SLP Andrea Gregg works with 2-year-old Abby Lawson at The Johns Hopkins Hospital in Baltimore, MD. Gregg is helping Abby, who has been using her cochlear implant since July 2006, to connect meaning to the speech sounds she is detecting through the implant.

Multichannel cochlear implants (CIs) were approved by the U.S. Food and Drug Administration (FDA) in 1990 for use in children with profound hearing loss ages 2-17 years (Clark, 2000). Because early CI systems yielded relatively modest levels of open-set speech understanding (Staller et al., 1991), clinicians were understandably conservative in determining CI candidacy. Since then, advances in the design of CI systems have yielded substantial improvements in outcomes and a concomitant broadening of candidacy criteria (Balkany et al., 2002; Staller et al., 2002).

Current FDA guidelines permit cochlear implantation in children aged 12-23 months with profound hearing loss, and in children ages 2 years or older with severe to profound hearing loss (ASHA, 2004). Pediatric candidacy is based on a number of factors, including the degree of hearing loss; demonstration of minimal benefit from amplification; enrollment in therapy that promotes auditory skill development; and absence of medical contraindications. To date, approximately 120,000 individuals worldwide have been implanted with one of three FDA-approved CI systems.

Evaluating Benefits

In clinical practice, CI candidacy is determined on a case-by-case basis. As pointed out in the recent ASHA Technical Report (ASHA, 2004), clinicians must consider whether an individual will receive more communication benefit from a CI than from a hearing aid. Whenever possible, auditory threshold measurements, as well as outcomes from word and sentence testing, should be used to determine candidacy. However, many young children lack the cognitive or linguistic skills to participate in speech recognition testing. For such children, CI candidacy may be determined primarily on the basis of aided speech detection thresholds after an appropriate hearing aid trial. Because auditory detection levels with a CI should be in the 20 dB-30 dB HL range across the audiometric frequencies of 500-4,000 Hz, cochlear implantation may be warranted if thresholds obtained with a well-fit hearing aid do not provide comparable sound detection. Although detection levels do not guarantee spoken word recognition, detection is fundamental to speech recognition ability. It is important to obtain individual ear measurements to understand how each ear contributes to audition and whether the differences between ears are significant to consider when choosing one for implantation.

Today the expectations for speech and language development in children who receive a CI are higher than ever before (Moog, 2002). As a result, indications for cochlear implantation are being expanded within the population of children who represent "traditional" CI candidates (i.e., those with severe-to-profound hearing loss) as well as to new populations of children, such as those with auditory neuropathy/dys-synchrony or those with additional disabilities.

Bilateral Cochlear Implantation

Traditionally, CI surgery has been performed in one ear to preserve residual hearing in the non-implanted ear for hearing aid use or future CI technology. However, the success of unilateral implantation in traditional CI candidates has led many clinicians, surgeons, and patients/parents to consider bilateral implantation. Approximately 4,600 individuals (almost 4%) have received bilateral cochlear implants—a number that is on the rise. Because many CI candidates now have some residual hearing prior to implantation, clinicians must consider whether bilateral implantation will yield more benefit than the combined use of a CI and a hearing aid on the non-implanted ear.

Recent studies have evaluated the performance of children who wear a CI on one ear and a hearing aid on the other, often referred to as bimodal listening. Results indicate that the two inputs can be combined successfully and may provide a form of binaural hearing (Ching et al., 2001; Litovsky et al., 2006). In children using a CI and hearing aid compared with a CI alone, significantly better scores have been reported with bimodal listening for speech recognition in the presence of babble (Ching et al., 2001). Speech recognition in quiet and in noise in differing locations also shows significantly better scores for children using bimodal listening, although not all advantages of binaural hearing have been demonstrated (Ching et al., 2006). CI outcomes can be greatly improved in the presence of noise when small amounts of low-frequency hearing are available; thus, the possible benefits of amplification in the second ear must be investigated for each child (Holt, Kirk, Eisenberg et al., 2005).

Studies in children using two sequentially placed CIs indicate that localization acuity measures, such as the minimal audible angle (MAA), are improved for a majority of children, although not all, when compared with use of a single CI (Litovsky et al., 2006). The majority of children perform better on an MAA measure in the first implanted ear compared with the second. Some children tested over a one- to two-year period show improved performance with further implant experience (Litovsky et al., 2006). Within subjects, bilateral results on speech recognition measures in noise indicate improved performance compared with the unilateral condition in children (Kuhn-Inacker et al., 2004; Peters et al., 2004), although the degree of improvement varies across subjects and studies, as do the measures and test conditions. It is evident that a large number of variables must be considered in the evaluation of bilateral CIs for children.

Second Ear Candidacy

Candidacy decisions for implantation of the second ear again occur on a case-by-case basis. Evaluation of auditory detection with a hearing aid in the second ear and, if possible, evaluation of speech recognition performance should be taken into consideration. As more information about the results of bilateral implantation becomes available, candidacy decisions will include comparisons with children who have received bilateral CIs.

A critical component in the assessment of bilateral performance appears to be the evaluation of speech recognition under more real-life listening conditions. Scientists at Washington University have recently investigated a clinical speech recognition protocol to evaluate bilateral benefit in adult CI recipients that incorporates varied signal-to-noise levels, differing background noise types including talker babble, and multiple speakers with differing rates of speech and dialects (Firszt et al., 2006). In real-life listening situations and throughout a child's day, conversation occurs in the presence of background noise and with more than one talker. Thus, speech recognition requires the ability to locate and follow multiple talkers and attend to meaningful speech targets while ignoring interfering noise, tasks that are best accomplished with binaural hearing. Assessment tools and conditions that are more appropriate for evaluation of binaural hearing in children are critically needed.

The amount of auditory stimulation in each ear needed for the development of binaural hearing in children is unknown. In addition, the potential benefit or loss of benefit related to the timing of a second implant is not clear. While bilateral CIs could strengthen the bilateral pathways during development and allow for improved binaural hearing, it is likely that there is a group of children at risk for poor benefit from a second CI. Despite the theoretical advantages of binaural hearing, there could be peripheral or central degeneration that results in binaural interference rather than binaural improvements. The long-term benefit and identification of factors that contribute to the acquisition of binaural hearing with bilateral implants in children need further investigation. Systematic study of children who receive bilateral CIs will provide for more informed recommendations concerning if, and when, to implant the contralateral ear.

Auditory Neuropathy/Dys-synchrony

Persons with auditory neuropathy/dys-synchrony (AN/AD) display dys-synchronous peripheral neural responses in the presence of evidence of outer hair cell function (Starr et al., 1996). Potential underlying mechanisms in AN/AD patients include an absence of inner hair cells in the cochlea, compromised synaptic connections between the inner hair cells and the auditory nerve, or axonal loss or demyelination of the auditory nerve itself. Changes in neural firing patterns as a result of abnormalities at any of these sites could account for reduced or dys-synchronous responses (Starr, 2001). The majority of patients with AN/AD report hearing difficulty characterized by difficulty understanding speech, particularly in background noise. Studies of auditory processing using psychophysical measures show that AN/AD patients have particular difficulty with the timing (temporal) characteristics of sound (Zeng et al., 1999, 2001).

AN/AD patients show variation along several dimensions, including sound awareness, the presence of other sensory or motor neuropathies outside of the auditory system, age of onset, and residual ability to utilize auditory information. Scientists at Kresge Hearing Research Laboratory as well as others have observed that patients distribute along a continuum in relation to their functional communication ability. A few patients at one extreme exhibit no overt delays or auditory complaints until adulthood; they obtain enough information auditorily without intervention to develop speech and language. Though these patients generally demonstrate residual speech recognition ability in quiet, they still report considerable difficulty in noise and present a dys-synchronous auditory brainstem response (ABR). At the other extreme are patients with AN/AD who exhibit a total lack of sound awareness. Between these two groups is the largest group of patients we have seen: they demonstrate inconsistent auditory responses, and manage the best in quiet and the poorest in noise. Their audiograms are inconsistent with other test results. The ABR is always dys-synchronized and middle-ear muscle reflexes (MEMRs) are absent.

CIs are an important management option for patients with AN/AD. We and others have observed much success with CIs in AN/AD patients who lack sound awareness and those who are in the large middle group of patients described above. Post-implant neural response telemetry, electrical ABR, and MEMRs demonstrate presence of neural synchrony and are comparable to responses in typical CI patients (Shallop et al., 2001). Comparisons of performance in CI children with and without AN/AD have demonstrated comparable performance (e.g., Trautwein et al., 2000; Peterson et al., 2003). Patients at the end of the continuum, where speech and language development proceeds without intervention, may not be immediate candidates for CIs. Of course, this may change in the event that their hearing status changes or as demands on their auditory abilities become more challenging.

Candidacy for a CI is not related to audiometric pure-tone thresholds in AN/AD. Patients with AN/AD have been implanted with pure-tone thresholds ranging from typical hearing to mild hearing loss to profound hearing loss. Otoacoustic emissions, MEMRs, and our increased understanding of auditory physiology and genetics provide methods of understanding and delineating auditory function beyond that possible with an audiogram alone (e.g., Spoendlin, 1979; Moore et al., 2000; Hogan & Turner, 1998, Berlin et al., 2002). AN/AD diagnosis confirms the importance of physiologic methods and the test battery.

The critical element in the management of infants and children with AN/AD is language development. Our experience has taught us that visual communication methods (e.g., cued speech, sign language, signed English) are necessary for language development. The use of auditory information alone for learning language is, in our experience, unsuccessful without cochlear implantation. Visual communication methods are recommended prior to cochlear implantation, or if a CI is not chosen as a management option.

Children with Additional Disabilities

The Gallaudet Research Institute reported that one-third of children in the United States with some degree of hearing loss also have additional disabilities (Holden-Pitt & Diaz, 1998). These prevalence data indicate that a number of children referred to CI centers will come from this population, especially at CI programs affiliated with tertiary care pediatric hospitals. Some surgeons have elected to implant children with multiple disabilities with an FDA-approved CI system in the hope of improving their quality of life (Waltzman et al., 2000). It is important to examine CI outcomes in this population so that parents and professionals can make informed decisions about CI candidacy.

Only a few investigations have studied CI outcomes in children with multiple disabilities (e.g., Hamzavi et al., 2000; Waltzman et al., 2000); several have reported increased social connectedness and environmental awareness or improved speech perception and production skills following cochlear implantation. Pyman et al. (2000) retrospectively compared the speech and language development of pediatric CI recipients who demonstrated cognitive and/or motor delay with those who had no additional disabilities. Their results suggested that the presence of additional impairments tended to slow aspects of speech and language development involving higher-level speech processing abilities; few children with additional disabilities acquired open-set word recognition even after four years of CI use.

Because the array of disabilities represented in the previous investigations typically was quite broad, it is difficult to determine the effects of each disability in addition to deafness. Holt and Kirk (2005) examined the speech and language development of two groups of pediatric CI recipients. One group had mild cognitive delay in addition to deafness; the other group had no additional disabilities. As a group, the children with cognitive delay had similar post-implant speech perception skills but poorer language skills than the comparison group of recipients with no additional disabilities. These results are in partial agreement with other investigations showing a relationship between nonverbal IQ and speech/language development in children with CIs (Dawson et al., 2002; Geers et al., 2002, 2003).

The previous studies suggest that children with CIs who have additional disabilities may attain a wide range of communication skills ranging from simple sound detection to oral/aural communication. In many cases, outcomes were based on the number and extent of the additional disabilities (ASHA, 2004). However, even children who demonstrate minimal communication gains on objective speech and language measures may receive benefits from a CI, such as increased environmental awareness and "connectedness" to family members and others (Waltzman et al., 2000). More work is needed to better define benefit in "nontraditional" populations of pediatric CI recipients. By following the progress of children over time, we will be better informed about reasonable post-implant expectations for CI recipients with multiple disabilities.