Anaesthesia for major middle ear surgery

Key points.

• •The function of the middle ear is to transmit sound energy from the environment to the inner ear.
• •The facial nerve runs through the middle ear cavity and is at risk during middle ear surgery (MES).
• •Facial nerve monitoring can be used to lower the chance of iatrogenic injury and is affected by neuromuscular blocking drugs.
• •A clear intraoperative field is important for successful MES.
• •The use of total i.v. anaesthesia can be advantageous for MES.

Learning objectives.

By reading this article you should be able to:

• •Explain the functional anatomy of the middle ear and its relationship to the facial nerve.
• •List the common indications for middle ear surgery.
• •Discuss the anatomical and technological basis of cochlear implantation.
• •Describe methods by which anaesthetists can improve the intraoperative surgical field.

The middle ear is an anatomically complex region in which surgery may be required to treat a variety of conditions. Access to the middle ear is also required for the placement of multichannel cochlear implants, a technology that has revolutionised the treatment of sensorineural hearing loss. This article will review the anatomy, pathology, and surgical considerations relevant to major middle ear procedures, and the implications for the anaesthetist.

Applied anatomy

The tympanic cavity of the middle ear is an air-filled space located in the petrous part of the temporal bone. The tympanic membrane forms the majority of its lateral wall and separates it from the external acoustic meatus (Fig 1). Sound waves travel into the external ear causing the tympanic membrane to vibrate. These movements are in turn carried through the middle ear by the three auditory ossicles. The malleus is the first of these and is attached to the tympanic membrane by its handle. Its head articulates with the incus, which in turn articulates with the stapes. The base of the stapes lies in the oval window on the medial wall of the tympanic cavity where it transmits its energy to the perilymph of the inner ear. Beneath the oval window is a second membrane-covered opening to the inner ear called the round window. The purpose of the round window is to move in response to pressure changes in the perilymph, thereby allowing the perilymph to move inside the inner ear and stimulate the hair cells of the cochlea.

Fig. 1.

Anatomy of the ear. The ear is divided into three parts. The outer ear, middle ear, and inner ear. The tympanic membrane separates the outer ear from the middle ear (courtesy of J. McAllister).

The tympanic cavity is connected to the nasopharynx via the pharyngotympanic (Eustachian) tubes, which aerate the middle ear and equalise its pressure with that of the atmosphere. Eustachian tube dysfunction appears to be an important factor in the development of middle ear infections (otitis media). The tympanic cavity also connects to a second air-filled space (the mastoid air system) via the aditus to the mastoid antrum in its posterosuperior wall. Middle ear infections therefore usually result in some degree of mastoid inflammation, but this only causes clinically apparent mastoiditis in a small proportion of cases. The tympanic cavity can be divided anatomically into the tympanic cavity proper (mesotympanum) and the attic, which is situated superior to the tympanic membrane.

The primary function of the middle ear is to offset the losses that would otherwise occur because of the differing acoustic impedances of air and cochlear fluid. The mechanical arrangement of the middle ear system allows a gain of up to 27 dB. This mainly results from the relative difference in the surface areas of the tympanic membrane and the oval window. A person with healthy hearing can detect sounds with frequencies between 20 and 20,000 Hz, although the auditory system is most sensitive in the frequency range of 500–4000 Hz.¹

A pure tone audiogram, typically used to quantify hearing loss, assesses hearing between 250 and 8000 Hz. The sound energy levels required to create auditory perception at each frequency are compared with averaged normative data. In the presence of hearing loss, the raised level of energy required for detection of a pure tone sound is expressed using the logarithmic dB scale. A 30 dB hearing loss indicates that the sound level had to be raised by 30 dB above normal levels to be perceived. This is equivalent to increasing the sound pressure intensity by 1000 times above normal threshold levels. In general terms, a 30 dB hearing loss indicates mild hearing loss, whereas a 100 dB hearing loss indicates profound hearing loss.¹

Sensory nerve supply

The sensory innervation to the ear is complex and variable. Important nerves include the auriculotemporal nerve, a branch of the mandibular division of the trigeminal nerve. The auriculotemporal nerve is a major contributor of sensation to the external acoustic meatus and the external surface of the tympanic membrane. The external auditory meatus also receives innervation from the greater auricular nerve (from the cervical plexus) and the auricular branch of the vagus nerve (Arnold's nerve). Sensation to the middle ear cavity, including the medial surface of the tympanic membrane, is mainly from branches of the glossopharyngeal nerve (Supplementary Fig 1).

Facial nerve

During its course from the brainstem to the stylomastoid foramen, the facial nerve passes through the temporal bone and tympanic cavity, enclosed within the bony facial canal (also termed the fallopian canal). The nerve runs posterior to the cochlea to reach the geniculate ganglion and then enters the middle ear cavity on its medial wall, above the oval window. It passes in a posteroinferior direction, first on the medial wall and later on the posterior wall where it exits to run vertically downward in the anterior wall of the mastoid process.

The facial canal can be dehiscent (i.e. having a deficient bony covering) in the middle ear cavity. This is most common in the segment superior to the oval window. The dehiscence may be either microscopic or macroscopic. These and other variants are more common in patients with middle ear pathology and can increase the risk of damage to the nerve during surgery.²

Blood supply

The tympanic membrane, middle ear, and mastoid air system receive their arterial supply from several small arteries, most of which come from branches of the external carotid artery (Supplementary Fig 2). Important vessels include the anterior tympanic artery and superior tympanic branch of the middle meningeal artery, both of which arise from the maxillary artery. The posterior tympanic branch of the stylomastoid artery and the inferior tympanic branch of the ascending pharyngeal artery also make significant contributions. A smaller contribution also comes from the internal carotid via the caroticotympanic arteries.

Venous drainage is via the pteryoid plexus and the superior petrosal sinus, which ultimately drain to the external and internal jugular veins respectively.

Surgical considerations

Common middle ear diseases

A frequent indication for middle ear surgery is cholesteatoma. A cholesteatoma is a cystic mass of stratified squamous epithelium in the middle ear or mastoid air space which constantly produces keratin. Although not neoplastic, it is locally destructive in a similar fashion to a slow-growing benign tumour. Cholesteatomas can erode bone and damage or destroy nearby structures in the middle ear such as the ossicles or facial nerve. In extreme cases, cholesteatomas can erode through the base of the skull to cause intracranial infections.

The cell biology of cholesteatomas is complex, and there are several competing theories regarding their pathogenesis.³ Most appear to originate from the attic or the posterosuperior part of the tympanic membrane. A CT scan (Fig 2) is useful in decision making for major middle ear surgery.

Fig. 2.

Axial computerised tomography scan demonstrating an opacified left mastoid cavity compatible with cholesteatoma (courtesy of S.D. Rejali).

Other common indications for middle ear surgery include otosclerosis (fixation of the foot plate of the stapes with reduction of sound transmission to the cochlea), and complications of chronic suppurative otitis media, such as non-healing perforations of the tympanic membrane (Fig 3), tympanosclerosis, and ossicular chain disruption.

Fig. 3.

Tympanic membrane of left ear with a large central perforation (courtesy of S.D. Rejali). (a) Annulus (thickened peripheral rim). (b) Umbo (central part of tympanic membrane). (c) Incus bone seen through the tympanic membrane. (d) Handle of malleus. (e) Large central perforation exposing the middle ear.

Major surgical procedures

The main aims of surgical intervention include ensuring a ‘safe’ ear to minimise the risk of future complications, to prevent discharge from the ear, and to improve or maintain hearing function. An overview of the more common procedures is given in Table 1. Combinations of procedures may be undertaken (e.g. atticoantrostomy with tympanoplasty).

Table 1.

Common middle ear procedures.

Procedure	Description
Cochlear implant	Placement of a multichannel electrode in the scala tympani through the medial wall of the middle ear. Used in patients with severe and profound sensorineural hearing loss. It partially replaces the function of cochlea transforming sound energy into electrical signals.
Myringoplasty	Isolated repair of the tympanic membrane (also known as a type 1 tympanoplasty). A suitable graft, often temporalis facia, is placed and acts as a scaffold over which new epithelium will grow to repair the hole. Inlay or underlay techniques can be used.
Tympanoplasty	Myringoplasty combined with a repair of chronic middle ear changes to eradicate any chronic infection and restore middle ear function. This can be performed endoscopically.
Ossicular chain reconstruction	Repair (ossiculopasty) or partial replacement of one or more bones of the ossicular chain.
Canal wall up mastoidectomy (includes cortical mastoidectomy and combined approach tympanoplasty)	Removal of bone and disease from the mastoid air system, with preservation of the posterior/superior wall of the external acoustic meatus. In general, this gives a better chance of good functional outcome but a higher chance of disease recurrence. The middle ear can be accessed via this route to remove disease and repair structures inside it.
Canal wall down mastoidectomy (includes atticotomy, atticoantrostomy, modified radical mastoidectomy, and radical mastoidectomy)	Removal of bone and disease from the mastoid air system along with the posterior/superior wall of the external acoustic meatus. The hollowed-out mastoid may be reconstructed or if left open forms part of the external ear canal.
Atticotomy	Removal of disease from the attic, performed for limited attic disease.
Atticoantrostomy	Performed for removal of more extensive disease, starting from the attic and proceeding into the mastoid antrum.
Stapedectomy	Removal of part of the stapes and placement of a prosthesis to improve hearing.
Stapedotomy	Fenestration of the stapes footplate to improve hearing

There are a number of surgical approaches to tympanic membrane repair. The traditional approach is via the external acoustic meatus (transcanal/per meatal approach), but more extensive procedures may require exposure using an endaural or retro-auricular incision. Middle ear procedures are increasingly being performed using endoscopic techniques. In comparison with the operating microscope, the endoscope provides a wide field, high resolution, and a magnified view of the surgical field. For tympanoplasty, this allows a good exposure of the tympanic membrane, without the need for an incision in the ear canal. The endoscope also has the added benefit of enabling the surgeon to look around corners. However, dissection can be challenging as one hand is needed to hold the camera. With a microscopic approach, both of the surgeon's hands are free for dissection, but visualisation of deeper recesses in the middle ear is limited. Retraction of soft tissues and additional drilling to expose the target pathology are therefore needed.

Laser surgery

The physics and safety aspects of medical laser use have been covered previously in this journal.⁴ Uses of laser in middle ear surgery include cholesteatoma removal, where potassium-titanyl-phosphate (KTP) lasers can lower the risk of disease recurrence and may improve hearing outcomes.⁵ Lasers (often CO₂ lasers) are also used for footplate fenestration in primary stapedotomy, with lower risk reported compared with conventional techniques.⁶

Facial nerve monitoring

The incidence of facial nerve injury during middle ear surgery is low, at around 0.1%, but can have a profound psychosocial impact on the patient when it does occur.² Besides sharp and blunt direct trauma, heat generated from high speed burrs can also damage the nerve, without direct contact being made.

The primary means of facial nerve preservation is by a meticulous surgical technique, but intraoperative nerve monitoring is thought to further reduce the chance of an iatrogenic injury.⁷ EMG responses of the muscles of facial expression, most commonly the orbicularis oris, orbicularis oculi, or both, are used. The potential difference between two intramuscular electrodes placed at the start of surgery is measured, with changes in voltage indicating muscle activity and therefore facial nerve activity. If the voltage across the electrodes exceeds a certain set threshold (which can be adjusted manually and is typically in the range of 100 μV), an audible alarm is generated to alert the surgeon.

In addition, a monopolar stimulator probe can be used to stimulate the facial nerve directly to confirm nerve position, function, or continuity. The current at the electrode tip can be adjusted manually. Typical waveforms used during stimulation are square waves of 100 ms duration at a frequency of 4 Hz. A current of 0.1 mA is often sufficient at the cerebellopontine angle to stimulate the bare facial nerve, whereas higher currents (e.g. 0.5–2 mA) may be required to stimulate the nerve when covered with fibrous tissue, granulation tissue or bone. Above 3 mA direct contact or stimulation of the facial nerve may cause nerve damage. New drills can have the stimulator function built into the handset, so that an alarm is triggered if the burr nears the facial nerve during drilling.

Cochlear implantation

Cochlear implants are typically used in patients with severe and profound sensorineural hearing loss. Implants rely on the fact that in most congenital and acquired cases of sensorineural hearing loss, there is sufficient residual cochlear nerve function for it to be directly electrically stimulated. In children with congenital causes, the aim is to perform the procedure before the age of 30 months to give the best opportunity for speech development.⁸

In the first stage of implantation, a cortical mastoidectomy is typically performed through which the facial recess is opened, providing a view of the medial wall of the middle ear. A flexible multichannel electrode is then passed into the scala tympani, which is the distal end of the perilymph-containing channel of the cochlea. This can be done either via the round window or by drilling a separate hole through the medial wall of the middle ear (a cochleostomy).⁹ The electrode array runs from a receiver-stimulator unit which, once positioned, is completely internal to the patient (Fig 4).

Fig. 4.

Components and location of a cochlear implant device (courtesy of J. McAllister).

The electrode array consists of multiple electrodes within a single insulating cover. Each individual wire terminates at a particular part of the electrode's surface. Each wire can thus stimulate a different area of the cochlea, allowing differential frequency perception. The external parts of the system are an integrated microphone, speech processor, and battery unit worn behind the ear. The role of the external unit is to convert sound into a digital code which is transmitted, via a coil held in place with a small magnet, to the internal receiver-stimulator unit.

Many modern cochlear implants also incorporate a telemetry function that can be used to measure the electrophysiological response of the distal portion of the cochlear nerve to stimulation. The implant generates the stimuli and transmits the responses measured to the audiologist. The neural response telemetry (NRT) system (Cochlear Ltd, Sydney, Australia) is an example of such a system.¹⁰ This information can be used to confirm correct placement of the electrode array during surgery and to aid with the programming of the device.

X-rays and measurement of the electrically elicited stapedius reflex threshold (ESRT) are used less frequently to check the placement of a cochlear implant during surgery.¹¹ During ESRT measurement, electrical stimulation of the electrode array results in contraction of the ipsilateral stapedius muscle, with the threshold being the lowest level of current at which stapedius muscle contraction is seen.

As in other procedures involving the middle ear, the facial nerve is at risk and intraoperative nerve monitoring is used during cochlear implant surgery.

Considerations for anaesthesia

Preoperative evaluation

Patients may have a considerable degree of hearing loss, making history taking and communication difficult. Hearing aids, an environment with low ambient noise, positioning oneself closer to the patient's better ear, and facing the patient (making sure your face is well lit to aid lip reading) all help. In adults, the presence of cardiovascular or neurovascular disease requires careful assessment. The surgeon should be informed of patients in whom a hypotensive technique would not be suitable because of excessive risk.

Children presenting for cochlear implant may have a congenital syndrome as the cause of their sensorineural hearing loss, such as Stickler or Klippel–Feil syndrome.¹¹ These conditions are often associated with a difficult airway and require careful evaluation. Congenital sensorineural deafness can be associated with prolongation of the QT interval, either in isolation or as part of the Jervell and Lange–Nielsen syndrome (JLNS), so preoperative ECG screening is recommended.

General anaesthetic techniques

Some procedures can be carried out under local anaesthesia with conscious sedation, such as stapes surgery, tympanoplasty, and mastoidectomy.¹² However, the overwhelming majority of cases in UK practice are performed under general anaesthesia (GA). Reasons for using GA include patient and surgeon preference, avoidance of pain from local anaesthetic injections, and a reduced potential for sudden patient movements, which can compromise the surgery.

Tracheal intubation is often undertaken on account of limited access to the patient's head, long procedure times, frequent use of controlled ventilation to ensure normocapnoea, and occasional need for intraoperative repositioning of the head. In suitable patients though, a supraglottic airway device (SAD) can be an acceptable alternative. SADs have the advantage that neuromuscular blocking agents can be easily omitted and they also minimise the chances of coughing during emergence, which can affect the surgical repair.

When tracheal intubation is undertaken, careful use of neuromuscular blocking agents is required, as these drugs reduce the signal obtained by EMG facial nerve monitors. A single dose of an intermediate-acting agent (e.g. atracurium 0.5 mg kg⁻¹) can be given at induction of anaesthesia, provided that this initial dose is matched carefully to the patient's lean body weight and further ‘top up’ doses are avoided. The effects will typically subside before facial nerve monitoring is required, which is usually at least 30–45 min from induction of anaesthesia. Quantitative neuromuscular monitoring at the ulnar nerve, calibrated at induction, is recommended to confirm this. Although full return of neuromuscular function is desirable, a small degree of residual block (e.g. a train-of-four count of 3 or 4) does not appear to affect the ability of the facial nerve monitor to function.¹³ An alternative option is to avoid neuromuscular blockade entirely; however, neuromuscular blockade-free intubation techniques in adults can increase the risk of haemodynamic instability, poor intubating conditions, and postoperative airway discomfort.¹⁴

Nitrous oxide is typically avoided because of its ability to diffuse into the non-compliant middle ear cavity faster than nitrogen diffuses out. This is of little relevance during surgery, as the cavity will be open to air, but once closed a positive pressure will result. The ability of the Eustachian tubes to equalise pressure differences between the middle ear and atmosphere appears limited during GA, even in those without middle ear pathology.¹⁵ When nitrous oxide is discontinued, diffusion back into the blood can result in sub-atmospheric middle ear pressures; these pressure changes can potentially compromise an ossicular chain repair or tympanoplasty. Of note, these pressure changes also occur to a lesser extent when volatile anaesthetics are used without nitrous oxide, with desflurane causing greater changes compared with isoflurane.¹⁶

Total i.v. anaesthesia (TIVA) techniques are frequently used for middle ear surgery and give an enhanced quality of recovery. When a target-controlled infusion of remifentanil is used, extubation can be smooth, minimising the effect of coughing on middle ear pressure. In cochlear implant surgery, TIVA does not suppress the ESRT, unlike volatile anaesthetics, which do so in a dose-dependent manner leading to overestimation of the ESRT. TIVA is therefore preferred when the ESRT is to be measured.¹⁷ The type of anaesthesia has no effect on other electrophysiological measurements such as NRT.

Complex middle ear surgical procedures may take several hours, so patients should arrive at the operating theatre with an empty bladder and excessive administration of i.v. fluids should be avoided. Temperature maintenance is easily achieved with a full body forced-air warming device, and mechanical thromboprophylaxis measures should be used. Non-invasive arterial blood pressure measurement is sufficient in most cases, but it is worthwhile placing the cuff on the contralateral arm, as its repeated inflation and deflation may otherwise interfere with fine surgical tasks. In some procedures (e.g. stapedectomy), a portion of a superficial vein is taken as a graft, so venous cannulation should be undertaken at a different site to that selected by the surgeon.

Optimisation of the surgical field

Bleeding into the operative field can compromise middle ear surgery by making identification of anatomical structures difficult and is a risk factor for cholesteatoma reccurence.¹⁸ Bleeding occurs predominantly from the microvasculature, so arterial pressure, small vessel tone, and venous pressure are all important factors that affect the degree of bleeding.

Induced arterial hypotension has been used to improve operative conditions, but even profound hypotension does not always guarantee a good surgical field. This was demonstrated in an early case series where systolic arterial pressures of 30–70 mm Hg were induced, and a number of patients still had troublesome bleeding.¹⁹ Conversely, normal arterial pressures do not always preclude an adequate field, and the surgeon's satisfaction with the field does not necessarily correlate with the systolic arterial pressure.²⁰ For these reasons the need for, and safety of, induced arterial hypotension has been a matter of ongoing debate. This is particularly so for procedures such as middle ear surgeries, where blood loss is minimal and there is no secondary gain of reducing transfusion requirements. A pragmatic approach is therefore to aim initially for a MAP of approximately 80% of the patient's baseline MAP. If the surgical field is not adequate, the blood pressure can be reduced further in suitable patients. There are many methods available for inducing arterial hypotension; these are outlined in Table 2.

Table 2.

Common pharmacological methods for induced arterial hypotension. ICP, intra cranial pressure; MAC, minimum alveolar concetration; NMB, neuromuscular blockade; PONV, postoperative nausea and vomiting; TCI, target-controlled infusion.

Drug/technique	Mechanism of action	Suggested starting dose for adults	Advantages	Disadvantages
Increasing volatile agent concentration	Vasodilation Decreased myocardial contractility	Increase end tidal concentration by 0.2–0.3 MAC	Decreases cerebral oxygen requirements	Prolongs emergence Increases cerebral blood flow and ICP at high concentrations Peripheral vasodilation may result in increased bleeding
Remifentanil	Profound analgesia (μ receptor agonist) Reduction in cardiac output by decreasing heart rate	Increase plasma site target concentration by 1–2 ng ml−1	Aids tracheal tube tolerance without NMB Short half life Reduces concentration of volatile or propofol required for anaesthesia Can be continued during emergence for smooth extubation	Risk of bradycardia Small risk of opiate induced chest wall rigidity (rare in adults) Postoperative hyperalgesia (controversial)
Increasing concentration of TCI propofol	Vasodilation Decreased myocardial contractility	Increase plasma or effect site target concentration by 0.5–2 μg ml−1	Reduces ICP Maintains cerebral flow–metabolism coupling	Prolongs emergence Decreases cochlear blood flow (unlike sevoflurane)
Esmolol	Reduced heart rate (cardioselective β1 receptor antagonist)	Slow bolus of 0.5 mg kg−1 followed by infusion of 50–200 μg kg−1 min−1	No effect on emergence time Quick onset and offset	Bronchoconstriction in susceptible patients Negative inotropic effects at high doses Needs to be given by infusion
Labetolol	Vasodilation via α1 antagonism Reduction in heart rate via β antagonism (non-selective)	Slow bolus of 10–20 mg Can be repeated if needed after 5 min	No effect on emergence time Relatively quick onset and offset	Bronchoconstriction in susceptible patients Risk of myocardial failure at high doses Longer onset and duration than esmolol β2 mediated vasodilation
Magnesium	Non-competitive N-methyl-d-aspartate (NMDA) receptor antagonist Inhibits presynaptic release of acetylcholine at the neuromuscular junction	Slow bolus of 50 mg kg−1 over 5 min Can be followed by infusion of 15 mg kg−1 h−1 if required	Improves postoperative analgesia Reduces shivering Reduced PONV Reduces anaesthetic agent requirements Bronchodilator	Slower onset and can be difficult to titrate compared with some other agents Delayed emergence Potentiates neuromuscular blockade Postoperative hypotension
Dexmedetomidine25	α2 adrenoceptor agonist	Infusion of 0.5 μg kg−1 h−1	Decreases PONV Maintains cerebral flow–metabolism coupling Decreases anaesthetic requirements Can be continued during emergence for smooth extubation	Expensive Not widely available in the UK and unlicensed for use outside critical care units May cause bradycardia (contraindicated in patients with second or third-degree heart block) Potential for hypertension during loading and cessation Can cause drug-induced hyperthermia

A low upper body venous pressure can be aided by avoiding the use of tracheal tube ties, extreme lateral rotation of the head, and excessive administration of i.v. fluids. A 10° reverse Trendelenburg position is also beneficial for this reason, and has the added benefit of reducing arterial pressure to the ear by 2 mm Hg for each 2.5 cm of vertical height above the heart.²¹ Positive pressure ventilation will raise intrathoracic pressure and thus impede venous return, but this can be limited by shortening inspiratory times and using a low or no-PEEP setting where possible.

Overall though, positive pressure ventilation is beneficial to surgical conditions, as it avoids the moderate hypercapnia that occurs in patients breathing spontaneously. Acute hypercapnia causes an increase in sympathetic nervous activity, which can raise arterial pressure; it can also cause an acidosis, which leads to vasodilatation in most vascular beds. Controlled ventilation, aiming for an arterial CO₂ concentration of approximately 4.5–5 kPa, will therefore attenuate both of these effects. It should be noted though that after the induction of controlled arterial hypotension, end-tidal CO₂ tension tends to decrease, because of an increase in ventilation/perfusion mismatch, whereas arterial CO₂ stays relatively unchanged. This is of practical relevance, as attempting to normalise the end-tidal CO₂ concentration by reducing minute ventilation will result in a higher arterial CO₂ than anticipated. In general, therefore, the same ventilation settings should be maintained as the MAP decreases.

The particular anaesthetic agent used may also have an effect on the quality of the surgical field. One study showed that operating conditions were superior when a propofol–remifentanil TIVA technique was used, compared with an isoflurane–nitrous oxide–fentanyl technique.²²

Postoperative nausea and vomiting

Middle ear surgery is associated with a high incidence of postoperative nausea and vomiting (PONV). Contributory factors include surgery in younger patients; longer procedure times compared with other surgeries; direct stimulation of the vestibular system by drilling adjacent to the inner ear; and suction–irrigation (a caloric vestibular stimulant).

Avoidance of PONV is desirable, as it is unpleasant for patients and can potentially affect the surgical repair if there is ongoing retching or vomiting. Ondansetron and dexamethasone are both effective prophylactic agents, and efficacy is increased when they are given in combination.²³ TIVA also reduces early PONV.²² Prescription of a rescue antiemetic of a different class, such as cyclizine, is recommended for the postoperative period. In patients suffering from vestibular symptoms, which are not uncommon after cochlear implantation or middle ear surgery, betahistine (a structural analogue of histamine with H₁ agonist and H₃ antagonist activity) may also be useful.²⁴

Postoperative management and pain relief

When operating under GA, surgeons will typically infiltrate an adrenaline-containing local anaesthetic solution, with the aim of limiting bleeding and aiding postoperative analgesia. This, combined with paracetamol, NSAIDs (if appropriate), and oral opioids as required, usually provides good postoperative analgesia. Topical local anaesthetics can also be applied. One such measure practiced at the authors' institution is to pack the external ear canal with absorbable haemostatic material (Spongostan™ Special, Ferrosan Medical Devices, Soeborg, Denmark) soaked in levobupivacaine 0.5%.

Other possible minor postoperative complications include a change in taste and transient facial palsy (the incidence of which may be increased by the infiltration of high volumes of local anaesthetic drugs). Rare complications include cerebrospinal fluid leak, meningitis, and persistent facial nerve palsy. The majority of patients are discharged home on the day of surgery, or after a short hospital stay of 12–24 h.

Summary

The ideal anaesthetic for middle ear surgery is one that optimises the surgical field without excessive arterial hypotension and allows for intraoperative monitoring of facial nerve function. In addition, it should minimise the chances of excessive coughing on emergence from anaesthesia and PONV. A successful surgical outcome is helped if these aims are achieved.

Declaration of interest

The authors declare that they have no conflicts of interest.

MCQs

The associated MCQs (to support CME/CPD activity) will be accessible at www.bjaed.org/cme/home by subscribers to BJA Education.

Acknowledgements

The authors thank Mr S.D. Rejali, consultant ENT surgeon, University Hospitals Coventry & Warwickshire NHS Trust and Mr C. Coulson, consultant ENT surgeon, University Hospitals Birmingham NHS Foundation Trust for providing the clinical figures and advice offered in the preparation of this manuscript. We are also grateful to Mr Jason McAllister, Graphic Designer, University Hospitals Coventry & Warwickshire NHS Trust for help with the illustrations.

Biographies

Charles Pairaudeau FRCA is a specialty registrar in anaesthesia at University Hospitals Coventry and Warwickshire NHS Trust.

Cyprian Mendonca MD FRCA is a consultant anaesthetist at University Hospitals Coventry and Warwickshire NHS Trust and honorary associate professor at Warwick Medical School. He has interests in medical education, difficult airway management and anaesthesia for ENT and maxillofacial surgery.

Matrix codes: 1A02, 2A03, 3A02

Appendix A. Supplementary data

The following is/are the supplementary data to this article:

Supplementary Fig 1.

Nerve supply to the ear (courtesy of Mr J McAllister). 1. Branch of glossopharyngeal nerve 2. Branch of vagus nerve 3. Auriculotemporal nerve 4. Greater auricular nerve.

Supplementary Fig 2.

Blood supply to the middle ear (courtesy of Mr J McAllister). A. Right lateral view of the skull. B. Illustration to demonstrate blood supply to the middle ear. 1. External carotid artery 2. Posterior auricular artery 3. Internal jugular vein 4. Stylomastoid artery 5. Posterior tympanic artery 6. Facial Nerve 7. Stapedius muscle 8. Superior tympanic artery 9. Tensor tympani muscle 10. Middle meningeal artery 11. Anterior tympanic artery 12. Maxillary artery 13. Inferior tympanic artery. 14. Ascending pharyngeal artery.