J Am Acad Audiol 10 : 14-25 (1999)
Hearing Aid Outcome Measures for Children
Patricia G. Stelmachowicz*
Abstract
The provision of appropriate amplification for a young hearing-impaired child is critical as the
aided speech signal will be used for the development of speech and language . The purpose
of this paper is to provide an overview of the complex issues surrounding the documentation of hearing aid outcomes in the pediatric population . In the first two sections of the paper,
the unique characteristics and needs of the pediatric population and factors complicating the
measurement of outcome are described in detail . The third section provides a review of literature on existing outcome measures for children and the fourth section is devoted to a
discussion of alternative approaches . The final section is an overview of clinical and research
needs in the area of hearing aid outcome measures for children .
Key Words : Children, hearing aid, hearing loss, outcome
Abbreviations : AAI = Aided Audibility Index; ABR = auditory brainstem response ; AI =
Articulation Index; APHAB = Abbreviated Profile of Hearing Aid Benefit; CL = compression
limiting ; COSI = Client-Oriented Scale of Improvement; DSL= Desired Sensation Level; HPIC
= Hearing Performance Inventory for Children ; LDL = loudness discomfort level; LIFE =
Listening Instrument for Education ; PC = peak clipping ; SIFTER = Screening Instrument for
Targeting Educational Risk ; WDRC = wide dynamic range compression
umerous approaches are available to
evaluate the efficacy of a particular
N hearing aid fitting for adults . Objective measures of benefit typically have been
conducted in a structured clinical or laboratory
setting and have focused on speech recognition
under a variety of test conditions . It has been
suggested that this type of approach may not produce results that are highly correlated with the
hearing aid benefit experienced in everyday listening situations (Haggard et al, 1981 ; Berger
and Hagberg, 1982 ; Oja and Schow, 1984). This
is not too surprising, given the wide range of listening environments that a typical hearing aid
user may experience during a day. It is also
likely that the available clinical measures of
speech recognition may not be sensitive enough
to detect differences in performance produced by
various types of signal processing .
To circumvent the limitations of these objective clinical measures, a wide variety of subjective measures of hearing aid benefit have been
*Boys Town National Research Hospital, Omaha,
Nebraska
Reprint requests : Patricia G . Stelmachowicz, Boys
Town National Research Hospital, 555 North 30th Street,
Omaha, NE 68131
14
developed (see Weinstein [19971 for a review of
this literature) . The vast majority of these are
self-assessment inventories that attempt to
quantify hearing aid benefit across a range of
everyday listening conditions . This approach
has high content validity and studies have shown
good test-retest reliability for adults (Cox and
Alexander, 1995) . In recent years, these types of
metrics have gained increased use in clinical
settings to document hearing aid benefit and to
evaluate the efficacy of different types of signal
processing .
For adults and older children, a wide variety of objective and subjective metrics are available to document outcome. The various tests can
be used in isolation or in combination with
other tests in order to answer a particular clinical question . However, much of the methodology and many of the metrics typically used with
adults in the assessment of hearing aid benefit are not appropriate for use with infants and
young children . Advances in the early identification of hearing loss have increased the need
to make decisions early in life that are dependent upon demonstrating or failing to demonstrate device or treatment efficacy (e .g ., sensory
aid decisions, communication mode decisions) .
Appropriate amplification will contribute to the
speech and language acquisition process by
Hearing Aid Outcome Measures for Children/Stelmachowicz
supporting the child in extracting the rules of
phonology and syntax. To provide the best possible amplification device and take advantage
of technological advances, there is a critical
need for valid and reliable outcome measures
for this population .
This paper is divided into five sections . In
the first two sections, the unique needs of the
pediatric population and factors complicating the
measurement of outcome will be described in
detail . The third section provides a review of literature on existing outcome measures for children . The fourth section will be devoted to a
discussion of alternative approaches to document outcome . The final section will provide an
overview of clinical and research needs in this
area .
results (Boothroyd, 1991) . Subjective feedback
from the child will be minimal and informal observations by parents or other professionals often cannot be easily interpreted in terms of cause and
effect .
Numerous physical differences exist as well .
Both the ear canal resonance (Kruger, 1987 ;
Kruger and Ruben, 1987) and the real-ear-to-coupler difference (Nelson Barlow et al, 1988 ; Feigin et al, 1989 ; Moodie et al, 1994) have been
shown to vary as a function of age within the first
few years of life . It is also likely that the average speech input level to the hearing aid microphone will be substantially greater for infants
and toddlers than for adults due to proximity dif-
PEDIATRIC POPUIATION :
UNIQUE CHARACTERISTICS
ions and philosophies regarding the electroacoustic requirements for infants and young
children . Some investigators have suggested
that prescriptive hearing aid procedures developed for adults can be applied to young children (Byrne and Ching, in press ; Jordt et al, in
press) . These investigators argue that the revised
National Acoustics Laboratory (NAL) approach
AND UNIQUE NEEDS
he process of fitting hearing aids to an
T infant or young child is qualitatively dif-
ferent than the adult case in a number of ways .
Although newborn screening programs have
enabled audiologists to identify hearing loss
within the first few days of life, precise quantification of the degree and configuration of
hearing loss may not be possible for many
months . However, since recent data have shown
that children with hearing loss who are involved
in early intervention prior to 6 months of age outperform children who are not involved in intervention until later (Moeller, in press ; YoshinagoItano et al, in press), it is not considered reasonable to delay amplification until the child is
capable of providing a comprehensive behavioral audiogram (Pediatric Working Group,
1996) . For infants and older children with multiple disabilities, it is often necessary to select
amplification characteristics based largely upon
tone-burst auditory brainstem response (ABR)
thresholds . Because only a limited amount of
audiologic information is available early on,
hearing aid fitting schemes that require more
than audiologic thresholds cannot be used .
Even when behavioral measures can be
obtained, the results differ substantially from
the adult case . Test reliability may be affected by
developmental issues, such as level of cooperation,
attentiveness, and/or middle ear disease. Children's speech perception skills need to be measured within the confines of their receptive
language abilities, which may limit the procedures
that can be used and complicate interpretation of
ferences between parent and child in the first few
years of life (Stelmachowicz et al, 1993) .
At the present time, there are differing opin-
(Byrne and Dillon, 1986), which is based on the
amplification of all speech bands to equal loudness, should be appropriate for hearing-impaired
individuals of all ages .
Other investigators maintain that the amplification needs of infants and young children
may differ from those of adults who generally
acquire hearing loss in later life (Stelmachowicz, 1991 ; Seewald et al, 1996). Specifically, the
latter investigators feel that the goal of amplification may be different for young children in
that the hearing aids will be used for the development of language, speech perception, and
speech production . Many of the available prescriptive procedures were based upon the preferred hearing aid characteristics of adult users
or were validated in this group. Young children
with hearing loss have not yet developed the linguistic and world knowledge that supports
speech perception for adults in the form of contextual and redundancy cues .
Nittrouer and Boothroyd (1990) investigated
the effect of context on the perception of four-word
sentences in 4- and 5-year-old normal-hearing
children and adults . The subject's task was to
repeat both meaningful and nonmeaningful sentences presented at two different signal-to-noise
(SIN) ratios . While the addition of context
improved scores for both groups, the children
showed significantly smaller improvements in
15
Journal of the American Academy of Audiology/Volume 10, Number 1, January 1999
perception compared to the adults . These results
suggest that, under adverse listening conditions,
adults may be able to use contextual cues to "fill
in the blanks" in a way that is not possible for
young children. It follows that a less than optimal hearing aid fitting for an adult may not
affect the perception of speech as much as would
be the case for a child who is attempting to learn
speech and language through audition .
In addition to these semantic differences, a
number of investigators have shown that young
children may differ from adults in their ability
to use certain types of acoustic cues . Nozza et al
(1991) compared the discrimination thresholds
of two speech contrasts (/ba/ vs /da/ and /ba/ vs
/ga/) between 7- to 11-month-old infants and
adults . The infants required stimulus intensities
of 25 to 28 dB more than the adults in order to
reach a criterion level of performance. It also
appears that children tend to use slowly changing dynamic or temporal cues, whereas adults
tend to rely more on static spectral cues (Morrongiello et al, 1984; Nittrouer and StuddertKennedy, 1987 ; Nittrouer, 1992).
Some investigators (Gatehouse, 1992, 1993)
have shown that the speech perception performance of adult hearing-impaired listeners
improves over time with repeated exposure to
amplified speech . To date, there have been no systematic studies of acclimatization in young children . The existence of this phenomenon in
children could have a potential impact on the
ability to compare different types of signal processors during the hearing aid selection process . In
addition, the use of vastly different types of processing schemes in multiple-memory hearing
aids will alter the signal that is received by the
child. Studies to determine the magnitude of
acclimatization in the pediatric population and
the clinical significance of the effects are needed .
The issue of which prescriptive procedure,
if any, is optimal for young children has not
been resolved . Ross and Levitt (1997) point out
that, for many hearing-impaired individuals,
virtually any type of amplification device will
provide an improvement over unaided listening.
As such, objective and/or subjective comparisons of unaided and aided benefit in no way
guarantee that a hearing aid fitting is optimal.
For young children, it may be extremely difficult,
if not impossible, to ensure that a particular
set of hearing aid characteristics is substantially better than other alternatives . A small
number of studies have been conducted with
preschool and school-aged children in which use
gain or preferred gain was compared to the gain
16
recommended by various prescriptive procedures with varying results (Snik and Hombergen, 1993 ; Ching et al, 1994, 1996; Snik and
Stollman, 1995 ; Snik et al, 1995). These types
of studies are difficult to interpret because "use
gain" will be highly dependent upon the rationale and procedures used to fit the hearing aid
initially. Similarly, "preferred gain" may be
biased by the characteristics of the initial fitting
and the effects of acclimatization; if the child is
accustomed to listening with particular frequency/gain characteristics, this response may
be preferred over other alternatives in a pairedcomparison task . In the absence of a consensus
on how to determine optimal hearing aid characteristics for children, the need for valid and
reliable outcome measures that can be applied
on an individual basis is critical .
PEDIATRIC POPULATION :
FACTORS COMPLICATING THE
MEASUREMENT OF OUTCOME
W
hat should be used as an outcome measure?
Potential outcome measures include audi-
tory awareness, audibility of speech, speech
intelligibility, accuracy of speech production,
rate of language acquisition, loudness discomfort, and social development . The metric selected
will depend on both the age of the child and the
degree of hearing loss and will be influenced by
the age at which hearing loss identification
occurs, concomitant middle ear problems, the
existence of additional handicapping conditions,
developmental factors, and the child's home and
educational environments . The selected outcome measures also will depend upon the clinical question to be addressed . The question may
be relatively broad, such as : What is the best
hearing aid for a particular child? Or, given current hearing aid technology, more specific questions may be asked : Is a binaural or monaural
fitting most appropriate for a child with an
asymmetric hearing loss? Should a directional
or omnidirectional hearing aid be recommended?
How much high-frequency gain should be recommended for a child with a precipitous hearing loss? Should compression limiting (CL) or
wide dynamic range compression (WDRC) be
used for a child with a moderate-to-severe hearing loss? Under what conditions should a multimemory device be used and what hearing aid
characteristics should be programmed into each
memory? What real-ear saturation response
will produce levels that are both safe and comfortable? Obviously, each of these questions may
Hearing Aid Outcome Measures for Children/Stelmachowicz
require a different kind of outcome measure or
some combination of measures .
Over what time period should outcome measures be made? With children, one would not
expect to see immediate changes in speech production or language skills following the introduction of amplification or a change in hearing
aid characteristics . Repeated auditory experiences or acclimatization (Gatehouse, 1993) also
may be necessary in order to observe changes in
speech perception . In most cases, longitudinal
monitoring may be required to identify changes
in performance that can be attributable to amplification (e .g., improved speech production of a
now audible sibilant) . The time frame for changes
in behavior or performance is likely to be longer
than is typical for an adult hearing aid user .
How can ongoing developmental changes be
separated from changes related specifically to
amplification? If it is necessary to use a longitudinal approach to document outcome, it may
be difficult to separate the effects of normal
development or the influence of other treatments (e .g., speech and language therapy) from
the effects related to hearing aid use.
At what developmental age will outcome
measures developed for adults be valid for use
with children? For some subjective metrics it may
be sufficient to alter the vocabulary, language
level, and described situations to create an
appropriate assessment tool . However, auditory
goals and the child's ability to provide subjective
feedback can be expected to change over time in
a way that generally does not occur for adults .
Thus, for most types of outcome measures developed for the pediatric population, it will be necessary to develop a series of metrics to reflect
changes in performance as the child matures .
CURRENT OUTCOME MEASURES
Objective Measures of Outcome
For adults and older children, the most commonly used objective outcome measure is speech
recognition . A wide variety of tests are available,
measuring recognition of stimuli ranging from
nonsense syllables to words and sentences . The
materials, SIN ratio, and response format (closed
set, open set) can be tailored to suit the specific
objectives for each case .
Unfortunately, the situation is very different for young hearing-impaired children . Under
the best of circumstances, it is not possible to
obtain reliable measures of speech recognition
for children less than 3 years of age. Beginning
at about 3 years of age, the most commonly used
clinical tests require the child to select the appropriate response from among a small set of pictures (e .g ., Ross and Lerman, 1970 ; Katz and
Elliott, 1978) . While the words and pictures typically are within the repertoire of children in the
3- to 5-year age range, for hearing-impaired
children, it is often impossible to separate factors such as auditory experience, participation
in speech and language therapy, educational
environment, and phonologic development from
the true residual capacity of the damaged cochlea
(Boothroyd, 1991 ; Boothroyd et al, 1996) . That
is, a young child's previous experiences will
influence performance on this type of test and
are likely to complicate the interpretation of
results . If the goal of a test is to determine how
well a child is performing overall, these interactions may not be a problem . If the intent is to
assess the efficacy of a particular signal processing scheme, however, a valid test of auditory
speech perception capacity should be devoid of
lexical, syntactic, and semantic constraints .
While nonsense syllables meet this goal, the
use of these stimuli with young children is problematic in that it is difficult to provide an appropriate visual reference (e .g ., written words,
pictures) for each test item . An imitative test
format has been developed for use in research
studies with children as young as 3 years of age
(Boothroyd, 1991 ; Boothroyd et al, 1996) . In its
current form, this procedure requires post-test
editing of speech samples and scoring by a group
of normal-hearing listeners ; thus, it would not
be applicable in a typical clinical setting .
Recently, Boothroyd (1991) has developed a
PC-based video game (VIDSPAC) for evaluating
the perception of various acoustic contrasts in
children as young as 3 to 4 years of age . The program allows the child to select a colorful character to "present" the test stimuli and a wide
range of speech pattern contrasts can be evaluated (e .g., vowel height, final consonant voicing) .
The child's task is to respond when a change in
a string of stimuli is detected (/ta/, /ta/, /ta/, /da/,
/ta/). Following each block of trials, an animated
scene is displayed as a reward. In addition to the
well-designed graphics, which are engaging for
young children, the program has other features
that are important for this population. The interstimulus interval can be varied during the experiment according to the child's needs and
summary statistics provide an account of both
hits and false positive responses . This type of test
should be extremely useful in extending the
lower age limit at which valuable estimates of
17
Journal of the American Academy of Audiology/Volume 10, Number 1, January 1999
speech perception can be obtained from hearingimpaired children .
For older children, a much wider range of
objective tests is available. These include tools
that measure the perception of nonsense syllables, single words, sentences, and story comprehension . Special purpose materials are
available for children with varying degrees of
hearing loss and for auditory-alone versus auditory/visual perception. (See The Pediatric Working Group [19961 for a review of current speech
tests for this population.) As previously mentioned, it is important to note the limitation of
speech recognition as an outcome measure for
both children and adults . While these objective
tests can provide a quantitative estimate of performance that can be compared to expected
goals, it is unlikely that these measures will be
sensitive enough to compare different hearing
aid characteristics.
Subjective Measures of Outcome
To date, only a small number of subjective
tests are available for children and most of these
have been developed for use in educational settings exclusively. The Hearing Performance
Inventory for Children (HPIC), developed by
Kessler et al (1990), is a self-assessment instrument for children (ages 8-14 years) that samples
a child's perceived communication difficulties in
a variety of academic environments. The primary
goal of this test is to develop a personal profile
of difficult academic communication situations
to aid in the design of an individualized management program . The test consists of 31 test
items depicting a variety of typical classroom listening situations (e .g., hearing the teacher when
his/her face is not visible) . For each test item, a
picture and a verbal description are given and
the child's task is to rate communication difficulty on a 5-point scale . In the development of
the HPIC, test-retest measures were obtained
on 15 normal-hearing and 7 hearing-impaired
children, but no large-scale studies have been
reported using this metric .
The Listening Inventories for Education
(LIFE) were developed by Smaldino and Anderson (1997), also for use in the classroom. The
LIFE consists of three inventories, one for the
student and two for the teacher. The Student
Appraisal of Listening Difficulty is intended to
be used in a pre- and post-test format in order
to document changes following intervention . It
has 15 test items that are picture cards depicting different common listening situations in the
18
classroom, chosen to be sensitive to the expected
benefits of various forms of classroom amplification (e .g ., hearing aids, personal FM systems,
soundfield FM systems) . The child's task is to
report the level of difficulty associated with each
of the 15 different situations . Children in grades
1 to 3 use a 3-point scale and older children use
a 5-point scale. The Teacher Appraisal of Listening Difficulty is designed to be given after
intervention and consists of 16 questions related
to specific areas of improvement in either behavior or communication. The Teacher Opinion and
Observation List consists of four items in an
open-ended format . Test-retest reliability of the
student questionnaire was assessed with 19
normal-hearing third and fourth graders over a
1- to 3-week period . No significant differences
were observed between the scores for any of the
test items. These inventories are flexible and can
be used in a variety of ways (e .g., document efficacy of amplification or classroom teaching methods, inservicing of teachers, self-advocacy) .
The Screening Instrument for Targeting
Educational Risk (SIFTER) (Anderson, 1989)
was developed to identify school-aged children
who are at risk for educational failure due to
hearing loss and thus in need of further evaluation . This 15-item test is completed by the
child's teacher and covers five content areas :
academics, attention, communication, class participation, and school behavior. Teachers use a
5-point scale to evaluate a child using items
such as "How distractible is the student in comparison to his/her classmates?" The total score
in each content area is categorized as either a
pass, fail, or marginal outcome. A preschool version (3 years-kindergarten) of this test has also
been developed (Anderson and Matkin, 1996) .
While this test was designed as a screening tool,
it can be used to assess the efficacy of various
forms of intervention in the classroom (e .g., preferential seating, soundfield amplification, FM
systems, personal amplification) .
Recently, Kopun and Stelmachowicz (1998)
adapted the Abbreviated Profile of Hearing Aid
Benefit (APHAB), developed by Cox and Alexander (1995), to be appropriate for children in the
10- to 15-year age range. Children with both
moderate and severe hearing losses showed a
pattern of responses across the four subtests of
the APHAB that was similar to the aided adult
data reported by Cox and Alexander and
test-retest reliability over a 2- to 3-month period
yielded correlation coefficients in the 0.83 to
0.95 range. Parents also were given the test in
order to assess the parent's perception of the
Hearing Aid Outcome Measures for Children/Stelmachowicz
child's communication problems . Interestingly,
the correlations between the child and parental
scores on all four subscales were quite low. That
is, for this group of subjects, the parents and chil-
skills develop . For an infant with a newly identified hearing loss, initial goals may include
auditory awareness and alerting to familiar
environment signals . By 9 to 12 months of age,
dren did not agree with respect to communication difficulties . No obvious trends were observed
and it is unclear whose impression best reflects
the degree of difficulty actually encountered . It
is possible that the parents may have noticed
communication difficulties that were not apparent to the child . It is also possible that parents
of children in this age range may have had limited opportunities to observe their child in the
variety of communication situations described
more specific goals may be appropriate such as
differentiating certain sounds or words . This
type of approach may require clear and structured guidelines for parents and other caregivers to ensure that their observations are
in this questionnaire . These results raise an
important issue regarding the development of
subjective measures for use with the pediatric
population . Can parents, caregivers, or other
individuals provide a valid estimate of communication difficulties via observation?
The Meaningful Auditory Integration Scale
(Robbins et al, 1991) is a parent report measure
designed for use with profoundly hearingimpaired children . This tool consists of 10 questions that include items on device acceptance and
use, awareness of familiar environmental
sounds, and the ability to distinguish speech
from nonspeech stimuli . Parents use a 5-point
scale in responding to each test item . This test
is most appropriate for young children with profound hearing loss as ceiling effects are observed
for children with greater residual hearing .
It may be possible to modify some existing
subjective tests for use with young children .
Moeller (1993) has suggested that modification
of the Client-Oriented Scale of Improvement
(COSI) described by Dillon et al (1991, 1997)
might be appropriate for infants and toddlers .
In its original form, the COSI is designed to
allow the patient to select up to five situations
in which communication is difficult . Following
a trial period with amplification, the patient
estimates the degree of change relative to the
unaided situation using a 5-point scale . This is
an efficient way to assess outcome because the
questionnaire is individualized to focus on the
specific goals of the patient . Because the goals
differ from case to case, however, findings cannot be easily grouped across subjects . Poten-
tially, this approach may prove to be very useful
for young hearing-impaired children because
individual situations and goals are likely to differ substantially across children, even when
they have similar hearing losses . Furthermore,
goals are likely to change relatively often in the
first few years of life as speech and language
unbiased and focused appropriately on the specific goal .
ALTERNATIVE APPROACHES
t should be obvious that the documentation
of hearing aid efficacy in the pediatric popu-
lation is extremely complex . For a variety of
reasons, it is unlikely that a single outcome
measure or a single approach to the problem
could provide a means by which to determine efficacy in all cases . The following alternative
options are described in detail along with the
assumptions and rationale for each approach .
Studies of Device Efficacy
For most hearing aid fittings, a wide range
of options (e .g ., frequency response, gain, circuit
type) are available and it would be impossible
to entertain all options during the evaluation
process . Studies of device efficacy with welldefined groups of subjects can help provide
guidelines to indicate when certain types of
technology may be most appropriate . For exam-
ple, laboratory studies of adults with mild-tomoderate hearing loss have shown that
directional microphones can improve the SIN
ratio by 5 to 7 dB over omnidirectional microphones (Hawkins and Yacullo, 1984; Valente et
al, 1995) . If a particular hearing aid circuit or
characteristic is shown to be efficacious for a
given group of hearing-impaired individuals,
then it can be assumed that the feature would
be efficacious in individual cases.
To date, only a few studies of this type have
been conducted with children . Christensen and
Thomas (1997) compared the performance of
peak clipping (PC), CL, and WDRC in 9- to
14-year-old children with mild-to-moderate hearing loss and found significantly better speech
recognition with WDRC and CL compared to
the PC circuit. Significant differences in perception were not observed between the WDRC
and CL circuits . In the second part of this study,
a multiple-memory device was programmed
with the three circuit types and the children
19
Journal of the American Academy of Audiology/Volume 10, Number 1, January 1999
wore the hearing aids for 2 months, recording
preferred memory settings in a diary. WDRC was
preferred most often in noise and CL was preferred most in quiet . The children reported subjective benefit from the use of multiple memories
in different listening environments . Seewald et
al (1997) measured speech recognition and aided
loudness growth in a group of adolescents using
both linear and WDRC hearing aids . They found
better speech recognition over a wider range of
input levels with WDRC compared to linear processing . In addition, WDRC produced loudness
growth functions that closely approximated normal loudness growth .
Gravel et al (in press) investigated the efficacy of dual-microphone technology for children
in the 4- to 6- and 7- to 11-year age ranges . For
the perception of single words, the directional
microphone advantage was similar for both
groups (an effective improvement in SIN ratio
of -4 .8 dB). For sentence materials, the younger
group demonstrated an advantage of 4.3 dB and
the older group showed an advantage of approximately 5.2 dB . These values are lower than
those reported for adults (approximately 8.0
dB) using the same type of dual-microphone
technology (Valente et al, 1995).
Laboratory studies of device efficacy in specific populations can play an important role and
have some advantages over individual measures
of outcome. In these types of studies, the necessary variables can be controlled, thus facilitating the interpretation of results.
Audibility of Speech :
Predictive Measures
While factors such as frequency or temporal resolution and abnormal growth of loudness
may play a role in speech perception, numerous
studies have suggested that the primary factor
responsible for speech recognition problems in
individuals with hearing loss is the reduction in
signal audibility (Pavlovic, 1984 ; Humes et al,
1987 ; Zurek and Delhorne, 1987 ; Humes, 1991).
As such, measures of the audibility of speech
under various aided conditions may provide a
basis for determining efficacy of a particular
hearing aid fitting. Both the Desired Sensation
Level (DSL) fitting procedure for linear hearing
aids (Seewald, 1992) and DSL i/o (Cornelisse et
al, 1995) for nonlinear hearing aids provide a
graphic display of the audibility of average
speech across frequency. As shown in Figure 1,
this display can provide a visual reference to aid
20
Speech SPLogram
14
120100Heal Ear SPL
(dB)
s0-
so4020-a
0 '-TTr--rT--I
._4000 __8000
125
_ 250
500
-1000
2000__ .
Frequency (Hz)
Figure 1 Graphic display from DSL i/o (Cornelisse et
al, 1995). The open squares represent auditory thresholds, the filled triangles represent the predicted upper
limit of comfort, and the three dashed curves show the
target values for soft, average, and loud speech.
in the finetuning of a hearing aid fitting. In this
figure, the open squares represent auditory
thresholds, the filled triangles represent the
predicted upper limit of comfort, and the three
dashed curves show the target values for soft,
average, and loud speech . If target values can
be achieved, it can be assumed that speech will
be audible, yet comfortable.
It is also possible to provide a quantitative
estimate of audibility. The most commonly used
method is the Articulation Index (AI) (American
National Standards Institute, 1969), which provides a number ranging from 0 (inaudibility) to
1.0 (audibility of the entire spectrum). A number of investigators have suggested various
adaptations of the AI to allow this approach to
be applied to aided listeners under clinical conditions (Mueller and Killion, 1990 ; Hou and
Thornton, 1994; Stelmachowicz et al, 1996). The
Hou and Thornton (1994) and Stelmachowicz et
al (1996) approaches provide aided AI values that
incorporate a range of assumed input levels to
the hearing aid. The Situational Hearing Aid
Response Profile also provides a visual display
of audibility and the aided dynamic range for various listening conditions (Stelmachowicz et al,
1996). Figure 2 shows a display of the audibility of speech under four listening conditions for
a child with a moderate hearing loss using this
approach . The Xs represent auditory thresholds, the asterisks show the hearing aid maximum output. The cross-hatched region shows the
portion of the speech spectrum that is audible
Hearing Aid Outcome Measures for Children/Stelmachowicz
and the associated aided audibility index (AAI)
is displayed . Such an approach can provide a
quantitative way to compare audibility across
hearing aids or across situations .
It is important to note that the aided audiogram or functional gain measures often cannot
provide a valid representation of the audibility
of speech at typical input levels (Macrae and Frazier, 1980 ; Macrae, 1982 ; Stelmachowicz and
Lewis, 1988). This is particularly true for wide
dynamic range hearing aids where the measured gain varies as a function of input level.
Paired-Comparison Measures
A number of investigators have used a
paired-comparison approach to evaluate differences perceived in various hearing aid characteristics by children . Eisenberg and Levitt (1991)
investigated the feasibility of using a paired-comparison task to select hearing aids for children .
Children with mild to moderately severe hearing loss judged the clarity of a children's story
that had been filtered with nine different frequency-gain shapes . They found the technique
to be feasible for hearing-impaired children as
young as 6 .5 years of age . Ching et al (1994) used
a paired-comparison procedure with a group of
21 children (ages 6-19 years) with severe to
profound hearing losses . The purpose of the
study was to evaluate the reliability and sensitivity of intelligibility judgments under audiovisual and auditory-alone conditions . Results
showed that the method could be used reliably
with subjects as young as 6 years of age and that
the mode of presentation had little effect on the
results . Recently, Stelmachowicz et al (in press)
used a paired-comparison procedure to assess
quality differences between speech processed
by either CL or PC . The frequency/gain and output characteristics were selected to reflect a
typical state-of-the-art hearing aid fitting ; thus,
the acoustic differences across stimuli were subtle compared to many earlier studies of this
nature . Normal and hearing-impaired subjects
in three age ranges (7-9 years, 10-12 years,
and adults) were tested. While the normal-hearing adults and the 10- to 12-year-old children
were able to distinguish between the stimuli, the
youngest normal-hearing group and both groups
of hearing-impaired children were relatively
insensitive to these acoustic differences . These
investigators suggested that the reliability and
validity of paired-comparison measures may
depend upon the magnitude of the physical differences between the stimuli to be compared .
That is, it may only be possible to use this technique with young children when the perceptual
contrasts to be compared are relatively large .
Loudness Measures
While many prescriptive hearing aid methods provide recommended real-ear saturation
levels, these levels are based on average data .
Hawkins et al (1987) have suggested that hearing aid rejection is often based on loudness discomfort. A small number of studies have been
reported in which measures of loudness discomfort level (LDL) have been attempted in
young children . Kawell et al (1988) evaluated
hearing-impaired children in the 7- to 14-year
age range and reported across-session intrasubject standard deviations that ranged from a
low of 3.0 dB at 500 Hz to 5.8 dB at 3000 Hz . Stuart et al (1991) also evaluated 7- to 14-year-old
hearing-impaired children and their within-session intrasubject standard deviations were
slightly lower (3 .4-4 .9 dB). Compared to adult
data, these values are within 1 dB of the standard deviations reported by Cox (1981) but are
4 dB greater than those reported by Hawkins et
al (1987) . Macpherson et al (1991) evaluated
LDLs in a younger group of normal-hearing
children (3-5 years) and concluded that only
children with mental ages > 5.0 years could perform this task reliably.
Teghtsoonian (1980) applied a cross-modality matching technique to loudness scaling in 4to 12-year-old normal-hearing children and
adults . Children were asked to match the loudness of a 1000-Hz tone to the length of a
retractable tape . The slope of the function relating intensity to length was invariant across all
ages, but the intercept differed significantly
between the 4 and 6 year olds and the older
subjects . That is, at a given intensity, the younger
subjects selected a shorter length than did the
older subjects . By 8 years of age, both the slope
and the intercept approximated adult values .
Since developmental changes did not alter the
slope of the functions, the author concluded that
cross-modality matching can be used successfully
in children as young as 4 years of age . Despite
the clinical use of loudness growth measures to
assist in the fitting of nonlinear hearing aids for
adults, no systematic studies with young hearing-impaired children have been reported .
21
Journal of the American Academy of Audiology/Volume 10, Number 1, January 1999
250
FREQUENCY (Hz)
500
1000
2000
4000
8000
125
250
140
FREQUENCY (Hz)
500
1000
2000
4000
8000
500
1000
2000
4000
8000
120
100
80
60
40
20
0
125
250
500
1000
2000
4000
8000
125
250
Figure 2
Graphic display from the Situational Hearing Aid Response Profile (Stelmachowicz et al, 1996). Xs represent auditory thresholds, asterisks show the maximum output, and the cross-hatched region shows the audible
region
of speech . The shaded region (lower left panel) denotes peak clipping . An aided audibility index (AAI) is displayed in
each panel .
CLINICAL INNOVATION
AND RESEARCH NEEDS
Speech Recognition Tests
There is a great need for more sophisticated
approaches to the measurement and quantification of speech recognition abilities in young
children . A developmentally appropriate battery of measures that considers the symbolic
and attentional demands of young children is
needed . While interactive video technology may
be promising with children > 3 years of age, the
clinical measurement of speech recognition dur22
ing infancy is problematic. Laboratory studies
using visual reinforcement techniques (Kuhl,
1983) have shown that normal-hearing infants
are capable of recognizing a variety of acoustic
differences . At present, these techniques have not
been adapted for use in clinical settings .
Many of the current tests that are available
for children > 3 years of age are somewhat outdated both in terms of the pictorial representations and vocabulary. Within this age range,
there is also a need for tests that would focus on
the type of speech recognition problems experienced by children with mild hearing loss or
hearing loss restricted to the high frequencies.
Hearing Aid Outcome Measures for Children/Stelmachowicz
Subjective Inventories
As noted earlier, a small number of inventories have been developed for use with schoolaged children in academic settings . There is a
need for additional inventories that focus on
specific listening situations outside the classroom . Studies are needed to determine the lower
age limit at which pictorial representation and/or
simple descriptions of listening situations (e .g.,
listening to television with other children talking) can provide both valid and reliable data .
Parental Inventories
Parental inventories are widely used and
well accepted in a number of fields (e .g ., speechlanguage pathology, psychology, pediatrics) in
order to document developmental milestones,
abnormal behavior, and a variety of other factors . Indeed, portions of many current developmental metrics include test items that may be
used to document auditory behaviors . Unfortunately, no single metric has been developed that
focuses exclusively on the auditory behaviors of
infants and young children and on the expected
benefits to be derived from amplification . These
types of measures may be the only alternative
for use with infants and toddlers . There is a
need to develop clearly focused measures that
will allow us to infer some probabilistic cause and
effect relationship with amplification .
Modify Existing Subjective
Procedures (e .g., Loudness
Growth, Quality Judgments,
Preferred Frequency Response)
While some progress has been made in this
area, more studies are needed to determine if
procedures designed for adults can be modified
to provide valid and reliable data in young children . The degree of success for any particular
procedure is likely to be influenced by the degree
of hearing loss, cognitive development, and the
child's attention span, as well as the design features of the task and the stimuli used .
Identify Objective Correlates
As the age at which the identification of
hearing loss and the initiation of amplification
decreases, the need for objective measures of performance increases . There is a need to identify
objective measures that correlate highly with the
perceptual measures that can only be obtained
reliably from adults and older children . For
example, ABR thresholds have been shown to
agree well with behavioral results (Stapells et
al, 1995), thus allowing hearing aids to be fitted
within the first few weeks of life . It may be possible to use electrophysiologic measures to provide suprathreshold measures of auditory
function . Boothroyd (1991) has suggested that
it may be possible to develop objective tests,
using evoked potential techniques, to measure
differential responses to various acoustic contrasts, thus providing an estimate of the auditory capacity for speech perception.
SUMMARY
T
he documentation of hearing aid efficacy in
young children is a complex issue con-
founded by a variety of variables . While some
progress has been made in recent years, there
is a great need for additional work in this area .
It is likely that a battery of measures, adapted
to certain age ranges, will be required . Parent
report measures may have a significant role to
play with the youngest children with more
reliance on direct measures as children mature .
In test development, both the symbolic and
attentional constraints of children must be con-
sidered . It also will be important to develop test
items that focus on the specific benefits to be
expected from various forms of amplification if
children are to take advantage of emerging technologies .
Acknowledgment. Portions of the work reported in this
paper were supported by a grant from the National
Institute on Deafness and other Communication Disorders
(NIH-NIDCD Grant P-60 DC00982) . The author wishes
to thank Kathryn Beauchaine, Brenda Hoover, Dawna
Lewis, and Richard Seewald for their helpful comments
on this paper.
REFERENCES
American National Standards Institute. (1969) . Methods
for the Calculation of the Articulation Index . (ANSI 53 .51969 .) New York : Acoustical Society of America.
Anderson K . (1989) . Screening Instrument for Targeting
Education Risk (SIFTER) . Austin, TX : Pro-Ed .
Anderson K, Matkin N. (1996) . Screening Instrument for
Targeting Educational Risk in Preschool Children (Age
3-Kindergarten) (Preschool S.I.FTE .R .) . Tampa, FL :
Educational Audiology Association .
Berger KW, Hagberg EN . (1982) . Hearing aid users attitudes and hearing aid usage . Monographs in Contemporary Audiology 3 .
23
Journal of the American Academy of Audiology/ Volume 10, Number 1, January 1999
Boothroyd A. (1991). Speech perception measures and
their role in the evaluation of hearing aid performance .
In : Feigin J, Stelmachowicz P, eds. Proceedings of the
1991 National Conference of Pediatric Amplification.
Boys Town National Research Hospital, Omaha, NE,
1991,77-91 .
Boothroyd A, Eran O, Hanin L. (1996) . Speech perception and production in children with hearing impairment .
In : Bess FH, Gravel JS, Tharpe AM, eds. Amplification
for Children with Auditory Deficits . Nashville, TN : Bill
Wilkerson Center Press, 55-74.
Byrne D, Ching T. (1997) . Optimising amplification for
hearing-impaired children: I. Issues and procedures. Aust
J Educ Deaf 3:21-28 .
Byrne D, Dillon H . (1986) . The National Acoustic
Laboratories' (NAL) new procedure for selecting the gain
and frequency response of a hearing aid. Ear Hear
7 :257-265 .
Ching T, Newall P, Wigney D. (1994). Audio-visual and
auditory paired comparison judgments by severely and
profoundly hearing impaired children : reliability and frequency response preferences. Aust JAudiol 16 :99-106.
Ching T, Newall P, Wigney D . (1996) . Frequency response
and gain requirements of severely and profoundly hearing impaired children . Aust JAudiol 18 :99-101 .
Christensen LA, Thomas TE . (1997) . The Use of MultipleMemory Hearing-Aid Technology in Children. Presented
at the 2nd Biennial Hearing Aid Research and
Development Conference, September 1997, Bethesda,
MD .
Cornelisse LE, Seewald RC, Jamieson DG . (1995). The
input/output formula : a theoretical approach to the fitting of personal amplification devices. JAcoust Soc Am
97 :1854-1864 .
Cox RM. (1981). Using LDLs to establish hearing aid limiting levels . Hear Instr 32 :16-20.
Cox RM, Alexander GC . (1995) . The Abbreviated Profile
of Hearing Aid Benefit. Ear Hear 16 :176-186 .
Dillon H, James A, Ginis J. (1997) . Client Oriented Scale
of Improvement (COSI) and its relationship to several
other measures of benefit and satisfaction provided by
hearing aids . J Am Acad Audiol 8 :27-43 .
Dillon H, Koritschoner E, Battaglia J, Lovegrove R, Ginis
J, Mavrias G, Carnie L, Ray P, Forsythe L, Towers E,
Goulias H, Macaskil F. (1991) . Rehabilitation effectiveness I: assessing the needs of clients of anational hearing
rehabilitation program. Aust JAudiol 13 :55-65 .
Eisenberg LS, Levitt H. (1991) . Paired comparison judgments for hearing aid selection in children . Ear Hear
12 :417-430 .
Feigin JA, Kopun JG, Stelmachowicz PG, Gorga MP
(1989). Probe-tube microphone measures of ear-canal
sound pressure levels in infants and children . Ear Hear
10 :254-258 .
Gatehouse S. (1992). The time course and magnitude of
perceptual acclimatization to frequency responses : evidence from monaural fitting of hearing aids . JAcoust
Soc Am 92 :1258-1268 .
Gatehouse S . (1993). Role of perceptual acclimatization
in the selection of frequency responses for hearing aids .
J Am Acad Audiol 4 :296-306 .
24
Gravel J, Fausel N, Liskow C, Chobot J. (in press) .
Children's speech recognition in noise using dual microphone hearing aid technology. Ear Hear.
Haggard MP, Foster JR, Iredale FE . (1981) . Use and benefit of postaural aids in sensory hearing loss. Scand Audiol
10 :45-52 .
Hawkins DB, Walden BE, Montgomery A, Prosek RA.
(1987) . Description and validation of an LDL procedure
designed to select SSPL90 . Ear Hear 8:162-169 .
Hawkins DB, Yacullo WS . (1984) . Signal-to-noise advantage of binaural hearing aids and directional microphones
under different levels of reverberation. J Speech Hear
Disord 49 :278-286 .
Hou Z, Thornton AR . (1994) . A model to evaluate and
maximize hearing aid performance by integrating the
articulation index across listening conditions. Ear Hear
15 :105-112 .
Humes LE . (1991) . Understanding the speech-understanding problems of the hearing impaired. J Am Acad
Audiol 2 :59-69 .
Humes LE, Dirks DD, Bell TS, Kincaid GE . (1987) .
Recognition of nonsense syllables by hearing-impaired
listeners and by noise-masked normal hearers . JAcoust
Soc Am 81 :765-773 .
Jordt J, Ching T, Byrne D, Birtles G, Dillon H, Flora RD,
Hill M, Battaglia J, Kirievsky L. (1997) . Optimising amplification for hearing impaired children : II . Aided speech
audiogram. Aust JEduc Deaf 3:29-33 .
Katz DR, Elliott LL . (1978) . Development of a New
Children's Speech Discrimination Test. Paper presented
at the American Speech-Language-Hearing Association
Convention, November 1978, Chicago, IL .
Kawell M, Kopun J, Stelmachowicz PG. (1988) . Loudness
discomfort levels in children . Ear Hear 9 :133-136 .
Kessler AR, Giolas TG, Maxon AB . (1990) . The Hearing
Performance Inventory for Children (HPIC): Reliability
and Validity. Poster presented at American SpeechLanguage-Hearing Association Convention, November
1990, Seattle, WA .
Kopun JG, Stelmachowicz PG . (in press) . The perceived
communication difficulties of children with hearing loss .
JAm Acad Audiol.
Kruger B . (1987) . An update on the external ear resonance in infants and young children. Ear Hear 8 :333-336.
Kruger B, Ruben AJ. (1987) . The acoustic properties of
the infant ear. A preliminary report . Acta Otolaryngol
103 :578-585 .
Kuhl PK (1983) . The perception of speech in early infancy:
four phenomena. In : Gerger S, Mencher G, eds. The
Development ofAuditory Behavior . NewYork : Grime and
Stratton, 187-218.
Macpherson BJ, Elfenbein JL, Schum RL, Bentler RA.
(1991) . Thresholds of discomfort in young children . Ear
Hear 12 :184-190 .
Macrae J . (1982) . The validity of aided thresholds . Aust
JAudiol 4 :48-54 .
Macrae J, Frazier G. (1980). An investigation of variables
affecting aided thresholds . Aust JAudiol 2 :56-62 .
Moeller MP. (1993) . Auditory Learning: Efficacy and
Evaluation . Presented at Developments in Pediatric
Audiology: Assessment and Amplification, September
1993, Omaha, NE .
Hearing Aid Outcome Measures for Children/Stelmachowicz
Moeller MP. (1998). Early intervention of hearing loss in
children . In : Bess FH, ed . Fourth International
Symposium on Childhood Deafness . Nashville, TN : Bill
Wilkerson Center Press, 305-310.
Moodie KS, Seewald RC, Sinclair ST. (1994) . Procedure
for predicting real-ear hearing aid performance in young
children. Am JAudiol 3:23-31 .
Children with Auditory Deficits . Nashville, TN : Bill
Wilkerson Center Press, 161-191.
Smaldino J, Anderson K (1997) . Development of the
Listening Inuentory for Education . Presented at the
Second Biennial Hearing Aid Research and Development
Conference, September 1997, Bethesda, MD .
Morrongiello BA, Robson RC, Best CT, Clifton RK . (1984) .
Trading relations in the perception of speech by 5-yearold children . J Exp Child Psychol 37 :231-250 .
Snik AF, Hombergen GC . (1993). Hearing aid fitting of
preschool and primary school children. An evaluation
using the insertion gain measurement. Scand Audiol
22 :245-250 .
Mueller HG, Mllion MC . (1990). An easy method for calculating the articulation index. Hear J 43 :14-17 .
SnikAF, Stollman MH . (1995) . Measured and calculated
insertion gains in young children . Br JAudiol 29 :7-11.
Nelson Barlow NL, Auslander MC, Rines D, Stelmachowicz PG . (1988) . Probe-tube microphone measures in
hearing-impaired children and adults . Ear Hear
9 :243-247 .
Snik AF, van den Borne S, Brokx JP, Hoekstra C. (1995) .
Hearing aid fitting in profoundly hearing-impaired children . Comparison of prescription rules. Scand Audiol
24 :225-230 .
Nittrouer S. (1992) . Age-related differences in perceptual effects of formant transitions with syllables and
across syllable boundaries . J Phon 20 :1-32.
Nittrouer S, Boothroyd A. (1990) . Context effects in
phoneme and word recognition by young children and
older adults . JAcoust Soc Am 87 :2705-2715 .
Nittrouer S, Studdert-Kennedy M. (1987) . The role of
coarticulatory effects on the perception of fricatives by
children and adults . JAcoust Soc Am 30 :319-329 .
Nozza RJ, Rossman RNF, Bond LC. (1991) . Infant-adult
differences in unmasked thresholds for the discrimination of consonant-vowel syllable pairs . Audiology
30 :102-112 .
Oja GL, Schow RL . (1984) . Hearing aid evaluation based
on measures of benefit, use, and satisfaction . Ear Hear
5:77-86 .
Pavlovic CV (1984) . Use of the articulation index for
assessing residual auditory function in listeners with
sensorineural hearing impairment . J Acoust Soc Am
75 :1253-1258 .
Pediatric Working Group . (1996) . Amplification for infants
and children with hearing loss . Am JAudiol 5 :53-68 .
Robbins AM, Renshaw J, Berry S. (1991) . Evaluating
meaningful auditory integration in profoundly hearingimpaired children . Am J Otol 12 (Suppl):144-150 .
Ross M, Lerman J. (1970) . A picture identification test
for hearing-impaired children . J Speech Hear Res
13 :44-53 .
Ross M, Levitt H. (1997) . Consumer satisfaction is not
enough : hearing aids are still about hearing. Semin Hear
18 :7-11.
Seewald RC. (1992) . The Desired Sensation Level method
for fitting children : version 3 .0 . Hear J 45 :36-41 .
Seewald R, Cornelisse L, Jenstad L, Moodie S, Shantz J,
Pumford J. (1997) . Rationale for WDRCApplications in
Pediatric Fittings. Presented at Developments in Pediatric
Audiology: Current Issues in Assessment and
Management, October 1997, Omaha, NE .
Seewald RC, Moodie KS, Sinclair ST, Cornelisse LE .
(1996) . Traditional and theoretical approaches to selecting amplification for infants and young children . In : Bess
FH, Gravel JS, Tharpe AM, eds . Amplification for
Stapells DR, Gravel JS, Martin BA . (1995) . Thresholds
for auditory brain stem responses to tones in notched
noise from infants and young children with normal hearing or sensorineural hearing loss . Ear Hear 16 :361-371 .
Stelmachowicz PG . (1991). Current issues in pediatric
amplification . In : Feigin J, Stelmachowicz PG, eds.
Pediatric Amplification : Proceedings ofthe 1991 National
Conference. Omaha, NE : Boys Town National Research
Hospital, 1-17 .
Stelmachowicz PG, Kalberer A, Lewis DE . (1996) .
Situational hearing aid response profile (SHARP). In :
Bess FH, Gravel JS, Tharpe AM, eds. Amplification for
Children with Auditory Deficits . Nashville, TN : Bill
Wilkerson Center Press, 193-213 .
Stelmachowicz PG, Lewis DE . (1988) . Some theoretical
considerations concerning the relation between functional gain and insertion gain . J Speech Hear Res
31 :491-496 .
Stelmachowicz PG, Lewis DE, Hoover B, Keefe D. (in
press) . Objective and subjective effects of peak clipping
vs . compression limiting in normal and hearing-impaired
children. J Acoust Soc Am .
Stelmachowicz PG, Mace AL, Kopun JG, Carney E. (1993) .
Long-term and short-term characteristics of speech: implications for hearing aid selection for young children . J
Speech Hear Res 36 :609-620 .
Stuart A, Durieux-Smith A, Stenstrom R. (1991). Probe
tube microphone measures of loudness discomfort levels
in children . Ear Hear 12 :140-143 .
Teghtsoonian M . (1980) . Children's scales of length and
loudness : a developmental application of cross-modal
matching . J Exp Child Psychol 30 :290-307 .
Valente M, Fabry DA, Potts LG . (1995) . Recognition of
speech in noise with hearing aids using dual microphones .
J Am Acad Audiol 6:440-449 .
Weinstein BE . (1997) . Outcome measures in the hearing
aid fitting/selection process . Trends in Amplification
2 :117-137 .
Yoshinago-Itano C, Sedey AL, Coulter DK, Mehl AL . (in
press) . The impact of newborn hearing screening on language development. Pediatrics .
Zurek PM, Delhorne LA. (1987). Consonant reception in
noise by listeners with mild and moderate sensorineural
hearing impairment . J Acoust Soc Am 82 :1548-1559 .