rch
tic
rro
ect
M.
of N
31 Ju
11 S
deficits in TBI patients, including accuracy and subjectivity
report [4,42]. It has
Cognitive Brain Research 22 (21. Introduction
Following traumatic brain injury (TBI), survivors with
diffuse axonal damage or more focal lesions or haemor-
rhages often demonstrate patterns of executive function
deficits similar to patients with focal prefrontal lesions
[45,46,51,6,53,27,1,15,26,47]. The white matter of the
prefrontal cortex is known to be particularly vulnerable to
the effects of TBI [46,26]. These impairments in executive
control often evident in TBI patients can manifest them-
selves in numerous ways, including difficulties in self-
direction, planning, organisation, problem-solving, sustain-
ing attention, disinhibition, and self-monitoring. Goldberg
and Barr [19] implicate the executive functions of cognitive
self-monitoring and the comparison of the outcome of one’s
cognitive operations with self-awareness.
Self-awareness and realistic self-appraisal following
TBI are recognised as a significant factor in predicting
potential benefits from treatments, overall long-term out-
come, return to community living, and productive lifestyle
[2,33,11,31,32]. Prigatano and Schachter [34] propose that
patients who are unable to detect performance errors in tasks
have deficits in blocal awareness.QAnother form of awareness
deficit in frontal lobe patients is bglobal awarenessQ or
unawareness of disorder. TBI patients are classically more
aware of physical deficits than they are of cognitive, social, or
behavioural difficulties [34]. There are a number of meth-
odological difficulties involved in assessing global awarenessAbstract
Impaired deficit awareness is common following traumatic brain injury (TBI) and is a major obstacle to rehabilitation. We have previously
confirmed the presence of impaired error awareness in TBI using a highly discriminating go/no-go procedure. In the present study, we extend
this work to try to identify more closely the nature of the error awareness deficit using measures of electrodermal activity (EDA). Sixteen
participants with TBI and sixteen age-, sex-, and education-matched controls performed the Sustained Attention to Response Task (SART),
while EDAwas recorded. TBI detected significantly fewer errors compared to controls. EDAwas significantly attenuated for TBI participants
even to errors of which they were aware; error detection rates and EDA amplitude were also correlated. These findings suggest that poor
insight following TBI may result, in part, from impaired error processing abilities.
D 2004 Elsevier B.V. All rights reserved.
Theme: Disorders of the nervous system
Topic: Trauma
Keywords: Error awareness; Electrodermal activity; Traumatic brain injury; No-go taskResea
Poor insight in trauma
by impaired e
Evidence from el
Fiadhnait M. O’Keeffe, Paul
Department of Psychology and Trinity College Institute
Accepted
Available online0926-6410/$ - see front matter D 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.cogbrainres.2004.07.012
* Corresponding author. Tel.: +353 1 608 2684; fax: +353 1 608 1176.
E-mail address: Ian.Robertson@tcd.ie (I.H. Robertson).report
brain injury mediated
r processing?
rodermal activity
Dockree, Ian H. Robertson*
euroscience, Trinity College Dublin, Dublin 2, Ireland
ly 2004
eptember 2004
004) 101–112
www.elsevier.com/locate/cogbrainresof self, significant other, and clinician
been proposed that investigations into error processing,
detection, and correction (i.e., investigations into blocal

patients failed to respond similarly. Zahn and Mirsky [54]
have shown that persons with CHI show marked deficits in
BraiawarenessQ) may provide insights and explanations into the
more general phenomena of unawareness [20] and that an
intact process of error monitoring implies an intact awareness
of deficit [56].
Rabbitt [35] suggested, using purely behavioural data,
that error processing involves two mechanisms—an error
detection mechanism and a set of error correction mecha-
nisms. Recent functional magnetic resonance imaging
(fMRI) and event-related potential (ERP) studies have
supported this view, and have shown a dissociation between
error detection and behaviour correction. In an fMRI study,
Garavan et al. [16] showed that the detection of an error
involved medial–frontal activation, in particular the anterior
cingulate. Error correction, on the other hand—which
involved a change in mental set and establishing a more
appropriate task set—was shown to be accomplished by the
left dorsolateral areas of the prefrontal cortex. ERP studies
have also shown some dissociation between detection and
correction of errors. Two psychophysiological components
are thought to reflect aspects of these error-monitoring
processes—error negativity (Ne/ERN) and error positivity
(Pe) [12,18,13]. It is possible that the Ne may reflect some
manifestations of some error detection systems that compare
behaviour against some internal standard or goal, and that
the midline prefrontal cortex and the anterior cingulate have
also been indicated as generators of this response
[43,52,29,9,41]. Falkenstein et al. [13] suggested that the
Pe was related to the conscious processing of an error and
may be related to the adjustment of response strategies
following this error recognition.
A number of studies have shown error detection impair-
ments following TBI or frontal lobe lesions. Although not
conclusive, ERP studies have shown disrupted or unusual
Ne and Pe production in patients with lateral and medial
prefrontal lesions, and lesions of the anterior cingulate
region [43,50,17]. Hart et al. [20] studied awareness of
errors among people with TBI in naturalistic everyday
situations, such as making toast, packing a child’s school-
bag, and wrapping a gift. TBI patients showed less
awareness and corrected fewer errors than controls. It was
suggested that head injury may result in impairment in the
allocation of resources needed for effective execution and
monitoring of routine tasks. Stemmer et al. [43] recorded
error awareness behaviour on a flanker task trial based on
overt behaviour recorded by the researchers by documenting
behavioural signs, such as exclamations, whispered swear-
ing, and grimaces, during an ERP study of patients with
lesions to the prefrontal cortex. In their study, three of the
five patients showed inconsistent or mixed error awareness
behaviour.
Electrodermal activity (EDA) has been commonly used
as an index of psychological processing properties of
stimuli, such as significance, novelty or emotional rele-
vance, and effortful processing [55]. Lehrer et al. [21]
F.M. O’Keeffe et al. / Cognitive102showed that a group of closed-head injury (CHI) patients
showed less physiological reactivity, including attenuatedboth tonic and phasic electrodermal responding to mean-
ingful, significant, or demanding situations or stimuli. Their
results suggest that CHI has long-lasting effects on the
ability to mobilize activation during effortful, controlled
information processing. Patients with frontal lobe damage
also tend to display attenuated EDA responses to stimuli of
emotional or psychological significance [55]. They can
quite often recall disturbing images in detail but with a
distinct absence of feeling. Their lack of affective response
was interpreted as their descriptions representing semantic,
rather than emotional, associations [8,49]. Damasio [7]
refers to this distinction between factual knowledge and
emotional response as, bto know does not necessarily mean
to feel.Q
Based on previous studies showing that TBI patients
have poor error processing abilities [20,28] and evidence
that TBI and frontal lesions patients tend to show decreased
physiological response as measured by EDA to emotional
significant stimuli, we focus on the emotional response to
making an error in a paradigm that imposes cognitive
significance to a rare no-go target. We believe that this may
aid understanding of error monitoring in TBI patients, as it
is possible that the impairment of emotional processing in
TBI may be linked to their awareness deficits in error
detection and correction. We hope to identify a specific error
processing deficit in response to making an error on a
simple continuous performance task (the SART) and
compare the emotional responses of TBIs to controls on
making such an error.
The aim of the present study was to investigate the
performance of 16 TBI patients on a task of sustained
attention (SART), monitor their detection of errors while
they performed the task, and analyse their EDA response to
errors and correct responses. We compared their perform-
ance on the tasks and EDA responses to neurologically
healthy controls in order to determine whether error
awareness deficits were evident in impaired psychophysio-
logical responsiveness to errors.
2. Overview
Participants performed a test of sustained attention,
which has proven sensitivity to TBI, is cognitively simple,
and has been used to successfully demonstrate impaired
error awareness in a TBI population. This task has also beenskin conductance response (SCR), during cognitive tasks as
compared to controls. They suggested that this may have
been a result of a deficit in situation-appropriate modulation
of physiological activity in that group. The pattern of
physiological reactivity found in the control group was
interpreted as adaptive responses to specific tasks, whereas
n Research 22 (2004) 101–112shown to activate prefrontal regions of the brain in a number
of different studies [24,30,14]. EDA activity was recorded