New Zealand’s public cochlear implant criteria exclude most people with single-sided deafness because their other ear hears normally or near normally. Stronger evidence now supports a targeted funding pathway—but not implantation for every person with one-sided deafness.

In a quiet room, single-sided deafness can be deceptively easy to underestimate. A person may understand a conversation and appear to hear well.

Move that conversation into a café, workplace, car, classroom or family gathering and the limits become clearer. The person may struggle to separate a voice from background noise, miss speech coming from the deaf side or be unable to work out who is talking. Crossing a road or responding to an alarm can be more difficult when the direction of a sound is unclear.

This is not simply a loss of volume in one ear. It is the loss of the brain’s ability to compare sound arriving from two sides.

New Zealand’s current public funding criteria do not account for that distinction. They require severe-to-profound hearing loss in both ears. As a result, most people with single-sided deafness are excluded before their communication difficulties, tinnitus or ability to function in daily life are assessed.

The case for changing that rule is becoming stronger. It is not, however, a case for cochlear implantation on demand. Surgery carries costs and risks, outcomes vary, and CROS hearing aids or bone-conduction devices remain suitable for some people.

The realistic question is whether New Zealand should establish a separately funded, clinically targeted pathway for people likely to receive meaningful benefit.

New Zealand’s current rule excludes single-sided deafness by design

Disability Support Services currently requires a person to have severe-to-profound hearing loss in both ears, receive insufficient benefit from standard hearing aids and be assessed as likely to benefit from a cochlear implant.

Public funding is not limited to the operation. It includes assessment, the implant and processor, surgery, audiology, rehabilitation, maintenance, replacement devices and processor follow-up intended to support the implant throughout the recipient’s life. Adults are responsible for some batteries, repairs and spare parts. (Disability Support Services)

Single-sided deafness, or SSD, is generally defined as severe-to-profound hearing loss in one ear with normal or near-normal hearing in the other. By definition, most people with SSD cannot meet New Zealand’s requirement for bilateral hearing loss. (JAMA Network)

There is a legitimate reason for caution. Cochlear implantation requires surgical, audiological and rehabilitation capacity as well as long-term public expenditure. Funding an additional group from an unchanged budget could lengthen the wait for people with severe-to-profound loss in both ears, for whom a first implant may provide a substantial gain.

That would not be equitable. Any SSD programme should therefore be additional to—not carved out of—existing funding for conventional cochlear implant candidates.

But protecting access to first implants does not require New Zealand to retain an automatic exclusion forever.

Single-sided deafness is not “half a hearing problem”

Hearing with two ears allows the brain to compare tiny differences in the timing and loudness of sound. Those comparisons help a person locate its source, focus on speech and reduce the effect of competing noise.

With one functioning ear, the head itself can block or soften sound coming from the deaf side. A person can compensate by turning their head, choosing a particular seat, asking others to change places or relying on visual information. These strategies may work in a predictable setting. They are less effective when several people are speaking, voices are moving or the person needs to respond quickly.

Research consistently identifies difficulty understanding speech in noise and locating sound as central effects of SSD. Tinnitus is also common, although not universal. (JAMA Network)

It is therefore inaccurate to describe SSD as a straightforward 50% hearing loss. A single percentage cannot represent the loss of spatial hearing or the extra work required to follow a conversation through one ear.

A quiet-room hearing test also cannot fully reproduce a meeting, busy restaurant or workplace. Funding assessments need to examine how a person performs when speech and noise come from different directions.

What CROS and bone-conduction devices can—and cannot—do

A CROS hearing aid places a microphone on the deaf side and sends sound to the hearing ear. A bone-conduction device transmits sound through the skull to the functioning cochlea.

Both approaches can improve awareness of speech and other sounds coming from the deaf side. A conventional CROS does not require surgery, while bone-conduction systems may be worn externally or implanted. These are valuable options and may be the right choice for someone who does not want a cochlear implant, cannot receive one or obtains sufficient benefit from rerouting sound.

Their limitation is that both sides of the environment are still being delivered to one auditory pathway. They do not provide the brain with separate input from two ears, so sound localisation generally remains poor. (JAMA Network)

Their performance can also depend on where speech and noise are located. A major randomised trial found that CROS and bone-conduction devices helped in some arrangements, particularly when speech came from the deaf side. In another configuration, transferring noise from the deaf side to the functioning ear made speech understanding worse. (JAMA Network)

A cochlear implant takes a different approach. It supplies electrical input to the auditory nerve on the deaf side, giving the brain information from both directions.

That does not recreate normal acoustic hearing. Implant sound differs from natural hearing, takes time and practice to interpret, and requires surgery, programming and rehabilitation. (NIDCD)

What the strongest evidence says about cochlear implants for single-sided deafness

For many years, the evidence supporting cochlear implantation for SSD was encouraging but limited. Reviews found improvements in hearing and quality of life, but the underlying studies were generally small, observational, followed patients for only six to 12 months and carried a moderate-to-high risk of bias. (JAMA Network)

That evidence has now crossed an important line: from mainly favourable case series to a randomised comparison with two-year follow-up. It has not crossed into certainty for every candidate.

Speech in background noise

The strongest comparative evidence comes from the Dutch CINGLE trial. It included 120 adults with SSD and compared a cochlear implant, a bone-conduction device, CROS and no treatment.

At the beginning of follow-up, 28 participants had received a cochlear implant, 25 had selected a bone-conduction device, 34 were using CROS and 26 had chosen no treatment. After 24 months, the cochlear implant group performed better overall in tests of speech perception in noise and reported better hearing-specific speech outcomes than the other groups. (JAMA Network)

The results were not equally strong in every listening situation. When speech and noise both came from directly in front, the cochlear implant group improved by 1.3 decibels. The researchers noted that this was below the approximately three-decibel change generally considered noticeable and was therefore not clinically meaningful.

The greater advantages appeared when speech and noise were separated across space—the situations in which receiving information from both sides should matter most. (JAMA Network)

That distinction is important. A cochlear implant should not be presented as producing a large benefit in every noisy room. Its additional value is most apparent where the position of a speaker and surrounding noise would otherwise put the person with SSD at a disadvantage.

Locating sound

A separate analysis from the same trial found that the cochlear implant group had better objective sound-localisation accuracy and better self-reported spatial hearing at 24 months than the CROS, bone-conduction and no-treatment groups. CROS improved participants’ subjective spatial-hearing scores to a smaller degree but did not produce a consistent objective localisation improvement. (Sage Journals)

A 2026 meta-analysis combined 40 studies involving 667 participants. It found a clinically meaningful average reduction in localisation error after implantation. Performance still remained below that of people with normal hearing in both ears, and there was substantial variation between studies and recipients. (researchgate.net)

The conclusion should therefore be that cochlear implants can improve localisation—not that they make spatial hearing normal.

Tinnitus

Tinnitus can be a major part of SSD. In the tinnitus analysis from the CINGLE trial, the cochlear implant group recorded the largest and most stable average reduction over two years. Median scores fell by 23 points on the Tinnitus Handicap Inventory and by 60 points on a 100-point tinnitus-burden scale. The number of participants reporting no tinnitus increased from two at baseline to seven at 24 months. Bone-conduction devices also improved some tinnitus measures, although by a smaller amount. (Frontiers)

Tinnitus relief is not guaranteed and should not be promised. The study’s participants had, on average, mild-to-moderate tinnitus at baseline, and the outcomes were self-reported. (Frontiers)

It is also important not to count the speech, localisation and tinnitus findings as three independent randomised trials. They are different analyses of the same CINGLE participant group.

Do people continue using the implant?

Long-term use is one test of whether a device remains worthwhile after the novelty and intensive early follow-up have ended.

A 2026 Antwerp study reviewed 78 adults implanted for acquired unilateral severe-to-profound hearing loss, including 59 people with SSD. Data logging showed an average of 12 hours of processor use per day. Eight per cent stopped using their implants during follow-up, while others continued using them for most waking hours as far as 20 years after surgery. (Springer Link)

This is encouraging but should be interpreted carefully. It was a single-centre observational study involving stringently selected candidates, not a randomised population study. Its authors themselves stressed the importance of strict eligibility criteria.

In June 2026, the American Academy of Otolaryngology–Head and Neck Surgery endorsed cochlear implantation as an option for adults with unilateral or asymmetric sensory hearing loss. That position reflects the direction of current specialist practice in the United States. It is not, on its own, a funding decision for New Zealand. (AAO-HNS)

What people with one-sided deafness say is lost—and regained

Laboratory tests can count correctly repeated sentences or degrees of localisation error. They do not always show the work involved in managing ordinary communication.

In a qualitative study involving semi-structured interviews with 52 people with SSD from several countries, participants described increased reliance on family members, reduced autonomy, changes in communication within the family and uncertainty caused by inconsistent information from health professionals. (PubMed)

A 2024 qualitative study interviewed five adults before and after receiving a cochlear implant for SSD. Before implantation, participants described difficulty communicating in noise, anxiety, physical imbalance and reduced social participation. Afterward, they reported improvements in communication and participation, but some continued to be dissatisfied with aspects of the implant’s sound quality. (E-CSD)

Five participants are not enough to predict outcomes across a national programme. People who choose implantation may also be more motivated or more severely affected than those who decline it.

The value of qualitative evidence is different. It identifies consequences that a standard audiogram or generic health questionnaire may miss: dependence on another person, the effort of constantly repositioning oneself, avoiding difficult environments and being unable to respond confidently when sound comes from the deaf side.

These accounts should inform public assessment. They should not replace clinical evidence or economic analysis.

What the evidence does not prove

A balanced funding case must acknowledge several limitations.

First, a cochlear implant does not restore normal hearing. The brain must integrate natural sound from one ear with electrical sound from the other, and results differ between recipients.

Second, the major randomised trial selected adults whose deafness had lasted between three months and 10 years. It excluded people with abnormal cochlear anatomy and retrocochlear conditions. Its results cannot automatically be applied to every person with congenital, very long-standing or medically complex SSD. (JAMA Network)

Third, not everyone will use the device consistently. Even within the carefully selected long-term Antwerp group, eight per cent discontinued use. (Springer Link)

Fourth, surgery has risks. In the Dutch trial, one participant had their cochlear implant removed after experiencing unexplained pain. (Sage Journals)

Finally, CROS and bone-conduction devices should not be dismissed merely because they do not provide separate input to the deaf ear. They may offer sufficient improvement without cochlear implantation for some people. Patient preference, hearing needs, anatomy, duration and cause of deafness, tinnitus, rehabilitation capacity and likely daily use all matter.

The evidence supports access to specialist assessment. It does not support an assumption that every person with SSD should proceed to surgery.

Are cochlear implants for single-sided deafness cost-effective?

Clinical effectiveness does not automatically establish value for the taxpayer.

A public system must consider the entire care pathway: assessment, device, surgery, mapping, rehabilitation, maintenance, processor replacement and management of complications. It must also consider what else could be funded with the same money.

Economic studies of cochlear implantation for SSD have reached different conclusions because they use different costs, time horizons, utility measures and assumptions about work.

A 2024 Dutch lifetime model found that adult cochlear implantation was cost-effective from a societal perspective, with an estimated incremental cost-effectiveness ratio of approximately €3,493 per quality-adjusted life year compared with a bone-conduction device. The model’s favourable result depended heavily on including productivity.

When productivity costs were removed, the estimate rose to €32,341 per quality-adjusted life year compared with bone conduction and €23,767 compared with no treatment—both above the study’s €20,000 Dutch threshold. (PLOS)

A French analysis produced a very different short-term result. At six months, its estimated cost was €422,279 per quality-adjusted life year. Modelling the benefits over longer periods reduced that figure to €57,561 at 10 years, €38,006 at 20 years and €26,715 at 50 years. The estimate was more favourable for participants with severe tinnitus. (Springer Link)

Neither calculation can simply be converted into New Zealand dollars and treated as a local answer.

Together, they show what drives the economic case:

• the nationally negotiated price of the device and surgery
• the number of years a recipient is expected to use it
• whether benefits remain stable
• the severity of tinnitus and functional difficulty
• the likelihood of regular processor use
• work and study participation
• the utility measure used to value improvement
• the cost of lifelong audiological support and replacement processors.

The taxpayer case is therefore promising but unsettled. It is not yet possible to claim that funding cochlear implants for SSD would produce direct net savings for New Zealand.

It is equally unreasonable to assume that the only economic effect is the initial operation. If better spatial hearing helps a person remain effective at work, communicate without relying on another person or reduce the effects of severe tinnitus, those outcomes have value. New Zealand needs to measure that value rather than borrowing assumptions from Europe.

A realistic public funding pathway for New Zealand

The choice is not between funding a cochlear implant for every person with one deaf ear and retaining an absolute prohibition. A staged pathway could manage clinical and financial uncertainty.

Protect access to first implants. SSD funding should be additional and ring-fenced. It must not reduce implant numbers or extend waiting times for people with severe-to-profound loss in both ears.

Use targeted eligibility rather than automatic entitlement. A multidisciplinary team should consider adults with severe-to-profound sensorineural loss in one ear, suitable cochlear anatomy and an intact auditory nerve where the expected improvement is clinically meaningful.

Assess function, not only the quiet-room audiogram. Evaluation should include spatial speech-in-noise testing, localisation, hearing-specific quality of life, listening effort and tinnitus burden. Work, study, caring and safety requirements may be relevant, but employment should not become a condition that disadvantages older or disabled people.

Consider less invasive options properly. A CROS or bone-conduction trial can give candidates direct experience of alternatives. It should be used where clinically appropriate rather than imposed as a mechanical hurdle when the option is unsuitable.

Require realistic expectations and rehabilitation. Candidates need clear information about implant sound, variability, surgical risk, expected daily use and the time required to integrate electrical and acoustic hearing.

Fund the full care pathway. Public funding should cover the device, operation, mapping, rehabilitation, maintenance and scheduled processor replacement. Funding surgery without long-term support would create an incomplete service and increase the risk of non-use.

Negotiate nationally. A national procurement agreement should seek an SSD device price based on expected programme volume rather than accepting private retail costs.

Start with a staged programme and publish the results. New Zealand should collect speech-in-noise, localisation, tinnitus, listening effort, device-use and hearing-specific quality-of-life data. It should also track complications, employment and study outcomes, whānau support, long-term service costs and reasons for non-use.

Results should be reported by age, ethnicity, geography, cause of deafness and duration of auditory deprivation. Otherwise, a new pathway may disproportionately benefit people who already have access to specialist information and the confidence to advocate for themselves.

What about children with single-sided deafness?

New Zealand’s current paediatric criteria also require severe-to-profound hearing loss in both ears, so most children with SSD are excluded from Disability Support Services funding. (Disability Support Services)

Children require a separate policy rather than being added to an adult pathway. Their auditory system, language and classroom listening abilities are still developing, and the effect of prolonged deprivation in one ear may differ from acquired deafness in adulthood.

The American Academy of Otolaryngology–Head and Neck Surgery now supports implantation for appropriately selected children with unilateral sensorineural loss, while emphasising suitable anatomy and the importance of timing. (AAO-HNS)

The paediatric evidence is developing, but questions about age at implantation, congenital causes, cochlear nerve deficiency, long-term processor use and family commitment require distinct criteria. Children and adults should not be combined under one blunt funding rule.

Pindrop’s position

Adults and children with single-sided deafness should be eligible for consideration for a publicly funded cochlear implant when a multidisciplinary assessment finds that:

• the deaf ear is medically and anatomically suitable
• less invasive management provides insufficient functional benefit or is inappropriate
• a cochlear implant is likely to produce a meaningful improvement
• the candidate or family has realistic expectations
• the recipient is likely to use the processor and participate in rehabilitation.

Eligibility should not guarantee implantation. It should guarantee a fair assessment.

Funding must be separately resourced and include the full long-term service. It cannot come at the expense of people waiting for their first implant for severe-to-profound bilateral hearing loss.

This is not an argument that CROS and bone-conduction devices have no value. It is not a promise that cochlear implantation will eliminate tinnitus, create normal hearing or work equally well for everyone.

It is an argument against using one functioning ear as automatic proof that a person has no need.

One good ear is not two-ear hearing. New Zealand’s funding criteria should be able to recognise that—carefully, transparently and on the evidence.

Frequently asked questions

What is single-sided deafness?

Single-sided deafness is severe-to-profound hearing loss in one ear while the other ear has normal or near-normal hearing. It commonly affects the ability to locate sound and understand speech when background noise is present. (JAMA Network)

Can a cochlear implant help single-sided deafness?

For appropriately selected people, a cochlear implant can improve access to speech coming from the deaf side, spatial hearing, sound localisation and hearing-specific quality of life. It may also reduce tinnitus. It does not restore normal hearing, and individual outcomes vary. (JAMA Network)

Are cochlear implants for single-sided deafness publicly funded in New Zealand?

Most people with SSD do not qualify through Disability Support Services because the national criteria require severe-to-profound hearing loss in both ears. Some people may have other funding routes depending on the cause and circumstances of their hearing loss, but there is no general public SSD entitlement. (Disability Support Services)

Is a cochlear implant better than a CROS hearing aid?

It depends on the person and listening situation. CROS can improve access to sounds coming from the deaf side without surgery. A cochlear implant is more invasive but provides input to the auditory pathway of the deaf ear and has produced better overall spatial-hearing and speech-in-noise results in a two-year randomised trial. (JAMA Network)

Can a cochlear implant stop tinnitus in the deaf ear?

Some people experience a major reduction, and a smaller number report no tinnitus after implantation. Others receive limited or no relief. Tinnitus improvement should be presented as a potential benefit rather than a guaranteed outcome. (Frontiers)

‍

Evidence reviewed: July 2026.

Editorial note: This article discusses population evidence and public policy. Individual suitability for cochlear implantation requires assessment by a specialist cochlear implant team.