The ScreeLing: occurrence of linguistic deficits in acute aphasia post-stroke.

Abstract

To investigate the occurrence of semantic, phonological and syntactic deficits in acute aphasia with the ScreeLing after the establishment of its psychometric properties. To examine the relationship between these deficits and: (i) overall aphasia severity; and (ii) quality of Spontaneous Speech.
The reliability and validity of the ScreeLing was established by investigating 141 subjects with acute aphasia (2 weeks after stroke), 23 with chronic aphasia, and 138 healthy controls. In addition, the acute patients were assessed with the Token Test and a Spontaneous Speech rating (Aphasia Severity Rating Scale).
The ScreeLing was found to be valid and reliable for assessing the presence and severity of aphasia and linguistic deficits at 12 days after stroke. In 22.4% of the patients deficits were found in only 1 of the 3 linguistic levels; phonology was most frequently disturbed (16.3%), compared with semantics (2.7%), and syntax (3.4%). The number of impaired linguistic levels was related to aphasia severity: patients with a 3-level disorder had the lowest Token Test scores; patients with a selective phonological disorder had the highest Spontaneous Speech ratings. Phonology alone explained 54.6% of the variance in the Spontaneous Speech rating.
In the acute stage, linguistic-level deficits are already present independently of each other, with phonology affected most frequently.

Full text

Page 1

ORIGINAL REPORT
J Rehabil Med 2012; 44: 429–435
J Rehabil Med 44
© 2012 The Authors. doi: 10.2340/16501977-0955
Journal Compilation © 2012 Foundation of Rehabilitation Information. ISSN 1650-1977
Objective: To investigate the occurrence of semantic, pho-
nological and syntactic deficits in acute aphasia with the
ScreeLing after the establishment of its psychometric prop-
erties. To examine the relationship between these deficits
and: (i) overall aphasia severity; and (ii) quality of Sponta-
neous Speech.
Methods: The reliability and validity of the ScreeLing was
established by investigating 141 subjects with acute apha-
sia (2 weeks after stroke), 23 with chronic aphasia, and 138
healthy controls. In addition, the acute patients were as-
sessed with the Token Test and a Spontaneous Speech rating
(Aphasia Severity Rating Scale).
Results: The ScreeLing was found to be valid and reliable for
assessing the presence and severity of aphasia and linguistic
deficits at 12 days after stroke. In 22.4% of the patients defi-
cits were found in only 1 of the 3 linguistic levels; phonology
was most frequently disturbed (16.3%), compared with se-
mantics (2.7%), and syntax (3.4%). The number of impaired
linguistic levels was related to aphasia severity: patients with
a 3-level disorder had the lowest Token Test scores; patients
with a selective phonological disorder had the highest Spon-
taneous Speech ratings. Phonology alone explained 54.6% of
the variance in the Spontaneous Speech rating.
Conclusion: In the acute stage, linguistic-level deficits are al-
ready present independently of each other, with phonology
affected most frequently.
Key words: aphasia; language disorders; diagnosis; cerebrovas-
cular accident; screening test; semantics; phonetics.
J Rehabil Med 2012; 44: 429–435
Correspondence address: Hanane El Hachioui, Erasmus MC
University Medical Center, Department of Neurology, Room
EE 2291, PO Box 2040, 3000 CA Rotterdam, The Netherlands.
E-mail: h.hachiouiel@erasmusmc.nl
Submitted September 30, 2011; accepted December 13, 2011
INTRODUCTION
The prognosis of aphasia after stroke depends largely on its
initial severity (1–3), but other factors may also play an impor-
tant role (2, 4). Regression models have so far explained only
part of the variance of the outcome of aphasia (2, 5), indicating
that other prognostic factors have not yet been discovered.
For example, the nature of the linguistic disorder may be an
important prognostic factor for aphasia outcome.
The only data available are about the frequencies of aphasia
subtypes in acute stroke, and these are inconsistent. For in-
stance, the reported incidence of Broca’s aphasia varies from
11% to 22% (5–7), probably due to the fact that classification
of aphasia is difficult in the acute stage. Many patients are
not classifiable according to classic aphasia syndromes (6)
and during the first weeks after stroke these syndromes tend
to change.
It has been reported that domain-specific cognitive func-
tions are good predictors for long-term cognitive outcome
(8). In addition, the prevalence of domain-specific cognitive
deficits in the acute stage after stroke has been established (9).
For aphasia, this information is unknown. In order to explore
whether the core linguistic components of language production
and comprehension, i.e. semantics, phonology and syntax, are
relevant prognostic factors, detailed information is first needed
on the nature and occurrence of linguistic-level deficits. This
information is lacking in the acute stage because of the lack
of tests providing a specific linguistic-level diagnosis suit-
able for administration in the early stages after stroke when
time-consuming tests are too much of a burden. The existing
screening tools for acute aphasia usually reflect the approach
taken in traditional aphasia test batteries that assess language
modalities such as comprehension and reading (10, 11), and
are not aimed at the linguistic-level deficits.
To the best of our knowledge, the only linguistic screening
test designed to assess the presence of aphasia and to dif-
ferentiate linguistic-level disorders in the acute stage is the
ScreeLing. In a small group study (n = 17) 30% of the patients
showed selective linguistic disorders on a research version of
this test (12). The test has been refined; less accurate subtests
have been replaced and, based on item analysis, further adjust-
ments have been made in order to enhance its clinical value
(see Appendix I for further details) (13).
Information on the occurrence of linguistic-level disorders
may be important for several reasons. Establishing the oc-
currence of linguistic deficits in the acute stage will provide
more insight into early recovery patterns. Discovering which
linguistic deficits are persistent and which may recover spon-
taneously provides a basis for the selection of additional, more
ThE SCREELINg: OCCURRENCE OF LINgUISTIC DEFICITS IN ACUTE
AphASIA pOST-STROkE
Hanane El Hachioui, MSc1, Mieke W. M. E. van de Sandt-Koenderman, PhD2,3,
Diederik W. J. Dippel, MD, PhD1, Peter J. Koudstaal, MD, PhD1 and Evy G. Visch-Brink, PhD1
From the Departments of 1Neurology, 2Rehabilitation Medicine, Erasmus MC University Medical Center and 3Rijndam
Rehabilitation Center, Rotterdam Neurorehabilitation Research (RoNeRes), Rotterdam, The Netherlands

Page 2

430
H. El Hachioui et al.
comprehensive assessment, which may result in a better guid-
ance during the treatment course. Furthermore, insight into the
occurrence of linguistic-level deficits in the acute stage may
be used to examine the impact of early linguistic profiles on
final outcome.
The aims of this study were: (i) to report on the psycho-
metric properties of the revised ScreeLing; (ii) to investigate
the occurrence of linguistic-level deficits in a large group
of patients with aphasia at 2 weeks after stroke; and (iii) to
determine the relationship between linguistic-level deficits
and overall aphasia severity, as well as the verbal abilities in
spontaneous speech.
METhODS
Participants
Acute aphasia patients. Patients were recruited from the stroke units
of 17 hospitals in the Netherlands, and screened by the local neurolo-
gist (based on clinical examination) and speech-language therapist
(SLT) (based on an interview). Inclusion criteria were: adult Dutch
native/near-native speaker (i.e. education in Dutch started from early
childhood and primary use of the Dutch language in everyday life);
aphasia after a first-ever intracerebral haemorrhage or infarction; and
testable with the ScreeLing (13) between 2 days and 2 weeks after
stroke (i.e. alert during the administration of the test and not too ill
to tolerate at least 15 min of the ScreeLing assessment; it was also
allowed to administer the 3 linguistic components in a maximum of
3 test sessions if completed within 2 consecutive days). Exclusion
criteria were: pre-stroke dementia (suspected or confirmed); severe
dysarthria; developmental dyslexia; severe impairment of vision and
hearing (based on the medical history and standard clinical examination
by the attending physician); illiteracy; and psychiatric disorder.
Chronic aphasia patients. Adult Dutch native/near-native speakers
(i.e. education in Dutch started from early childhood and primary use
of the Dutch language in everyday life) with aphasia after intracerebral
haemorrhage or infarction of at least 6 months who were testable with
the ScreeLing (13), were recruited from 10 treatment centres by their
SLT. Exclusion criteria were: dementia (suspected or confirmed); severe
dysarthria; developmental dyslexia; severe impairment of vision and
hearing (based on the medical history and standard clinical examination
by the attending physician); illiteracy; and psychiatric disorder.
Healthy control group. Native/near-native speakers of Dutch (i.e.
education in Dutch started from early childhood and primary use
of the Dutch language in everyday life) older than 18 years were
recruited by speech-language therapy Masters students from their
family and friends. Exclusion criteria were: cerebral disease; dementia
(suspected or confirmed); developmental dyslexia; severe impair-
ment of vision and hearing (based on an interview); illiteracy; and
psychiatric disorder.
This study was approved by the central medical ethics committee
of Erasmus MC University Medical Center and by the local ethics
committees of the participating centres. Informed written consent
was obtained from the participants and/or their close relatives prior
to their inclusion in the study.
Assessment
1. The ScreeLing investigates 3 linguistic levels (i.e. semantics, pho-
nology, syntax) with a maximum score for each level of 24, and a
maximum overall score of 72 (13) (see Appendix I). The ScreeLing
and the 3 linguistic levels were handled as continuous variables (i.e.
mean values and standard deviations (SD’s) are reported) in line with
the previous report on the research version of this test (12).
2. Spontaneous Speech was elicited in a 10-min semi-standardized
interview according to the Aachen Aphasia Test procedure (14) with
4 topics: the beginning and course of the disease; occupation; family
and housing conditions; and hobbies. This interview was evaluated
with the Aphasia Severity Rating Scale of the Boston Diagnostic
Aphasia Examination (15). This categorical variable is a 6-point
scale varying from 0 “no usable speech or auditory comprehension”
to 5 “minimal discernible speech handicap”.
3. The Token Test (36 items) is a well-known and well-validated test
to measure the presence and the severity of aphasia (16). The Token
Test score was handled as a continuous variable (i.e. mean values and
SD’s are reported) in line with the report on the 36 item-version we
used (16).
For the acute patients, the assessment comprised the complete set
of tests. The healthy control group was assessed with the ScreeLing
and the Token Test; the chronic patients were tested twice with the
ScreeLing with an interval of minimally 1 and maximally 2 weeks to
investigate the test re-test reliability.
Statistical analyses
First, we established the psychometric properties of the ScreeLing by
conducting reliability and validity analyses. We calculated the internal
consistency with Cronbach’s α in the acute patients and healthy con-
trols combined. The test-retest reliability was determined in the chronic
patient group using Bland-Altman plots. For the construct validity, we
compared the ScreeLing performance of the acute patients with that of
the healthy controls with independent samples t-tests. The diagnostic
accuracy of the ScreeLing and each of its 3 linguistic levels was deter-
mined by means of receiver operating characteristic (ROC) analysis.
The sensitivity and specificity were set at the optimal cut-off point.
In order to provide information on concurrent validity, correlation
analyses were conducted between the ScreeLing and the Token Test,
and between the ScreeLing and the Spontaneous Speech rating.
Secondly, differences in mean scores between the 3 linguistic levels
were examined separately for the acute patients and healthy controls
with paired samples t-tests, in order to establish whether the subtests
were equally complex for healthy speakers, and to investigate whether
the linguistic levels were equally impaired in aphasia. To obtain the oc-
currence of the linguistic-level disorders in the acute patients frequency
analyses were used. For establishing possible differences in aphasia
severity between subgroups of the acute patients (i.e. with a selective
linguistic-level disorder, a combined disorder, or a 3-level disorder), we
performed one-way ANOVA analysis and Kruskal-Wallis analysis. To
identify pairwise differences we conducted post-hoc multiple compari-
sons tests with Bonferroni correction and Mann-Whitney tests. Finally, to
determine the impact of the linguistic-level disorders in the acute patients
on Spontaneous Speech we used ordinal regression analysis.
All analyses were carried out with SPSS 15.0 (SPSS Inc., Chicago,
USA).
RESULTS
Between June 2007 and June 2009, 147 acute stroke patients
with aphasia were included. The complete assessment was
administered at 11.66 days (SD 2.10 days) after stroke. We
excluded 6 patients whose assessments could not be completed
within the time limits because no SLT was available for testing.
An additional 23 chronic patients (mean time after stroke 49.96
months, SD 95.62 months) were included between November and
December 2009. We included 138 healthy controls from April to
May 2007. Participants’ characteristics are shown in Table I.
The 3 groups were compared with Mann-Whitney tests. The
acute and chronic patients did not differ significantly for age
or education level. The healthy controls were younger than the
J Rehabil Med 44

Page 3

431
Occurrence of linguistic deficits post-stroke
acute patients (Z = –4.46, p < 0.001) and the chronic patients
(Z = –2.64, p = 0.008). Their education level was higher than
of the acute patients (Z = –3.51, p < 0.001) and the chronic
patients (Z = –2.95, p = 0.003).
Psychometric properties of the ScreeLing
The Cronbach’s α of the total ScreeLing and phonology was
0.95; of semantics and syntax was 0.93. These results show
high internal consistency for the total ScreeLing and for its
linguistic levels.
Test-retest reliability of the ScreeLing was examined in the
chronic group. Each patient was assessed at a mean interval
of 10 days (SD 3.16). The Bland-Altman plots illustrate high
agreement between the two assessments, indicating a high
stability of the ScreeLing over time (Fig. 1).
The comparison of the performances on the ScreeLing of
the acute patients and the healthy controls revealed an overall
significant difference on the total ScreeLing and its linguistic
levels (Table II).
A ROC analysis showed that the ScreeLing discriminates
accurately (0.94) between aphasic patients and healthy controls
(Table III). The optimal cut-off score for the total ScreeLing
was 68, i.e. patients scoring less than 68 were classified as
aphasic. This led to a sensitivity of 0.94 and a specificity of
0.81 with an overall correct classification of 0.88.
The ScreeLing and its linguistic levels correlated signifi-
cantly with the Token Test and the Spontaneous Speech rating
in the acute aphasic patients (Table IV). The Token Test showed
the strongest correlation with the overall ScreeLing score.
The Spontaneous Speech rating was most related to phonol-
ogy, as this is the only part of the ScreeLing that incorporates
language production. The high similarity and the significant
relationships between the ScreeLing and the other two aphasia
tests suggested a good concurrent validity.
Selective linguistic disorders
We examined possible differences in mean scores between the
3 linguistic levels with paired samples t-tests (Table II). In the
Table I. Baseline characteristics of the participants
Acute patients
(n = 141)
healthy controls
(n = 138)
Chronic patients
(n = 23)
Age, years, mean (SD) [range]
gender, n (%)
Female
Male
handedness (EhI), n (%)
Right-handed
Left-handed
Ambidextrous
Unknown
Level of education, n (%)
Unfinished elementary school
Elementary school
(Unfinished) Middle school
Sophomore high school or lower vocational education
Junior high school or middle vocational education
Senior high school or higher vocational education
University
Unknown
Type of stroke, n (%)
Infarction
haemorrhage
Both (infarction and haemorrhage)
Clinical localization of stroke, n (%)
Left hemisphere
Right hemisphere
66.61 (14.90) [19–96] 55.74 (20.83) [18–88] 67.96 (14.76) [29–89]
75 (53.2)
66 (46.8)
73 (52.9)
65 (47.1)
10 (43.5)
13 (56.5)
123 (87.2)
15 (10.7)
2 (1.4)
1 (0.7)
120 (87.0)
11 (8.0)
7 (5.1)
0 (0)
21 (91.3)
2 (8.7)
0 (0)
0 (0)
3 (2.1)
20 (14.2)
5 (3.6)
44 (29.1)
38 (27)
26 (18.4)
5 (3.5)
3 (2.1)
0 (0)
11 (8.0)
12 (8.7)
15 (10.9)
46 (33.3)
31 (22.4)
23 (16.7)
0 (0)
–
0 (0)
3 (13)
0 (0)
9 (39.1)
9 (39.1)
1 (4.4)
0 (0)
1 (4.4)
121 (85.8)
20 (14.2)
0 (0)
20 (87)
2 (8.7)
1 (4.3)
–
139 (98.6)
2 (1.4)
22 (95.7)
1 (4.3)
SD: standard deviation; EhI: Edinburgh handedness Inventory.
Table II. Construct validity: mean total ScreeLing and linguistic-level scores for the acute patients and healthy controls
Acute patients (n = 141)
Mean (SD) [SE]
healthy controls (n = 138)
Mean (SD) [SE]
Difference
Mean (95% CI)
p (independent samples
t-tests)
Semantics
phonology
Syntax
Total ScreeLing
19.22 (5.6) [0.47]
16.98 (6.06) [0.51]
17.96 (5.76) [0.49]
54.16 (16.14) [1.36]
23.63 (0.63) [0.05]
23.69 (0.63) [0.05]
23.53 (0.77) [0.07]
70.85 (1.38) [0.12]
4.41 (3.47–5.35)
6.71 (5.70–7.73)
5.57 (4.60–6.54)
16.70 (14.0–19.39)
< 0.001
< 0.001
< 0.001
< 0.001
SD: standard deviation; SE: standard error of the mean; CI: confidence interval.
J Rehabil Med 44

Page 4

432
H. El Hachioui et al.
healthy control group there was one small significant difference
between phonology and syntax (p = 0.038, 95% confidence
interval (CI) = 0.0 to 0.32), in favour of phonology. In the
acute group there was a significant difference between all 3
levels, i.e. between semantics and phonology (p < 0.001, 95%
CI = 1.49 to 2.99), between semantics and syntax (p < 0.001,
95% CI = 0.71 to 1.81), and between phonology and syntax
(p = 0.001, 95% CI = –1.56 to –0.40). The phonological level
showed the lowest scores; semantics scored the highest.
To ascertain the occurrence of linguistic-level deficits in the
first 2 weeks after stroke, we conducted frequency analyses
(Table V). Selective linguistic-level disorders occurred in
22.4% of the patients; they scored lower than 22 on one par-
ticular level, whereas their score on the other two linguistic
levels was normal, i.e. > 22. A selective phonological disorder
occurred most frequently (16.3%). These patients had a mean
phonology score of 19.71 (SD 1.99), a mean Token Test-score
of 27.20 (SD 5.41), and 79.2% of them had a high Spontaneous
Speech rating (score 4 or 5). Among the combined disorders of
two linguistic levels, the most frequent was the combination
of a phonological and syntactic deficit (13.6%). Patients with
this combination had a mean phonology score of 17.60 (SD
3.46), a mean syntax score of 18.50 (SD 2.8), a mean Token
Test-score of 23.55 (SD 5.52), and 30% of the patients had a
high Spontaneous Speech rating. A 3-level disorder was found
in approximately 39% of the patients; these patients had a
mean score for semantics of 13.88 (SD 5.08), for phonology
11.60 (SD 5.33), and for syntax 12.58 (SD 4.86). This group of
patients had a mean Token Test score of 10.06 (SD 6.89) and
only 26.3% had a high Spontaneous Speech rating.
Twenty-five patients did not have a disorder on any of the lin-
guistic levels. All had been judged as aphasic by their neurolo-
gist and speech-language therapist. The Token Test classified
8 of these patients as aphasic; according to the Spontaneous-
Speech rating 17 were aphasic, whereas according to the overall
Table III. ScreeLing and its linguistic levels: accuracy, sensitivity, and
specificity (n = 279)
Accuracy
Optimal
cut-off point Sensitivity Specificity
Semantics
phonology
Syntax
Total ScreeLing
0.79
0.94
0.87
0.94
22
22
22
68
0.94
0.93
0.91
0.94
0.56
0.83
0.74
0.81
Table IV. Concurrent validity: comparing ScreeLing with Token Test and
Spontaneous Speech rating (n = 141)
Token Test
(pearson)
Spontaneous
Speech rating
(Spearman’s)p
Semantics
phonology
Syntax
Total ScreeLing
0.79
0.80
0.85
0.88
0.58
0.73
0.67
0.73
< 0.001
< 0.001
< 0.001
< 0.001
Fig. 1. Bland-Altman plots (n = 23). SD: standard deviation.




J Rehabil Med 44

Page 5

433
Occurrence of linguistic deficits post-stroke
score of the ScreeLing one patient was aphasic. Four did not
have aphasia according to any of these measures.
Relationship between linguistic-level deficits and aphasia
severity
There was an overall significant difference (p < 0.001) in aphasia
severity, measured by the Token Test, between the subgroups
of patients with a selective phonological deficit, a combined
phonological and syntactic deficit, and a 3-level deficit (one-way
ANOVA analysis). Patients with a 3-level disorder were the most
severe (p < 0.001). Their mean Token Test score was significantly
lower than the mean Token Test score of the patients with a
selective phonological disorder (mean difference = –17.13, 95%
CI = –20.94 to –13.33) and of the patients with a combined pho-
nological and syntactic disorder (mean difference = –13.49, 95%
CI = –17.50 to –9.48). There was no difference in mean Token
Test score between the patients with a selective phonological
disorder and a combined disorder.
The subgroups of patients with a selective phonological
deficit, a combined phonological and syntactic deficit, and
a 3-level deficit, showed an overall significant difference in
the Spontaneous Speech rating with the Kruskal-Wallis test
(χ2 = 30.50, degrees of freedom = 2, p < 0.001). The selective
phonological disorder group showed that more patients had
high Spontaneous Speech ratings than in the group with a
combined phonological and syntactic disorder (Mann-Whitney
tests, Z = –3.30, p = 0.001), and the 3-level disorder group
(Z = –5.39, p < 0.001). There was no difference in Spontaneous
Speech ratings between the patients with a combined disorder
and a 3-level disorder.
Semantic, phonological and syntactic scores explained
56.3% of the variance of Spontaneous Speech in ordinal
regression analysis. Semantics and syntax did not contribute
significantly to this effect: phonology alone explained 54.6%
of the variance.
DISCUSSION
The ScreeLing proved to be a valid and reliable measure for
assessing semantic, phonological, and syntactic deficits in
acute aphasia after stroke. Selective linguistic-level disorders
occurred in 22.4% of the aphasic patients with phonology as
most frequently affected. The importance of assessing the 3
linguistic levels separately was further underlined by the find-
ing that they had a different impact on spontaneous speech.
In addition, patients with a selective phonological disorder
had the highest Spontaneous Speech ratings. The number of
linguistic-level disorders was related to the severity of aphasia,
measured with the Token Test; patients with impairments on
all 3 linguistic levels had the lowest Token Test scores.
Our study is the first report on the occurrence of linguistic-
level deficits in the acute stage in a large cohort of aphasic
stroke patients. In addition, the ScreeLing is the first thoroughly
evaluated linguistic-level screening test suitable for assessing
the presence and severity of the main linguistic-level deficits
in early aphasia. It even exceeds the overall sensitivity and
specificity of the well-known Frenchay Aphasia Screening Test
(FAST), which was reported to be the best out of 6 aphasia
screening tests (17).
Some aspects of the ScreeLing deserve mention. In the acute
stage patients are often too ill to be tested extensively, therefore
the ScreeLing has to be short and easy to administer at the bed-
side in a hospital as well as in a rehabilitation setting. Another
crucial aspect regards the decreasing time of hospitalization:
sufficient linguistic information should be available as soon
as possible to enable additional targeted assessment for an
adequate referral. For each linguistic level, we selected various
tasks that optimally capture each linguistic level, as it is not
clear which linguistic task best represents language processing
at the 3 linguistic levels. Not all well-known linguistic tasks
appeared to be suitable for the acute stage. For example, we
decided not to use non-word repetition even though this is
known to represent phonological processing (18). This task
appeared too much of a burden for acute patients.
An earlier version of the ScreeLing proved to have a high
sensitivity (86%) and specificity (96%) in discriminating
aphasic and non-aphasic acute stroke patients (12) (see Ap-
pendix I for the modifications). A limitation of the present
study is that we did not examine the discriminative power
of the ScreeLing in stroke patients with and without aphasia,
as our norm group was restricted to healthy controls. Even
though this is standard procedure in neuropsychological tests,
we will try to incorporate this aspect in our future research.
Another limitation is that our healthy controls were not age-
matched and education-matched with the acute and chronic
patient groups. In our future research, we will try to include
norm groups of healthy speakers and stroke patients without
aphasia who are age- and education-matched with the aphasic
patients. A final limitation with respect to the psychometric
properties of the ScreeLing is the rather low specificity of
semantics and syntax. In clinical practice, this would result
in a patient being incorrectly classified as having a semantic/
syntactic disorder. It is almost impossible for screening tests
to be both highly specific and highly sensitive. We preferred
optimal sensitivity in order to avoid misdiagnosing patients
with an actual semantic/syntactic disorder.
Our results demonstrate that differential assessment of
linguistic-level deficits is feasible at 2 weeks after stroke and
that the occurrence of selective linguistic disorders is not rare.
A selective semantic disorder was the least frequent and also
rarely occurred in combination with just one other linguistic-
Table V. Frequency of linguistic disorders (n = 141)
n (%)
Selective semantic deficit
Selective phonological deficit
Selective syntactic deficit
Semantic and phonological deficit
Semantic and syntactic deficit
Phonological and syntactic deficit
Semantic, phonological and syntactic deficit
4 (2.7)
24 (16.3)
5 (3.4)
2 (1.4)
4 (2.7)
20 (13.6)
57 (38.8)
J Rehabil Med 44

Page 6

434
H. El Hachioui et al.
level deficit. This means that if a patient has a deficit at the
semantic level, the phonological and/or syntactical levels will
also be affected. These findings support the notion that the se-
mantic level is the central level of language processing and is
involved in nearly all aspects of language (19). Damage to this
component is said to affect performance on any task requiring
comprehension or production of words (20).
Interestingly, we found that approximately 50% of the
variance in the Spontaneous Speech rating was explained by
the phonological level alone. This is not in line with previous
findings that semantic function contributes more to the vari-
ance of verbal communication than phonology (21, 22). These
results were obtained in more chronic stages, i.e. 3–5 months
(21) and 1–338 weeks after stroke (22). Verbal communica-
tion might be heavily influenced by phonological deficits in
the acute stage and more by semantic deficits in the chronic
stage. So far, data about the occurrence of linguistic deficits
are presently available only for the acute stage.
In a previous pilot study we found that phonology took sig-
nificantly longer to improve than semantics and syntax, i.e. up
to 4 months after stroke (23). The present study shows that in
acute patients the phonological level is affected most severely
and most frequently. Further insight into the recovery course
of the 3 linguistic levels is needed to evaluate the relevance of
our findings for treatment. In a current follow-up study we are
investigating the recovery of semantics, phonology and syntax
in the first year after stroke. We will address the occurrence of
the linguistic deficits at various time-points and their relation
to functional outcome. Furthermore, their additional prognostic
value will be investigated.
ACkNOWLEDgEMENTS
Financial support: This research was supported by the Netherlands Or-
ganization for Scientific Research (NWO) (grant number 017.002.083).
We thank the speech-language therapists who included and/or tested
the patients of the following centres:
Hospitals: Bronovo; Canisius-Wilhelmina; Deventer; gelre Zutphen;
hagaZiekenhuis; kennemer gasthuis; Clara; MCh; St Elisabeth; Ter-
gooiziekenhuizen; TweeSteden Tilburg; UMC St. Radboud; Bernhoven;
Rijnstate; Zevenaar; Diaconessenhuis Meppel.
Rehabilitation centres: Trappenberg; Waarden gorinchem; Leijpark;
Tolbrug; Rijndam; RMC groot klimmendaal; Sint Maartenskliniek;
Sophia Den haag; Sutfene Warnsveld.
Nursing homes: Afasie Trainingscentrum; Amaris Gooizicht; Regina
Pacis and Gelders Hof; Watersteeg Veghel; Zevenaar; Den Ooiman;
Hazelaar, Tilburg; Sint Jacob, Boerhaave; Cortenbergh; Zuiderhout;
Reggersoord; Sint Jozef, Deventer; Waelwick; Bieslandhof; Irene Dekker-
swald and Margriet.
Private practices and others: Bonnier-Baars; Mantelers-Nijssen;
Annika van hemert; Linda Thiadens.
The authors are also grateful to Siri Siepel for scoring the Spontaneous
Speech samples.
Conflicts of interest: E. G. Visch-Brink, W. M. E. van de Sandt-Koender man
and h. El hachioui receive royalties from the publication of the test
ScreeLing. The publisher has had no influence on the data collection,
methods, the interpretation of data, and the final conclusions.
REFERENCES
Lazar RM, Minzer B, Antoniello D, Festa JR, Krakauer JW,
Marshall RS. Improvement in aphasia scores after stroke is well
predicted by initial severity. Stroke 2010; 41: 1485–1488.
Pedersen PM, Vinter K, Olsen TS. Aphasia after stroke: type, se-
verity and prognosis. The Copenhagen aphasia study. Cerebrovasc
Dis 2004; 17: 35–43.
Laska AC, Hellblom A, Murray V, Kahan T, Von Arbin M. Aphasia
in acute stroke and relation to outcome. J Intern Med 2001; 249:
413–422.
Lazar RM, Speizer AE, Festa JR, Krakauer JW, Marshall RS.
Variability in language recovery after first-time stroke. J Neurol
Neurosurg psychiatry 2008; 79: 530–534.
pedersen pM, Jorgensen hS, Nakayama h, Raaschou hO, Olsen
TS. Aphasia in acute stroke: incidence, determinants, and recovery.
Ann Neurol 1995; 38: 659–666.
godefroy O, Dubois C, Debachy B, Leclerc M, kreisler A, Lille
Stroke P. Vascular aphasias: main characteristics of patients hos-
pitalized in acute stroke units. Stroke 2002; 33: 702–705.
Inatomi Y, Yonehara T, Omiya S, hashimoto Y, hirano T, Uchino
M. Aphasia during the acute phase in ischemic stroke. Cerebrovasc
Dis 2008; 25: 316–323.
Nys gM, van Zandvoort MJ, de kort pL, van der Worp hB, Jansen
BP, Algra A, et al. The prognostic value of domain-specific cognitive
abilities in acute first-ever stroke. Neurology 2005; 64: 821–827.
Nys gM, van Zandvoort MJ, de kort pL, Jansen Bp, de haan Eh,
kappelle LJ. Cognitive disorders in acute stroke: prevalence and
clinical determinants. Cerebrovasc Dis 2007; 23: 408–416.
Flamand-Roze C, Falissard B, Roze E, Maintigneux L, Beziz J,
Chacon A, et al. Validation of a new language screening tool for
patients with acute stroke: the Language Screening Test (LAST).
Stroke 2011; 42: 1224–1229.
Enderby PM, Wood VA, Wade DT, Hewer RL. The Frenchay Apha-
sia Screening Test: a short, simple test for aphasia appropriate for
non-specialists. Int Rehabil Med 1987; 8: 166–170.
Doesborgh SJ, van de Sandt-koenderman WM, Dippel DW, van
Harskamp F, Koudstaal PJ, Visch-Brink EG. Linguistic deficits in
the acute phase of stroke. J Neurol 2003; 250: 977–982.
Visch-Brink EG, van de Sandt-Koenderman WME, El Hachioui H.
ScreeLing. Houten: Bohn Stafleu van Loghum; 2010.
Graetz P, De Bleser R, Willmes K. [Dutch edition of the Aachener
Aphasie Test (AAT)]. Lisse: Swets & Zeitlinger; 1992 (in Dutch).
goodglass h, kaplan E. The assessment of aphasia and related
disorders. philadelphia: Lea and Febiger; 1972.
De Renzi E, Faglioni P. Normative data and screening power of a
shortened version of the Token Test. Cortex 1978; 14: 41–49.
Salter K, Jutai J, Foley N, Hellings C, Teasell R. Identification of
aphasia post stroke: a review of screening assessment tools. Brain
Inj 2006; 20: 559–568.
Nadeau SE. Phonology: a review and proposals from a connection-
ist perspective. Brain Lang 2001; 79: 511–579.
Patterson K, Shewell C. Speak and spell: dissociations and word-
class effects. In: Coltheart M, Sartori g, Job R, editors. The cog-
nitive neuropsychology of language. London: Lawrence Erlbaum
Associates publishers; 1987, p. 273–294.
Basso A. Aphasia and its therapy. New York: Oxford University
press Inc.; 2003.
Doesborgh SJC, van de Sandt-koenderman WME, Dippel DWJ,
van Harskamp F, Koudstaal PJ, Visch-Brink EG. The impact of
linguistic deficits on verbal communication. Aphasiology 2002;
16: 413–423.
Fucetola R, Connor LT, perry J, Leo p, Tucker FM, Corbetta
M. Aphasia severity, semantics and depression predict func-
tional communication in acquired aphasia. Aphasiology 2006;
20: 449–461.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
J Rehabil Med 44

Page 7

435
Occurrence of linguistic deficits post-stroke
El hachioui h, van de Sandt-koenderman MW, Dippel DW, koud-
staal PJ, Visch-Brink EG. A 3-year evolution of linguistic disorders
in aphasia after stroke. Int J Rehabil Res 2011; 34: 215–221.
Butterworth B, Howard D, McLoughlin P. The semantic deficit
in aphasia: the relationship between semantic errors in auditory
comprehension and picture naming. Neuropsychologia 1984; 22:
409–426.
Zhu Z, Zhang JX, Wang S, Xiao Z, huang J, Chen hC. Involvement
of left inferior frontal gyrus in sentence-level semantic integration.
Neuroimage 2009; 47: 756–763.
Howard D, Patterson KE. Pyramids and Palm Trees Test: A test
of semantic access from words and pictures. Bury St. Edmonds:
Thames Valley Co.; 1992.
Visch-Brink EG, Stronks DL, Denes G. [The Semantic Association
Test]. Lisse: Swets & Zeitlinger; 2005 (in Dutch).
Crutch SJ, Connell S, Warrington Ek. The different representa-
tional frameworks underpinning abstract and concrete knowledge:
23.
24.
25.
26.
27.
28.
evidence from odd-one-out judgements. Q J Exp psychol (Col-
chester) 2009; 62: 1377–1390.
pate DS, Saffran EM, Martin N. Specifying the nature of the pro-
duction impairment in a conduction aphasic: a case study. Lang
Cogn process 1987; 2: 43–84.
kay J, Lesser R, Coltheart M. psycholinguistic assessments of
language processing in aphasia. Hove: Lawrence Erlbaum As-
sociates Ltd; 1992.
Bastiaanse R, Maas E, Rispens J. [The Verbs and Sentences test:
WEZT]. Lisse: Swets & Zeitlinger; 2000 (in Dutch).
Thompson Ck, Shapiro Lp, Ballard kJ, Jacobs BJ, Schneider SS,
Tait ME. Training and generalized production of wh- and NP-
movement structures in agrammatic aphasia. J Speech Lang hear
Res 1997; 40: 228–244.
Chatterjee A, Maher L. grammar and agrammatism. In: Nadeau
SE, Gonzalez Rothi LJ, Crosson B, editors. Aphasia and language.
New York: The Guilford Press; 2000, p. 133–159.
29.
30.
31.
32.
33.
Semantics (24 items)
1. Word-picture matching (6 items); 6 photos of objects, 5
semantically related foils. A traditional task for semantic
processing (24). Example: gorilla, tiger, elephant, polar bear,
wolf, giraffe.
2.Identifying semantically anomalous sentences (6 items); choice
correct/incorrect. This requires recognizing the violation of
semantic selection restrictions (25). Example: “The ice chose the
wrong direction”.
3. Verbal semantic association (6 items); choice out of 4 words i.e.
1 correct, 2 distracters semantically related with the target word,
and 1 unrelated distracter. Differentiating between relevant and
irrelevant semantic features is required (26, 27). Example: letter:
chalk, paint, pen, grass.
4.Odd-word out (6 items); choice out of 4. The word that does
not fit into the same semantic category has to be selected (28).
Example: violin, siren, trumpet, piano.
phonology (24 items)
1. Repetition of words and phrases (6 items); to examine
phonological disorders in the output route. phonological
complexity is varied according to word length, consonant
clusters, identical vowels, phoneme-grapheme correspondence
(29). Example: “monopolie” (monopoly); “de excentrieke
antiekhandelaar” (the eccentric antique dealer).
2.Reading aloud words and phrases (6 items); level of complexity
matches the repetition task. phonological processing may
vary depending on the input route (30). Example: “macaroni”
(macaroni), and “de enthousiaste beroepsgoochelaar” (the
enthusiastic professional magician).
3. Equal/unequal judgment of spoken word pairs (6 items); choice
yes/no. A task to examine the phonological input route (30).
Example: “straat-staart” (street-tail).
4. Matching first phoneme of a spoken word with the grapheme
(6 items); choice out of 3. phoneme analysis and phoneme-
grapheme conversion is required (30). Example: “boek” (book):
g, k, b.
Syntax (24 items)
1. Sentence-picture matching (8 items); choice out of 3 or 4
photographs. The task requires syntactic comprehension,
including reversible sentences, subject-verb agreement, reflexive
verbs, passive sentences, prepositions, and verb tense (31).
Example: sentence “The man’s hair is being cut by the woman”;
3 pictures (i) “the man’s hair is being cut by the man”, (ii) “the
woman’s hair is being cut by the man”, (iii) “the man’s hair is
being cut by the woman”.
Wh-questions (4 items); photographed situation with a “Wh”-
question. “Wh”-questions require syntactic processing of the
non-canonical sentence construction (31–32). Example: “Wie
ziet dat hij een taartje pakt?” (Literally: “Who sees that he a cake
takes?”) The photograph depicts a man and woman talking, while
a boy takes a cake. The woman is looking at the boy.
Identifying syntactic incorrect sentences (6 items); choice correct/
incorrect. This requires processing of word order, subject-verb
agreement, auxiliaries, and conjunctions (31). Example: “Die
bloemen is veel te duur” (Those flowers is far too expensive).
Sentence completion with function words (6 items); choice out
of 4. Foils are well-known for addressing syntactic processing:
personal pronouns, arguments, prepositions, auxiliaries, and
different forms of verb tense or transitive/intransitive verbs (33).
Example: “De jongen geeft zijn vriendin...” (The boy gives his
girlfriend…) Naar de film (to the movie), parfum (perfume),
wandelen (hiking), van de chauffeur (of the driver).
For semantics and syntax all items are presented aurally as well as
visually in order to gain insight into the underlying linguistic disorder
independently of the input route.
This is a description of the final version of the ScreeLing, referred to
in this paper. In an earlier research version, all subtasks were validated
against the judgment of a linguist (12).
In this final version, the following phonological and syntactic
subtasks have been adapted. phonology 3 and 4 consisted of
respectively “reverse the word” (“pan” → “nap”) and lexical decision
(“cimputer”). Both were replaced after item analyses. Phonology 3
appeared to relate to a general capacity to perform the required action
rather than to intactness of phonological processing. The lexical
decision task was too easy; 80% of the patients performed perfectly.
To reduce chance level in Syntax 1, more foils were added to the
original choice of 2. In Syntax 3, selecting a grammatically correct
sentence out of 3 possibilities appeared to be too time-consuming.
Syntax 4, repetition of sentences with mainly function words, was
replaced by an easier-to-score task; this ensures that the test can also be
used by professionals from other disciplines, such as neurologists and
neuropsychologists.
2.
3.
4.
AppENDIX I. ScreeLing. The following requirements are met: short (30 min); suitable for bedside administration (1 booklet and a score sheet) by
various disciplines; vivid material (colour photographs, varying tasks); simple scoring system (right/wrong).
J Rehabil Med 44