15Principles of Complication Avoidance in Ear Surgery

FThe deficit you must never trade for hearing

Ossiculoplasty is an elective operation done to improvehearing in an ear that, in most cases, still works. That single fact governs everything about how the operation should be approached. The patient arrives with a conductive deficit and a moving face, a tasting tongue, and an inner ear that — whatever the air–bone gap — usually carries their useful hearing. The cardinal sin of the operation is to hand them back a smaller air–bone gap at the price of a deficit they did not have when they walked in: a facial palsy, a dead or worse-hearing ear, or a numb, metallic tongue. A brilliant reconstruction built over a new deficit is not a success; it is a different, and worse, injury.

Complication avoidance in ear surgery is therefore not a separate skill bolted on at the end. It is the frame around the whole procedure. The surgeon who reconstructs well but injures the facial nerve has failed; the surgeon who accepts a slightly imperfect prosthesis angle to keep the nerve and the labyrinth intact has succeeded, even if a second sitting is needed. Three structures sit close enough to the field to be injured by ordinary manoeuvres — the facial nerve, the inner ear, and the chorda tympani— and the entire discipline is built around protecting them [2016].

FThree neighbours, three deficits

It helps to name, plainly, what each structure does and what its injury costs. The facial nervemoves the face; for much of its course through the middle ear it is separated from the instrument by no more than a wafer of bone, and often by nothing at all. Its injury produces a facial palsy that the patient sees in every mirror — a private deafness is bearable in a way that a public palsy is not. The inner ear — cochlea and vestibule — carries the very hearing the operation is meant to improve, plus the sense of balance; its injury ranges from a high-frequency dip to a completely dead ear, often with disabling vertigo. The chorda tympani carries taste from the front two-thirds of the tongue across the upper middle ear, directly across the surgeon’s line of access; its injury gives altered taste, loss of taste, or an unpleasant metallic sensation.

Each of these injuries has its own characteristic mechanism in ossicular surgery, and each has its own simple, non-negotiable rule for avoidance. The rest of this module takes them in turn.

TProtecting the facial nerve

The relevant part of the facial nerve is the tympanic (horizontal) segment, which runs along the medial wall of the cavity to form the roof of the oval window — the exact destination of a total prosthesis. The bony canal that should cover it is naturally absent (dehiscent) in a substantial minority of ears, classically over the oval window, and is far commoner in cholesteatoma, in revision, and in long-standing disease. In those ears the nerve may lie bare and scarred into the disease the surgeon has come to clear. The commonest mechanism of iatrogenic palsy is mundane: a curette or a confidently advancing burr in the oval window niche, used without first confirming whether the white bony canal is intact or whether a soft pink tube of unprotected nerve lies there instead [2023, 2016].

The protective principles are old and few. Orient before you dissect: identify the lateral semicircular canal, cochleariform process, oval window and the line of the nerve before any instrument approaches the medial wall — disorientation, not bad luck, underlies most injuries [2016]. Match technique to the bone you find: over an intact canal a curette may be used with care; over a dehiscent or prolapsed nerve it must not, and is replaced by sharp tangential dissection that lifts disease off the nerve rather than pressing toward it. Forbid diathermy and sustained pressure on or beside the nerve. And keep the burr running parallel to the nerve, never end-on, with the lowest workable speed and copious irrigation. Above all, be willing to compromise the reconstruction for the nerve: accept a less-than-ideal prosthesis vector, choose a partial reconstruction onto a mobile stapes, or stage the operation rather than work a curette against an exposed seventh nerve [2016, 2023].

TProtecting the inner ear

The inner ear can be injured without ever being opened. The most underappreciated route is acoustic trauma from the drill. Classic measurements showed that every use of the burr exposes the operated cochlea to noise of around 100 dB and the opposite cochlea only 5–10 dB less — enough to account for some of the high-frequency sensorineural losses seen after tympanoplasty, and detectable even in the non-operated ear [1976, 2009]. The single most dangerous version of this is contact between a spinning burr and an intact, sound-conducting ossicular chain, which couples that energy straight into the labyrinth: a momentary touch of the burr on the incus can cause a permanent high-tone loss. The defences are to keep the burr off the chain, to disarticulate or protect the chain before drilling near it, to use lower speeds and diamond rather than cutting burrs as the inner ear is approached, and to irrigate generously to damp both heat and noise.

The second route is mechanical injury at the footplate. An over-long or over-tight prosthesis distends the annular ligament and loads the footplate, raising cochlear input impedance and risking subluxation, fistula and vertigo; the prosthesis must be measured to seat with a light press-fit, not jammed. Worse, undue force can fracture the footplate or tear the annular ligament, opening a perilymph leak directly into the vestibule. This is an intraoperative emergency: the reconstruction is abandoned, pressure is taken off the footplate, and the oval window is sealed at once with fascia, fat or perichondrium, with the ossiculoplasty staged for later. A “cartilage shoe” over the footplate can buffer a total prosthesis and protect against direct trauma [2016]. The guiding rule is constant: never force anything onto the footplate, and never apply diathermy at the oval window, where heat passes straight to the perilymph.

TThe chorda tympani: a quieter loss

The chorda tympani is the structure surgeons are most tempted to dismiss, and patients least expect to be warned about. It loops up into the middle ear and runs forwards between the incus and the malleus, crossing the upper mesotympanum directly along the line of access to the oval window. To reach the posterior ossicular chain the surgeon must work past it, and it is easily stretched against a retractor or divided by a slip of the instrument. The resulting deficit — altered or absent taste over the front of the tongue, and frequently a lingering metallictaste — is genuinely distressing to some patients, even if it is invisible to onlookers [2007].

Reassuringly, most dysgeusia is transient. In a prospective cohort, taste disturbance affected nearly half of patients at ten days but had fallen to under a quarter by four months and under one in ten by a year — the time course shown below. The nuance that matters for technique is that an initially healthy nerve that is deliberately divided tends to leave more lasting symptoms than one that is preserved, whereas a nerve already involved in disease may actually do worse if it is stretched than if it is cut cleanly [2024, 2007].

The practical policy follows from that evidence. Preserve the chorda when you safely can, mobilising it gently rather than holding it on stretch. If a healthy nerve genuinely obstructs safe reconstruction, divide it cleanly rather than leave it traumatised by traction. And — the part most often forgotten — counsel the patient beforehand that taste change is common, usually temporary, and occasionally lasting, so that an expected symptom does not become an unexpected grievance.

CThe discipline: habits, not heroics

Complication avoidance is not a set of rescue manoeuvres for the moment things go wrong; it is a set of habits practised when everything is going right. Five recur across every safe ossiculoplasty. First, operate in a dry, well-exposed field: a bloodless field and adequate canaloplasty are themselves safety measures, because injuries happen in blood-obscured niches and through inadequate access. Second, orient continuously to the fixed landmarks — lateral semicircular canal, cochleariform process, oval and round windows — so the nerve and the labyrinth are always located, never assumed. Third, respect the medial wall: tangential sharp dissection, no curette over dehiscent nerve, no diathermy or pressure at the oval window, and the burr always off the chain. Fourth, measure rather than force the prosthesis, accepting a partial or staged result over a dangerous one.

The fifth habit is the willingness to stop and stage. The decision to leave an ear with imperfect hearing rather than risk a permanent deficit is the mark of a mature otologist, not a timid one. A footplate fracture, an unexpectedly bare nerve, an ear that will not stay dry — each is a reason to seal, simplify and return another day. Adjuncts support this discipline but never replace it. Intraoperative facial nerve monitoring is a cost-effective aid that helps localise the nerve and warns when it is stimulated, and is especially valuable in revision and cholesteatoma surgery; but it detects stimulation, not safety, and a stable trace must never be read as licence to push, because a sharp transection can occur with almost no warning [2003]. Preoperative imagingflags the high-risk ear — the previously operated, the cholesteatomatous, the ear with a fistula — but cannot be trusted to exclude a dehiscent canal, so the operative finding always overrides the scan [2023, 2016]. The safest surgeons are not those with the best images or the cleverest gadgets, but those who assume the nerve is bare, keep the burr off the chain, never force the footplate, and are content to leave an ear with an honest result and an intact face.