materials and methods
A retrospective cohort analysis was performed on all adult patients who had surgical insertion of a BAHA Connect device at our unit between 2016 and 2021. Medical records were retrospectively reviewed in 2022 for patients who underwent just BAHA Connect operations during this five-year period. Data analysis was performed reviewing patient and operative factors affecting bone-conduction hearing device complication rates and frequency of revision procedures.
Patient factors including age, co-morbidities and smoking status were reviewed. Operative factors were also recorded, including grade of surgeon performing the procedure, operative technique, local or general anaesthetic, skin thinning, intra-operative complications, skin thickness and abutment length.
Post-operative complications such as skin complications and dislodgement of bone-conduction hearing devices were documented based upon clinic and audiology documentation. Skin complications were evaluated using Holgers classification (Table 1).Reference Holgers, Tjellström, Bjursten and Erlandsson16
Holgers classification for skin complications at bone-conduction hearing device sitesReference Steehler, Larner, Mintz, Steehler, Lipman and Griffith32
Data was recorded and analysed on Excel (Microsoft Corporation, Redmond). Statistical analysis was performed using a binary logistics regression model with p-value of 0.05 used to determine statistical significance.
Results and analysis
Cohort analysis
Of the 166 patients underwent surgical implant of a BAHA Connect device in the five-year study period (2016 to 2021), 99 patients were female and 67 were male. The youngest adult patient was 18 and oldest was 91, with an average age of 56 (± 16.5 years standard deviation). One hundred and four patients had no co-morbidities of note and nine were diabetic.
Complication rates and causes
Twenty-seven patients (16 per cent) had skin reactions after their initial bone-conduction hearing device procedure (Holgers grade 3 or 4). These cases included skin infection, inflammation and granulation. In addition, 15 cases (9.0 per cent) had skin overgrowth, and in 8 cases (4.8 per cent) the bone-conduction hearing devices dislodged. This breakdown can be found in Table 2.
Complications following initial bone-conduction hearing device procedure
Twenty-nine patients out of the total cohort (17.5 per cent) had at least one procedure thereafter. The majority of additional procedures were due to skin infection/inflammation with Holgers grade 3 or 4 (11 patients) and skin overgrowth (10 patients). Seven patients (4.2 per cent) had re-operation due to bone-conduction hearing device extrusion secondary to head trauma. One case (0.6 per cent) was due to failure of osseointegration, with no preceding trauma. This is highlighted in Table 3.
Indications for revision procedure
Holgers’ criteria were used to categorise the severity of adverse skin problems in bone-conduction hearing device patients following operation. The majority had grade 0 (60 per cent) in every follow-up within the timeframe of this study. The remainder included grade 1 (9 per cent), grade 2 (15 per cent), grade 3 (10 per cent) and grade 4 (6 per cent). All grade-4 patients had revision procedures to remove the implant and then re-implant after some time. Increasing Holgers score led to an increased risk of further procedures in theatre to replace or remove the abutment, which was statistically significant (p = 0.0005).
Additional factors and revision procedures
Sixteen per cent of cases performed under general anaesthetic required revision procedures, compared to 18 per cent of cases under local anaesthetic. Fourteen per cent of patients with the bone-conduction hearing devices abutment in the line of the incision (linear approach) required a revision procedure, whilst 18 per cent of patients where the abutment was placed separate to the incision required revision. The majority of cases (n = 139) were done placing the implant away from the incision site compared to 29 cases where the implant was placed in the line of the incision.
For grade of surgeon, 31 per cent of cases performed by registrar required revision, compared to 17 per cent by the consultant, and 9 per cent by the senior fellow. Eighty-five per cent of the cases were performed by a consultant only, and the odds of a revision procedure based on grade of surgeon were not statistically significant (p = 0.82).
Of the 10 patients who required procedures due to skin overgrowth, five patients had their abutment changed to a longer length in theatre whilst four patients had only skin removal with no change in abutment. One patient required a second revision procedure due to further skin overgrowth.
Intra-operative complications
Nine cases were not straightforward intra-operatively. Six of these cases were due to bleeding following the countersink drill, in which all six patients had a second hole drilled. Three other cases required a second countersink hole as the first attempt did not engage the implant. Of these nine cases, five required revision procedures. The reasons for revision procedure were due to the abutment falling out secondary to head trauma (two cases), skin infection (two cases) and overgrowth (one case). Only one case had skin thinning performed, which did not incur a revision procedure.
Discussion
Complication rates in the literature
There have been many published studies detailing the rates of complications following bone-conduction hearing device procedures as well as the reasons for these complications. Complication rates ranged from 6 to 67 per cent of patients in similar studies, although most papers have many fewer patients.Reference Hobson, Roper, Andrew, Rothera, Hill and Green7, Reference Shirazi, Marzo and Leonetti8, Reference Fontaine, Hemar, Charpiot and Debry17–Reference Wazen, Young, Farrugia, Chandrasekhar, Ghossaini and Borik27 There was heterogeneity on reporting complications, which may account for some of the variation in rates. In our study, we had a comparable complication rate of 29 per cent and rate of revision procedure of 17.5 per cent. This is one of the very few studies that looked at the complication rate of bone-conduction hearing devices in relation to several patient (age, sex, gender, smoking habit) and surgical (general anaesthesia versus local anaesthesia, grade of surgeon, incision type) factorsReference Hobson, Roper, Andrew, Rothera, Hill and Green7, Reference Fontaine, Hemar, Charpiot and Debry17, Reference House and JW20
Hobson et al. performed a similar review of bone-conduction hearing device complication rates in a tertiary UK unit.Reference Hobson, Roper, Andrew, Rothera, Hill and Green7 In their study, a complication rate of 23.9 per cent and revision procedure rate of 12.1 per cent were documented amongst a study with 144 patients. As in our study, the majority of complications referenced are skin complications. Skin overgrowth and infection or inflammation accounted for 55 per cent of total complications, compared to 68 per cent in our cohort. Furthermore, skin complications and bone-conduction hearing devices dislodgement accounted for roughly 90 per cent of indications of revision bone-conduction hearing device procedures in both studies. Following this study we have implemented induction antibiotics in all bone-conduction hearing device recipients (800 mg intravenous teicoplanin).
Interestingly, de Souza et al. performed a 10-year systematic review of complications of transcutaneous devices.Reference de Souza, Vallejos Riart, de Souza, de Brito and Bento28 The average complication rate of 21.3 per cent is comparable to our findings (29 per cent) as well as published complication rates in other percutaneous device papers, again with the majority being skin complications.Reference Hobson, Roper, Andrew, Rothera, Hill and Green7, Reference Shirazi, Marzo and Leonetti8, Reference Fontaine, Hemar, Charpiot and Debry17–Reference Wazen, Young, Farrugia, Chandrasekhar, Ghossaini and Borik27
Individual factors affecting complication rates
Operative technique
In this study, operative technique was divided into incisions in line with the implant (29 cases) or away from it (139 cases). No significance was found comparing the techniques, which is in keeping with other studies identifying no statistically significant differences between the placement of the incisions.Reference Miura, Ogando, Pasqualini, Neves and Lubianca Neto29–Reference Bezdjian, Smith, Gabra, Yang, Bianchi and Daniel33 Furthermore, these studies suggested that both techniques have good surgical outcomes however operative time was shorter when the implant was placed within the skin incision.Reference Miura, Ogando, Pasqualini, Neves and Lubianca Neto29
In this study, there was no difference in complication rates or revision surgeries whether the operation was performed under local or general anaesthesia. The rate of revision procedure was higher in surgical registrars (residents/doctors in training) compared to consultants (attending surgeon) as primary surgeon, but this was not statistically significant (p = 0.56).
Patient factors: skin complications
Holgers Criteria have been used to define the degree of adverse reaction of the skin.Reference Holgers, Tjellström, Bjursten and Erlandsson16 In general, grade 4 usually necessitates removal of the abutment, whilst lower grades can be managed conservatively, if possible. In this study, 18 patients were classified as Holgers grade 3. Forty per cent of these patients required removal of the abutment as they did not respond to conservative management, with the remainder settling with antibiotic treatment. One patient required admission for intravenous antibiotics, which ended up requiring abutment removal. Twenty-four per cent of grade-2 skin reactions had a revision procedure (6 out of 25 patients), with four cases due to skin overgrowth, one due to failure of osseointegration and one due to bone-conduction hearing device dislodgement.
The Holgers grading distribution of patients is known to vary between different studies.Reference Peñaranda, Garcia, Aparicio, Montes, Barón and Jiménez24–Reference Marfatia, Priya, Sathe and Mishra26, Reference Candreia, Birrer, Fistarol, Kompis, Caversaccio and Arnold34 Previous studies have identified that dark-skinned patients had a higher skin-reaction rate.Reference Candreia, Birrer, Fistarol, Kompis, Caversaccio and Arnold34, Reference Zeitler, Herman, Snapp, Telischi and Angeli35 Zeitler et al. published that an African American patient population had a significantly higher rate of skin-site complications compared to other ethnic groups.Reference Zeitler, Herman, Snapp, Telischi and Angeli35 This patient group also had higher prevenance of keloid scars, with all five African American patients with major skin complications having either keloid or hypertrophic scar formation, thus suggesting a potential correlation.Reference Zeitler, Herman, Snapp, Telischi and Angeli35 Other potential risk factors include higher body mass index and smoking.Reference Gardiner, Lavallee, Sima and Coelho36 These factors were not examined in this study, however the vast majority of patients were Caucasian.
One patient required skin thinning. The patient subsequently developed a Holgers grade-3 skin reaction that resolved with topical and oral antibiotic treatment.
Gardiner et al. noted that patients with a positive tobacco history had an increased risk of skin complications immediately after a percutaneous hearing implant procedure.Reference Gardiner, Lavallee, Sima and Coelho36 Additionally, active smokers had higher rates of skin complications in long-term follow up compared to non-smokers. In this study, 34 per cent of patients who had a revision procedure had a smoking history. This is in relation to 13 per cent in our population that smoke.37
Of the 14 patients 18–29 years old, 42 per cent required a revision procedure. Five out of the six patients were due to skin complications, and one due to bone-conduction hearing device dislodgement. This was comparably high with respect to the other age groups. Older age groups had complication rates of 13–14 per cent, apart from the 50–59 year-old age group which scored a 22 per cent complication rate. Statistical analysis highlighted that increasing age correlated to a lower rate of revision procedure, with this result being statistically significant (p = 0.03). Potential factors explaining this include older patients may be more likely to keep their wound clean, or less likely to participate in sports or activities which might result in dislodgement of their hearing aid. Furthermore, our study identified that females had lower rates of revision procedure than males (p = 0.009).
• Bone-conduction hearing devices surgery represents a simple intervention to boost audiological outcomes. Skin related complications are common and may lead to revision procedure
• Patient related factors such as age, smoking status and gender are likely to affect complication rate
• Identification of factors that increase likelihood of complications should be identified to inform patients and clinicians
• Surgical factors such as anaesthetic or surgical technique do not seem to affect compliction rate
Limitations
Our study was performed in a single tertiary UK centre; therefore outcomes may be limited to our single-site hospital practices. Ideally future studies should include multi-centre trials and standardise outcomes and methodology. A retrospective study was performed, threfore interpretation and reliance of documentation can lead to bias.
Conclusion
Percutaneous bone-anchored hearing aids (BAHA) can have complications that affect the patient’s ability to use the sound processor and require further procedures not infrequently. This study found that 29 per cent of patients suffered some form of minor or major complication after the operation. Most were skin related at the bone-conduction hearing device site. Of all bone-conduction hearing device patients, 17.5 per cent required a revision procedure, however 4 per cent were due to head trauma and not due to a bone-conduction hearing device-related complication. These figures are comparable to other studies.Reference Hobson, Roper, Andrew, Rothera, Hill and Green7, Reference Ricci, Della, Faralli, Longari, Gullà and Mansi18, Reference House and JW20, Reference Badran, Arya, Bunstone and Mackinnon22, Reference Wazen, Young, Farrugia, Chandrasekhar, Ghossaini and Borik27 Factors including local or general anaesthetic, incision technique and grade of surgeon did not appear to significantly influence the rate of a revision procedure. However, younger age and smoking history were identified as potential risk factors for bone-conduction hearing devices complications post insertion. Further studies are required to look at these potential risk factors to identify a causal link.
Acknowledgements
We are grateful to Polly Scutt, medical statistician at the University of Nottingham, for her help and advice with statistical analysis.
Funding
There are no funding or competing financial interests to declare.
