U.S. flag

An official website of the United States government

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Research Council (US) Committee on Disability Determination for Individuals with Hearing Impairments; Dobie RA, Van Hemel S, editors. Hearing Loss: Determining Eligibility for Social Security Benefits. Washington (DC): National Academies Press (US); 2004.

Cover of Hearing Loss

Hearing Loss: Determining Eligibility for Social Security Benefits.

Show details

7Hearing Loss in Children

Spoken communication is uniquely human. If the sense of hearing is damaged or absent, individuals with the loss are denied the opportunity to sample an important feature of their environment, the sounds emitted by nature and by humans themselves. People who are deaf or hard-of-hearing will have diminished enjoyment for music or the sound of a babbling brook. We recognize that some deaf and hard-of-hearing children are born to deaf parents who communicate through American Sign Language. Without hearing, these children have full access to the language of their home environment and that of the deaf community. However, the majority of deaf and hard-of-hearing children are born to hearing parents. For these families, having a child with hearing loss may be a devastating situation. The loss or reduction of the sense of hearing impairs children's ability to hear speech and consequently to learn the intricacies of the spoken language of their environment. Hearing loss impairs their ability to produce and monitor their own speech and to learn the rules that govern the use of speech sounds (phonemes) in their native spoken language if they are born to hearing parents. Consequently, if appropriate early intervention does not occur within the first 6-12 months, hearing loss or deafness, even if mild, can be devastating to the development of spoken communication with hearing family and peers, to the development of sophisticated language use, and to many aspects of educational development, if environmental compensation does not occur.

Hearing loss can affect the development of children's ability to engage in age-appropriate activities, their functional speech communication skills, and their language skills. Before we consider the effects of hearing loss on this development, we will review briefly the extensive literature on the development of speech and language in children with normal hearing. Although the ages at which certain development milestones occur may vary, the sequence in which they occur is usually constant (Menyuk, 1972).

This chapter discusses the nature of the emergence of communication skills in normally hearing children as well as the unique effects of early hearing loss and deafness on this process for infants and children. We give details of the special nature of assessments and rehabilitation strategies appropriate for infants and children with hearing loss and finally discuss how considerations for disability determination need to be tailored to the special needs of this population.


Children with Normal Hearing

Speech Skills

Infants begin to differentiate among various sound intensities almost immediately after birth and, by 1 week of age, can make gross distinctions between tones. By 6 weeks of age, infants pay more attention to speech than to other sounds, discriminate between voiced and unvoiced speech sounds, and prefer female to male voices (Nober and Nober, 1977).

Infants begin to vocalize at birth, and those with normal hearing proceed through the stages of pleasure sounds, vocal play, and babbling until the first meaningful words begin to occur at or soon after 1 year of age (Bangs, 1968; Menyuk, 1972; Quigley and Paul, 1984; Stark, 1983). Speech-like stress patterns begin to emerge during the babbling stages (Stark, 1983), along with pitch and intonational contours (Bangs, 1968; Quigley and Paul, 1984; Stark, 1983).

According to Templin (1957), most children (75 percent) can produce all the vowel sounds and diphthongs by 3 years of age; by 7 years of age, 75 percent of children are able to produce all the phonemes, with the exception of “r.” Consonant blends are usually mastered by 8 years of age, and overall speech production ability is generally adult-like by that time (Menyuk, 1972; Quigley and Paul, 1984).

Language Skills

Language studies have described vocabulary and grammatical development of children with normal hearing. Studies of grammatical development have focused on both word structure (e.g., prefixes and suffixes), termed “morphology,” and the rules for arranging words into sentences, termed “syntax.” Vocabulary development up to young adulthood is estimated at roughly 1,000 word families per year, with vocabulary size estimated at approximately 4,000-5,000 word families for 5-year-olds and 20,000 word families for 20-year-olds (see Schmitt, 2000, for discussion). A word family is defined as a word plus its derived and inflectional forms. Most morphological and syntactic skills are fully developed by the age of 5 years, and grammatical skills are fully developed by age 8 (Nober and Nober, 1977). By age 10 to 12, most children with normal hearing have reached linguistic maturity (Quigley and Paul, 1984). In summary, by age 4½ years, children with normal hearing are producing complex sentences. Although a majority of the speech sounds in English are mastered by age 4, and most of the grammatical categories by age 5, it is not until age 8 that a normally hearing child has fully mastered grammar and phonology and has an extensive vocabulary (Nober and Nober, 1977).

Children with Hearing Loss

A review of speech and language development in children with hearing loss is complicated by the heterogeneity of childhood hearing loss, such as differences in age at onset and in degree of loss; we review these complicating factors separately following a more general overview. Mental and physical incapacities (mental retardation, cerebral palsy, etc.) may also coexist with hearing loss. Approximately 25-33 percent of children with hearing loss have multiple potentially disabling conditions (Holden-Pitt and Diaz, 1998; McCracken, 1994; Moeller, Coufal, and Hixson, 1990). In addition, independent learning disabilities and language disabilities due to cognitive or linguistic disorders not directly associated with hearing loss may coexist (Mauk and Mauk, 1992; Sikora and Plapinger, 1994; Wolgemuth, Kamhi, and Lee, 1998). For example, Holden-Pitt and Diaz (1998) reported the following incidences of additional impairments in a group of children with some degree of hearing loss:


blind/uncorrected vision problem (4 percent),


emotional/behavioral problem (4 percent),


mental retardation (8 percent), and


learning disability (9 percent).

The coexistence of other disabilities with hearing impairment may impact the way in which sensory aids are fitted or the benefit that children receive from them (Tharpe, Fino-Szumski, and Bess, 2001). A recent technical report from the American Speech-Language-Hearing Association stated that pediatric cochlear implant recipients with multiple impairments often demonstrate delayed or reduced communication gains compared with their peers with hearing loss alone (American Speech-Language-Hearing Association, 2004).

In this chapter, we focus on speech and language development in children with prelingual onset of hearing loss (before 2 years of age) without comorbidity. However, it should be kept in mind that the presence of multiple handicapping conditions may place a child at greater risk for the development of communication or emotional disorders (Cantwell, as summarized by Prizant et al., 1990). In addition, these children may require adaptations to standard testing routines to accommodate their individual capacities.

Natural acquisition of speech and spoken language is not often seen in individuals with profound hearing loss unless appropriate intervention is initiated early. One of the primary goals in fitting deaf or hard-of-hearing children with auditory prostheses (hearing aid or cochlear implant) is to improve the ease and the extent to which they can access and acquire speech and spoken language. It should be kept in mind that the children under discussion typically are not born to deaf parents; those children may acquire American Sign Language as their native language.

Speech Skills

Speech and voice characteristics of persons who are deaf or hard-of-hearing are generally acknowledged to differ significantly from those of individuals with normal hearing (Abberton and Fourcin, 1975; Hood and Dixon, 1969; Monsen, 1974, 1978, 1983b; Monsen and Engebretson, 1977; Monsen, Engebretson, and Vemula, 1978, 1979; Nickerson, 1975; Nober and Nober, 1977; Stark, 1983; Wirz, Subtelny, and Whitehead, 1981). A congenital or prelingually acquired hearing loss reduces the intelligibility of talkers who are deaf or hard-of-hearing and impairs the production and tonal aspects of their speech (John and Howarth, 1965; Markides, 1970; McGarr and Osberger, 1978; Monsen, 1979; Osberger and Levitt, 1979; Smith, 1975).

Difficulties with speech sound production include problems with the articulation of vowels and consonants, such as substitutions, distortions, and omissions (Hudgins and Numbers, 1942; Zimmerman and Rettaliata, 1981); excessive use of a neutral vowel, such as schwa Image p2000b4b5g183001.jpg the unstressed vowel sound in the second syllable of the word “kitten” (Markides, 1970); lack of adequate differentiation between various vowels (Angelocci, Kopp, and Holbrook, 1964; Levitt and Stromberg, 1983); and failure to differentiate between voiced and voiceless consonant sounds, for example “b” and “p” (Calvert, 1962; Monsen, 1976; White-head, 1983). These problems are accompanied by a significantly slowed rate of general speech sound awareness (phonological development) in children with hearing loss (Subtelny, 1983). Although many talkers who are deaf or hard-of-hearing can correctly produce phonemes in isolation, they may still be unable to smoothly combine the phonemes in connected speech. Thus, reduced speech intelligibility can result.

Language Skills

Vocabulary knowledge in children with hearing loss may be age appropriate or reduced, with results showing large variability (Gilbertson and Kamhi, 1995; Seyfried and Kricos, 1996; Yoshinaga-Itano, 1994). In general, however, the rate of vocabulary growth is slowed, and may plateau prematurely (Briscoe, Bishop, and Norbury, 2001; Carney and Moeller, 1998; Davis, 1974; Davis, Elfenbein, Schum, and Bentler, 1986; Moeller, Osberger, and Eccarius, 1986). Word entries may have less breadth or flexibility of meaning (Moeller et al., 1986; Yoshinaga-Itano, 1994). In particular, nonliteral or abstract word usage may be impoverished. The dynamic time course of accessing the meanings of words may also be slowed (slowed lexical retrieval) in children with hearing loss, although again, large unpredictable variability among individuals occurs (Jerger, Lai, and Marchman, 2002). In concert with vocabulary development, grammatical knowledge is also reduced in children with hearing loss.

For example, in a sentence construction task, 14-year-old children who were deaf or hard-of-hearing performed similarly to 6- to 8-year-old children with normal hearing (Templin, 1966). In spoken language samples, the sentences of children who were deaf or hard-of-hearing were of shorter lengths with simpler sentence constructions and syntax (Brannon and Murray, 1966; Seyfried and Kricos, 1996). Sentences in the passive voice were not successfully comprehended or produced by about half of 17- to 18-year-old children who were deaf or hard-of-hearing (Power and Quigley, 1973). In studies of the morphological rules for different types of suffixes (e.g., -s as in sings and -er as in singer), children who are deaf or hard-of-hearing generally show inferior performance (Bunch and Clarke, 1978; Cooper, 1967; Elfenbein, Hardin-Jones, and Davis, 1994). The extent to which specific language skills are delayed versus deviant in the presence of childhood hearing loss continues to be pursued. It should also be noted that language proficiency is a strong predictor of reading achievement (Carney and Moeller, 1998). Thus, age-appropriate literacy skills typically are not observed in children with hearing loss and language problems.

Complicating Factors

One factor that influences the extent to which speech and language development is affected by hearing loss is the child's age when the loss occurs. Auditory deprivation early in life has serious consequences for subsequent development (Davis, 1965). In general, children with early, prelingual hearing losses more frequently display deficits in the respiratory, articulatory, and phonatory aspects of speech (Binnie, Daniloff, and Buckingham, 1982; de Quiros, 1980; Itoh, Horii, Daniloff, and Binnie, 1982; Nober and Nober, 1977). Three important periods for onset of hearing loss have been described by de Quiros (1980). Children whose hearing loss occurs during the first 2 years of life are considered prelingually deafened. If profound deafness occurs prior to 2 years of age and intervention is delayed, speech can be severely disturbed. Hearing losses that occur between ages 2 and 5 will result in the loss of speech skills unless appropriate sensory aids and aural rehabilitation are provided. Finally, de Quiros suggests that deafness occurring after age 5 can result in appropriate articulation; however, good articulation skills may deteriorate if sensory input is not reestablished with amplification. Again, these detrimental outcomes may be ameliorated if the child is provided with a sensory aid that can provide auditory access to the acoustic properties of speech.

Age at onset of hearing loss is not the only important prognostic indicator for speech and language development in children who are deaf or hard-of-hearing. A second critical factor is the degree of hearing loss. In a classic study, Hudgins and Numbers (1942) described an inverse relationship between articulatory errors and audiometric scores. Other authors have suggested that speech intelligibility decreases as the degree of hearing loss increases (Jensema, Karchmer, and Trybus, 1978; McGarr and Osberger, 1978; Monsen, 1983a). Similarly, Quigley and Paul (1984) reported that deficits in language comprehension and usage increase with degree of hearing loss. In spite of this overall trend, however, a general finding is that degree of hearing loss cannot perfectly predict speech and language abilities in individual children (Davis et al., 1986; Gilbertson and Kamhi, 1995; Mayne, Yoshinaga-Itano, and Sedey, 1998; Seyfried and Kricos, 1996).

Unilateral Hearing Loss

A study published in 1998 estimated that the prevalence of unilateral hearing loss in school-age children ranged from 6.4 to 12.3 per 1,000 and at that time there were 391,000 school-age children with unilateral hearing loss (Lee, Gomez-Marin, and Lee, 1998). There is a pervasive misunderstanding that unilateral hearing loss is of no consequence and that this problem can be disregarded. Consequently, children with unilateral hearing loss often receive no direct intervention, such as amplification or therapeutic services. However, research now has shown that children with unilateral hearing loss are disadvantaged. In particular, children with unilateral hearing loss have difficulty in understanding speech in noisy environments and are deficient relative to their peers in localization of a sound source (Bess, Tharpe, and Gibler, 1986). Another study found that 32 percent of children in a cohort with unilateral hearing loss failed a grade in school and were significantly delayed in language compared with a matched group of children with normal hearing (Klee and Davis-Dansky, 1986).

Age at Intervention

Because speech and language develop rapidly during the early years in children's lives (up to age 5), the importance of early intervention, including suitable amplification or cochlear implantation, can be seen. It is generally agreed that such intervention procedures are most effective when initiated as early as possible after the identification of the hearing loss (Silverman, 1983). According to Ling (1979), the motor skills required for speech can be learned at any time, but they are most likely to be transferred to the spontaneous level if children have not developed firmly established error patterns. Intervention techniques should be initiated at an early stage and should mirror the pattern of development in children with normal hearing (Ling, 1979).


Recent data from the Gallaudet Research Institute's annual survey indicate that approximately 51.2 percent of children and youth with hearing loss are white, 15.4 percent black, 24.5 percent Hispanic, 4.3 percent Asian-Pacific Islander, and 0.8 percent American Indian, with the rest falling under the “other” or multiethnic categories (Gallaudet Research Institute, 2002). About 54 percent are male and 46 percent are female. This survey represents a database of roughly 60 percent of children in the United States who are deaf and hard-of-hearing and is based on reports from educational programs in which these children are enrolled (Karchmer and Mitchell, 2003). Current racial/ethnic proportions now mirror those found in the United States (Holden-Pitt and Diaz, 1998).

The following is a condensation of information presented by Karchmer and Mitchell (2003). In terms of educational placement, 31.7 percent are in regular education settings, 12.6 percent are in resource rooms in these settings, 27.5 percent are in self-contained classrooms in regular education settings, 24.7 percent are in specialized schools, and the rest of the children (3.5 percent) are in other types of settings. In essence, two-thirds of all such children are now receiving at least some academic instruction in a regular classroom. Slightly more than 90 percent come from homes where only one spoken or written language (primarily English or Spanish) is used regularly. Self-contained educational settings have a larger percentage (almost 25 percent) of students from Spanish-speaking homes, almost twice the percentage found in the other three settings taken in aggregate. Special schools tend to enroll children with profound hearing impairments while self-contained classrooms serve students across the hearing spectrum, and regular school settings serve mostly those with less than severe degrees of hearing loss. The primary communication mode in educational settings is strongly related to degree of hearing loss. Specifically, those with profound losses are typically educated in programs that use signing or signing with speech, while students with milder hearing losses are most typically in speech-based programs. Students in regular education settings are least likely to have additional disabilities.

With regard to academic achievement, children who lose their hearing after learning English generally achieve higher scores on standardized tests, including reading, than those with hearing loss at younger ages. The exception is deaf children of deaf parents, who tend to have higher English-language achievement scores than those with limited access to linguistic interaction both inside and outside the home. It is important to keep in mind, however, that the range of results is considerable. Mathematics performance, while higher than for language-based achievement, is not equivalent to that for hearing peers.

In addition, students with hearing loss tend to demonstrate the same relative academic performance differences as their hearing peers across racial and ethnic groups. Karchmer and Mitchell (2003) emphasize the confound between race/ethnicity and lower socioeconomic status (SES), which makes it difficult to identify the impact of SES for students who are deaf and hard-of-hearing. Sex differences are minimal.

In the recent past, it has been reported that the vast majority of persons educated in deaf schools (95 percent) reach a reading age of only 9 years (Stern, 2001; Traxler, 2000). Reading achievement scores are reduced even for students with minimal sensorineural hearing loss (Bess, Dodd-Murphy, and Parker, 1998). Reading deficits are exacerbated by reduced vocabulary, as previously discussed. More important, however, are deaf children's deficits in phonological awareness. This is a crucial skill in the development of sound-symbol associations and consequently in reading ability. Paul (2003) emphasizes that deaf readers who use phonological codes for processing print do better than those who use nonphonological codes. Consequently, access to phonological information is critical. Reading and generalized linguistic difficulties can also be manifested in deficits in other areas of academics, including mathematics (Hyde, Zevenbergen, and Power, 2003) and science (McIntosh, Sulzen, Reeder, and Kidd, 1994).

Although children are not in the workforce, they do spend considerable time in school. It is known that poor classroom acoustics (Acoustical Society of America, 2000) exacerbate difficulties in development of speech perception and eventually contribute to language and cognitive problems (Nelson, Soli, and Seltz, 2002). Perceptual difficulties in children due to poor classroom acoustics are especially challenging for children with even mild to moderate hearing loss (Bess et al., 1998), especially if English is not their primary language (Nelson et al., 2002). Poor classroom acoustics can result from too much noise intruding into the classroom from outside, leading to a poor signal-to-noise ratio for speech communication (Soli and Sullivan, 1997). Many classrooms also are poorly designed in terms of controlling for reverberation time (Knecht, Nelson, Whitelaw, and Feth, 2002). As explained earlier, reverberation time is a measure of the amount of time that a sound remains in the room after the original sound source has ceased, due to the reflections within the room (Acoustical Society of America, 2000). Long reverberation times mean that sounds already produced can interfere with newly produced sounds, leading to low speech intelligibility. It is known that normally hearing children's auditory processing capabilities are adversely affected by long reverberation times (Johnson, 2000; Litovsky, 1997). Children with hearing loss are at risk in their ability to understand spoken communication in many schools. This difficulty can lead to reduced language and cognitive development.

A clinical entity known as central auditory processing disorder (CAPD) is a dysfunction in perceiving auditory signals that is not attributed to peripheral hearing loss (McFarland and Cacace, 1995). CAPD is believed to subsume specific language and reading disabilities in school-age children. The area of CAPD assessment and remediation is not considered in this report, because this disability by definition is not attributed to hearing impairment. Children suspected of this disorder would be evaluated more appropriately in the domain of developmental disabilities.

In summary, even mild hearing loss places children at risk for speech, language, and educational problems. A view expressed in the literature, however, is that strong familial support and early enrollment in high-quality intervention programs can increase the chances of successful speech and language outcomes in the presence of childhood hearing loss (Moeller, 2000; Yoshinaga-Itano and Apuzzo, 1998a, 1998b; Yoshinaga-Itano, Sedey, Coulter, and Mehl, 1998). Appropriate early intervention, for example, educational intervention, hearing aids, or cochlear implantation, can help to improve the performance of deaf children in areas of language and academic performance (Boothroyd and Boothroyd-Turner, 2002; Tomblin, Spencer, and Gantz, 2000). Next we turn to diagnosing and quantifying hearing loss in infants and children.


Hearing loss in infants is not obvious, and without specific measures to test for the condition it will go undetected for a significant period of time. As recently as 1996, the average age of identification of hearing loss in the United States was 30 months (Harrison and Roush, 1996). The means to evaluate infants for hearing loss with clinical tools, such as auditory brainstem response (ABR) and otoacoustic emissions (OAEs) have emerged in the past 20 years. However, until recently newborn hearing loss screening programs have existed in only a few hospitals for high-risk infants. In 1993, the National Institute on Deafness and Other Communication Disorders (NIDCD) of the National Institutes of Health issued a consensus statement (National Institutes of Health, 1993) calling for screening of all infants for hearing loss by 3 months of age. By 2000, the American Academy of Pediatrics (1999, 2000) endorsed universal newborn hearing screening for all infants prior to hospital release. Currently all 50 states have legislation either passed or pending to mandate universal newborn hearing screening, or they are conducting screening for most newborns without legislation.

Neonatal hearing screening programs have proven effective as the first step in early identification of infants with congenital hearing loss. These programs identify infants at risk for a mild or more serious hearing loss. It is important to note that screening identifies only which infants are in need of a more complete assessment to determine if hearing loss exists. Infants failing this screening require a diagnostic audiological assessment by a qualified audiologist to confirm the presence of hearing loss and determine the exact type and degree of hearing loss in each ear. Results of the diagnostic evaluation are used to determine the degree of disability, to determine eligibility for rehabilitation programs and financial assistance, and to form the basis for fitting of amplification and placement in appropriate educational settings.

Although the screening process for newborn hearing loss is excellent, it is not perfect. Children with mild hearing loss or hearing loss in restricted frequency regions may pass the screening. Some children develop significant permanent hearing loss after the newborn period, which is not detected by the screening. Infants with neural hearing loss, such as auditory neuropathy, will pass a screening if an OAE test alone is used for screening (Sininger, 2002). Infants with late-onset hearing loss will be missed in the newborn screening; and constant surveillance is needed by the medical community to find these infants and begin remediation as soon as possible (Joint Committee on Infant Hearing, 2000).

The screening and diagnostic testing process is designed to expedite intervention for children with hearing loss and maximize the opportunity to provide audition during critical learning periods (Sininger, Doyle, and Moore, 1999; Yoshinaga-Itano et al., 1998). The goal of early hearing detection and intervention programs is to identify hearing loss and begin intervention including fitting of hearing aids at or before 6 months of age.

Assessment of hearing loss in infants requires age-appropriate procedures. Infants under 6 months of age cannot give an accurate response to sounds at threshold levels, regardless of their ability to detect them. These infants require an audiological test battery based on objective physiological tests that reveal threshold-level responses, as well as information regarding the functioning of the peripheral auditory system.


General agreement exists in the United States regarding the essential elements of an appropriate protocol for diagnostic audiological assessments of infants and young children. An audiologist with appropriate state licensure or equivalent credentials must perform such assessments. The complete battery includes a history and parent interview, evaluation of middle ear function, OAE testing, and an age-appropriate assessment of auditory behaviors. The core of the diagnostic evaluation protocol is an estimate of audiometric thresholds using auditory brainstem response or other proven electrophysiological assessment with frequency-specific stimuli. According to the American Academy of Pediatrics (2000, p. 804):

Audiologists providing the initial test battery to confirm the existence of a hearing loss in infants must include physiological measures and developmentally appropriate behavioral techniques… For infants birth to six months of age, the test battery … must contain an electrophysiological measure of threshold such as ABR or other appropriate electrophysiological test using frequency-specific stimuli.

Threshold Audiometry

Unlike vision, human auditory sensitivity is adult-like within a few days of birth (Adelman, Levi, Linder, and Sohmer, 1990; Klein, 1984; Sininger, Abdala, and Cone-Wesson, 1997). Consequently, hearing loss degree and configuration are judged by the same standards for newborns as for adults.

The basic hearing evaluation for persons of any age is the pure-tone audiogram. Thresholds are also measured using speech stimuli. Establishing thresholds for tonal and speech stimuli by air and bone conduction using standard adult procedures is possible with children who have a developmental age of 4-5 years. Prior to that age, procedures must be modified to meet developmental demands. For all pediatric assessments, multiple-procedure test batteries are recommended to ensure the consistency of results.

Pure-tone or frequency-specific threshold tests in infants and children are classified either as physiological tests, in which a response is determined by some objectively measured change in physiological status, or behavioral tests, in which an overt response is elicited from children in response to sound and their responses are judged by an audiologist. Physiological tests do not actually measure perception of sound but can generally predict hearing thresholds or the range of hearing with a great deal of precision. The most valuable of these tests for threshold prediction for infants less than 6 months old is the ABR. A promising but as yet less proven technique for threshold prediction in these very young children is the auditory steady-state response (ASSR). Other physiological measures that correlate with hearing levels and support the test battery include tympanometry, acoustic middle ear muscle reflex, and OAEs.

Behavioral threshold tests for children under 5 are classified in Table 7-1. It is generally assumed that reliable behavioral responses to sound in infants under the developmental age of 6 months will be suprathreshold. By 6 months of age, normally developing children can be trained to respond to threshold-level stimuli using an operant conditioning paradigm known as visual reinforcement audiometry (VRA). Consequently, before 6 months of age, audiometric thresholds must be inferred from physiological tests such as ABR.

TABLE 7-1. Behavioral Tests of Hearing Threshold.


Behavioral Tests of Hearing Threshold.

However, during the 0- to 6-month age period, it is possible to obtain unconditioned responses to sound, such as a change in sucking behavior, startle reflex, or eye widening. This test paradigm is known as behavioral observation audiometry (BOA). These responses will be suprathreshold and cannot rule out mild or moderate hearing loss. BOA is nonetheless a valuable part of the test battery for infants under age 6 months to substantiate overall impressions.

Children with normal vision at the developmental level of typical 6-month-olds naturally turn their heads to find the source of an interesting sound. VRA takes advantage of that fact by reinforcing head turns with a pleasant visual stimulus, usually an animated toy that is lit to become visible for a short time following a head turn that is time-locked to the presentation of an auditory stimulus. Tones and speech can be used. The test must be administered quickly after appropriate conditioning to maintain the child's interest. A variety of visual reinforcers can be used to elicit head turns in response to near-threshold level stimuli. VRA can be administered using insert earphones for an ear-specific response or with a bone-conduction vibrator. If a child will not tolerate earphones, the stimuli can be presented through a speaker into the sound field of a sound-treated chamber. This procedure limits the conclusions of the tests to hearing in the better ear and cannot determine a unilateral hearing loss. Generally, normally hearing 6-month-old infants will respond to stimuli of 20 dB HL or better (Widen and O'Grady, 2002).

VRA may no longer hold the interest of children who have reached the developmental status of a 2-year-old. In that case, the children's interest can usually be maintained by involving them in a play activity. Play audiometry involves making a game of hearing sounds. Children respond to the sound presentation, for example, by dropping a block into a bucket or stacking a ring on a peg. Devices are available that dispense a tangible reward, such as a piece of candy or a token, when an appropriate response to sound is given. This is known as tangible reinforcement audiometry (TROCA). As long as the interest of the child can be maintained, these techniques will yield accurate audiometric threshold evaluations.

Otoacoustic Emissions

OAE tests are described in detail in Chapter 3. The objective nature of these measures and the ease with which they generally can be obtained make OAEs ideal for assessment of cochlear function in infants, toddlers, and all cooperative children. These measures are used in newborn hearing screening because the presence of a response can be interpreted as indicating normal cochlear function sufficient to indicate hearing levels generally better than 30 to 40 dB. These levels correspond to no more than mild hearing loss and are in line with screening levels for newborns.

As emphasized in Chapter 3, OAEs used to assess children must be used in a test battery to avoid common pitfalls. The presence of an OAE alone does not indicate normal hearing and the absence of a response can be due to factors other than cochlear dysfunction, such as middle ear dysfunction.

Auditory Brainstem Response

The ABR as described in Chapter 3 is an important part of any audiometric evaluation of infants and children. The ABR is present in infants as young as 28 weeks gestation (Starr, Amlie, Martin, and Sanders, 1977). ABR is stable in infants and is unaffected by sleep or sedation. Frequency-specific ABR is the test universally recommended to determine thresholds in infants who do not pass newborn hearing screening (Joint Committee on Infant Hearing, 2000). The accuracy of predicting the actual hearing threshold using ABR is quite good, generally within 10 dB (Sininger et al., 1997; Stapells, Picton, Durieux-Smith, Edwards, and Moran, 1990).

ABRs generated in infants for threshold prediction may require specific recording parameters that differ from those that are standard for use with adults. Infant ABRs are slower and later than those of adults and generally require a long analysis window of 20-30 ms and a lowered band-pass filter (30 to 1000 Hz). Infants' recordings, especially those made during natural sleep, may be noisy, requiring longer averaging time to achieve adequate signal-to-noise ratio for reliable response detection.

The ABR can sometimes be absent or severely abnormal, even when the inner ear is functioning well, due to auditory neuropathy. Children with this disorder will have an abnormal or absent ABR but usually will also show a present OAE. When this condition exists, neither the ABR nor the OAE can be used to predict hearing levels. The hearing loss in patients with auditory neuropathy can be of any degree, and their speech perception ability is severely disordered. Therefore, it is very important to include the measurement of OAEs in any assessment of hearing using ABR. If the ABR and OAE are disparate in the prediction of threshold, then auditory neuropathy must be suspected, and neither will be a good indicator of hearing level.

Auditory Steady-State Response

The ASSR, previously known as the steady-state evoked potential (SSEP), is another way of objectively assessing frequency-specific responses, and it is described fully in Chapter 3. For infants and children, the appropriate modulation frequency range is about 80-100 Hz. Long segments of these stimuli are presented and ongoing electroencephalographic activity is sampled and analyzed in the frequency domain. When the neural activity shows a preference for the modulation frequency over other frequencies in the analysis, it is assumed that the auditory system is responding to the carrier frequency.

One reservation about the use of ASSR for measurement of hearing in infants and children is the lack of good data on them (Stapells, 2002). Rickards has found that normally hearing infants may not have a reliable response below about 40 dB (Rickards et al., 1994). This would make it impossible to distinguish between mild hearing loss and normal hearing, which is critically important for determination of amplification needs. Data from Perez-Abalo and colleagues (Perez-Abalo et al., 2001) showed that although they were able to determine hearing loss in the severe and profound range, in general, there was only fair agreement between ASSR thresholds and hearing levels in children with hearing loss. Her data also showed that ASSR was unable to determine thresholds below 40 to 50 dB nHL (stimuli calibrated relative to thresholds in normally hearing individuals) in the children at any frequency. At this time it would be not be prudent to recommend the use of ASSR to determine hearing loss in infants and young children, especially those with mild and moderate hearing loss.

Immittance Audiometry

Immittance audiometry, as described in Chapter 3, is used to determine the physical status of the middle ear by assessing the mobility of the tympanic membrane. From these measures one can infer whether the membrane is intact or perforated, the pressure of the middle ear, and the functional status of the Eustachian tube. When middle ear function is impaired, for example due to reduced tympanic membrane mobility from otitis media with effusion, hearing sensitivity is reduced and generally the audiogram will show an air-bone gap.

Immittance audiometry is an objective measure requiring no overt response from patients, although cooperation for fitting of a pressuretight ear canal probe and a few minutes of sitting still are required. Consequently, immittance is an important part of any pediatric audiometric assessment, because of the possibility of limited cooperation for standard bone conduction measures. Also, the incidence of otitis media with effusion is quite high in infants and toddlers.

When tympanic membrane mobility is normal, the acoustic stapedial reflex can be used as an assessment of brainstem auditory system function. Children with moderate hearing loss or greater will show elevated reflex thresholds or no reflex. Auditory neuropathy will also eliminate the acoustic stapedial reflex.

Immittance measures in children are essentially the same as those used for adults with one exception. Based on the physical properties of the ear canal, the probe frequency recommended for neonates and young infants is 1000 Hz (Rhodes, Margolis, Hirsch, and Napp, 1999).


The ability to hear and understand speech can have a profound impact on all aspects of children's communication development and daily functioning. Clinicians must have available a wide array of age-appropriate outcome measures that allow them to target different aspects of communication skills (Kirk, 1999, 2000; Kirk, Pisoni, and Osberger, 1995). A battery of communication tests should be employed that includes measures of spoken word recognition, speech production, and receptive and expressive language abilities. The tests and administration procedures for each communication area are considered independently below. It may be necessary to select other tests or adapt the procedures for children who have other impairments in addition to hearing loss. For example, closed-set picture tests may be needed for children who cannot produce verbal or signed responses. Similarly, questionnaires completed by parents or caregivers may be used for children who cannot participate in structured testing situations.

Spoken Word Recognition

Tests utilizing speech stimuli can help to determine the extent to which a hearing loss affects the ability to perceive, recognize, and discriminate speech sounds. This information can be useful in diagnosing the type and severity of the hearing disorder, in assessing candidacy for sensory aid use (hearing aid versus cochlear implant), in informing aural rehabilitation strategies, and in estimating a child's listening abilities in real-world listening situations.

There are a number of factors that must be considered when selecting tests of spoken word recognition and interpreting their results. Characteristics that can impact performance on such measures include the child's chronological age, vocabulary and language level, speech production abilities, and cognitive abilities. Children with congenital or prelingually acquired hearing loss often are delayed in the development of speech and language skills and have restricted vocabularies. Thus, test measures with unfamiliar vocabulary may underestimate their auditory skills. Similarly, the task must be appropriate for the child's language and developmental abilities. If children do not have the cognitive ability to understand and perform a task, the testing will be invalid.

Methodological variables that impact performance outcomes include presentation level (Gravel and Hood, 1999), the method of stimulus presentation, the response format of the tests, the use of competing stimuli, and the sensory modality in which the speech signal is presented.

Recorded Versus Live Voice Stimulus Presentation

There are merits and drawbacks to the use of both live voice and recorded test administration. Live voice testing is problematic because the results depend, in part, on the characteristics of the person administering the test. Thus, results obtained with live voice stimulus presentation are not comparable across clinics or research centers unless talker equivalence can be demonstrated. Several clinicians and researchers have argued that consistency in presentation between listeners or over time can be maintained only through the use of recorded test stimuli (Carhart, 1965; Mendel and Danhauer, 1997). However, this is true only when a standard recording of a particular test is used. The use of different versions of a recorded test (i.e., those recorded by different talkers) can introduce as much variability in the acoustic signal as when two different talkers administer live voice tests. The advantages of live voice test administration are that it allows the examiner greater flexibility and often takes less time than using recorded materials. The use of live voice tests may be desirable for young children because the rate of stimulus presentation can be varied to ensure that children are attending to the task. In general, the use of recorded tests is preferred for assessing performance in older children, so that results can be compared across centers or testing intervals. There are no absolute age guidelines for converting from live voice to recorded test administration. It depends, in part, on the child's previous auditory experience and developmental level. It seems reasonable to attempt recorded testing by the time children are 5 years old. In fact, this age guideline for recorded test administration has been adopted in recent clinical trials of cochlear implant use in children.

Open-Set Versus Closed-Set Test Formats

Open-set tests are those in which listeners theoretically have an unlimited number of response possibilities. That is, no alternatives are provided from which to select a response. Closed-set tests restrict the listener to one of a fixed number of possible responses (e.g., as in a multiple-choice test). Closed-set tests reduce vocabulary and cognitive demands inherent in the speech perception task; they often are used to assess speech recognition in listeners with reduced language skills or limited speaking and writing abilities. However, closed-set tests may not adequately evaluate the perceptual processes that support word recognition in daily life.

In contrast, open-set tests are thought to better reflect the types of communication demands encountered in natural listening situations. For example, performance on open-set tests of spoken word recognition is influenced by cognitive (top-down) processing, just as is speech comprehension in the real world. Cognitive processing is facilitated by an individual's general knowledge (including vocabulary and linguistic knowledge), as well as by expectations based on the situational or linguistic context of the speech event.

Sometimes clinicians or researchers wish to evaluate an individual's sensory capabilities without the influence of cognitive or linguistic factors (see Tyler, 1993) in order to obtain information about the speech cues that are well conveyed by the individual's sensory aid. This information may be helpful in planning therapeutic intervention. Closed-set tests of word or nonsense syllable recognition can be used to assess perception of speech features in the absence of cognitive influences; these tests typically use foils that are acoustically or phonetically similar to the target item. Such closed-set tests of speech-feature perception also can be used to assess the performance of children who have limited auditory-only speech understanding.

Sources of Variability in the Speech Signal

In daily listening, children must cope with a great deal of variability in the speech signal introduced by different talkers, different speaking rates, different dialects, and competing noise, to name a few factors. Although clinical tests of spoken words in quiet yield estimates of children's speech understanding in optimal listening situations, they may not accurately estimate performance in daily living in which there are many sources of competing noise. Whenever possible, it is best to evaluate word recognition using tests containing multiple talkers or competing stimuli.

Multimodal Spoken Word Recognition

Information in the speech signal is conveyed through both the auditory and the visual modalities in face-to-face conversation. Children with similar degrees of hearing loss can differ greatly in their ability to understand speech through listening alone, and also in their ability to integrate auditory and visual speech information (Lachs, Pisoni, and Kirk, 2001). Assessing perceptual performance in the auditory-only, visual-only, and auditory + visual modalities provides information about how well speech is understood through listening alone, and also about the enhancement in speech perception obtained when auditory and visual cues are combined. Furthermore, assessing performance in the auditory + visual modality may better estimate real-world performance.

Measures for Preschool-Age Children

For the very youngest children, subjective communication scales, such as the Infant-Toddler Meaningful Auditory Integration Scale (IT-MAIS) (Zimmerman-Phillips, Robbins, and Osberger, 2000) have been used to assess auditory development. This scale assesses a variety of meaningful auditory behaviors used in everyday life. There are 10 questions. Parents are asked to respond on a scale from 0 to 4 indicating how often their child demonstrates a particular behavior. A 0 indicates that the child never demonstrates a given listening behavior, and a 4 indicates that the child always demonstrates a listening behavior. Thus the total score on the scale can range from 0 to 40.

By the time children reach the age of 3 years, they may be able to participate in formal spoken word recognition testing. Many measures of spoken word recognition have been developed over the years. We briefly present information on some of the more commonly used assessment tools that currently are available. (Interested readers may wish to review the work of Jerger, 1983, for a more thorough description and a comparison of the psychophysical characteristics of pediatric speech tests.) All of the tests described below require the administration of a standard number of test items across children. Other tests, such as the Early Speech Perception Test—Low Verbal Version (ESP-Low Verbal) (Moog and Geers, 1990) and the Early Speech Perception Test (Moog and Geers, 1990) follow a hierarchical approach, in which children move on to successive test items only after they achieve a minimum score on a previous level of the test. Although such tests can be useful in informing aural habilitation programs, the committee's view is that a more consistent test administration procedure is preferable for the purposes of determining disability in children.

The Northwestern University-Children's Perception of Speech Test (NU-CHIPS) (Elliott and Katz, 1980) is a closed-set test of speech discrimination. The test consists of 50 monosyllabic words that can be depicted in simple pictures and fall within the vocabulary of very young children. This is a closed-set picture-pointing task with four pictures per page. The test is intended for children who demonstrate a vocabulary recognition age of at least 2.5 years as measured by the Peabody Picture Vocabulary Test (Dunn and Dunn, 1997).

The Pediatric Speech Intelligibility Test (PSI) (Jerger and Jerger, 1984; Jerger, Lewis, Hawkins, and Jerger, 1980) was developed for children as young as 3 years of age for evaluating a range of auditory disorders. Both monosyllabic words and sentence materials were generated from words produced by children with normal hearing between 3 and 7 years of age. The test consists of 20 words depicted on four picture plates with five items per plate, and 10 sentences depicted on two five-picture plates. The child responds by pointing to the word or sentence that is presented. Jerger, Jerger, and Abrams (1983) found that test-retest reliability of the PSI word test was high for both children with normal hearing and children with hearing loss (r = .92). No significant differences were noted among the four lists of monosyllabic words. The reliability of performance on the PSI sentences also was high (r > .82) for children with normal hearing and those with hearing loss.

The Multisyllabic Lexical Neighborhood Test (MLNT) is an open-set word recognition test whose theoretical underpinnings are the Neighborhood Activation Model (NAM) of spoken word recognition (Luce and Pisoni, 1998). The NAM proposes that words are organized into “similarity neighborhoods” based on their frequency of occurrence (i.e., how often words occur in the language) and the density (i.e., acoustic-phonetic similarity) of words within the lexical neighborhood. The MLNT consists of two lists of 24 two- and three-syllable words containing vocabulary that is suitable for children age 3 years and older. On each list, half of the words are lexically “easy”; that is, they occur often in the language of young children and have few acoustically phonetically similar words with which they can be confused. The remaining “hard” words on the test have the opposite lexical characteristics.

The MLNT can be administered either via live voice or in recorded form. Children respond by repeating the word they perceive. Their responses are scored as the percentage of lexically easy and hard words that are correctly identified. One also can derive a total composite percentage-correct score. It is typical to observe better performance on the easy than the hard words. This suggests that children are sensitive to the acoustic-phonetic similarity among words and that they use this information to encode, store, and retrieve words from their mental lexicon. Use of the MLNT can provide diagnostic information regarding spoken language processing. For example, if children show very large differences in the ability to recognize easy and hard words, it would suggest that they are unable to make fine acoustic-phonetic distinctions among words. Test-retest reliability and interlist equivalency are high (Kirk, Eisenberg, Martinez, and Hay-McCutcheon, 1999).

Measures of Spoken Word Recognition for School-Age Children

The Word Intelligibility by Picture Identification (WIPI) test (Ross and Lerman, 1970) was developed to assess spoken word recognition in children with limited vocabularies who could not read. The test consists of six-picture plates that can be used to evaluate recognition of four different lists of 25 monosyllabic words. Children must make relatively fine acoustic-phonetic distinctions to correctly identify the target word from among the five foils on each plate. The WIPI is recommended for children with mild to moderate hearing loss ages 5-6 and children with severe hearing loss ages 6-7. Ross and Lerman (1970) found good interlist equivalency and test-retest reliability for this measure.

The Lexical Neighborhood Test (LNT) (Kirk et al., 1995) is an open-set monosyllabic word recognition test. Test motivation, administration, and scoring are identical to the MLNT. The LNT consists of two lists of 50 monosyllabic words. Half of the items on each list are lexically hard and half are lexically easy. As for the MLNT, interlist equivalency and test-retest reliability are high.

The Phonetically Balanced Kindergarten Word Lists (PB-K) test (Haskins, 1949) was developed to assess open-set monosyllabic word recognition. The test consists of four lists of 50 monosyllabic words that are phonetically balanced (that is, the sounds in the test items occur in the same proportion in which they occur in spoken English). This test has been the most widely used measure of isolated word recognition in clinics and research centers throughout the United States for more than 50 years. Because the requirement of phonetic balancing within lists constrains test item selection, some words on the PB-K may not be familiar to young children with limited vocabularies (Kirk, Sehgal, and Hay-McCutcheon, 2000). The test can be administered via live voice or recordings. Children respond by repeating the item heard. Responses are scored as the percentage of words or phonemes correctly identified.

The Hearing in Noise Test for Children (HINT-C) was developed by Nilsson, Soli, and Gelnett (1996), based on the work of Bench, Bamford, and colleagues (Bench and Bamford, 1979; Bench, Bamford, Wilson, and Clift, 1979). This test consists of a number of lists of 12 sentences each. Vocabulary was selected to be familiar to children age 6 and older. A recorded version of the test is used to test children in quiet or in the presence of speech-spectrum-shaped noise. The test has been administered in two ways. In one method, the reception threshold for sentences is determined adaptively with the speech and speech-spectrum noise presented from 0° azimuth (directly in front of the patient) and with the noise at 90° (left of the patient) and 270° (right of the patient) relative to the speech signal (0°). In another method, the percentage of words correctly identified in quiet and in noise are determined.

The Common Phrases Test (Robbins, Renshaw, and Osberger, 1995b) was developed to assess the understanding of familiar phrases used in everyday situations. This test was motivated by the idea that children would better be able to recognize familiar phrases than monosyllabic words in an open-set format. Furthermore, such a test has greater face validity than an isolated word test. The test consists of 6 lists of 10 sentences; separate lists can be administered in the auditory-only, visual-only, and auditory + visual modality. Children may respond by correctly repeating all of the words in the sentence or by responding appropriately to a question. Performance is scored as the percentage of sentences correctly identified.

Speech Production Outcome Measures

A battery of tests also is needed for assessing speech production skills in children who are deaf or hard-of-hearing. One common approach is to evaluate vowel and consonant production in a variety of tasks ranging from imitation of nonsense syllables through elicited and spontaneous productions in words and sentences (Tobey, Geers, and Brenner, 1994). Speech tasks that use imitation require children to reproduce an examiner's utterance accurately. Speech tasks that elicit productions (such as picture naming or object description) require children to produce target sounds in the absence of a model from the examiner, in this case, yielding information about the children's speech and expressive language abilities. Finally, spontaneous samples provide a representative sample of the child's usual speech and expressive language skills. When assessing speech production skills, it is common for the examiner to score children's responses on-line, or to record the responses for later transcription by another clinician. This type of scoring may be influenced by the examiner's familiarity with the children or with the speech of other talkers who are deaf or hard-of-hearing.

Language Tests

Researchers have applied several different strategies to examine the same types of behaviors in children with hearing loss. One school of investigators (Geers and Moog, 1994) has developed and applied measures specifically for children with hearing loss, such as the Grammatical Analysis of Elicited Language. Other schools (e.g., Robbins, Osberger, Miyamoto, and Kessler, 1995a) have applied tests developed for children with normal hearing, such as the Reynell Scales of Language Development, using caution in the interpretation of results. Still others (Moeller, 1988) have argued for integrating formal and informal measures, prudently incorporating standardized tests with informal measures, such as following directions, narrative (re)production, spontaneous language samples, etc. Many tests of language development are available. The committee does not recommend any particular test for use in SSA disability determination.


When children are born with a hearing loss, prosthetic devices can be used to amplify the environmental sounds to levels that make them audible (see Chapter 5). If the hearing loss is quite severe, acoustic amplification, such as that provided by hearing aids, may not have enough power to provide audibility, especially for soft speech sounds. In those cases, surgically administered cochlear implants can be used to sample environmental sounds, transduce sounds into a series of electrical impulses, and present these representations of the sound to the auditory nerve directly through the inner ear. Fitting of prosthetic devices to children is a complex process. When care is taken to fit appropriately and follow with educational intervention to train them in the use of the altered signal and to augment development of communication, the communication ability and auditory development of children can be dramatically improved.

Hearing Aids

When hearing loss is identified in infants or toddlers, hearing aids can and should be fitted as soon the degree and type of hearing loss is known in each ear. Hearing aids can be adjusted or programmed to provide a frequency response that is appropriate for the hearing loss of the individual and to provide the appropriate amount of amplification based on the degree of loss for sounds of specific frequency. The amount of amplification and maximum power can also be adjusted in most aids to be appropriate for the hearing loss and the acoustic environment. The characteristics of the hearing aids are selected or programmed to match the severity and frequency response of the child's hearing loss. Consequently, any hearing aid fitting begins with the diagnostic audiological assessment discussed earlier.

Styles of Hearing Aids

Infants and young children are generally fitted with a behind-the-ear (BTE) model of hearing aid. These devices hang behind the pinna, and amplified sound is routed to the ear canal via tubing ending in a custom earmold. Infants' and toddlers' pinnas grow very rapidly, requiring new earmolds on a regular basis to ensure proper acoustic coupling between the device and the child.

Exceptions to the use of BTE aids in children do occur. In rare instances, a young child may need a “body style” hearing aid. These aids are housed in a body-worn case, which is attached to the custom fit earmold via a cord and external receiver. These devices are used only in special cases, such as when a bone-conduction device is needed for a child with no external pinnas on which to attach a BTE or when an ear-worn device is not practical for a bedridden child. Older children and adolescents may be able to use in-the-ear (ITE) or in-the-canal (ITC) hearing aids, depending on the degree of hearing loss.

Special Features and Assistive Devices

Children in learning situations both at home and in classrooms need as clear and quiet a signal as possible. Assistive listening devices such as frequency modulation (FM) or induction coil systems (described in Chapter 5) are often quite beneficial to children and are usually recommended to be used alone or in conjunction with personal amplification devices. Hearing aids equipped with direct audio input connectors can be adapted for this application. Infants and children may not have the sophistication to participate in loudness discomfort measures or to manipulate a volume control to avoid loud sounds. Their amplification systems will require flexibility to set loudness growth and limiting and to allow for manipulation of such parameters as the child learns to respond. Wide dynamic range compression (WDRC) may be a particularly important feature of amplification for an infant or child on whom loudness discomfort levels are not obvious. Amplification devices fitted to children should have flexibility to adjust the gain and frequency response to accommodate changes in the external ear and ear canal due to growth and to respond to any possible changes in hearing sensitivity over time. Because of their flexibility, digitally programmable aids may be advantageous for use by infants and children.

Hearing Aid Candidacy

Infants and children who are in the process of developing language and speech have a critical need to hear. Children with permanent, bilateral average hearing loss of 20-25 dB (or greater) should be evaluated for the use of amplification. Children with unilateral hearing loss should be considered for fitting with amplification whenever possible. Those with a profound unilateral loss that is not helped by conventional aids may receive benefit from an FM system or contralateral routing of signal (CROS) hearing aid, in which a microphone is placed at the nonfunctioning ear and the signal is routed to the contralateral, normally functioning ear.

Fitting and Verification of Amplification

Proper fitting of amplification begins with determination of audiometric thresholds. As in all audiometric procedures, modifications must be made to adult procedures to accommodate the capacity of infants to provide feedback. Prescriptive formulae, such as the NAL-NL1 or DSL (see Chapter 5), should be used to determine the appropriate hearing aid characteristics based on the children's audiometric thresholds, and they should be modified based on the individual acoustic characteristics of each child's ear. Procedures to determine the appropriateness of the fitting compare the prescribed performance to the actual acoustic output. These comparisons can be measured from the child's ear (real ear measures) or in a hearing aid coupler with appropriate corrections made based on actual measurements from that child's ear, called real ear to coupler difference (RECD) (Moodie, Seewald, and Sinclair, 1994). These corrections to coupler measures are especially important when fitting infants and small children to avoid overamplification because of increased sound levels in tiny ear canals.

Outcomes of Hearing Aid Use

For linguistically sophisticated adults, the goal of amplification use is the enhanced perception of speech in a variety of listening environments. For infants or young children, for whom speech has not yet acquired meaning, the audibility of sounds in a wide frequency range is the initial primary goal. Consistent exposure to sounds in environmental context is the basis for development of auditory neural networks that will ultimately be responsible for the organization of auditory information and learning about complex patterns of sounds including speech. If the infant or toddler with hearing loss can be provided with consistently audible speech through amplification, he or she should be able to learn to recognize speech, distinguish voices, and develop spoken vocabulary. Children with significant hearing loss will not be able to monitor their own voice and speech sounds unless these are audible to them. Other more rudimentary skills also are developed through auditory experience, such as awareness and recognition of environmental sounds and development of the ability to find sound-emitting objects in space (sound localization). As children develop, speech perception ability is monitored as a metric of aided performance.

Stelmachowicz (1999, p. 16) describes the most common outcome measures used in children who use amplification as “auditory awareness, audibility of speech, speech intelligibility, accuracy of speech production, rate of language acquisition, loudness discomfort and social development.” Normal developmental time-courses exist for many of these skills in all children, as discussed earlier in this chapter. Such skills can be assessed only after they are expected to be present, regardless of the hearing status. For example, there are few measures of speech recognition for children under 3 years of age. For older children, it is important to distinguish the effects of amplification on outcome measures from normal developmental patterns and from the effects of the therapy and early intervention that necessarily accompany the fitting of amplification.

Surveys and Inventories for Pediatric Hearing Outcome Measures. Hearing loss is now often identified in very young infants due to the advent of newborn hearing screening. Few outcome measures can be used directly with these children. Consequently, it is increasingly common for professionals to use questionnaires or surveys of parents, caregivers, or teachers to gain insights into the early development of auditory behaviors. As children age, more direct objective measures can be taken if allowances for normal developmental patterns of ability are taken into account. However, questionnaires and inventories can be very useful for pre- and postintervention assessment over time, for example in assessing the use of amplification or devices such as FM systems.

Among parent-teacher survey tests is the Screening Instrument for Targeting Educational Risk (SIFTER) (Anderson, 1989), used by educators for school-age children. A preschool version is also available (Anderson and Matkin, 1996). The Meaningful Auditory Integration Scale (Robbins, Renshaw, and Gerry, 1991) and the IT-MAIS infant-toddler version (Zimmerman-Phillips et al., 2000) are parent inventories used generally with children who demonstrate severe to profound hearing loss. These scales probe for hearing aid use, acceptance, and basic auditory development.

The Listening Inventories for Education (LIFE) (Smaldino and Anderson, 1997) is a classroom inventory that includes student and teacher appraisals of listening difficulty and a teacher opinion and observation list.

Kopun and Stelmachowicz (1998) have used an adapted version of the Abbreviated Profile of Hearing Aid Benefit (APHAB) (Cox and Alexander, 1995) successfully with children ages 10-15 with hearing loss as well as with their parents. Discrepancies were found between the children's assessments and those of their parents, which points out the need for development and validation of tools that can be used with children.

Many other functional assessments based on report of behaviors are being developed for assessing outcomes in children with hearing loss. For example, the Functional Auditory Performance Indicators (FAPI), from Stredler-Brown and Johnson (2001), address many functional areas, including localization, discrimination, and short-term auditory memory. It can be used over time to map a given child's progress. Each of these inventories has strengths in evaluating certain aspects of auditory development and may best be suited to a particular environment or degree of hearing loss.

Direct Measures for Toddlers and Young Children. To be appropriate for infants and toddlers, measures of speech perception ability that can be used to evaluate adequacy of amplification must be modified to be age appropriate for both the task and the perceptual skills of the children. If speech stimuli are used, the influence of the children's linguistic capacity on speech perception must be considered carefully. For example, it is easier to perceive words in one's vocabulary or to perceive a sentence when one has the grammatical capacity to construct such a sentence. Eisenberg and Dirks (1995) looked at children's ability to judge speech clarity using paired comparison and category rating tasks. They found that by 5 years of age, children could make reliable clarity judgments of distorted speech, especially with the paired comparison method. Dawson, Nott, Clark, and Cowan (1998) evaluated a test procedure using a play paradigm for assessing the ability of young children to discriminate between pairs of speech sounds. They found that 82 percent of 3- and 4-year-old children and 50 percent of 2-year-old children could perform this task and would reliably indicate when two stimuli were discriminated. Boothroyd and colleagues have developed an imitative speech perception test that allows for evaluation of speech feature perception in toddlers and has the added attraction of allowing a comparison of visual, auditory, and auditory + visual perception ability (Kosky and Boothroyd, 2003). This simple task (IMSPAC) can be used with toddlers as young as 3 years of age. To date, we know of no studies that have used these measures to evaluate the outcome of amplification in children.

Age at Amplification. Both Yoshinaga-Itano et al. (1998) and Moeller (2000) have found that language outcomes in children with hearing loss are significantly related to age at intervention. Specifically, both studies found that if infants are enrolled in an early-intervention program for deaf and hard-of-hearing children before 1 year of age (Yoshinaga-Itano found 6 months to be critical and Moeller found 11 months), there is a significant positive effect on the child's later language ability. Conversely, delayed intervention leads to poorer language outcomes in children. However, in neither study was the intervention necessarily tied to amplification fitting. Consequently, although conventional wisdom based on consistent widespread clinical observations supports the value of fitting amplification as soon as a hearing loss is diagnosed (Joint Committee on Infant Hearing, 2000; Pediatric Working Group, 1996), published data on communication outcomes (language, speech perception, speech production) that confirm the utility of early amplification are not currently available.

Device Efficacy and Features in Children—Audibility. The outcomes of children using amplification can be measured in many ways. At the most basic are tests that evaluate the audibility of sounds when using a device. The most important factor is the audibility of sounds within the frequency range that comprises speech signals. In addition, Stelmachowicz, Pittman, Hoover, and Lewis (2002) emphasize that infants and children may have specific additional needs for adequate audibility of frequencies considered to be high in the speech spectrum (above 3000 Hz). She points out that children who are learning language need to be able to hear and perceive these very high frequency sounds that include “s” because of its importance in denoting possession, plurality, and verb tense. Also, infants and children are often listening to female speech, which is higher in frequency content than male speech.

Assessing the audibility of sounds has become a routine procedure in the process of fitting and evaluating amplification for infants and children (Pediatric Working Group, 1996; American Academy of Audiology, 2003). Procedures such as the desired sensation level input/output (DSL i/o) (Cornelisse, Seewald, and Jamison, 1995) or NAL NL1 (Dillon, Birtles, and Lovegrove, 1999) allow comparison of amplified signals to target prescription levels. Appropriate signals, such as speech or speech-shaped noise, are directed to the hearing aid, and the amplified output is measured in the ear canal of the infants or children under optimal conditions, or in a standard hearing aid coupler with corrections appropriate for estimating sound levels that would occur in infants' or toddlers' ears. Applied stimulus type and level can be manipulated to simulate real listening conditions, and the characteristics of the amplification can be adjusted to maximize the response in a particular infant's ear and to document the audibility of the signal, based on the hearing loss of each child. Scollie et al. (2000) have provided support for the 4.1 version of the DSL by demonstrating that the DSL target levels for amplified speech are very similar to the preferred listening levels of children with hearing loss.

Aided audibility can be also be quantified by the articulation index (American National Standards Institute, 1969). The articulation index is a number from 0 to 1 which approximates the percentage of the speech spectrum that is audible. Stelmachowicz and colleagues have developed a system to provide a visual display of the audibility of speech at various input levels and to take into account the amplification and output-limiting characteristics of the hearing aid. This system is known as the situational hearing aid response profile (SHARP) (Stelmachowicz, Lewis, Kalberer, and Cruetz, 1994). Presently, the above measures of audibility in an aided situation are used clinically as initial indicators of aided performance in infants and children, or a sort of early outcome measure.

Measuring Outcomes of Hearing Aids and Features. Standard measures of hearing aid fitting often include determining speech recognition scores for words or sentences with the speech presented in the context of background noise. Often the level of the noise that can be tolerated relative to the level of the speech is the metric of performance. These measures can be used only with older children (over about age 7). Pettersson (1987) demonstrated that children with hearing loss ages 8 to 20 performed better on speech perception when using their own amplification devices than when evaluated with amplified speech presented through an audiometer. She attributes much of this finding to the fact that the children had accommodated to their amplification and were familiar with the amplification characteristics.

School-age children with all degrees of hearing loss have been shown to have an advantage in listening through hearing aids while using a directional microphone (Gravel and Hood, 1999; Hawkins, 1984; Kuk, Kollofski, Brown, Melum, and Rosenthal, 1999). Gravel et al. (1999) noted that the directional advantage is greatest for older children with more advanced language skills.

FM Systems. Several studies have shown that a significant advantage is afforded to children with hearing loss for perceiving speech in real-life situations when an FM transmission system is used. These advantages have been demonstrated in the classroom (Boothroyd and Iglehart, 1998; Hawkins, 1984; Pittman, Lewis, Hoover, and Stemachowicz, 1999) and to a lesser degree in the home (Moeller, Donaghy, Beauchaine, Lewis, and Stelmachowicz, 1996). Most studies point out that while speech perceived from a distance may be enhanced using FM systems, the need for local microphones must be addressed in order to allow children to monitor their own voices and speech as well as to interact with classmates.

Outcomes with Minimal and Unilateral Hearing Loss. There is clear evidence that children with even unilateral or mild hearing loss are at a significant disadvantage when listening in noisy situations, such as classrooms (Crandell, 1993). However, there are few outcome data on the use of amplification with these children. Kenworthy, Klee, and Tharpe (1990) compared the use of a CROS hearing aid and personal FM devices on children with unilateral hearing loss in a variety of listening conditions. They found an advantage to the CROS aid but only in limited conditions, whereas the personal FM provided uniform advantage for speech perception in these children. However, device fitting for children with unilateral hearing loss is not universally accepted as a clinical procedure. English and Church (1999) reported that only 27 percent of children with unilateral hearing loss are using amplification in the classroom.


Determining Candidacy for Cochlear Implants

Cochlear implants are approved by the FDA for use in children as young as 12 months of age. For very young children, measures of unaided and aided speech detection are the primary determinants of cochlear implant candidacy. In addition, the IT-MAIS (Zimmerman-Phillips et al., 2000) has been used to assess auditory development in relation to cochlear implant candidacy. That is, candidacy is determined by a lack of development of auditory skills over a 3-6 month time span. For older children, spoken word recognition tests are an important part of candidacy determination and post-implant assessment of benefit. Recent clinical trials have employed open-set measures of word recognition, such as the LNT (Kirk et al., 1995) and measures of sentence recognition, such as HINT-C (Nilsson et al., 1996) to determine candidacy in school-age children. For preschool-age children, the open-set MLNT (Kirk et al., 1995) and less difficult closed-set tests, such as the ESP Test (Moog and Geers, 1990), have been utilized to determine candidacy and monitor long-term communication skill development.

Cochlear Implant Communication Outcomes in Children

Spoken Word Recognition

In early investigations, children who used previous generations of cochlear implant systems demonstrated significant improvement in closed-set word identification but very limited open-set word recognition (Miyamoto et al., 1989; Staller, Beiter, Brimacombe, Meckelenburg, and Arndt, 1991). The introduction of newer processing strategies yielded greater speech perception benefits in children, just as in adults. Like adults (see Chapter 5), children with cochlear implants demonstrate a wide range of postimplant communication abilities (Gantz, Tyler, Woodworth, TyeMurray, and Fryauf-Bertschy, 1994; Miyamoto et al., 1991; Staller et al., 1991; Waltzman, Cohen, and Shapiro, 1995). The majority of children with current cochlear implant devices achieve at least moderate levels of open-set word recognition.

For example, Cohen, Waltzman, Roland, Staller, and Hoffman (1999) reported word recognition scores for a group of 19 children that ranged from 4 to 76 percent words correct with a mean of 44 percent words correct. Osberger, Barker, Zimmerman-Phillips, and Geier (1999) reported average scores of approximately 30 percent correct on a more difficult measure of isolated word recognition in children with the Clarion cochlear implant. More recently Geers, Brenner, and Davidson (2003a) reported average word recognition scores of 50 percent correct when the stimuli were presented auditorily only, and scores of 80 percent when the children had access to both auditory and speech-reading cues.

The rate of development of auditory skills following surgery for a cochlear implant seems to be increasing as cochlear implant technology improves and as children are implanted at younger ages (Cohen et al., 1999; Osberger et al., 1999; Young, Carrasco, Grohne, and Brown, 1999). In contrast to adults with postlingual deafness, children's postimplant speech perception skills continue to develop over a relatively long time course. Continued improvements with increasing device use have been noted in children who have used their devices for as long as 5 to 8 years (Tyler, Fryauf-Bertchy, Gantz, Kelsay, and Woodworth, 1997).

Further evidence of the success of cochlear implants in children with congenital or early-acquired deafness was demonstrated by Tyler, Rubenstein, Teagle, Kelsay, and Gantz (2000). They used sentence materials to compare the performance of such children with that of adults with a later onset of deafness. Tyler et al. found that 60 percent of the adults and 70 percent of the children tested correctly identified at least half of the words presented in sentences. These data suggest that the auditory system of children with congenital or prelingually acquired hearing loss will develop spoken language processing abilities similar to those of adults who had the benefit of learning language through a normal auditory channel.

Several investigators have compared the performance of children with cochlear implants to that of their peers who use conventional amplification. The speech perception abilities of pediatric cochlear implant recipients met or exceeded those of their peers with unaided pure-tone average thresholds ≥ 90 dB HL who use hearing aids (Meyer, Svirsky, Kirk, and Miyamoto, 1998; Svirsky and Meyer, 1999).

Demographic Factors

A number of demographic factors have been shown to influence speech perception and spoken word recognition in children with cochlear implants. Early results suggested better speech perception performance in children deafened at an older age with a corresponding shorter period of deafness (Fryauf-Bertschy, Tyler, Kelsay, and Gantz, 1992; Osberger, Todd, Berry, Robbins, and Miyamoto, 1991; Staller et al., 1991). Those children who lost their hearing after acquiring spoken language (usually becoming profoundly deaf after age 2 years) are categorized as postlingually deafened. This group of children is able to use multichannel cochlear implants in a manner similar to adults with postlingual deafness. In most instances, the speech perception skills of children with postlingual deafness improved over a few months and were at the upper end of adult performance curves for sentence and monosyllabic word understanding (Gantz et al., 1994). It is most likely that the short duration of deafness as well as a young and adaptive central nervous system was responsible for the outstanding results in this group of cochlear implant recipients.

When only children with prelingual deafness are considered, age at onset of hearing loss does not appear to significantly impact speech perception and spoken word recognition outcomes (Miyamoto, Osberger, Robbins, Myers, and Kessler, 1993). It is clearly evident that earlier implantation yields superior cochlear implant performance in children (Fryauf-Bertschy, Tyler, Kelsay, Gantz, and Woodworth, 1997; Illg et al., 1999; Lenarz et al., 1999; Nikolopoulos, O'Donoghue, and Archbold, 1998; O'Donoghue, Nikolopoulos, Archbold, and Tait, 1999). Although the critical period for implantation of congenitally or prelingually deafened children has not been determined (Brackett and Zara, 1998), preliminary evidence suggests that implantation prior to age 3 years may yield improved results (Waltzman and Cohen, 1998; Waltzman et al., 1995, 1997). Finally, the variables of communication mode and unaided residual hearing also influence speech perception performance (Cowan et al., 1997; Hodges, Ash, Balkany, Scholffman, and Butts, 1999; Osberger and Fisher, 1998; Zwolan et al., 1997). Oral children and those who have more residual hearing prior to implantation typically demonstrate superior speech understanding.

Speech Intelligibility and Language

Improvements in speech perception are the most direct benefit of cochlear implantation. However, if children with cochlear implants are to be fully integrated into the hearing world, they must also acquire the language of their surrounding community and be able to produce it intelligibly. The speech intelligibility and language abilities of children with cochlear implants improve significantly over time (Allen, Nikolopoulos, and O'Donoghue, 1998; Moog and Geers, 1999; Svirsky, Robbins, Kirk, Pisoni, and Miyamoto, 2000; Waltzman et al., 1995) and, on average, exceed those of their age- and hearing-matched peers with hearing aids (Tomblin, Spencer, Flock, Tyler, and Gantz, 1999; Svirsky, 2000; Svirsky et al., 2000). Tobey and her colleagues reported average speech intelligibility scores of greater than 65 percent for a large group of children with cochlear implants (Tobey, Geers, Brenner, Altuna, and Gabbert, 2003). That is, more than half of what they said could be understood by listeners who were not familiar with the speech of deaf talkers. This is far greater than previous reports for children with profound deafness who used hearing aids (Smith, 1975). Speech intelligibility and spoken language acquisition are significantly correlated with the development of auditory skills (Moog and Geers, 1999; Tomblin et al., 1999). Although a great deal of variability exists, the best pediatric cochlear implant users demonstrate highly intelligible speech and age-appropriate language skills. These superior performers are usually implanted at a young age and are educated in an oral/aural modality (Moog and Geers, 1999).

The relationship between language development and literacy (reading and writing skills) in children with prelingual deafness who use a cochlear implant was investigated by Spencer, Barker, and Tomblin (2003). Children with cochlear implants scored within one standard deviation of their age-matched peers on measures of language comprehension, reading comprehension, and writing accuracy. The researchers reported that the children with cochlear implants differed from children with normal hearing in their ability to utilize correct grammatical structures and verb forms. This information should be useful in designing appropriate educational models for children with cochlear implants.

A comprehensive investigation of the factors that influence cochlear implant outcomes recently was completed by Geers and her colleagues (Geers, 2003; Geers and Brenner, 2003; Geers et al., 2003a; Geers, Nicholas, and Sedey, 2003b; Tobey et al., 2003). The primary goal of this study was to identify the impact of educational factors on cochlear implant outcomes. Because the characteristics of the family and the cochlear implant recipient also contribute to postimplant outcomes, these characteristics were carefully documented and controlled in the data analyses. The study evaluated postimplant outcomes in 181 children who were 8-9 years of age and who were implanted by the time they were 5 years of age. Performance outcomes included measures of speech perception, speech production, language, and reading skills development. The predictors of performance were very similar for the development of speech perception and speech production abilities. These included a higher nonverbal intelligence quotient (IQ), longer use of the newest generation of speech processing strategies, a fully active electrode array and higher electrical dynamic range, and a greater growth of loudness on the electrodes. When these characteristics were controlled, educational placement did have a significant impact on cochlear implant outcomes (Geers et al., 2003a; Tobey et al., 2003). That is, children who were educated in settings that emphasized the development and use of speaking and listening skills had the best prognosis for spoken language processing. Predictors of good language skills included a higher nonverbal IQ, smaller family size, higher SES, and being female (Geers et al., 2003b). These factors similarly impacted language development for children who used oral communication and those who used total communication. Finally, predictors of reading success included a higher nonverbal IQ, higher SES, being female, and later onset of deafness within the period from birth to 36 months (Geers, 2003). When the variance due to these factors was controlled, the development of good reading skills was associated with mainstream educational placement, use of a current generation of speech processor strategy with a wide dynamic range, a longer memory span, and the use of phonological coding strategies. Geers (2003) reported that reading competence was best predicted by language competence and speech production abilities.

In summary, cochlear implant technology and candidacy criteria have evolved greatly over the past 20 years. Today, patients with severe to profound deafness as young as 12 months old may be implanted. With earlier implantation and improved cochlear implant systems come continued increases in the benefits of cochlear implantation. Although wide variability in outcomes is noted, cochlear implant use by children with severe to profound hearing loss promotes the development of speaking and listening skills and the development of a spoken language system beyond what previously could be achieved with hearing aids. Children who are implanted at a young age and use oral communication have the best prognosis for developing intelligible speech and age-appropriate language abilities. However, it is also helpful to note that data from a Gallaudet Research Institute survey returned by 439 parents of children with cochlear implants indicate that roughly 50 percent continue to use sign language with their children as a supplement to spoken language (Christiansen and Leigh, 2002).

The potential for optimal speech and language development in young children is greatly influenced by parent intervention. This requires considerable time commitment on the part of motivated parents, directed toward ongoing activities that reinforce spoken language, as well as extensive speech and auditory therapy (Christiansen and Leigh, 2002). Generally, parents believe that their children's communication skills improve following implant (Christiansen and Leigh, 2002; Kluwin and Stewart, 2000). While this improvement does result in improved socialization with hearing peers, obstacles related to speech intelligibility and receptive understanding of what is communicated as well as attitudes of hearing peers continue to be ongoing factors (Bat-Chava and Deignan, 2001; Boyd, Knutson, and Dahlstrom, 2000). Psychological difficulties following implant, as reported by parents, are generally associated with getting the implant closer to adolescence and not being happy with it (Christiansen and Leigh, 2002).


Disability Determination Tests and Criteria

Action Recommendation 7-1. The recommended criteria for determination of disability in children who are deaf or hard-of-hearing are presented in tabular format in Table 7-2. To be considered disabled, children under 3 years of age must meet the criterion for hearing level only. Children older than 3 years must meet the pure-tone average criterion and either the criterion for deficit in speech perception or the criterion for delay in language. Children with marked mental retardation who cannot be evaluated for speech perception or language should be considered disabled if the hearing level criterion alone is met.

TABLE 7-2. Summary of Recommended Disability Determination for Children Who Are Deaf or Hard-of-Hearing.


Summary of Recommended Disability Determination for Children Who Are Deaf or Hard-of-Hearing.

Action Recommendation 7-2. Speech perception tests are administered in quiet using recorded test materials at 70 dB SPL (based on peak levels or the equivalent in dB HL if available). Presentation of the speech perception test is via sound field using personal amplification or cochlear implant if such is used by the child. A preliminary check of the functioning and appropriateness of the cochlear implant or amplification should precede the aided testing (see Chapter 5). If no device is used by the child, testing is performed unaided.

Action Recommendation 7-3. In general, average hearing levels will be determined from thresholds at 500, 1000, 2000, and 4000 Hz (PTA 5124). This differs from the PTA of 500, 1000, and 2000 Hz (PTA 512) used for adults. When thresholds are obtained with ABR, a minimum of two frequencies, one low (500 to 1000 Hz) and one high (2000 to 4000 Hz), can be used to determine average hearing level. It should be noted that when conditions indicate that auditory neuropathy may be present (absent ABR and normal OAE), it will not be possible to determine hearing thresholds by ABR. In those cases, disability should be presumed unless or until proven otherwise by age-appropriate behavioral testing (see Table 7-1). Behavioral threshold testing by VRA should be possible once the infant is older than 6 months of developmental age.

Action Recommendation 7-4. The committee carefully considered these guidelines in the preparation of recommendations. In contrast to the suggested adult standards for PTA hearing level, we are suggesting that the degree of hearing loss that is considered disabling in infants and children is 35 dB HL or greater before the age of 6 years, 50 dB from 6 to 12 years, and 70 dB from 12 to 18 years of age. We have selected these criteria for the reasons stated below.

As emphasized in the text of this chapter, a loss of hearing sensitivity can have a more detrimental effect on infants and children who are in critical learning periods for speech, language, and general communication ability than on their adult counterparts. School-age children depend on communication skills for all means of learning. Development of communication skills may be the most important task for an infant because it provides the basis for almost all subsequent learning.

The committee chose 35 dB HL as the minimum hearing loss criterion on the basis of studies that have documented significant delays in speech production, language, verbal intelligence, and associated areas of learning (Briscoe et al., 2001; Davis et al., 1986) in children with hearing loss, including those with mild loss. The strongest evidence that even mild bilateral hearing loss is debilitating to young children comes from the endorsement of groups like the National Institutes of Health (1993), the American Academy of Pediatrics Task Force on Newborn and Infant Hearing (1999), and the American Academy of Pediatrics (2000), that endorse programs for early detection of mild hearing loss of 30 to 40 dB because of evidence that this degree of loss will cause significant communication and educational delays.

The needs for intervention for children with hearing loss are particularly acute in the infant and preschool period, when peak gains are attained in language and speech. Particularly if language skills are developing well, the elementary-school-age child should be able to tolerate slightly more hearing loss and the criterion level for disability is adjusted to moderate (50 dB) rather than mild. It should be noted that at this age and older, assessments of speech perception and language skills are also recommended as part of the disability determination evaluation. Although the hearing level criterion for disability is less stringent at this age for a child than for an adult, it should be noted that a child with a moderate hearing loss whose speech perception skills and language ability are normal for his or her age may not be considered disabled. As these communication skills emerge, more emphasis is placed on them and the criterion for hearing level is raised.

Finally, we recommend a 70 dB HL criterion for high school students, which is less strict than the 90 dB recommended for adults. The child of high school age with a hearing loss is significantly challenged by the hostile acoustic environment in which he or she must function and learn and by the increased social-emotional pressure that accompanies this period of development. Therefore, we do not recommend the adult standard of 90 dB HL average hearing loss until the child has reached 18 years of age. We recommend that speech perception be evaluated in the auditory mode that is optimal for individual children. In most instances, the optimal mode will be binaural, well-fitted amplification or a cochlear implant. We recognize, however, that some children may not use amplification devices for a variety of reasons. The determination of disability in these children should be without amplification. A guiding principle is that disability determination should reflect the real-world situation of the child.

It is clear that the degree of hearing loss alone is not a perfect predictor of functional communication competencies. Therefore, the determination of disability requires assessment of a variety of communication skills. We chose the domains of speech perception and language processing because they are directly affected by childhood hearing loss and provide an important foundation for spoken language communication, reading, and multiple facets of educational achievement.

Given the SSA criterion, it would be ideal to select and administer standardized measures that yield normative scores allowing comparable cutoff criteria for both types of tests. At present, this may be possible for language testing but not for speech perception testing.

Standardized language measures provide normative data for children as a function of their age at time of testing. That is, within a given age range, normally developing, hearing children exhibit a normally distributed range of performance scores. By using normative data, we can determine where on the normal distribution an individual child's performance lies. Standard language assessment scores are normalized to a mean of zero and a standard deviation of 1. Because abnormally poor performance occurs in only one tail of the distribution, we use a standard score of > = 1.64 units below the age norm to indicate performance below the 5th percentile.

In contrast, spoken word recognition instruments containing age-appropriate vocabulary and response tasks should yield a skewed distribution with expected scores of 90-100 percent in normally developing, hearing children. This homogeneity of performance on speech perception tasks makes it impossible to derive a standard score for performance. Thus, we have taken a different approach in recommending guidelines for determining marked or extreme impairment in the areas of speech perception performance. We specified as the criterion for abnormality, scores that are significantly different from 90 percent correct (P <= 0.05). The percentage score deemed abnormal will vary based on the number of items in the test as indicated in Table 7-3. Table 7-4 summarizes the characteristics of the spoken word recognition measures described in this chapter.

TABLE 7-3. Cutoff Scores Determined to Be Significantly < 90 Percent at p <= 0.05.


Cutoff Scores Determined to Be Significantly < 90 Percent at p <= 0.05.

TABLE 7-4. Description of Speech Recognition Tests for Children.


Description of Speech Recognition Tests for Children.

The committee has not recommended assessment of speech perception in noise for children, although to do so in our view would increase the sensitivity of the testing for disability. Such testing was not recommended because of the lack of appropriately standardized tests across the age groups of interest. When tests exist, they may not be readily available to audiologists. With young children, test time and cooperation often are limited, and these test scores are not often available for scrutiny. However, checks for functional disability in Step 3 testing should detect the child's significant difficulty hearing in noise. As standardized testing protocols to assess this function become more available, this information should be incorporated into a complete assessment of ability of children with hearing loss.

Action Recommendation 7-5. In addition to the administration of standard audiometric and language tests, the evaluation for determination of disability for children who are deaf or hard-of-hearing should include a checklist to be completed by the audiologist. The checklist presented in Box 7-1 will ensure that information needed for test interpretation is available to SSA. For each checklist item, a response in the shaded box indicates that the response is invalid or needs explanation, as discussed below.

Box Icon

BOX 7-1

Pediatric Checklist for Audiological Evaluation for Disability Determination. Comments from audiologist: ______________________________________________________________

Recommendations for Further Research

This chapter reveals many unresolved issues with respect to determination of disability status for infants and children with hearing loss. Persons involved in the care and management of these children rely on published laboratory and clinical science for direction. Many of the questions raised could be answered with appropriate scientific investigation.

We recommend that the SSA partner with other research funding agencies for whom these questions are also relevant, such as the U.S. Department of Education and the National Institutes of Health, particularly the NIDCD and the National Institute of Child Health and Human Development. Issues that are particularly relevant to SSA are emphasized below. The top three listed here are considered highest priority research aims for SSA purposes; the others listed are not in order of priority.

Research Recommendation 7-1. There is a distinct need for standardized speech perception measures for infants and children that take into account developmental age and degree of hearing loss. Such tests should be developed in English as well as in other languages spoken in homes across the United States. When possible, such tests should incorporate evaluation of perception both in quiet and with relevant competing messages to simulate real-life situations, such as classrooms.

Research Recommendation 7-2. Often children with similar audiograms have very different higher-level auditory abilities, such as spoken word recognition. There are even less-direct relationships between severity of hearing loss and such long-term outcomes as educational achievement, vocational status, and overall psychosocial development. It is important to identify other factors that contribute to these outcomes in persons who have been deaf or hard-of-hearing since early childhood. Specifically, it is important to determine how the following factors influence long-term outcomes for children who are deaf or hard-of-hearing:

  • complexities of linguistic environment in the home and in the educational setting, such as multilingual environments, signed or spoken instruction.
  • the complexities of educational intervention, including the types of intervention, age at inception and duration, educational setting (mainstream, self-contained, home), and training of intervention specialists.

Research Recommendation 7-3. Many studies have documented the expected outcomes for children using cochlear implants, yet documenting benefits and outcomes of amplification use in children is a more complex task, and few controlled studies exist. More prospective studies of children using amplification are needed to determine related outcomes for communication, socialization, and educational achievement. For example, better studies are needed to determine how factors specific to the amplification fitting process and specific to the child influence such outcomes as language development, speech production, perception development, and social development. Examples of factors surrounding amplification fitting include type of aid fitted, features used such as directional microphones, feedback suppression, type of compression, or specific fitting formula selected. Factors specific to the child that could influence outcomes include the child's age at fitting, degree of hearing loss, and other disabling conditions.

The following research aims may do less to refine disability criteria for SSA but are extremely important in terms of understanding auditory issues relevant to communication development in infants and children who are deaf and hard-of-hearing.

Research Recommendation 7-4. Little is known about the interaction between vision and audition in the development of communication skills. Children with hearing loss often rely heavily on visual perception of speech (speech reading or speech feature cueing) to supplement their auditory capacity. Large individual differences may occur in the reliance on visual information when auditory information is incomplete. Currently, there are no standards for measuring visual, auditory, or combined perception of speech. There is a great need for the development of standard clinical measures that incorporate auditory and visual assessments.

Research Recommendation 7-5. There is a need to understand the effects of slight and unilateral hearing loss on the development of communication skills, educational performance, and social adjustment. Studies are needed to reveal the true nature of the dysfunction these children experience, especially in educational settings, and the possibilities for intervention that may help to mitigate such disturbances.

Copyright 2005 by the National Academy of Sciences. All rights reserved.
Bookshelf ID: NBK207837


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...