CI Atlas · Candidacy & Evaluation · Module 02

2How the criteria evolved

Who counts as a cochlear-implant candidate has changed more in forty years than almost anything else in the field. When multichannel implants arrived, candidacy was an almost absolute bar: only adults with total or near-total deafness — pure-tone averages of 100 dB or worse and no open-set speech understanding even with the most powerful hearing aids — were considered. Every revision since has loosened it. As implant outcomes climbed past what poorly-aided ears could achieve, the audiometric threshold fell, the amount of residual aided speech a candidate could keep rose toward half or more, and the whole field moved away from hearing sensitivity as the sole criterion toward a richer measure of functional hearing. This module traces that widening, because understanding it is the key to recognising a modern candidate.

FThe starting line

In the early 1980s the rule was strict: a candidate needed a total or near-total sensorineural loss — pure-tone averages of 100 dB or greater, amplified thresholds that failed to reach 60 dB, and no open-set speech recognition with powerful, best-fit hearing aids. The implant was a last resort for those who had essentially nothing to lose.

Candidacy began as a near-absolute bar: only those with no usable open-set speech through powerful aids qualified. As implant outcomes climbed past those of poorly-aided listeners, the audiological criterion was relaxed — first to limited aided discrimination, then to up to ~50% (a large UK study showed implant benefit differs little up to that level), and for selected devices ~60% aided sentences. The field also shifted from sensitivity alone toward functional hearing. Each loosening let in patients who would once have been turned away. Schematic.

FTThe bar comes down

As experience accumulated, mean speech scores with implants kept exceeding the aided results of people with lesser impairments. So the criteria expanded — first to allow limited aided discrimination, then up to about 50% sentence understanding, and for selected devices around 60%. A large UK study was pivotal: it showed implant effectiveness differs little between candidates with no aided speech discrimination and those scoring up to 50%, which justified relaxing the line.[2004]

CFrom sensitivity to function

Just as important as the numbers was a change of kind. Early criteria leaned on hearing sensitivity alone — the pure-tone average. The field moved toward metrics of functional hearing: what the patient can actually do with speech, in the best-aided condition, at realistic levels. The audiogram sets the scene, but aided speech understanding became the decisive measure — the subject of the next two modules.

CWhy it keeps moving

The boundary is still moving outward. Better electrodes and processing, hearing-preservation surgery, and a steady accumulation of evidence that patients outside the old criteria still benefit have pushed candidacy toward less severe losses and into new categories entirely — single-sided deafness, electric-acoustic, auditory neuropathy (Module 9).[2021] The practical lesson for the clinician is humility about yesterday's cut-offs: a patient who “would not have qualified” a decade ago may well qualify today.

TCA patchwork of labels and payers

“The criteria” are not one line but a moving patchwork. In the US the three manufacturers' FDA labels differ from one another— Cochlear's adult criterion is the most permissive (it accepts a sloping high-frequency loss), Advanced Bionics requires severe-to-profound loss, and MED-EL specifies a 70 dB three-frequency average — so teams in practice work to the least restrictive label across devices. Those labels are guidance, not mandate: off-label implantation is legal on an evidence base, though it is then not reimbursable by CMS-regulated plans (Module 6).[2020] International CE-marked criteria are looser and often word-based, explicitly including single-sided deafness and prelingual or perilingual adults the US tables do not foreground. The test materials moved too — from HINT sentences (which ceiling) toward AzBio sentences and CNC words.[2017] The upshot: a patient can qualify under one device or insurer and fail under another.

There is no single candidacy line. US FDA labels differ by manufacturer — Cochlear's adult criterion is the most permissive, MED-EL's the strictest — and teams effectively adopt the least restrictive label across devices. CMS sets its own, tighter coverage criterion, and European CE-marked labels are looser and often word-based, explicitly including single-sided deafness and prelingual adults. So the very same best-aided score can make a patient a candidate under one device or payer and a non-candidate under another. Thresholds shown are illustrative and change over time. Schematic.

Case 11.2 · 'He wouldn't have qualified before'

A patient with a 55% best-aided sentence score is referred. A trainee recalls that years ago candidacy required essentially no open-set speech, and assumes the patient cannot qualify.

How should the trainee's assumption be corrected?

Self-assessment — Module 22 questions

Question 1 · Foundation

How have cochlear-implant candidacy criteria changed since the early 1980s?

Question 2 · Trainee

Beyond the numbers, what conceptual shift accompanied the loosening of criteria?

Tracked locally in your browser — see /progress for the dashboard.