5Prognostic Factors Revisited: What Drives Outcome

FThe ear, not the implant

Ask a beginner why one ossiculoplasty hears beautifully and the next, done with the same prosthesis by the same surgeon, barely improves, and the answer usually reaches for the implant— the brand of titanium, the height of the shaft, the material. The accumulated evidence says almost the opposite. The single most important idea in modern ossiculoplasty is that the result is set far more by the ear the prosthesis sits inthan by the prosthesis itself. A reconstruction is only ever as good as the space, the mucosa and the ossicular scaffolding around it. Once a competent prosthesis is chosen, swapping it for a different competent prosthesis rarely changes the outcome; changing the ear’s biology and anatomy changes everything.

This is not a fringe opinion but the consistent finding of the named prognostic systems — Dornhoffer and Gardner’s OOPSindex, Kartush’s Middle Ear Risk Index (MERI), and Black’s SPITEmethod — each of which was derived by asking which variables actually correlate with hearing success across hundreds of operations [2001, 1994, 1992]. They cluster the winners into a handful of groups: the health of the mucosa and the degree of aeration; which ossicular remnants survive, above all the malleus; the disease and surgical history, especially how many times the ear has been operated; and eustachian-tube function. The prosthesis, notably, is not near the top of any of these lists. The explorer below lets you walk that hierarchy and see why each tier sits where it does.

Two non-factors deserve naming up front, because they trip up trainees. The size of the preoperative air–bone gap is not a prognostic factor: it measures the defect to be repaired, not the chance of repairing it, and a large gap in a favourable ear is exactly where the operation offers the most to gain. And patient age in adults, by itself, is a weak predictor compared with the environmental and ossicular variables. Keep returning to the principle: prognosis lives in the combination of dominant ear factors, not in the gap or the implant.

FMucosa and aeration: the dominant pair

At the very top of every list sit the middle-ear mucosa and aeration, and they are really two faces of one requirement: the reconstructed chain must live in a healthy, air-filled, low-impedance space. Mucosa that is oedematous, polypoid or fibrotic does not merely look unhealthy — it physically fixes and loads the prosthesis, tethers it with adhesions, and can re-grow to splint the chain months after surgery. A poorly ventilated cavity is just as damaging: as middle-ear pressure falls, the drum retracts, stiffness rises, and the whole tympano-ossicular system transmits sound less efficiently, collapsing eventually onto the reconstruction and the promontory.

The physics here are unforgiving and explain why aeration outranks the implant. A completely non-aerated middle ear can produce 35–55 dB of conductive loss even when the ossicular chain is anatomically intact, and a minimum of roughly 0.3–0.5 mL of air is needed to keep ossicular coupling within about 10 dB of normal. Mehta and colleagues demonstrated the principle directly in perforated ears: conductive loss rose as the middle-ear and mastoid air volume fell, while being largely independent of where the perforation sat [2006]. The lesson is stark — no prosthesis, however elegant, can outrun a stiff, retracted, poorly ventilated space.

This is also why the surgeon’s most valuable prognostic work often happens beforethe prosthesis is even chosen: clearing disease, restoring a healthy mucosal lining, opening the oval and round window niches, and protecting ventilation pathways. Schuring’s much-quoted prediction — that the future of ossiculoplasty would rest more on solving the “ancillary” problems of eustachian-tube dysfunction, cholesteatoma control and mucosal regeneration than on reconstruction technique — is really a statement about this hierarchy of determinants.

TOssicular remnants and the malleus

After the environment, the surviving ossicular remnants carry the most weight, and one landmark dominates: the malleus handle. When present it gives the prosthesis lateral support, holds it away from the promontory, and aligns the vector of sound transmission toward the footplate, preserving a fragment of the native lever. The effect is large and durable. In long-term multivariate analysis a present malleus made a successful result roughly six times more likely at six months, and at five years it was the only factor that remained an independent predictor of success [2006]. Reconstructing onto the malleus, rather than terminating a prosthesis directly on the drum, is one of the few technical choices that genuinely moves the prognostic band.

The stapes superstructure behaves differently, and this surprises trainees. Its presence or absence dictates the type of prosthesis — a partial replacement (PORP) rests on an intact stapes head, while a total replacement (TORP)must span to the footplate — but it is not a strong independent outcome predictor in its own right. In the OOPS derivation the stapes superstructure status did not independently predict outcome once mucosa, drainage, malleus and surgical history were accounted for [2001]. The familiar finding that PORPs “do better” than TORPs is largely a reflection of the ears they are used in: a titanium series found ABG closure within 20 dB in about 86% of PORP versus 55% of TORP cases, but the difference tracked the superstructure and the disease that destroyed it, not a property of the prosthesis [2010]. The prosthesis title follows the anatomy.

TSurgery count, disease and ventilation

The third dominant cluster is the ear’s history, and the most underrated single variable in it is the number of prior operations. Every previous procedure leaves scar, distorts the anatomy, depletes the ossicular remnants, and — importantly — selects for disease that was hard to control in the first place. Prior surgery was an independent adverse predictor in the OOPS index [2001], and a revision, canal-wall-down ear sits in a measurably worse band than an otherwise identical primary one. OOPS was in fact the first staging system to formally incorporate canal-wall status, a variable the earlier indices had omitted despite its proven significance [2001]. Counting operations is a quick, powerful triage.

Active drainage and infectionform the next tier. A wet ear — graded historically by Bellucci from a dry mucosal ear through to persistent otorrhoea with mucosal disease — both signals ongoing pathology and predicts a less stable result, which is why staging to achieve a dry, healed ear first so often lifts the prognostic band [1973]. Beneath all of this sits the eustachian tube. Chronic dysfunction keeps the ear under-aerated, perpetuates effusion and atelectasis, and is one of the commonest reasons a technically perfect reconstruction drifts back over years — the “E” that Black singled out in SPITE [1992]. These factors compound: the prototype of the prognostically poor case is a revision, canal-wall-down, draining ear with diseased mucosa and a non-functioning eustachian tube, and in such an ear reconstruction alone rarely buys a large hearing gain whatever prosthesis is used.

CWhy prosthesis choice ranks low

If the environment and the ossicular scaffold dominate, where does that leave the prosthesis? Genuinely useful, but interchangeable within a wide competent range. The commercial implants span weights from a few milligrams to over forty and a variety of materials — titanium, hydroxyapatite, porous polyethylene, hybrid composites, and autograft incus — yet comparative studies repeatedly show similar hearing results and extrusion rates across them once they are well coupled and well positioned. When the OOPS, MERI and SPITE systems were built, prosthesis material did not emerge as a dominant independent driver; the “P” in SPITE refers chiefly to prosthesis coupling and stability, not to brand [1992, 1994]. The implant’s job is to bridge the gap competently and then get out of the way of the biology.

This reframes how a clinician should spend effort. The decisions that move outcome are not which catalogue number to order but the ones that change the ear’s band: staging a hostile ear to heal before reconstructing, incorporating the malleus whenever the anatomy allows, restoring aeration and clearing window-niche obstruction, and being honest enough to defer or declinereconstruction in an ear whose environment makes a meaningful gain unlikely. Choosing between two adequate prostheses for such an ear is rearranging the deckchairs. Material does matter at the margins — for extrusion risk it is mitigated by cartilage interposition, and a biologically inert, stable implant is preferable — but it is a second-order decision made after the ear has been placed in its prognostic band.

CReading the indices honestly

The named indices are the practical embodiment of this hierarchy, and they earn their keep by turning a vague sense that “this is a difficult ear” into a comparable, evidence-grounded score. MERI weights otorrhoea, ossicular status, granulation or effusion and prior surgery into a single figure stratified from normal to severe disease, and its revisions have even added factors such as smoking [2001]. OOPS, derived purely statistically, retained only the variables that independently predicted outcome — and notably discarded some “obvious” ones, including the stapes superstructure and cholesteatoma [2001]. Use them to structure planning and counselling and to audit your own results against a common yardstick.

But read them with humility. When OOPS and MERI are tested head-to-head as predictors of an individualear’s hearing result, their discrimination is only modest — receiver-operating-characteristic areas around 0.55 for MERI and 0.64 for OOPS at one year, with OOPS edging ahead but neither approaching a strong predictor [2021]. A score stratifies groupswell and the individual ear only loosely. None of the indices can see the fibrosis, the footplate mobility, the precise quality of the malleus or the live behaviour of the eustachian tube that you assess at the microscope and on follow-up — the very things that sit at the top of the hierarchy. The mature position is therefore to let the dominant determinants — mucosa, aeration, ossicular remnants and surgery count — drive the plan, to use the indices to discipline and communicate it, and to treat the choice of prosthesis as the last and lightest decision in the chain.