Cochlear Implantation: The Most Misunderstood and Under-Recommended Treatment Option in Audiology?

By Sarah Sydlowski

This article is a part of the September/October 2022, Volume 34, Number 5, Audiology Today issue. 

Cochlear Implant Candidacy

Cochlear implantation is commonly viewed as a treatment of last resort. When benefit from hearing aids is limited or non-existent, a cochlear implant (CI) is commonly accepted as a reasonable treatment option. Between 1984 and 2015, a total of 170,252 U.S. adult and pediatric patients underwent cochlear implantation, accounting for 240,056 devices (iData Research, 2016; Nassiri et al, 2022). These patients represent only a fraction of the more than eight million annual hearing aid users in the United States (iData Research, 2016; Nassiri et al, 2021). There are substantially fewer implant recipients, but is this disproportionate uptake of CIs a reflection of clinical reality or is it indicative of an outdated system? 

Cochlear implantation initially was approved in the middle of the 1980s by the Food and Drug Administration (FDA) for use in post-lingually deafened adults with profound bilateral sensorineural hearing loss (SNHL). At the time, this approved indication aligned well with the clinical evidence, which was based on recipients with little to lose by trying the initially experimental device. Bilateral profound hearing loss with no aided speech-recognition ability was deemed appropriate for candidacy. 

Over time, ‘clinical candidacy’ and ‘coverage’ by a payer have become confusingly conflated.

As insurance payers noted the benefit of the first implant technology capable of restoring a human sense, most payers (including the Centers for Medicare and Medicaid Services (CMS)) adopted coverage plans that allowed patients with bilaterally profound hearing loss access to CI. As new implants were developed, manufacturers sought updated labeling indications, eventually expanding candidacy criteria to the parameters that many professionals cite as standard CI candidacy criteria: moderate to profound bilateral SNHL with aided sentence recognition in the ear to be implanted of less than 50 percent and less than 60 percent in the contralateral ear and when bilaterally aided (Cochlear, 2022). Frustratingly, the coverage determination adopted by CMS fell short of this clinical standard, requiring best aided open-set sentence recognition of less than 40 percent (CMS, 2005). 

Over time, “clinical candidacy” and “coverage” by a payer have become confusingly conflated by referring providers who are unaware that advancements in technology and increasingly positive outcomes outpace both FDA labeling and many coverage policies. Contemporary clinical practices and outcomes are advancing at a rapid pace and FDA criteria and private insurance coverage are not expanding quickly enough to keep up (Zwolan and Basura, 2021; Moses and Friedmann, 2021). 

Today, the misalignment among clinical candidacy, FDA labeling, and payer coverage policies confounds providers. Many providers base clinical decision-making on labeling and coverage instead of evidence-based clinical outcomes, prompting an unknown number of potential CI candidates to forgo candidacy evaluation or implantation that may be very appropriate. 

The approximate U.S. CI use rate likely falls between 2 percent, using contemporary clinical candidacy criteria, and 13 percent, using traditional candidacy criteria (Nassiri et al, 2022; iData Research, 2016). Thus, although the annual proportion of CI recipients to new audiometric candidates has increased for patients with less severe degrees of hearing loss, the total population of untreated individuals who may benefit from CI continues to rise (Nassiri et al, 2022). 

Case Scenario

Hearing aid verification confirmed her hearing aids were programmed to National Acoustic Laboratories (NAL) targets and were reasonably fit. Her audiogram and excellent aided sentence recognition (TABLE 1) do not meet FDA-labeled indications for a CI. As a result, most audiologists would likely recommend hearing aids in this scenario. Yet, her handicap scores for both tinnitus and hearing suggest (TABLE 2) she perceives substantial difficulty. Further discussion reveals she has tried and returned numerous pairs of hearing aids. 

Her aided word- and sentence-recognition in noise are poor. This patient is a prime example of why expanding CI candidacy is necessary.

CIs are a reliable and effective intervention for the majority of adults who have limited speech-perception abilities. On average, adults with CIs may achieve post-operative speech recognition of 54 percent for words, 74 percent for sentences, and 50 percent correct for speech in noise (Boisvert et al, 2020). Further, implantation of ears with less severe degrees of hearing loss and more measurable residual hearing also results in higher post-implant speech recognition and improved quality of life (Dowell, 2016; Birman and Sanli, 2020; Gifford et al, 2010). The expansion of candidacy criteria over the last decade has been most notable in three domains: (1) aided word instead of sentence recognition, (2) ear-specific instead of best-aided scores, and (3) increased degrees of residual hearing.

Words, Instead of Sentences, Identify the Best Candidates

Many of the largest and most experienced implant programs support the use of ear-specific word recognition, rather than audiometric thresholds or aided sentence recognition in the best aided condition (Deep et al, 2021; Gifford et al, 2010; Sladen et al, 2017; Sydlowski et al, 2022). In more contemporary regulatory reviews for short electrode arrays and unilateral hearing loss, manufacturers received labeling based on ear-specific, aided word understanding, but, in my experience, these contemporary criteria tend to be less well-known by referring providers and less commonly cited in payer-coverage policies. Specifically, they describe candidacy based on word understanding in the ear to be implanted of up to 60 percent and up to 80 percent in the contralateral ear. Residual hearing can be normal out to 1000 Hz, precipitously dropping to severe-profound high frequency SNHL. 

CIs are a reliable and effective intervention.

Word recognition is preferable compared to sentences because sentences contain context and assess higher level processing skills (e.g., working memory). Conversely, word-recognition testing de-emphasizes processing skills and the use of context to enhance performance.

Ear-Specific Determination of Candidacy

The consideration of individual ears has also opened the door for recommending CIs in patients with varying degrees of hearing loss or normal hearing in the non-CI ear (Arndt et al, 2011; Galvin et al, 2019; Jakob et al, 2021; Firszt et al, 2018). Two manufacturers have approved labeling for single-sided deafness (SSD) and asymmetric hearing loss (AHL). However, both labels limit aided word-recognition scores to less than 5 percent in the ear to be implanted. 

The benefit of CIs in patients with SSD and AHL may extend well beyond patients with profound hearing loss in the implanted ear (Sydlowski et al, 2022). Even when the contralateral ear has normal hearing sensitivity, there is well-documented evidence of consistently improved outcomes and high rates of long-term CI adoption and use by recipients (Deep et al, 2021).

Presence of Residual Natural Hearing

The presence of measurable residual hearing in an ear under evaluation for CI may concern patients and audiologists. However, improved electrode arrays and surgical techniques may allow for hearing preservation. Today, most CI candidates with residual hearing in the implanted ear (including even normal thresholds through 1000 Hz) have successful outcomes with CI (Gantz et al, 2016; Dunn et al, 2020), in part because of preserved low-frequency residual hearing and the possibility of electric-acoustic stimulation (Gantz et al, 2018). The American Academy of Otolaryngology-Head and Neck Surgery (AAO-HNS) defines functionally relevant residual hearing as a low-frequency pure-tone average (125, 250, 500 Hz) less than 80 dB HL (Adunka et al, 2018) and recommends ongoing evaluation and management of acoustic hearing in patients exhibiting responses at this level. 

Improved electrode arrays and surgical techniques may allow for hearing preservation.

Clinical Candidacy, Not Coverage

CIs are effective among older adults (older than 65 years) and improve social isolation, depression, and cognitive performance (Castiglione et al, 2016; Gurgel et al, 2022). Yet, older adults who are Medicare Part B beneficiaries must currently wait longer and have more severe degrees of hearing loss before cochlear implantation will be covered (Bernhard, 2021; Schafer et al, 2021). Specifically, Medicare coverage requires that potential candidates score less than 40 percent on recorded open-set sentences in the best-aided condition, which is far more restrictive than both FDA labeling and clinical best practices. 

Some private insurers continue to cite CMS guidelines for CI coverage, restricting access to cochlear implantation until there is no other viable option. Although most private payers allow for pre-authorization and approval outside of published coverage guidelines, Medicare Part B does not allow such a practice, limiting eligible recipients to those who meet their strict requirements for coverage and facilitating the confused understanding of providers that coverage is equivalent to candidacy. This limitation results in an access disadvantage for Medicare beneficiaries. Patients with military insurance were 13 times more likely to pursue surgery as compared to patients with Medicare (Tolisano et al, 2019).

Between 2014 and 2050, the Medicare population is projected to grow from 54 to 93 million beneficiaries (Congressional Budget Office, 2014; Neuman et al, 2015). The number of adults older than age 60 who are eligible for conventional CI is expected to double by 2060 to almost 4 million (Goman et al, 2018), making it imperative to align coverage policies with best clinical practices. Otherwise, the denominator (potential candidates) will substantially increase while the numerator (implant recipients) will remain fixed, a phenomenon that is already beginning (Nassiri et al, 2022).  

In response to a request for reconsideration of the national coverage determination submitted by Zwolan and Buchman (2020), CMS is considering proposed expansion of coverage for implantation for adults with aided sentence-recognition scores of up to 60 percent (an increase from 40 percent). Several studies have specifically considered outcomes for patients with sentence-recognition scores greater than 40 percent and less than 60 percent, the results of which support implantation of candidates with greater degrees of pre-operative sentence-recognition ability than those currently covered under CMS guidelines (Mudery et al, 2017; Zhang and Coelho, 2018; Zwolan et al, 2020). Evidence in the literature and open public comments (CMS, 2022) is currently being reviewed and a proposed decision memo is due September 1, 2022. 

It is time for coverage to catch up with clinical best practices.

Consequences of Expanding Candidacy Criteria and CI Under-Utilization

Expanding candidacy criteria and the reconsideration of factors that prevent CI access, including coverage determination policies, will increase access for patients who would benefit from CIs. Returning to the previously discussed case, this patient was recommended cochlear implantation based primarily on her poor aided word understanding. Her difficulty hearing in noise and significant self-reported handicap contributed to the recommendation. 

Post-CI (TABLE 1), her aided speech recognition improved significantly, particularly for words in quiet and sentences in noise. Importantly, her bimodal performance (CI + hearing aid) exceeds her post-operative CI only and pre-operative bilateral hearing aid scores. Perhaps most notably, her perceived hearing and tinnitus handicaps were greatly reduced (TABLE 2) and she is once again participating in the activities she enjoys. Despite these excellent outcomes, it is likely that she would have been overlooked as a CI candidate (Nassiri et al, 2022) because she did not meet conventional candidacy criteria. 

It is time for coverage to catch up with clinical best practices. At the very least, it is imperative that all providers of hearing care understand the differences between “clinical candidate” and “coverage” and refer for CI evaluation accordingly. When an estimated 87 to 98 percent of the individuals who can benefit from this technology are not receiving it (Nassiri et al, 2022; iData Research, 2016), it is time to ask why this is happening and to reverse the trend. 

CIs are not a last resort. Using ear-specific word-recognition testing is more likely to identify appropriate candidates. Residual hearing is not a contraindication to CI and may actually improve the prognosis for benefit. When Medicare and other payers align their coverage policies to clinical best practices, more candidates will be able to access this life-changing technology. Until these principles are universally understood, expansion of CI candidacy falls short of reaching the patients who can benefit.