12Immediate Postoperative Complications

FThe early window: what can go wrong, and when

Most ossiculoplasties are uneventful, and the great majority of patients go home the same day with nothing more than a packed ear and an instruction to keep it dry. But the first days after surgery are when the characteristic complications of middle-ear work declare themselves, and recognising them early — and knowing which are benign and which are warnings — is part of safe practice. The structures at risk are precisely those that crowd the few cubic millimetres of the operative field: the facial nerve running through the medial wall, the chorda tympani arcing unprotected across the cavity, the inner ear a fraction of a millimetre beyond the stapes footplate, and the graft and prosthesis that have just been assembled and are not yet healed into place.

It helps to think in terms of timing. Some problems appear at once — bleeding, an immediate facial weakness, vertigo on first moving. Others surface over days, such as a delayed facial palsy or the gradual realisation of a taste disturbance. And a few — prosthesis extrusion, graft failure, fibrosis — begin in this early window but only become obvious weeks later. This module concentrates on the immediate events: how to tell a benign postoperative symptom from a true complication, and what to do first. Across all of them the same principle recurs: a symptom that is mild and settling is usually reassurance, while one that is severe, persistent, or progressive is an alarm that calls for senior assessment.

FBleeding and the early wound

In absolute terms the blood loss of ossiculoplasty is trivial — this is microsurgery, not major surgery — and serious postoperative haemorrhage is rare. The bleeding that matters intra-operatively is the capillary ooze that, invisible to the naked eye, fills the magnified field and obscures the very structures the surgeon must judge. That is why a bloodless field is engineered before and during the operation with head-up positioning, adequate anaesthetic depth, topical vasoconstrictors and, when needed, controlled hypotension. After surgery, the relevant questions become different: is there reactionary bleeding from the canal or the post-auricular wound, and is the ear forming a haematoma?

A small amount of blood-stained dischargefrom a packed canal in the first day or two is expected and self-limiting. The patient should be warned of it so it does not cause alarm. What is not expected is brisk fresh bleeding, an expanding post-auricular swelling, or a tense haematoma under the flap — the last of these can lift a graft, harbour infection, and threaten the reconstruction, and may need evacuation. In practice, meticulous haemostasis at closure, a well-applied dressing, and avoidance of straining, nose-blowing and heavy lifting in the early days are what keep the wound quiet. Patients on antiplatelet or anticoagulant drugs deserve specific perioperative planning, since they tilt the balance toward both intra-operative ooze and postoperative bleeding.

FTaste disturbance and the chorda tympani

The chorda tympani nerve carries taste from the anterior two-thirds of the ipsilateral tongue and parasympathetic secretomotor fibres to the submandibular and sublingual salivary glands. It branches from the facial nerve, then loops forward across the middle ear, passing between the long process of the incus and the handle of the malleus — right through the working space of ossiculoplasty — before leaving anteriorly. Crucially, over this course it has no bony covering, so it is exposed to retraction, desiccation, heat and division during dissection. It is, in short, one of the most frequently injured nerves in all of surgery [2012].

Injury produces a metallic or altered taste, reduced taste (hypogeusia) or its loss (ageusia) on one side of the tongue, and sometimes a sensation of oral dryness or a burning tongue. Patients often find it more bothersome than clinicians anticipate. The reassuring news is that most early disturbance improves: in a prospective cohort, taste disorders were reported by 42.7% at ten days, 23.3% at four months, and only 9.2% at one year [2024]. A counter-intuitive but well-established teaching point is that deliberate, clean division of a chorda that is genuinely in the way is often tolerated better than leaving it stretched and traumatised; persistent dysgeusia is more common after a previously healthy nerve is transected than after it is preserved, so the nerve should be protected when it can be, and divided cleanly when it cannot [2012, 2024]. The trajectory of recovery is shown below.

TFacial palsy: immediate versus delayed

Facial weakness is the complication every otologist fears, and the single most useful discriminator is timing. The facial nerve runs through the operative field along the medial wall of the middle ear, and the most vulnerable part is the tympanic (horizontal) segmentas it passes above the oval window and the cochleariform process, where the bony canal is often dehiscent. In the classic series of iatrogenic injuries, mastoidectomy was the commonest procedure but tympanoplasty accounted for a notable minority, and — the sobering figure — in 79% of cases the injury was not recognised at the time of operation [1994]. That is exactly why disciplined postoperative recognition matters.

An immediate palsy seen in the recovery room has two very different explanations. The benign one is local-anaesthetic diffusion onto the nerve: a temporary chemical block that resolves within a few hours as the agent clears. The serious one is intraoperative nerve injury— stretch, heat from a drill or laser, or partial or complete transection. The two are separated by waiting: if a dense palsy persists once any local anaesthetic has clearly worn off, structural injury must be assumed, the operative findings reviewed, and prompt re-exploration and repair or decompression considered. A delayedpalsy is a different animal altogether. It appears days after an uneventful operation — mean onset around eight days — and is attributed to oedema in the fallopian canal or reactivation of latent herpes or varicella virus in the geniculate ganglion. The nerve is structurally intact, and the prognosis is excellent: in a meta-analysis the overall incidence was about 0.65% and more than 95% of patients recovered completely [2019]. Management is therefore conservative — document the House–Brackmann grade [1985], give corticosteroids with antiviral cover, protect the eye, and observe. The triage tool below sets the three scenarios side by side.

Two practical safeguards reduce the chance of an unrecognised injury. The first is documenting facial function in recovery, every time, so that any deviation is noticed against a known baseline. The second is keeping neuromuscular blockade light or absent when facial-nerve monitoring is in use, so the warning signal is preserved during dissection. When weakness is found, grading it formally on the House–Brackmann scale gives a reproducible anchor for follow-up and for deciding whether function is recovering or deteriorating [1985].

TVertigo and the threatened inner ear

Some unsteadiness after middle-ear surgery is normal. Handling the ossicular chain transmits energy to the inner ear, the labyrinth is stimulated, and many patients feel briefly dizzy or nauseated in the first hours; this usually settles within a day. The art lies in separating that benign, settling vertigo from the kind that signals an inner-ear emergency. Severe, persistent vertigo with spontaneous nystagmus, especially when accompanied by a new sensorineural drop in hearingor tinnitus, is not something to wait out — it demands urgent assessment [2023].

Several mechanisms produce this alarming picture. An over-long prosthesis intruding into the vestibule contacts and stimulates the saccule, producing sound-induced vertigo and, if it pistons against the inner ear, sensorineural loss. A perilymph fistula — a leak around a footplate that has been disturbed or fenestrated — causes fluctuating hearing and positional vertigo. Direct labyrinthine injury, from excessive force on a stapes that turns out to be fixed or from violating the footplate, is the most feared, since it can cause a permanent dead ear. The general lesson is mechanical restraint: never force a fixed ossicle, size the prosthesis so it couples without intruding, and treat a triad of vertigo, nystagmus and sensorineural loss as a reason to act rather than to observe. Postoperative nausea and vomiting deserves a mention here too, both because vestibular stimulation makes ear surgery a high-risk operation for it and because retching raises middle-ear pressure and can disturb a fresh reconstruction; multimodal antiemetic prophylaxis is routine.

CEarly graft and prosthesis failure

The reconstruction itself can fail early. Graft failure — a tympanic membrane graft that does not take, lateralises, or blunts at the anterior angle — is favoured by infection, poor eustachian-tube function and negative middle-ear pressure, haematoma under the graft, and technical issues such as inadequate anterior support. A hostile, infected or poorly aerated middle ear is the recurring villain, which is why a quiescent, dry ear is sought before definitive surgery and why a haematoma or early infection in this window is treated seriously. The clinical clues are persistent discharge, a new perforation, or a graft that looks unhealthy at the first dressing change.

Prosthesis extrusion begins as an interface problem in this early period even though it usually completes later. A rigid prosthesis head loading the drum directly concentrates pressure on the epithelium, which can erode and let the prosthesis migrate laterally; the telltale sign is a small perforation with the white head presenting at the drum surface. Reported extrusion of titanium prostheses averages around 5%, with a wide range, and is notably higher in children[2023]. The single most effective preventive measure is to interpose a cartilage shield between the prosthesis head and the tympanic membrane, which spreads the load and lowers extrusion to roughly 1–2%; it is now standard practice for exactly this reason [2023]. The comparison below makes the effect plain.

Prosthesis displacementis the other early mechanical failure: a prosthesis that is too short, too loose, or poorly coupled can slip off the stapes capitulum or footplate, presenting as an unchanged or worsened air–bone gap once swelling and packing have settled. Both extrusion and displacement teach the same lesson about technique — stable coupling, the right length and tension, cartilage protection at the drum, and a healthy interface — and both are reasons the early postoperative review and the first audiogram are not mere formalities but the moment when an under-performing reconstruction is caught. Set against the rarity of catastrophic events, these interface failures are the everyday quality issues of ossiculoplasty, and attention to them in the early window is what separates a good hearing result from a disappointing one.