Introduction
Bone-anchored hearing aids (BAHAs) were first developed in the 1970s. These replaced the conventional bone conduction aid, which involved wearing a tight, visible headband. A BAHA uses the principle of bone conduction and osseointegration to transfer sound vibrations to a functioning inner ear. It consists of a permanent titanium implant, and removable abutment and sound processor.1
Informed consent for surgery requires a discussion of the procedure and its associated risks, the alternatives, and reported complications. In BAHA surgery, risks commonly discussed during the consent process are infection, soft tissue hypertrophy, skin graft or flap failure, osseointegration failure, and the need for further surgery.Reference Tjellstrom, Hakansson and Granstrom2, Reference Lloyd, Almeyda, Sirimanna, Albert and Bailey3 The Department of Health ‘Reference guide to consent for examination or treatment’ recommends that the consent process includes discussion of the common complications and that consent is tailored according to individual patient risk factors.4 In light of the Chester v Afshar law case findings,5 the Reference Guide also recommends providing patients with information regarding all significant risks, even if the risks are small (less than 1 per cent chance of occurrence).
Bone-anchored hearing aid surgery is a relatively uncommon procedure performed at a few specialist centres. Subsequently, there is a paucity of information regarding rare complications, which may also be under-reported. A literature review of Pubmed, Embase and Medline, using the multi-field search terms ‘bone anchored hearing aid’ or ‘bone anchored hearing device’ and ‘complications’, revealed one case of cerebrospinal fluid (CSF) leak and seizure,Reference Chan, Pantelides and Kim6 two cases of intra-cerebral abscessReference Scholz, Eufinger, Anders, Illerhaus, Konig and Schmieder7, Reference Deitmer, Krassort and Hartmann8 and one case of epidural haematoma.Reference Mesfin, Perkins, Brook, Foyt and German9Although there is widespread awareness amongst otolaryngologists regarding the potential risk of CSF leak, it is not commonly discussed during the consent process.
We discuss a case of a CSF leak in a patient undergoing bone-anchored hearing aid surgery. This is an uncommon but serious complication, with the potential to lead to significant morbidity.
Case report
The report concerns a 47-year-old man, who had previously undergone bilateral tympanomastoid surgery for chronic suppurative otitis media with cholesteatoma. He had bilateral severe mixed hearing loss and was therefore a longstanding user of a behind-the-ear hearing aid. Despite regular and aggressive mastoid cavity management, the patient suffered from recalcitrant, bilateral aural discharge.
Following joint audiology and ENT assessment, the patient was deemed suitable for BAHA implantation. In August 2012, he underwent a right-sided, single-stage BAHA implantation procedure.
The implant site was carefully marked 55 mm posteriorly from the ear canal, in line with the top of the pinna. After split-skin graft fashioning and soft tissue resection, the site for the fixture was prepared using a guide drill (starting with a 3 mm depth drill and, following palpation of the fundus of the hole, proceeding to a 4 mm drill). At this time, there was no evidence of a CSF leak, and the bony inner table of the cranium was intact. However, following the commencement of the counter-sink drilling to 4 mm, a CSF leak was noted. The counter-sink drill bit was therefore changed from a 4 mm to a 3 mm bit and the hole carefully fashioned; no obvious tear or breach of the dura was seen on inspection of the site. A 3 mm abutment was subsequently inserted into the bone and the CSF leak ceased. After skin closure with 5-0 Vicryl Rapide sutures, an Allevyn® Ag dressing with Aureocort® cream was applied and a healing cap was placed on the patient.
The patient was given a dose of intravenous cefuroxime (1.5 g) at the time the CSF leak was identified (peri-operatively). Post-operatively, the ward staff were made aware of the CSF leak and asked to monitor for a raised temperature and signs of neurological deficit.
The patient remained stable overnight, as indicated by baseline observations. He developed no temperature or neurological sequelae. He was discharged home 1 day post-operatively and advised to watch for a temperature, headache, stiff neck or rash.
On out-patient review one week later, the patient remained well, with a clean, dry wound. No other complications were noted.
Discussion
The standards for obtaining valid consent for patient treatment are presented in the General Medical Council (GMC) ‘Consent guidance: patients and doctors making decisions together’ publication. For an informed decision, the discussion between the doctor and patient must include: the nature of the operation (the intended benefits and indications), and the potential risks and complications. These should include common or rare but serious complications.10 In this case, the patient was informed about the risks of infection, bleeding, graft failure and abutment loosening.
• Informed consent requires a discussion between the doctor and patient regarding the benefits and risks of surgery, including common and rare but serious complications
• Bone-anchored hearing aid (BAHA) surgery poses a cerebrospinal fluid (CSF) leak risk; subsequent meningitis is rare but serious
• A CSF leak should be discussed with potential BAHA recipients during the informed consent process
• Awareness of rare complications aids prompt and appropriate treatment
• Pre-operative computerised tomography scanning should be considered in certain individuals to aid implant placement
A CSF leak alone can be managed peri-operatively without difficulty. However, the potential for the development of meningitis makes a CSF leak a rare but potentially life-threatening complication. The GMC guidance states that it is our duty to inform patients of this level of risk prior to surgery, even if the complication is extremely rare. Should we therefore inform patients about the possibility of a CSF leak, and the potential risks of intra-cerebral abscess and epidural haematoma? Or is it sufficient to inform patients simply of a generic ‘risk of infection and bleeding’?
The companies Cochlear11 and Oticon12 have both produced surgical guidelines that mention CSF leak as a possible complication; however, no reference is made as to how frequently CSF leak can occur. For patients with good bone density, they suggest placing the implant to seal the CSF leak. If the bone is too thin, the implant should be re-sited, after filling the leak with soft tissue or bone wax. In our case, the CSF leak that occurred peri-operatively was managed by placing the abutment to seal the CSF leak, administrating intravenous antibiotics and conducting neurological observations. Those who undergo elective intracranial neurosurgical procedures are routinely given single-dose intra-operative prophylactic antibiotics, and there is good evidence to suggest that such a conservative approach is sufficient.Reference Moorthy, Sarkar and Rajshekhar13
Routine BAHA surgery involves drilling a very narrow hole. This makes the identification and protection of the dura more difficult. Novakovic et al. analysed temporal bone computerised tomography (CT) scans to establish outer calvarial thickness prior to BAHA system insertion.Reference Novakovic, Meller, Makeham, Brazier, Forer and Patel14 They measured calvarial thickness 55 mm posterior to the superior canal wall in the axial and coronal plane, at vertical levels from +1 to +5 cm above the superior canal wall. They found that 55 per cent of temporal bones had at least one site that was thinner than 4 mm and 21 per cent had at least one site thinner than 3 mm. Thickness in the coronal plane was greatest at +1 cm and thinnest at +4 cm above the superior canal wall.Reference Novakovic, Meller, Makeham, Brazier, Forer and Patel14 Because of the inherent exposure to radiation in CT scanning, and the low incidence of a CSF leak in BAHA surgery, the scanning of all patients is not recommended. However, one might consider scanning individuals who are at a higher risk of CSF leak, such as children, and adults with known reduced calvarial thickness.
Current guidelines from Cochlear15 and Oticon16 suggest siting the implant 50–55 mm from the ear canal in a line running from the top of the pinna or just below. If our patient were to have a second BAHA fitted, it would seem sensible to perform a CT scan to confirm calvarial thickness.
