Brain Injury, June 2009; 23(6): 585–594
Rehabilitation of the central executive of working memory after
severe traumatic brain injury: Two single-case studies
CLAIRE VALLAT-AZOUVI1,2,3, PASCALE PRADAT-DIEHL2,3,4,
& PHILIPPE AZOUVI2,3,5,6
1UGECAM-Antenne UEROS, Ho ˆpital Raymond Poincare ´, Garches, France,2INSERM U731, Paris, France,
3Faculte ´ de me ´decine, laboratoire de Physiologie et physiopathologie de la motricite ´, UPMC, Paris, France,
4AP-HP, Ho ˆpital de la Salpetrie `re, Service de Me ´decine Physique et de Re ´adaptation, Paris, France,5AP-HP, Ho ˆpital
Raymond Poincare ´, Service de Me ´decine Physique et de Re ´adaptation, Garches, France, and6Universite ´ de Versailles
Saint Quentin, France
(Received 19 September 2008; revised 17 March 2009; accepted 31 March 2009)
Background/objective: A deficit of the central executive of working memory is a frequent finding in patients with severe
traumatic brain injury (TBI). The objective of the present study was to assess the efficacy of a rehabilitation programme of
the central executive after severe TBI.
Method: An experimental single-case design was used in two patients with remote severe TBI suffering from an isolated
central executive deficit. Outcome was assessed with specific working memory tests (spans, Brown Peterson, n-back), non-
specific cognitive tasks requiring working memory (dual-task, arithmetic solving problem), an ecological questionnaire to
assess generalization to everyday life and non-target tasks not requiring working memory, to assess the specificity of the
Results: Performance was stable on two baseline sessions before therapy. For both patients, an improvement was found for
target measures, mainly for central executive tasks, and for the questionnaire on attention failures in everyday life.
In opposition, no change was found for non-target measures.
Discussion: Improvement was not seemingly related to spontaneous recovery, nor to re-test effects. This study suggests that
specific cognitive training may improve the central executive of working memory in patients with remote severe TBI.
Keywords: Cognitive rehabilitation, executive functioning, attention, traumatic brain injury
Patients with severe traumatic brain injury (TBI)
frequently suffer from a deficit in central executive
functions of working memory [1–3]. The most widely
used model [4, 5] assumes that working memory can
be divided into at least three components involved in
temporary maintenance and manipulation of infor-
mation. The central executive is an attentional
control system, while the phonological loop and the
visuo-spatial sketchpad are two modality-specific
slave systems. The central executive is involved in
dual-task processing and in control-type functions
under non-routine conditions. It is related to
dorsolateral prefrontal activation .
There have been relatively few studies addressing
in a systematic way the different sub-systems of
working memory after severe TBI. Early studies
reported poor performance after severe TBI on
backward digit span  or on the Brown-Peterson
paradigm . A few individual cases of selective
impairment of the central executive after TBI have
been reported [9, 10]. Several studies addressed
Correspondence: Philippe Azouvi, Service de Me ´decine Physique et de Re ´adaptation, Ho ˆpital Raymond Poincare ´, Garches, 92380 France.
Tel: þ33147107074. Fax: þ33147107073. E-mail: email@example.com
ISSN 0269–9052 print/ISSN 1362–301X online ? 2009 Informa Healthcare Ltd.
dual-task processing, which is one of the key
functions of the central executive [11, 12]. The
main results were that TBI patients perform nor-
mally on divided attention tasks that can be carried
out relatively automatically, while they are impaired
performed under high time-pressure, including sub-
stantial working memory load or requiring executive
control [11, 13–23]. Other studies, using different
experimental tasks, such as the n-back task or the
paced auditory serial addition test, also found
evidence for an impairment of working memory
after TBI [1, 24–26]. Working memory deficits have
also been found in children with severe TBI [27–29].
Recently, the different sub-systems of working
memory were systematically assessed in a group of
30 patients with subacute-remote severe TBI .
Only marginal group differences were found regard-
ing the two slave systems, while patients performed
poorly on most central executive tasks, particularly
those requiring a high level of controlled processing.
Taken together, the different studies summarized
above all suggest that severe TBI is associated with
an impairment of executive aspects of working
However, there has been little research on reha-
bilitation of working memory after TBI. Training of
working memory has been found useful in a few
other conditions, such as children with ADHD
[30, 31], stroke [32, 33] or in a patient with a
brain tumour in the left temporal lobe .
Significant training-induced changes in brain activity
have also been found in a functional imaging study in
healthy individuals . To the authors’ knowledge,
only two studies on rehabilitation of working
memory after TBI have been reported [36, 37].
The former study reported successful working
memory rehabilitation in five patients with mild
TBI. The latter study  was a group study of
rehabilitation of central executive deficits in nine
patients with TBI of various degree of severity.
Rehabilitation lasted ?4 weeks and relied on the
Paced Auditory Serial Addition Test (PASAT) or on
two related procedures. Patients, as a group,
improved significantly in all cognitive functions
dependent on the central executive, but not in
those functions not thought to tap this system. A
significant improvement was also found on psycho-
social outcome measures. The aim of the present
study was to assess the effectiveness of a compre-
hensive programme of rehabilitation addressing the
different components of working memory, according
to Baddeley’s  model, in patients with remote
severe TBI suffering from central executive deficits.
A part of this programme (that regarding verbal
working memory) was recently found effective in
a case study of a left hemisphere stroke patient with
more complex tasks,
a selective deficit of verbal working memory . A
single-case design was used, including an assessment
of the generalization to daily life activities.
A 28-year old right-handed man sustained a TBI in
May 1997. He previously worked as a financial
manager (education duration: 15 years). Initial
Glasgow Coma Scale (GCS) score was 12, but his
condition worsened in the following hours. CT scan
showed small haemorrhagic lesions in the left poste-
rior ventricular horn, the left parietal lobe and in the
upper brainstem. Coma lasted a few days and post-
traumatic amnesia (PTA) 1 week (coma duration was
not explicitly recorded in the acute medical charts).
A first neuropsychological assessment was carried out
2 months after the injury, which revealed marked
attention and short-term memory deficits, while
long-term memory and executive functions appeared
to be relatively well preserved. He was discharged
form the hospital and received a traditional neurop-
sychological rehabilitation in the community. He
went back to work as a financial manager 6 months
post-injury. However, he experienced difficulties that
did not tend to decrease with time and he lost his job.
He was then referred to the community re-entry
facility (French UEROS programme ), 3 years
after the injury. At that time, his spontaneous
complaints were as follows: difficulty taking notes,
concentrating, memorizing information in the short-
term, holding a conversation and mental fatigue. He
felt that he needed major mental effort to maintain an
efficient level of attention to his everyday tasks. A first
baseline neuropsychological assessment was carried
out that showed a dramatic impairment of working
memory, associated with a normal performance on
tasks not requiring working memory (that will be
referred to as non-target measures, including mea-
sures of speed of processing, selective attention,
memory). Detailed performance will be presented
thereafter, see Results section.
and visual long-term
A 30-year old right-handed man sustained a severe
TBI in February 2001. He previously worked as a
cook (education duration: 9 years). Initial GCS
score was 8. CT scan showed right prefrontal and
temporal contusions. Coma duration was 10 days
and PTA duration 1 month. He was referred to the
facility 30 months post-injury, to assess his ability to
return to work. He had few spontaneous complaints
and seemed to be partly unaware of his difficulties.
C. Vallat-Azouvi et al.
He had a mild logorrhoea and inappropriate
euphoria. When questioned about possible beha-
viour/personality changes, he nevertheless admitted
that he was more irritable. The first baseline
attention and working memory deficits, while other
cognitive functions were within the normal range
(see Results section and Tables for a detailed
presentation of patients’ performance).
A specific cognitive training of the central executive,
the phonological loop and the visuo-spatial sketch-
pad was used. Training addressed both storage and
processing of information in working memory.
Twelve different tasks were used, each one arranged
in a difficulty hierarchy order organized to follow two
criteria, capacity and level of processing. Capacity
was related to the amount of information (e.g.
number of words, syllables or phonemes). Level of
cognitive processing was related to stimulus com-
plexity (e.g. frequency of usage, concreteness).
Stimuli were presented, one at a time, at a regular
speed (1 per second) and responses were given
Training was given twice a week, during 8
(Patient 1) or 6 (Patient 2) months. Each session
lasted 1 hour. The difficulty level was progressively
adapted to the patients’ performance, starting at n-1
backward digit span level. Each level was trained
until the patient succeeded in 90% of the trials.
Before starting the therapy, task feasibility was
assessed in 10 control subjects, matched to the
patients for age, sex and education. All tasks were
completed at ceiling by control subjects. Eight tasks
addressed verbal aspects of working memory, the
four others the visuo-spatial sketchpad. A descrip-
tion of the 12 tasks can be found in the Appendix.
A single-case methodology was used . Although
patients were respectively 30 and 39 months post-
injury, to control for spontaneous recovery, two
baseline measures were taken, with a 3 months inter-
test interval. The outcome measures were obtained
just after the end of the therapy. A 3-month follow-
up was planned, but could be completed only for
patient 1, because patient 2 had returned to work at
that time and did not wish to come back for testing
again. To control for re-test effects, for all cognitive
tasks, three parallel versions were designed and
outcome measures were always different from the
tasks used for the therapy.
To reduce as far as possible an examiner’s bias,
therapy and assessment were performed by different
Both patients were informed of the aim of this
study. Data were obtained in compliance with the
Considering the fact that formal statistics were
difficult to conduct on such a small sample, data
have been analysed by comparison of each individ-
ual patient to the performance of healthy subjects.
Some of the outcome measures (that will be
presented in the next session) were selected from
the computerized Test for Attention Performance
[40, 41]. This allowed a direct comparison of each
patient’s performance to the published norms of this
battery. However, this study also used as outcome
measures experimental working memory tasks, spe-
cially designed for the purpose of this study and for
other related studies in the department [3, 33], for
which there were no published norms available.
Consequently, for these experimental tasks, and for
the attention questionnaire, patients’ performance
was compared to a group of 10 carefully matched
controls. To take into account the fact that the two
patients had different education levels, two different
control groups were used. These controls were
selected among a larger database from the depart-
ment (unpublished data). Each group (n¼10) was
carefully matched, in terms of age, education
duration and handedness (right-handed) to each
age¼25.8 years, SD¼3.3, range: 23–32, education
duration: mean¼13.9 years, SD¼1.5, range¼10–
15; patient 2: mean age¼33.3 years, SD¼5.0,
range: 22–40, education duration: mean¼9.5 years,
SD¼1.1, range¼8–11). Controls could not be
matched for gender (each group included five
men and five women), but gender was not assumed
Four sets of outcome measures were used, including
non-target measures to assess the specificity of
Cognitive tasks specifically designed to assess the different
components of working memory
Digit and visuo-spatial spans.
ward digit and visuo-spatial spans were assessed,
with five trials for each span length. The test
ended when a patient failed five consecutive trials
percentage of correct responses for each span
Forward and back-
Rehabilitation of working memory
Central executive tasks.
. Brown-Peterson paradigm. The Brown Peterson
processing of information. This task was used
both in the verbal and visuo-spatial modalities
[9, 42, 43]. In the verbal modality, patients were
asked to recall consonant trigrams after three
delays (5, 10 or 20 seconds) with or without an
interfering task (motor task, articulatory suppres-
sion and digit addition). The visuo-spatial Brown-
Peterson task used a procedure similar to the
Corsi Block-tapping test . It consisted of nine
white 2-cm cubes fastened in a random order to a
blackboard. At each trial, the examiner tapped
two blocks in a pre-determined order. Patients
were required to repeat the same sequence after
three delays (5, 10 or 20 seconds), first without
distractor, then with an interfering task. The
interfering task was designed in order to tap
visuo-spatial functions: subjects were asked to
reproduce manual postures made by the exam-
iner. The measures were the percentage of correct
responses for each recall delay (out of five trials)
and under each experimental condition.
. N-back. The 2-back ‘working memory’ sub-test of
Performance [40, 41, 45] was used. Numbers
were presented on the computer screen at a
1-second rate. Patients were required to press a
key when a number matched a number presented
2-back in the sequence. Measures were median
Reaction Time and number of errors. For this
task, that was not given to the control group,
available standardized norms were used [40, 41].
Patients’ scores were compared to the percentile
of the standardization sample (i.e. the percentage
of healthy subjects obtaining a performance sim-
ilar or below the patient).
Test for Attentional
Non-specific tasks requiring working memory.
tasks that were not specifically designed to assess
working memory, that were not trained, but that
nevertheless include a high working memory load.
. Dual-task. Although dual-task processing is usu-
ally considered as one of the key functions of the
central executive , recent studies suggested that
it might be dissociable from other functions of the
central executive [46, 47]. Moreover, it was not
specifically addressed by the therapy. This is the
reason why it was considered here as a ‘non-
specific’ task. The divided attention sub-test of
the Test for Attentional Performance [40, 41, 45]
again compared patient’s performance to standar-
dized norms. Two simultaneous choice reaction
times tasks (one visual and one auditory) were
given. The visual task consists of crosses that
appear in a random configuration in a 4?4
matrix. Patients had to detect whether the crosses
form the corner of a square. The auditory task
includes a regular sequence of high and low beeps.
Patients had to detect an irregularity in the
sequence. The dual task was preceded by a
practice session, in order to control that patients
were able to perform single tasks without diffi-
culty. Measures were number of omissions.
m-solving was assessed with 15 problems of
increasing complexity, including items from the
The Rating Scale of Attentional Behaviour (RSAB)
 was used to assess the consequences of cogni-
tive impairments on daily-life activities. This ques-
attentional failures in everyday life and has been
found valid and sensitive to TBI. It was used here
because there is no published questionnaire specif-
ically designed to assess working memory-related
difficulties in daily life and because attention failures
are closely related to the kind of problems due to
deficits of the central executive of working memory
(e.g. difficulty in dual-task situations). The RSAB
includes 14 questions, rated on a 5-point Likert scale
(range: 0–4). Higher scores indicate more difficulties
in everyday life and the maximal score is 56. The
RSAB was used here as a self-questionnaire by the
These measures were assumed not to require work-
ing memory and were not expected to improve after
therapy. Non-target measures included attentional
and long-term memory tests. Attentional tests not
including working memory load were a simple visual
reaction time, a go–no go selective attention task,
both from the Test for Attentional Performance [40,
41, 45], and the Trail Making Test (TMT) .
Long-term verbal memory was assessed with a word
recognition task, specifically designed for this trial.
Thirty words were presented orally to the patients
and had to be recognized 20 minutes later, mixed
with 60 distractors. The score was the number of
words correctly recognized. A high number of words
was used in this task in order to minimize the risk of
a ceiling effect. Visual long-term memory was
assessed with Complex Figure recall .
C. Vallat-Azouvi et al.
Cognitive tasks specifically designed to assess the different
components of working memory
Digit and visuo-spatial spans.
was slightly impaired and improved after therapy
(Table I). Patients performed below the range of
matched controls at span length 6 (Patient 1) or 5
(Patient 2). After training, a dramatic improvement
was found for patient 1, who reached the upper
range of controls’ performance at levels 6 and 7.
Improvement was maintained at follow-up. For
patient 2, improvement was less dramatic. His
baseline performance was below or at the lower
range of matched controls for span lengths 5 and 6.
After therapy, he reached the upper range of
controls’ performance at both these levels.
Backward digit spans were severely impaired in
patient 1, who could not reach level 4 on any
baseline session. He dramatically improved after
Forward digit span
therapy, reaching the range of controls’ performance
at levels 4–6. Improvement was maintained at
follow-up. Patient 2 performed slightly better on
baseline, where he obtained a performance within
the lower range of matched controls. He also
improved after therapy (Table II).
Baseline forward and backward visuo-spatial spans
were withinnormal limits
Consequently, results will not be presented in
Central executive tasks.
. Verbal Brown-Peterson task. Performance in the two
easiest conditions (without interference and with
motor task) was nearly perfect for both patients in
the two baseline sessions. However, performance
was markedly impaired in the two more demand-
ing conditions. With articulatory suppression,
Patient 1 performed far below the normal range
Table I. Forward digit spans. Correct responses (%) for patients and 10 matched controls (for control subjects results are mean percentage
and SD in brackets).
Table II. Backward digit spans. Correct responses (%) for patients and 10 matched controls (for control subjects results are mean
percentages and SD in brackets).
Rehabilitation of working memory
on baselines at the 10- and 20-second delays.
After therapy, he dramatically improved, reaching
a nearly perfect performance, that was maintained
at follow-up. Patient 2 performed just below the
normal range and his performance improved after
therapy, reaching the range of matched controls
(Table III). In the most difficult condition, with
digit addition as interfering task, Patient 1 expe-
rienced major difficulties on baselines (nearly 0%
improved after therapy and on follow-up, reach-
ing the upper range of matched controls. Patient 2
performed just below the controls’ range at
baseline and reached the upper range of controls’
performance after therapy (Table IV).
. Visuo-spatial Brown-Peterson task. Both patients
(100% hits) without
interference, but fell clearly below the controls’
range under the interference condition (Table V).
Performance was stable before treatment and
improved after training, reaching the upper
range of controls’ performance.
. N-back task. Both patients performed poorly on
the 2-back task at both baseline sessions, with a
high number of errors (6–11 omissions out of 15
targets, corresponding to a performance at or
below the 10th percentile of the standardization
sample). The number of errors decreased and
reached a normal range (<3 misses, >Percentile
50) after treatment and at follow-up (Patient 1).
Non-specific tasks requiring working memory
poor and stable (4–10 errors out of 30 targets,
<percentile 10). After treatment, both patients
improved, reaching the normal range (<2 errors).
Improvement was maintained at follow-up for
Both patient’s baseline performance was
impaired in both cases (correct responses on the
two successive baselines: Patient 1: 50% on the
sessions, patient 2¼40% and 50%). Post-treatment,
both patients reached a perfect performance, similar
to that of controls (100% correct responses).
Ecological questionnaire (RSAB)
Results, displayed in Table VI, were different in the
two patients. Patient 1 was highly aware of his
difficulties and his baseline complaint score on the
RSAB was high, largely above the controls’ range.
Complaints dramatically decreased after treatment
and on follow-up, reached the normal range.
Table III. Verbal Brown-Peterson task, with articulatory suppres-
sion as interfering task. Correct responses (%) for patients and 10
matched controls (for control subjects results are mean percen-
tages and SD in brackets).
5 seconds10 seconds 20 seconds
Table IV. Verbal Brown-Peterson task, with digit addition
as interfering task. Correct responses (%) for patients and
10 matched controls (for control subjects results are mean
percentages and SD in brackets).
5 seconds10 seconds20 seconds
Table V. Visual Brown-Peterson task, interfering condition.
Correct responses (%) for patients and 10 matched controls (for
control subjects results are mean percentages and SD in brackets).
5 seconds10 seconds20 seconds
C. Vallat-Azouvi et al.
The case was different for the second patient. His
complaint score was just above the controls’ range
with the RSAB, suggesting that he was not fully
problems were actually reported by his close
relationships). Complaint score, that was stable on
the two baseline sessions, nevertheless dropped by
?50% after training.
Baseline performance on most non-target measures
(simple reaction times, trail making test A and B and
long-term memory) was within normal limits in both
cases (percentiles 31–90 of standardization samples
depending on the tests). The only test that showed a
performance at the lower limit of the normal range
was the selective attention task (go–no go) for
Patient 1 and on the first baseline session only
(percentile 14). Performance remained perfectly
Particularly, it should be noted that verbal recogni-
tion, that was designed to minimize ceiling effect,
remained stable across successive testing sessions in
both patients (?50% hits, at the level of control
The two patients included in the present study both
suffered from a relatively isolated impairment of the
central executive of working memory. As compared
to controls carefully matched for age and education
duration, they performed poorly on working memory
tasks, mainly those requiring executive control.
Indeed, digit spans were only slightly impaired and
visual spans were normal, while central executive
tasks wereall severelyimpaired.This deficit
appeared to be isolated, as indicated by a normal
performance on tests of verbal and visual long-term
memory, speed of processing and selective attention.
The deficit was chronic and stable. Indeed, patients
were, respectively, 30 and 39 months post-trauma
and stability of performance was controlled on two
successive baseline assessments, with a 3-month
interval. Finally, the deficit had clinically significant
consequences upon daily-life functioning, particu-
larly on vocational reintegration. These difficulties
were clearly documented in the first patient with the
RSAB. In the second patient, the questionnaire was
less conclusive, probably due to some degree of
Rehabilitation, focusing on storage and processing
aspects of working memory, lasted several months.
Training tasks were arranged in a progressive hier-
archical order. Post-therapy assessment showed an
Improvement was particularly marked for central
executive tasks, for which both patients reached the
medium or upper range of controls’ performance
after training. There was also an improvement of
related but untrained tasks (dual-task and arithmetic
problem-solving). More importantly, there was also
a decrease of attention failures in everyday life, as
assessed with the RSAB. In addition, after this
experimental rehabilitation, patients returned to full-
time work at their previous level. By contrast, there
was no improvement of non-target measures (of
attention and long-term memory) that were not
trained and were not expected to improve. As
previously mentioned, a problem with non-target
tasks was that some of them were performed at
ceiling, leaving little range for improvement (indeed,
these patients were selected precisely because they
suffered from a relatively isolated deficit of working
memory). However, performance was not at ceiling
on two tasks (go–no go and verbal recognition
memory) that were not modified after rehabilitation.
Hence, improvement in central executive functions
could not readily be attributed to spontaneous
recovery, non-specific cognitive stimulation or re-
test effects. Indeed, the deficits were stable before
the experimental rehabilitation started and improve-
ment was limited to the domain that was specifically
addressed by the treatment. In addition, to control
for any re-test effect, parallel forms of tests were used
in the different testing sessions and outcome mea-
sures were always different from the tasks used for
the therapy. These results suggest that the training
resulted in a specific improvement of the central
executive of working memory, with a generalization
to untrained tasks and a transfer to everyday life
The present results confirm and extend findings
from a previous study using a part (verbal training
Table VI. Ecological questionnaire. RSAB: Rating Scale of
Attentional Behaviour . Higher scores indicate more difficul-
ties in everyday life. For control subjects data are mean and range
RSAB, total (/56)
Rehabilitation of working memory
tasks) of the present rehabilitation programme in a
patient with a left parietal stroke . However, this
latter patient mainly suffered from a deficit of verbal
aspects of working memory, while in the present
study the main core of deficit was the central
These results are also in accordance with two
previous studies on rehabilitation of the central
executive after TBI [36, 37]. However, these two
latter studies differed from the present study in
several aspects. The Cicerone  study only
concerned patients with mild TBI, who represent a
quite different issue. In the Italian study ,
severity of injury was more heterogeneous (five
severe). Also, training mainly relied on the PASAT
or related procedures that only tap one aspect of the
central executive. In contrast, this training pro-
gramme relied on a large number of tasks tapping
different aspects of central executive functions.
There are nevertheless similarities between these
two earlier studies and the present one: they all rely
on the same theoretical model and the training
arranged in a hierarchic order, by increasing working
memory load (or time pressure).
There are a few limitations to the present study.
From a methodological point of view, the main
problem was that, for practical reasons, assessment
could not be blinded. As a consequence, a bias could
not be completely excluded. However, as previously
stated, to reduce such risk as far as possible,
assessment and therapy were performed by different
investigators. Moreover, the dramatic improvement
observed on some central executive tasks (such as
the Brown-Peterson or the dual-task) was not likely
to be due to a placebo effect or to an examiner’s bias.
The second limitation was that only two baseline
measures were given before starting the therapy.
Stability of performance on three or more baseline
However, these patients were more than 2 years
post-injury and their difficulties in everyday life did
not tend to improve with time, making spontaneous
recovery very unlikely. The third limitation is the
small number of patients. Whether these results can
be generalized to other patients remains to be
demonstrated. This is a problem common to all
single-case studies. However, it is generally agreed
that single-case designs are appropriate to test
precisely an individual patient’s response to treat-
ment and to tease out effects of other recovery
factors from effects of intervention strategy .
Moreover, single-case designs also offer the oppor-
tunity to assess clinical as well as statistical sig-
nificance, while in group studies clinical significance
is rarely considered. However, it is acknowledged
that the small sample size renders the present results
preliminary and further studies in a larger sample of
subjects are needed to assess to what extent these
results generalize to other patients with remote
severe TBI. The fourth limitation was related to
the nature of the training tasks. Training was based
on an intensive and long-duration programme,
requiring a high motivation from the patient.
Patient 2 appeared to have only partial awareness
of his difficulties, but nevertheless agreed to partici-
pate. However, it must be acknowledged that this
programme may be difficult to conduct in patients
with major anosognosia
In conclusion, the present study suggested an
effectiveness of a specific therapy focusing on central
executive aspects of working memory and its gener-
alization to daily-life activities in patients with severe
TBI. Such therapy can easily be used in a conven-
tional rehabilitation setting. These encouraging
results, obtained in single-case studies, support the
need for a study in a larger sample of patients.
or severe behavioural
We would like to thank the two patients for their
participation to the study and the neuropsychology
students who contributed to rehabilitation and/or
Declaration of interest: The authors report no
conflicts of interest. The authors alone are respon-
sible for the content and writing of the paper.
1. McAllister TW, Flashman LA, Sparling MB, Saykin AJ.
Working memory deficits after traumatic brain injury: cate-
cholaminergic mechanisms and prospects for treatment—a
review. Brain Injury 2004;18:331–350.
2. Serino A, Ciaramelli E, Di Santantonio A, Malagu S,
Servadei F, Ladavas E. Central executive system impairment
in traumatic brain injury. Brain Injury 2006;20:23–32.
3. Vallat-Azouvi C, Weber T, Legrand L, Azouvi P. Working
memory after severe traumatic brain injury. Journal of the
International Neuropsychological Society 2007;13:770–780.
University Press; 1986.
5. Baddeley A. Fractionating the central executive. In: Stuss DT,
Knight RT, editors. Principles of frontal lobe function.
New York: Oxford University Press; 2002. pp 246–260.
6. Collette F, van der Linden M. Brain imaging of the central
executive component of working memory. Neuroscience and
Biobehavioral Reviews 2002;26:105–125.
7. Brooks DN. Long and short term memory in head injured
patients. Cortex 1975;11:329–340.
8. Stuss DT, Ely P, Hugenholtz H, Richard MT, Larochelle S,
Poirier CA, Bell I. Subtle neuropsychological deficits in
C. Vallat-Azouvi et al.
9. Van der Linden M, Coyette F, Seron X. Selective impair-
ment of the ‘central executive’ component of working
memory: a single case study. Cognitive Neuropsychology
10. Allain P, Etcharry-Bouyx F, Le Gall D. A case study of
selective impairment of the central executive component of
working memory after a focal frontal lobe damage. Brain and
11. Park NW, Moscovitch M, Robertson IH. Divided attention
impairments after traumatic brain injury. Neuropsychologia
12. Leclercq M, Azouvi P. Attention after traumatic brain injury.
In: Leclercq M, Zimmermann P, editors. Applied neuropsy-
chology of attention. Hove, UK: Psychology Press; 2002.
13. Brouwer WH, Ponds RWHM, Van Wolffelaar PC, Van
Zomeren AH. Divided attention 5 to 10 years after severe
closed head injury. Cortex 1989;25:219–230.
14. Spikman J, van Zomeren AH, Deelman BG. Deficits of
attention after closed-head injury: slowness only? Journal of
Wolffelaar PC. Central executive aspects of attention in
subacute severe and very severe closed head injury patients:
16. McDowell S, Whyte J, D’Esposito M. Working memory
impairments in traumatic brain injury: evidence from a dual-
task paradigm. Neuropsychologia 1997;35:1341–1353.
17. Stablum F, Leonardi G, Mazzoldi M, Ulmita C, Morra S.
Attention and control deficits following closed head injury.
18. Vilkki J, Virtanen S, Surma-Aho O, Servo A. Dual task
performance after focal cerebral lesions and closed head
injuries. Neuropsychologia 1996;34:1051–1056.
19. Brouwer WH, Withaar FK, Tant MLM, Van Zomeren AH.
Attention and driving in traumatic brain injury: a question of
20. Mangels JA, Craik FI, Levine B, Schwartz ML, Stuss DT.
Effects of divided attention on episodic memory in chronic
traumatic brain injury: a function of severity and strategy.
21. Azouvi P, Jokic C, Van der Linden M, Marlier N, Bussel B.
Working memory and supervisory control after severe closed
head injury. A study of dual task performance and random
22. Leclercq M, Couillet J, Azouvi P, Marlier N, Martin Y,
Strypstein E, Rousseaux M. Dual task performance after
severe diffuse traumatic brain injury or vascular prefrontal
23. Azouvi P, Couillet J, Leclercq M, Martin M, Asloun S,
Rousseaux M. Divided attention and mental effort after
severe traumatic brain injury. Neuropsychologia 2004;42:
24. Bublak P, Schubert T, Matthes-von Cramon G, von
Cramon Y. Differential demands on working memory for
guiding a simple action sequence: evidence from closed-head-
injured subjects. Journal of Clinical and Experimental
25. Perlstein WM, Cole MA, Demery JA, Seignourel PJ,
Dixit NK, Larson MJ, Briggs RW. Parametric manipulation
of working memory load in traumatic brain injury: behavioral
Neuropsychological Society 2004;10:724–741.
26. Christodoulou C, DeLuca J, Ricker JH, Madigan NK,
Bly BM, Lange G, Kalnin AJ, Liu WC, Steffener J,
Diamond BJ, Ni AC. Functional magnetic resonance imaging
of working memory impairment after traumatic brain injury.
Journal of Neurology, Neurosurgery and Psychiatry 2001;71:
27. Levin HS, Hanten G, Zhang L, Swank PR, Ewing-Cobbs L,
Dennis M, Barnes MA, Max J, Schachar R, Chapman SB,
Hunter JV. Changes in working memory after traumatic brain
injury in children. Neuropsychology 2004;18:240–247.
28. Levin HS, Hanten G, Chang CC, Zhang L, Schachar R,
Ewing-Cobbs L, Max JE. Working memory after traumatic
brain injury in children. Annals of Neurology 2002;52:82–88.
29. Chapman SB, Gamino JF, Cook LG, Hanten G, Li X,
Levin HS. Impaired discourse gist and working memory in
children after brain injury. Brain and Language 2005;97:
30. Klingberg T, Forssberg H, Westerberg H. Training of
working memory in children with ADHD. Journal of
31. Klingberg T, Fernell E, Olesen PJ, Johnson M, Gustafsson P,
Dahlstrom K, Gillberg CG, Forssberg H, Westerberg H.
Computerized training of working memory in children with
ADHD—a randomized, controlled trial. Journal of the
American Academy of Child and Adolescent Psychiatry
32. Westerberg H, Jacobaeus H, Hirvikoski T, Clevberger P,
Ostensson ML, Bartfai A, Klingberg T. Computerized
working memory training after stroke—a pilot study. Brain
33. Vallat C, Azouvi P, Hardisson H, Meffert R, Tessier C,
Pradat-Diehl P. Rehabilitation of verbal working memory
34. Duval J, Coyette F, Seron X. Rehabilitation of the central
executive component of working memory: a re-organisation
approach applied to a single case. Neuropsychological
35. Olesen PJ, Westerberg H, Klingberg T. Increased prefrontal
and parietal activity after training of working memory. Nature
36. Cicerone K. Remediation of ‘working attention’ in mild
traumatic brain injury. Brain Injury 2002;16:185–195.
37. Serino A, Ciaramelli E, Santantonio AD, Malagu S,
Servadei F, Ladavas E. A pilot study for rehabilitation of
central executive deficits after traumatic brain injury. Brain
38. Le Gall C, Lamothe G, Mazaux JM, Muller F, Debelleix X,
Richer E, Joseph PA, Barat M. [Outcome of the Aquitaine
Unit for Evaluation, Training and Social and Vocational
Counselling (UEROS) at 5-year follow-up in young adults
with brain damage]. Annales de Readaptation et de Medecine
39. Wilson BA. Rehabilitation of memory. New York: The
Guilford Press; 1987.
40. Zimmermann P, Fimm B. Batterie de tests neuropsycholo-
giques pour l’e ´valuation des troubles de l’attention. 1.02c,
Traduction franc ¸aise P North, M Leclercq. Freiburg: Vera
Fimm, Psychologische Testsysteme; 1992.
41. Zimmermann P, Fimm B. Test for attentional performance
(TAP). English version. Herzogenrath: Psytest; 1995.
42. Brown J. Some tests of the decay theory of immediate
memory. Quarterly Journal of Experimental Psychology
43. Peterson LR, Peterson MJ. Short term retention of individual
verbal items. Journal of Experimental Psychology 1959;58:
Rehabilitation of working memory
44. Lezak MD. Neuropsychological assessment. New York:
Oxford University Press; 1995.
45. Zimmermann P, Fimm B. A test battery for attentional
performance. In: Leclercq M, Zimmermann P, editors.
Applied neuropsychology of attention. London: Psychology
Press; 2002. pp 110–151.
46. Miyake A, Friedman NP, Emerson MJ, Witzki AH,
Howerter A, Wager TD. The unity and diversity of executive
functions and their contributions to complex ‘frontal lobe’
tasks: a latent variable analysis. Cognitive Psychology 2000;
47. Asloun S, Soury S, Couillet J, Giroire JM, Joseph PA,
Mazaux JM, Azouvi P. Interactions between divided atten-
tion and working-memory load in patients with severe
traumatic brain injury. Journal of Clinical and Experimental
48. Wechsler D. WAIS-III: Manuel de l’Echelle d’Intelligence de
Wechsler pour Adultes, 3e `me e ´dition. Paris: Les Editions du
Centre de Psychologie Applique ´e; 1997.
49. Ponsford J, Kinsella G. The use of a rating scale of attentional
50. Reitan RM. Validity of the Trail Making Test as an indicator
of organic brain damage. Perceptual and Motor Skills 1958;8:
51. Taylor E. Psychological appraisal of children with cerebral
deficits. Cambridge, MA: Harvard University Press; 1959.
52. Wilson BA. Single-case experimental designs in neuropsy-
Experimental Neuropsychology 1987;9:527–544.
Appendix: Training tasks used for
. Reconstitution of words from oral spelling: the patient
had to reconstruct and pronounce a word that was
spelled out by the therapist.
. Reconstitution of words from oral spelling with a letter
omitted: this is the same task as the previous one,
but with a letter omitted, replaced by a ‘bip’.
. Oral spelling.
. Odd or even number of letters in a word: the task was
to find whether the number of letters in a word
heard was odd or even.
. Reconstitution of words from syllables: the therapist
presented cluttered syllable words that the subject
had to reorder and pronounce.
. Alphabetic way: the therapist presented a letter and
a calculation (addition or subtraction) and the
patient had to find the target letter following the
increasing or decreasing alphabetical order (for
example, Bþ6¼H; E?2¼C).
. Word sorting in alphabetic order: the patient was
required to sort a series of concrete words into
. Acronyms: the task was to find out the word made
from the initial phonemes of a list of words.
. 2-D mental imagery (1): the task required patients
to move mentally from one position to another on
an imaginal chessboard.
. 2-D mental imagery (2): the task required patients
to move mentally from one key to another on an
imaginal calculator keyboard.
. 3-D mental imagery: the task required patients to
move mentally from one position to another on an
. Visual n-back: three types of stimuli were used:
playing cards; figures; geometrical forms. Patients
were asked to say whether or not an item matched
the item presented 1-, 2- or 3-back in the
C. Vallat-Azouvi et al.