Content uploaded by Agnes Weill Chounlamountry
Author content
All content in this area was uploaded by Agnes Weill Chounlamountry on Oct 23, 2018
Content may be subject to copyright.
Brain Injury, October 2007; 21(11): 1165–1174
Rehabilitation of word deafness due to auditory analysis disorder
CATHERINE TESSIER, AGNES WEILL-CHOUNLAMOUNTRY,
NADINE MICHELOT, & PASCALE PRADAT-DIEHL
AP-HP, Groupe Hospitalier Pitie
´-Salpe
ˆtrie
`re, Service de Me
´decine Physique et Re
´adaptation, Paris, France; INSERM,
U731; UPMC, 47 Boulevard de l’ho
ˆpital, 75013 Paris
(Received 19 February 2007; revised 7 July 2007; accepted 8 July 2007)
Abstract
Background: Word deafness refers to an inability to understand spoken words despite intact hearing. In a cognitive
approach, word deafness could be explained by a deficiency at the lower perceptive level of the auditory process. The
impairment of the auditory analysis system would explain a disorder of identification of speech sounds. Only few studies
addressed rehabilitation of central auditory processing and have described therapy focused on phoneme discrimination.
Objective: To determine whether a specific auditory analysis rehabilitation addressing phoneme discrimination and phoneme
recognition may improve oral comprehension and communication.
Method: A single-case experimental design was used in a 65 year-old woman, with word deafness consecutive to a cerebral
infarction which occurred 10 months before. Verbal naming, written expression and written comprehension were normal.
Verbal comprehension, repetition and phoneme discrimination and recognition were impaired. In terms of cognitive model
of auditory processing, the patient showed impairment of the auditory analysis system affecting verbal comprehension. A
computerized rehabilitation of auditory analysis system was carried out in two consecutive tasks: phoneme discrimination
and phoneme recognition. Errorless learning therapy was used, with a difficulty hierarchy practised from the easier to the
most difficult phoneme and systematic visual cues which were progressively delayed and suppressed. This study tested the
efficacy and the specificity of this therapy on the addressed tasks (phoneme discrimination and recognition), related tasks
(oral comprehension and repetition), independent tasks (recognition of environmental sounds) and daily life
(questionnaire).
Results: The phoneme discrimination and recognition impairment was stable over 4 months before therapy. After therapy,
phoneme discrimination (p<0.001) and phoneme recognition (p<0.0001) were improved. The improvement was specific
to verbal sounds recognition, while non-verbal sounds recognition was unchanged. An improvement occurred for repetition
(p<0.05) and oral comprehension (p<0.01). The communication disability decreased (p<0.05).
Conclusion: In a case of word deafness, this study demonstrates not only the efficacy of a specific phoneme processing
therapy but also its efficacy in the improvement of higher level of cognitive treatment such as oral comprehension and
its transfer in daily life. The role of errorless therapy using systematic visual cues and difficulty hierarchy must be
underlined.
Keywords: Aphasia, cognitive rehabilitation, traumatic brain injury, stroke, disability
Introduction
Word deafness [1] is a central auditory disorder
[2, 3] defined as an inability to understand spoken
words, despite intact hearing. This deficiency can
occur after stroke [4–7] or traumatic brain injury [8]
and is a cause of severe disabilities in daily life.
Central auditory disorders [2, 3] are classical
neurological features. They were thought to be due
to bilateral temporal lobes lesions, but recent studies
have highlighted the role of brainstem lesions or
subcortical lesions. Studies have demonstrated
that central deafness [9], word deafness [10, 11]
Correspondence: Dr Pascale Pradat-Diehl, Service de Me
´decine Physique et de re
´adaptation, Pr Perrigot Ho
ˆpital de la Salpe
ˆtrie
`re, 47–83 boulevard de
l’ho
ˆpital, 75651 Paris, France. Fax: 33 1 42 16 11 48. E-mail: pascale.pradat@psl.aphp.fr
ISSN 0269–9052 print/ISSN 1362–301X online ß2007 Informa UK Ltd.
DOI: 10.1080/02699050701559186
Brain Inj Downloaded from informahealthcare.com by Universite Pierre & Marie Curie on 05/27/13
For personal use only.
and environmental sound recognition disorders [12]
may arise after lesions of the ascending
auditory pathway [13, 14]. Brainstem lesions [11]
or subcortical lesions [8, 10, 12] partly disconnect
the auditory cortex from the auditory nucleus and
disrupt the initial auditory process.
In a pure word deafness, the impairment of word
oral comprehension is isolated, but in a clinical
practice word deafness is frequently associated to
other central auditory disorders (central deafness or
auditory agnosia) or to language disorder and
aphasia. In a cognitive approach, the model of
auditory word comprehension [15] (Figure 1)
involves three components responsible for three
different aspects of word recognition and compre-
hension. At the lower level of the auditory process,
the identification of speech sounds is attributed to an
auditory analysis system. This perceptive level can be
assessed using phoneme discrimination or recogni-
tion. The next levels are the auditory input lexicon,
where the spoken forms of known words are stored,
and the semantic system storing word’s meaning.
This model provides a framework for analysing
disorders of spoken word comprehension (1).
Word deafness could be due to a deficit at the
lexical level (word form deafness) but also to
deficiencies at the auditory system analysis level
(word sound deafness). Indeed, a disorder at a low
level of treatment of the auditory analysis system
with impairment in discrimination and recognition
of phonemes would result in affecting processing of
the word form at the lexical level and could be
responsible for apparent ‘higher’ level deficits such
as word deafness. If so, analytic rehabilitation at the
lower level would improve these ‘higher’ level
deficits and improve word comprehension.
Only few studies addressed rehabilitation of
central auditory processing. The authors have
previously [16] reported a case of improvement in
communication behaviour after central deafness
therapy using auditory detection and phoneme
discrimination. However, in this study, neither
assessment nor therapy was systematic. This clinical
study was the base of the present study. In the
literature, the authors used two main strategies. On
one hand, compensatory strategy involving lip read-
ing was used to reduce disability following severe
cerebral deafness [5, 6, 17, 18]. On the other hand,
authors have described a therapy focused on
phoneme discrimination [4, 6]. Morris et al. [4]
described an aphasic patient with a disorder at an
early stage of auditory analysis, associated to a
semantic disorder and to a severe oral language
production deficit. The deficit of auditory analysis
was demonstrated with failure to discriminate pairs
of CVC items, in which when the pair was different
the consonant was changed by one or two distinctive
feature (minimal pairs). The rehabilitation focused
on auditory discrimination at a phonemic level,
using lip-reading and was based on minimal pair
contrasts. Exercises were phoneme-grapheme
matching, phoneme discrimination and matching
of auditory word with picture, written word. The
patient was given immediate feedback of the
accuracy of his attempt and in case of incorrect
response lip reading and hand signs were provided to
ensure the performance. Patient showed improve-
ment on tests of phoneme discrimination and a trend
of improvement for the other auditory comprehen-
sion or repetition task.
Cognitive rehabilitation refers to the theoretical
framework of cognitive models and has been largely
DEFICIT FUNCTION
Auditory presentation of words
Word-sound
deafness AUDITORY ANALYSIS
SYSTEM
Identify speech
sounds
Word-form
deafness
Identify familiar
word
AUDITORY INPUT
LEXICON
Word meaning deafness
Semantic deficit SEMANTIC SYSTEM Activate word
meanin
g
s
Figure 1. Model for auditory processing and auditory word comprehension [15], showing the disorders associated with impairment at each
level. The lower level, the ‘auditory analysis system’ identifies speech sounds in the speech wave. The phoneme discrimination and
recognition can be located at this level. Impairment in comprehension of words or sentences may result from disorders of this lower level of
processing.
1166 C. Tessier et al.
Brain Inj Downloaded from informahealthcare.com by Universite Pierre & Marie Curie on 05/27/13
For personal use only.
debated for several years [19–22]. Its role in disorder
assessment, identification of the level of impairment
and in therapy efficacy assessment is now obvious.
Particularly, identification of perceptive disorders at
a low level of treatment provides a framework for
rehabilitation programmes. Previous studies had
made this choice. In a case of rehabilitation for a
pure alexia, the authors [23] described the attempt
to remediate the peripheral level of reading during
which the visual input is encoded and analysed. In
rehabilitation of word deafness, it can be proposed
that a rehabilitation specifically addressing the lower
level of the auditory analysis system would improve
the access to the ‘higher’ level of treatment and so
improve word comprehension. It was also the goal of
Morris et al. [4] and Maneta et al. [6]. In their cases,
word deafness was explained by a deficit at a low
level of treatment of the phoneme process. The
therapy was elaborated to address specifically this
low level of treatment and not directly comprehen-
sion or repetition.
Recently, there has been great interest in errorless
learning as an intervention technique. Trial and
error approaches are avoided, because the act of
producing an error could strengthen this incorrect
production. In this approach, the task is manipulated
to reduce errors and to enhance appropriate produc-
tions. Errorless learning has been described to be
more efficient than effortful therapy in amnesic
patients [22, 24, 25]. Efficacy of errorless rehabilita-
tion was shown in aphasic patients [26], especially
for anomia [27] but also for phoneme discrimina-
tion. Decreasing assistance [28], the sequence of
visual cues being arranged from the strongest to the
weakest can be chosen in order to avoid errors.
Cueing is a classical tool in rehabilitation. It is widely
used in attention and visual neglect rehabilitation
[29–31].
The objective of this study was to determine
whether a specific auditory processing rehabilitation
could improve communication abilities in a case of
auditory analysis disorder. This study proposes the
hypothesis that in a case of word deafness with
auditory analysis disorder (i) a restorative approach
based on the cognitive model of auditory processing
and (ii) a therapeutic method using systematic
cueing avoiding errors could improve phoneme
processing abilities and subsequently communica-
tion impairment.
Patient and methods
Clinical report
The patient was a 65-year-old right-handed, native
French speaking woman. She received an 8th grade
education and was a retired society manager.
She previously suffered a slight presbycusis. Six
months before the first clinical assessment, she
suffered a cerebral infarction. The MRI showed
multiple lesions located in the brainstem on the right
lateral part of the medulla, the pons and mesence-
phalum, bilaterally in the internal capsule and the
periventricular white matter. These lesions indicate
bilateral subcortical vascular disease. She presented
with mild cerebellar syndrome. She walked indepen-
dently and she was living alone. She did not present
with visual field disorder, nor visual acuity impair-
ment. She presented with moderate diplopia due the
right 6th nerve palsy.
The patient complained about difficulties for
verbal comprehension (‘I don’t hear some syllables,
I ask people to repeat’) mainly on telephone, music
listening (‘Music is only noise’), voices recognition
(‘It’s hard for me to recognize voices’) and environ-
mental sounds discrimination (‘I hear the noise, but
I cannot recognize the phone ring’).
The audiogram showed mild hearing (40 db) loss
due to presbycusis and an increase of hearing level
for the high frequency sounds. The thresholds of
speech intelligibility were more increased than those
of pure tone sensitivity. Brainstem Auditory Evoked
Potentials showed an increase of interwave duration
I–III on the right ear, which indicates an impairment
of the conduction in the lower part of the right
brainstem. Interweave duration III–V on the right
ear was irregularly increased (two out of four).
Interweave durations I–III and III–V were normal on
the left ear. Cortical evoked potential and sensitivity
for pure tones did not show cortical impairment.
Comprehensive aphasia and phoneme perception
assessments were performed 10 months after the
stroke, 1 week before the beginning of the therapy
(T
2
in Table I and Figure 4). A previous assessment
Table I. Results of the assessments performed before and after
therapy. Results are expressed in percentage of errors.
T
1
T
2
T
3
T
4
Phoneme perception
Discrimination/25 19 19 25 25
Recognition/25 13 16 21 25
Sentences oral comprehension
With visual matching/25 17 24
Without visual matching/25 11 23
Repetition
Syllables/25 14 21
Words/25 22 24
Sentences/25 20 24
Communication
Conversation/24 22 24
Telephone/21 10 18
Environmental sounds recognition
Sound lot to/48 14 12
Rehabilitation of word deafness due to auditor y analysis disorder 1167
Brain Inj Downloaded from informahealthcare.com by Universite Pierre & Marie Curie on 05/27/13
For personal use only.
of phoneme perception had been performed 4
months before (T
1
), during the first examination.
Aphasia examination was performed with a widely
used French aphasia battery [32]. Verbal naming
(25/25), verbal and written expression and written
comprehension (25/25) were intact, showing
preserved lexical and semantic systems. In contrast,
verbal comprehension and repetition were impaired.
The tests (Table I) of oral repetition and compre-
hension were performed without using lip-reading.
The normal score for each task is 25.
Verbal repetition test: The patient scored 14/25
in syllables verbal repetition, 22/25 in words verbal
repetition and 20/25 in sentences oral repetition.
Verbal comprehension test: In the sentence-
picture matching test, the patient heard a sentence
and was required to point to the corresponding
picture among four. The patient scored 17/25. In
sentences comprehension task without matching
with a picture, yes–no questions were asked, such
as: ‘Is it colder in summer than in winter?’ ‘Are there
more days in 1 year than in 2 months?’ and questions
were asked after two short texts had been heard. The
patient’s performances were 11/25.
Assessment of phoneme perception (T
2
, Table I)
was processed through the computer used for
rehabilitation [33]. There was no visual help on the
computer as well as no lip-reading available. The
phonemes had the same intensity. This test was
normalized on five controls matched for age and
social level to the patient. Two tests were proposed:
(i) phoneme discrimination task, the response required
was to tell whether the two phonemes were similar or
different by pointing to written signs ‘same’ or
‘different’ (Figure 2(c)). Thirteen pairs of vowel
items and 12 pairs of consonant-vowel (CV) items
were proposed. In CV pairs, the vowel was always
the same:/a/. When the pair was different, the
consonant changed by only one distinctive feature.
The patient scored 19/25; the mean result of control
was 24.8 (SD ¼0.4)/25; and (ii) phoneme recognition
task, nine vowel items and 16 CV items with a/a/
vowel were produced. Twenty-five written letters
were presented on the computer screen. After
hearing the vowel or CV stimuli, the patient was
asked to point to the corresponding letter on the
screen. The patient scored 13/25; the mean result of
control was 24.25 (SD ¼0.95)/25.
Communication abilities. This study used the
‘Echelle de Communication Verbale de Bordeaux’
(Aphasic patients communication scale) [34, 35].
This scale includes closed questions concerning
patient’s behaviour in communication situations
and uses a four-point scale, ranging from 0 (severe
disability) to 3 (normal state). Eight questions
address daily life communication for oral conversa-
tion and the patient scored 22/24, seven questions
address telephone communication and the patient
was particularly deeply impaired (10/21).
In the environmental sound test, non-verbal
sounds recognition were assessed with a game
currently used in clinical practice (sound lotto).
Sounds produced by the use of objects were
recorded on a tape. A sound was played and the
patient had to point to the appropriate object among
four pictures. The patient scored 34/48 (a matched
control subject with presbycusis scored 46/48). This
patient suffered an auditory agnosia.
In summary, the patient suffered disorder in
repetition and in comprehension limited to verbal
presentation which matches definition of a word
deafness. This disorder was explained by an impair-
ment in phoneme discrimination and recognition
corresponding to an auditory analysis disorder in
reference to the Ellis model [15]. The better
performance in word repetition than in syllable
repetition can be explained by a top-down treat-
ment, the intact lexicon providing help to under-
stand word but not syllables.
The word deafness was not pure, it was associated
to an auditory agnosia but not to an aphasia.
With the hypothesis that impairment in oral
repetition and comprehension might be explained
by the deficit in phoneme processing, it was decided
to set up a specific therapy addressing the auditory
analysis disorder. Auditory agnosia was not
addressed by therapy.
The patient was informed of the experimental
nature of the therapy and gave her informed consent
to participate.
Description of the therapy
The experimental therapy began 10 months after the
neurological event and was carried out twice a week.
First, phoneme discrimination therapy was con-
ducted on the failed phoneme pairs until discrimina-
tion normalization. Then, phoneme recognition
therapy was conducted and addressed the phonemes
that were still not recognized.
A specific rehabilitation of the auditory analysis
system deficit was proposed. Specific phoneme
discrimination training was performed, followed by
specific phoneme recognition. Only phonemes failed
by the patient during discrimination or recognition
tasks were trained. One chose a difficulty hierarchy
organized following two criteria. First, the phoneme
complexity was increased. This difficulty hierarchy
was based on the perceptive difficulties of patients
suffering a central auditory disorder: vowels before
consonants, open back vowel/a/before close back
vowels/i/[36]. The phoneme presentation was
1168 C. Tessier et al.
Brain Inj Downloaded from informahealthcare.com by Universite Pierre & Marie Curie on 05/27/13
For personal use only.
arranged from the easier to the more difficult.
Secondly, help brought by visual cues was progres-
sively delayed and then suppressed. Visual cues
consisted of phoneme’s visual representation
(the phoneme/a
˜/has the visual representation ‘an’
on the screen) (Figure 3) and flickering of an area of
the screen. Visual help was progressively faded.
During the first step, the visual representation of the
(a)
(b)
(c)
Figure 2. Computer screen for the phoneme discrimination therapy at the first and second steps (a), at the third step (b) and the fourth
step (c). The dotted lines illustrate the flickering. (a) The visual representation of one phoneme is on the left, the signs ‘equal’ or ‘different’
are in the central area and the visual representation of two phonemes are on the right. At the first step of difficulty, the visual representations
appear successively. The visual representation of the first phoneme appears and flickers on the left side during its auditory presentation
(/pa/), then the symbols equal and different appear and finally the visual representation of two phonemes appears on the right side. The
visual representation of the second phoneme heard flickers during its auditory presentation (/ta/). At the second step the flickering is
delayed after the beginning of the auditory presentation. The patient is asked to point the target phoneme in the right area of the screen. In
this example the phonemes heard are different. (b) On the third step, the first phoneme’s visual representation appears only on the left side
of the screen and there is no flickering. After having heard the successive auditory presentation of the two phonemes, the patient is asked to
point to the sign ‘equal’ or ‘different’. (c) On the fourth step, there is no more phoneme visual representation but only successive auditory
presentation of the two phonemes. The task is to point to the sign ‘equal’ or ‘different’.
Rehabilitation of word deafness due to auditor y analysis disorder 1169
Brain Inj Downloaded from informahealthcare.com by Universite Pierre & Marie Curie on 05/27/13
For personal use only.
target item was flickering during all the auditory
phoneme occurrence, during the second step the
flickering was delayed after the beginning of the
auditory phoneme and during the third step there
was no flickering. Each level of difficulty based on
phoneme difficulty and visual help was practised
until it could be completed successfully. After each
level of difficulty was mastered, the patient could
begin to practice the next one.
Beside these common features, phoneme discri-
mination and phoneme recognition therapies dif-
fered on several points.
Tasks of phoneme discrimination therapy were
based on 25 phoneme pairs (13 pairs of vowel items
and 12 pairs of CV items in which the vowel was
always/a/). The pair was similar (/pa/, pa/) or
different (/pa/, /ta/ ). In this latter case the phoneme
differed by only one distinctive feature (p/t). During
consecutive auditory presentation of two phonemes,
the patient had to answer whether they were similar
or not. The computer screen was divided in three
areas: in the left area was the visual representation of
one phoneme/p/, in the right one was the visual
representation of two phonemes/p/and/t/. In the
central area were the signs ‘equal’ or ‘different’
(Figure 2). There were four steps of difficulty. On
the first step, phoneme’s visual representations
appeared consecutively on the left and then on the
right of the computer screen. During all auditory
presentation of the phoneme, the visual representa-
tion of corresponding heard phonemes was flicker-
ing. On the second step, the flickering appeared only
after a delay. On these two first steps, the task was to
point to the target phoneme in the right area of the
screen and the answer could be based both on the
visual help and on the auditory items. On the third
step, only the first phoneme’s visual representation
appeared on the left side of the screen, there was no
visual representation of the second phoneme heard
and there was no flickering. On the fourth step, there
was no more phoneme visual representation. On
these latter steps there was only a successive auditory
presentation of the two phonemes and the task was
to point to the sign ‘equal’ or ‘different’.
In tasks of phoneme recognition therapy the
computer screen presented the visual representations
of 25 phonemes (Figure 3). There were three steps
of difficulty. On the first step the visual representa-
tion of the target item was flickering during the
occurrence of the phoneme, on the second step it
was flickering with a delay and on the third step there
was no flickering. The patient was asked to point to
the phoneme she had heard.
This systematic use of visual cues was made easier
by the computerized form. The computer enables
combined various visual cues and control of their
occurrence delay. It also enables control and
reproducing of phoneme’s physical parameters.
Experimental device
A single-case ABCA experimental design was used,
where A refers to the baseline without therapy, B to
Phonemic perception
17
8
00 0
48
36
16 16
0
4
17
0
10
20
30
40
50
60
T1 T2 T3 T4 T5
Errors in %
Discrimination Recognition
Therapy
of
discrimination
Pre therapy Post-therapy
Therapy of recognition
Figure 4. Graphic representation of patient performance in
phoneme discrimination and recognition before and after therapy.
Five assessments were performed. Two assessments were
performed before the beginning of therapy 4 months before
therapy (T
1
) and 1 week before therapy (T
2
). A third assessment
was carried out between the phoneme discrimination and
recognition therapies (T
3
) and two other assessments were
performed just after the end of the therapy (T
4
) and 1 month
after its end (T
5
). The results are expressed in percentage of
errors.
Figure 3. Computer screen for the phoneme recognition therapy.
Visual representation of 25 phonemes. On the first step the visual
representation of the target item is flickering during the
occurrence of the phoneme auditory presentation, on the
second step it is flickering after a delay and on the third step
there is no flickering. The patient is asked to point to the phoneme
she heard.
1170 C. Tessier et al.
Brain Inj Downloaded from informahealthcare.com by Universite Pierre & Marie Curie on 05/27/13
For personal use only.
the phoneme discrimination therapy, C to the
recognition therapy.
To demonstrate efficacy and specificity of this
rehabilitation, two assessments were performed
before therapy began, 4 months before therapy
(T
1
) and 1 week before therapy (T
2
) to ensure
there was no improvement due to spontaneous
recovery. A third assessment was carried out
between phoneme discrimination and recognition
therapies (T
3
). Two other assessments were
performed after therapy, (i) just after the end of
therapy (T
4
) and (ii) 1 month later (T
5
) in order to
control stability of the performances after therapy.
This study, thus, obtained pre-therapy (T
1
–T
2
) and
post-therapy (T
4
–T
5
) baselines. It assessed (1)
performances on auditory analysis addressed by
therapy, i.e. auditory processing (phoneme discrimi-
nation, phoneme recognition), (2) the transfer to
non-trained tasks, i.e. other verbal tasks (verbal
comprehension and repetition) or non verbal tasks
(environmental sound recognition), (3) the commu-
nication handicap using the ‘Echelle de
Communication Verbale de Bordeaux’ (Aphasic
patients communication scale) [34, 35]. The specific
assessment of phoneme discrimination and recogni-
tion was conducted five times. Verbal repetition and
comprehension, environmental sound recognition
and disability in daily life were only assessed before
(T
2
) and after (T
4
) therapy.
McNemar Chi-square tests were conducted to test
the results’ significance. The significance level was
set at 0.05.
Results (Table I; Figure 4)
Pre-therapy baseline: between the first and the
second assessment before therapy (T
1
–T
2
), the
computerized assessment of auditory processing
indicated stability of phoneme discrimination and a
non-significant modification of phoneme recognition
(McNemar;
2
¼0.19; 1 df; p¼0.66).
After 10 therapy sessions (T
3
) there was no error on
the phoneme discrimination test (25/25). Percentage
of accuracy for phoneme discrimination changed
from 76% before therapy (T
1
&T
2
) to 100% after
therapy (T
3
,T
4
,T
5
). Discrimination of different
phonemes was acquired at different rates, from one
to four sessions. For example, one session was
sufficient to acquire/m/?/n/, two sessions for/J/?/I/.
Some phonemes were particularly difficult to
discriminate and required three sessions (/a/?/o/; /
G/?/e/) and even four sessions (/a/?/a
˜/; /u/?/i/; /b/?/
d/; /o/?/o
˜/; /t/?/k/). During rehabilitation sessions,
MA sometimes repeats the phonemes she heard and
only secondly the answer. Before discrimination
therapy percentage of accuracy in phoneme
recognition was 64% (T
2
). After discrimination
therapy (T
3
) this percentage was 84% with only four
phonemes non-recognized; /t/, /d/, /o
˜/ and /a
˜/. After
two sessions of phoneme recognition therapy, pho-
neme recognition was normal (100% at T
4
). Post-
therapeutic assessment (T
5
), 1 month after the end of
therapy, demonstrated stability of the results for
phoneme discrimination and recognition. This study
grouped the performances on phoneme discrimina-
tion and recognition (i) on T
1
and T
2
, as perfor-
mances before therapy and (ii) T
4
and T
5
,as
performances after therapy. The improvement of
performances after therapy in phoneme discrimina-
tion (McNemar;
2
¼10.1; 1 df; p¼0.001) and
phoneme recognition (McNemar;
2
¼16.4; 1 df;
p¼0.0001) was significant.
Improvement of repetition of syllable, words and
sentences was significant (McNemar;
2
¼5.76;
1 df; p¼0,16), as was improvement of oral
comprehension with visual cues from 68% to 96%
(McNemar;
2
¼4; 1 df; p¼0.04) and of oral
comprehension without visual cues from 44% to
92% (McNemar;
2
¼7.56; 1 df; p<0.01).
Communication in daily life was improved and this
improvement was statistically significant for con-
versation in daily life and on the telephone
(McNemar;
2
¼5.06; 1 df; p¼0.02). On the
opposite, performance in environmental sound
recognition, which was not addressed by therapy,
was not modified.
Discussion
This patient suffered a central auditory disorder with
phoneme discrimination and recognition disorder,
impairment for verbal comprehension and disability
for communication in daily life. In a clinical
classification these symptoms correspond to word
sound deafness [1, 37]. In the cognitive model [15]
(Figure 1) this impairment in phoneme processing
corresponds to a disorder at the lower perceptive
level of treatment: the auditory analysis system
(Figure 1). It leads to failure of access to higher
levels (lexical and semantic) and can be sufficient to
explain impairment of oral comprehension without
disorders of the auditory input lexicon or of the
semantic system [38, 39]. In this patient, brain
damage did not involve the temporal lobes, unlike
lesions classically described in word deafness [3],
but it was located in the brainstem and in the white
matter [10, 11] which partly disconnect the auditory
cortex from the auditory nucleus and disrupts the
initial auditory process.
In reference to the theoretical framework of
central auditory processing and of cognitive rehabili-
tation with errorless learning, one carried out a
Rehabilitation of word deafness due to auditor y analysis disorder 1171
Brain Inj Downloaded from informahealthcare.com by Universite Pierre & Marie Curie on 05/27/13
For personal use only.
specific therapy of these auditory analysis disorders
addressing the lower level of auditory processing in
order to improve verbal comprehension.
This therapy significantly improved phoneme
discrimination and phoneme recognition. These
improvements cannot be explained by spontaneous
recovery, since therapy began 10 months after the
initial stroke, after the main period of spontaneous
recovery [40–42]. Furthermore, there was no sig-
nificant improvement between the two assessments
during pre-therapy baseline.
The improvement was not limited to the trained
auditory tasks, but was significant in oral repetition
and comprehension which were not directly trained
during therapy. The results showed generalization to
verbal comprehension and repetition, which involve
auditory processing. The improvement was signifi-
cant in comprehension without visual matching
tasks, which is an especially difficult task in word
deafness because of the lack of visual help.
Generalization of phoneme discrimination and
recognition improvement to language components
involving auditory processing—comprehension and
repetition—supports the hypothesis that remediation
of low level auditory processing could improve the
access to higher levels (semantic or syntactic).
Furthermore, this improvement was also found in
verbal communication during daily life activities.
It should be underlined that the telephone commu-
nication was significantly improved after therapy,
although this condition has great difficulties because
of the total lack of visual help such as facial
expression and lip reading and that comprehension
on telephone is a severe disability in central auditory
disorders in general and for this patient in particular.
This generalization to non-trained tasks and transfer
to daily life activity demonstrates the efficacy of this
therapy and its relevance for patients with verbal
deafness.
There was previous evidence that phoneme dis-
crimination skills could be improved by therapy.
Morris et al. [4] performed a single-case study with
the aim to improve phonemic discrimination using
phoneme pair discrimination tasks. After therapy,
the subject improved significantly on phoneme pair
discrimination tests and in repetition, but only a
trend toward improvement in oral comprehension
was reported. This improvement in phoneme dis-
crimination was not found in another study [6],
in which therapy programme using phoneme dis-
crimination with lip-reading cues did not allow
changes in phoneme discrimination tasks. The
difference of severity of the impairment may explain
difference between Maneta et al.’s [6] study results
and those obtained by Morris et al. [4] and here.
Maneta et al.’s patient did not perceive any
phonemic contrast, thus practising that contrast
can not help. In contrast, Morris et al.’s patients as
well as this patient were still able to discriminate
some phoneme differences even if their perfor-
mances were not normal.
Morris et al.’s [4] therapy and this one are close,
but it is worthwhile to understand the differences.
Indeed, Morris et al.’s patient and this patient share
several common points. Both suffered cerebrovas-
cular disease, with subcortical brain lesions. The
severity of the phoneme discrimination and recogni-
tion impairment seems to be similar in both studies
and they benefited of the same amount of therapy
(12 sessions of 1 hour). As in this therapy, exercises
began with items where the patient was just
beginning to fail and gradually became increasingly
difficult; they used phonemes discrimination and
recognition of one phoneme. However, there was
one difference between the patient disorders and two
main differences between these therapies. This
patient presented word deafness but no aphasia
and had no other language disorder reasons to be
impaired in oral repetition or comprehension. In
contrast, Morris et al.’s patient presented with lexical
and semantic disorders which can worsen difficulties
for repetition and comprehension and decrease the
impact of phoneme discrimination or repetition
remediation on these exercises. First, they used
lexical exercises addressing a higher level on the
cognitive model whereas the protocol used here was
only a low level of treatment exercises. Morris et al.
[4] used auditory word-picture matching or correct/
incorrect judgement and auditory word-written word
matching. Secondly, the reference to an errorless
design was more systematic in this study and so use
of the cues was different. In Morris et al.’s study, the
cues were lip reading and sometimes hand signs.
During therapy, lip-reading was allowed during the
first sessions and was secondly reduced. However,
when the patient failed in one exercise, the item was
given again using lip-reading and hand sign. In
contrast, this study used systematic cueing and its
progressive fading, in order to facilitate the task at
each level of difficulty. This study provided a
‘decreasing assistance’ [28], the sequence of visual
cues being arranged from the strongest to the
weakest. This design was voluntarily chosen in
order to avoid errors, as far as possible. The results
reinforce those of Morris et al. [4] to demonstrate
efficacy of phoneme discrimination rehabilitation.
Therapy specificity must be questioned. In this
study, the improvement in phoneme processing and
oral comprehension and repetition was specific to
verbal sounds. There was no improvement in
environmental sounds recognition after this verbal
therapy. This result demonstrates the specificity of
this therapy for verbal sounds. Word deafness was
not associated with aphasia disorder as in Morris
1172 C. Tessier et al.
Brain Inj Downloaded from informahealthcare.com by Universite Pierre & Marie Curie on 05/27/13
For personal use only.
et al.’s [4] study, therefore a stability of language
tests can not be demonstrated. This study did not
find a specific effect of phoneme discrimination
therapy and phoneme recognition therapy. It
observed a generalization of the phoneme discrimi-
nation improvement to the phoneme recognition. It
is suggested that there is no specific effect of
discrimination and recognition therapy but a com-
plementary effect of both therapies because of the
same mechanism explaining both disorders.
The main aim of this study was to assess efficacy
and specificity of this rehabilitation programme. The
analysis and the rehabilitation of this auditory
disorder were based on the cognitive model of
auditory processing and auditory word comprehen-
sion proposed by Ellis et al. [15] and the efficacy and
specificity of the rehabilitation performed allow
discussion of this model.
The specificity of this rehabilitation addresses the
hypothesis of modularity in audition processing [3].
Some authors [13, 43] argued that word deafness
and verbal sound processing were rarely pure and
were frequently associated to environmental sound
agnosia and to amusia and argued that, at a
perceptive level, disorders of word comprehension,
of environmental sounds and musical perception
could be caused by disordered complex sound
perception. On the opposite, Polster and Rose [3]
proposed a modular architecture of auditory proces-
sing analogous to models of visual processing. MA
presented with associated impairments, but she
specifically improved in verbal processing which
reinforce the hypothesis of the modularity of this
process [3].
In reference to the cognitive model (Figure 1), the
efficacy of the rehabilitation addressing the lower
stage (auditory analysis) improving comprehension
and repetition at a higher level (lexical and semantic)
is coherent with this serial model [15]. Furthermore,
one proposes two modifications to this model. First,
the better performance in word repetition than in
syllable repetition observed before rehabilitation can
be explained by a top-down treatment, the intact
lexicon providing help to understand words but not
syllables [44, 45]. Secondly, this study proposes a
larger place for the phonological buffer, which was
shown to be dissociated from the auditory analysis
system [44]. The behaviour of the patient during
exercises of rehabilitation, using repetition of the
phonemes she heard, suggests a role of short-term
verbal memory and of the phonological loop of
working memory [46] in this training. The two
components of the phonological loop, the
phonological store (or phonological buffer) in
which representation of verbal material are held
and the sub-vocal rehearsal mechanism which serves
to refresh the contents of the phonological store, are
involved in processing and maintenance of verbal
information. The hypothesis is that this phonological
loop is normal and can be used to enhance the
training of phonemes. Unfortunately, this study did
not assess working memory before and after rehabi-
litation. This hypothesis should be addressed in a
further study.
Conclusion
In a case of verbal deafness, this study demonstrates
not only the efficacy of a specific phoneme proces-
sing rehabilitation addressing the lower perceptive
level of auditory analysis, but also its efficacy in the
improvement of higher level of cognitive treatment
such as comprehension, which confirms the cogni-
tive model and the generalization of the therapy.
The role of the errorless method of therapy must be
underlined. However, these results need to be
demonstrated in other patients with a pure deficit
on this low level of auditory treatment and,
furthermore, to aphasic patients presenting both
central auditory and language disorders.
Acknowledgements
The authors wish to express their gratitude to Rose
Katz, Catherine Morin, Mathilde Chevignard and
Dominique Mazevet for reading and commenting
upon the manuscript. We thank the anonymous
reviewers for their helpful comments. This work was
supported by Assistance Publique des Ho
ˆpitaux de
Paris and by grants from Institut National pour la
Sante
´et la Recherche Me
´dicale (CNEP INSERM
93CN42).
References
1. Franklin S. Dissociations in auditory word comprehension,
evidence from nine fluent aphasic patients. Aphasiology
1989;3:189–207.
2. Eustache F. De l’agnosie auditive a
`la neuropsychologie de
l’audition. Revue de Neuropsychologie 1993;3:353–356.
3. Polster PR, Rose SB. Disorders of auditory processing,
evidence for modularity in audition. Cortex 1998;34:47–65.
4. Morris J, Franklin S, Ellis AW, Turner JE, Bailey PJ.
Remediating a speech perception deficit in an aphasic patient.
Aphasiology 1996;10:137–158.
5. Wee J, Menard MR. ‘Pure word deafness’, implications for
assessment and management in communication disorder. A
report of two cases. Archives of Physical Medicine and
Rehabilitation 1999;80:1106–1109.
6. Maneta A, Marshall J, Lindsay J. Direct and indirect
therapy for word sound deafness. International Journal of
Language and Communication Disorders 2001;36:91–106.
7. Plasencia PM, Dorado JI, Serrano Rodriguez JM, Sellan C.
Neuropsychological evidence for ‘word-meaning deafness’
in a Spanish-speaking patient. Brain & Language 2006;
97:214–218.
Rehabilitation of word deafness due to auditor y analysis disorder 1173
Brain Inj Downloaded from informahealthcare.com by Universite Pierre & Marie Curie on 05/27/13
For personal use only.
8. Wirkowski E, Echausse N, Overby C, Ortiz O, Radler L. I
can hear you yet cannot comprehend: a case of pure word
deafness. Journal of Emergency Medicine 2006;30:53–55.
9. Musiek F, Charrette L, Morse D, Baran A. Central deafness
associated with midbrain lesion. Journal of the American
Academy of Audiology 2004;15:133–151.
10. Shivashankar N, Shashikala HR, Nagaraja D, Jayakumar PN,
Ratnavalli E. Pure word deafness in two patients
with subcortical lesions. Clinical Neurology and
Neurosurgery 2001;103:201–205.
11. Vitte E, Tankere F, Bernat I, Zouaoui A, Lamas G,
Soudant J. Midbrain deafness with normal brainstem
auditory evoked potentials. Neurology 2002;58:970–973.
12. Tanaka Y, Nakano I, Obayashi T. Environmental sound
recognition after unilateral subcortical lesions. Cortex
2002;38:69–76.
13. Griffiths TD, Rees A, Green GGR. Disorders of human
complex sound processing. Neurocase 1999;5:365–378.
14. Griffiths TD. Central auditory processing disorders. Current
Opinions in Neurology 2002;15:31–33.
15. Ellis AW, Franklin S, Crerar A. Cognitive neuropsychology
and the remediation of disorders of spoken language. In:
Humphreys G, Riddoch J, editors. Cognitive neuropsychol-
ogy and cognitive rehabilitation. HillsdaleNJ: Laurence
Erlbaum Associates; 1994. pp 287–315.
16. Ducarne B, Tessier C. Etude longitudinale et re
´e
´ducation de
la perception acoustique auditive d’un cas de surdite
´centrale.
Glossa 1992;30:10–17.
17. Roch Lecours A, Lhermitte F. Re
´e
´ducation des troubles de la
compre
´hension du langage oral. L’aphasie. Paris:
Flammarion; 1979. pp 578–584.
18. Shindo M, Kimikata K, Tanaka Y. Speech
discrimination and lip reading in patients with word deafness
or auditory agnosia. Brain and Language 1991;40:301–403.
19. Baddeley A. A theory of rehabilitation without a model of
learning is a vehicle without an engine, a comment on
Caramazza and Hillis. Neuropsychological Rehabilitation
1993;3:235–244.
20. Caramazza A, Hillis A. For a theory of remediation of
cognitive deficits. Neuropsychological Rehabilitation
1993;3:217–234.
21. Mitchum CC, Berndt RS. The cognitive neuropsychological
approach to treatment of language disorders.
Neuropsychological Rehabilitation 1995;5:1–16.
22. Wilson BA. Editorial, How do we know that rehabilitation
works? Neuropsychological Rehabilitation 1993;3:1–4.
23. Behrman M, McLeod J. Rehabilitation for pure alexia,
efficacy of therapy and implication for models of normal
word recognition. Neuropsychological Rehabilitation 1995;
5:149–180.
24. Glisky E. Acquisition and transfer of word processing skill by
an amnesic patient. Neuropsychological Rehabilitation
1995;5:299–318.
25. Komatsu S, Mimura M, Kato M, Wakamatsu N, Kashima H.
Errorless and effortful processes involved in the learning of
face-name associations by patients with alcoholic Korsakoff’s
syndrome. Neuropsychological Rehabilitation 2000;
10:113–132.
26. Fillingham J, Hodgson C, Sage K, Lambon Ralph M. The
application of errorless learning to aphasic disorders: A review
of theory. Neuropsychological Rehabilitation 2003;
13:337–416.
27. Fillingham J, Sage K, Lambon Ralph M. Treatment of
anomia using errorless learning. Neuropsychological
Rehabilitation 2006;16:129–154.
28. Riley G, Heaton S. Guideline for the selection of a method of
fading cues. Neuropsychological Rehabilitation 2000;10:
133–149.
29. Robertson I, Gray J, Prentland B, Waite L. Microcomputer-
based rehabilitation for unilateral left visual neglect, a
randomized controlled trial. Archives of Physical Medicine
and Rehabilitation 1990;71:663–668.
30. Ladavas E, Menghini G, Ulmita C. A rehabilitation study of
hemispatial neglect. Cognitive Neuropsychology 1994;
11:75–95.
31. Pradat-Diehl P, Marchal F, Durand E, Mazevet D,
Taillefer C, Migeot H, Azouvi P. Indic¸age visuel et
re
´entrainement a
`l’exploration visuelle chez les patients
he
´mine
´gligents. In: Perennou D, Brun V, Pe
´lissier J, editors.
Les syndromes de ne
´gligence spatiale. Paris: Masson; 1998.
pp 235–243.
32. Ducarne B. Examen de l’aphasie. Paris: Editions du Centre
de Psychologie Applique
´e; 1989.
33. Tessier C, Weill-Chounlamountry A. Inte
´re
ˆt de l’outil
informatique dans la the
´rapie re
´e
´ducative des TAOC. In
Entretiens d’Orthophonie de Bichat. Paris: Expension
scientifique franc¸aise; 2000. pp 39–45.
34. Darrigrand B, Mazaux JM. Echelle de Communication
Verbale de Bordeaux. Isbergues: OrthoEditions; 2000.
35. Darrigrand B, Mazaux JM, Dutheil S, Koleck M, Pradat-
Diehl P. Evaluer la communication de la personne
aphasique dans la vie quotidienne, proposition d’une e
´chelle
de communication verbale. In: Mazaux JM, Brun V, Pe
´lissier
J, editors. Aphasie. Re
´e
´ducation et re
´adaptation des aphasies
vasculaires. Paris: Masson; 2000. pp 88–94.
36. Ducarne B. Re
´e
´ducation se
´miologique de l’aphasie. Paris:
Masson; 1986.
37. Auerbach SH, Allard T, Naeser M, Alexander MP,
Albert ML. Pure word deafness, analysis of a case with
bilateral lesions and a defect at the prephonemic level. Brain
1982;105:271–300.
38. Blumstein S, Baker E, Goodglass H. Phonological factors in
auditory comprehension in aphasia. Neuropsychologia
1977;15:19–30.
39. Praamstra P, Hagoort P, Maassen B, Crul T. Word deafness
and auditory cortical function, a case history and hypothesis.
Brain 1991;114:1197–1225.
40. Pedersen PM, Jorgenson HS, Nakayama H,
Raaschou HO, Olsen TS. Aphasia in acute stroke,
incidence determinants, and recovery. Annals of
Neurology 1995;38:659–666.
41. Pedersen P, Vinter K, Olsen T. Aphasia after stroke: Type,
severity and prognosis. The Copenhagen aphasia study.
Cerebrovascular Diseases 2004;17:35–43.
42. Robey RR. A meta-analysis of clinical outcomes in the
treatment of aphasia. Journal of Speech, Language, and
Hearing Research 1998;41:172–187.
43. Pinard M, Chertkow H, Black S, Peretz I. A case study of
pure deafness word, modularity in auditory processing.
Neurocase 2002;8:40–55.
44. Martin RC, Breedin SD, Damian MF. The relation of
phoneme discrimination, lexical access, and short term
memory: A case study and interactive activation account.
Brain and Language 1999;70:437–482.
45. Jacquemot C, Dupoux E, Pallier C, Bachoud-Levi AC.
Comprehending spoken words without hearing phonemes: A
case study. Cortex 2002;38:869–873.
46. Baddeley A. Working memory. New York: Oxford University
Press; 1986.
1174 C. Tessier et al.
Brain Inj Downloaded from informahealthcare.com by Universite Pierre & Marie Curie on 05/27/13
For personal use only.
