13Reading the Evidence: How Well Do Other Implantable Hearing Devices Work?

FWhat we measure, and why a single number misleads

Every implantable hearing device is judged on the same family of outcomes, but each measures something different. Aided thresholds tell us how soft a sound the device makes audible; functional gain is the difference between the unaided and aided threshold, a direct measure of how much louder the world has become. Neither, however, tells us whether speech is understood. For that we need speech-recognition scores, measured separately in quiet and in background noise, because a device can restore audibility yet leave a patient struggling the moment a room becomes busy.

Two further outcome domains matter for specific indications. Sound localisation and speech-in-noise with spatially separated talkers test the binaural benefit that only matters when the goal is to restore a missing ear, as in single-sided deafness or bilateral fitting. Finally, self-report questionnaires capture the lived experience: daily device use, listening effort, and disease-specific quality of life. A device can post a handsome functional gain and still sit in a drawer, which is why subjective and usage outcomes are reported alongside the audiogram, not instead of it.[2022][2022]

TThe pooled numbers: functional gain and speech across device classes

For active transcutaneous bone-conduction implants the meta-analytic picture is reasonably consistent. Pooled functional gain in conductive and mixed losses sits around 28 to 32 dB, lifting many users from a moderate handicap into the near-normal range, with word-recognition scores improving substantially in quiet. A single-centre cohort of the same device class reported sound-field thresholds falling from roughly 60 dB HL before surgery to the low-to-mid 30s dB HL afterwards, a gain that remained stable beyond a year. Active devices appear to outperform passive transcutaneous designs particularly in the high frequencies, where skin attenuation most penalises a passive magnetic link.

Active middle-ear implants such as the floating-mass-transducer device show the same pattern from a different starting point. In sensorineural loss they significantly improve speech in noise and sound quality for patients who could not tolerate conventional aids, and coupling the transducer to the round or oval window extends the same benefit to conductive and mixed losses. Reported word-recognition gains are large in selected cohorts, but the comparator is usually the unaided ear or the patient’s own failed hearing aid, not a head-to-head device trial. The complication rate across these active classes is reassuringly low, typically in the low-to-mid teens of percent and dominated by minor events.[2019][2022][2024][2022][2017]

CHead-to-head data where it exists: the single-sided-deafness story

The cleanest comparative evidence in this whole field comes from single-sided deafness, where a randomised controlled trial pitted cochlear implantation against a bone-conduction device and a contralateral-routing-of-signal aid in the same patients. The result reframes how we counsel: only the cochlear implant improved sound localisation and reduced tinnitus burden, because only it restores genuine binaural input. The bone-conduction device and the routing aid relieve the head-shadow effect by moving sound across to the better ear, but they cannot recreate the interaural timing and level cues the brain needs to place a sound in space.

This trial is the exception that proves the rule. For most device choices we do not have randomised head-to-head data; we have registry series and single-arm cohorts, each enrolling a different mix of conductive, mixed and single-sided patients. The practical lesson is to match the outcome to the goal. If the goal is simply to make a deaf ear audible across the head, a bone-conduction device or routing aid suffices and the localisation deficit is expected. If the goal is to rebuild spatial hearing, the evidence points past these devices toward implantation of the deaf cochlea itself.[2021][2022]

CThe limits of the evidence, and how to counsel honestly

Most of what we know about these devices rests on small, single-centre, retrospective cohorts with short follow-up and no blinding. Inclusion criteria vary wildly: one series of an active bone-conduction implant might mix conductive, mixed, malformation and single-sided cases together, making the pooled functional gain a weighted average of clinically distinct problems. Publication bias favours the successful series, and the comparator is usually no device or a failed hearing aid rather than the realistic alternative the patient is actually choosing between. Heterogeneity, not effect size, is the dominant limitation.

Honest counselling therefore frames outcomes as ranges, not promises. Tell the patient that an active bone-conduction or middle-ear implant will very likely restore audibility and improve speech in quiet, that satisfaction and daily-use rates in published series are high, and that serious complications are uncommon. But be explicit that speech-in-noise gains are more modest and variable, that a single-sided device will not restore the ability to tell where a sound came from, and that the evidence base, while encouraging, cannot promise an individual result. The strongest predictor of a happy outcome remains a correct indication, not the device’s brochure figures.[2022][2022][2017]