Early Impaired Self-Awareness After Traumatic Brain Injury
Mark Sherer, PhD, Tessa Hart, PhD, Todd G. Nick, PhD, John Whyte, MD, PhD,
Risa Nakase Thompson, PhD, Stuart A. Yablon, MD
ABSTRACT. Sherer M, Hart T, Nick TG, Whyte J,
Thompson RN, Yablon SA. Early impaired self-awareness
after traumatic brain injury. Arch Phys Med Rehabil 2003;84:
168-76.
Objectives: To evaluate predictors of early impaired self-
awareness after traumatic brain injury (TBI); to examine inter-
relationships of the perceptions of patient, clinician, family,
and significant other of how patients are functioning after TBI;
and to determine how early impaired self-awareness helps to
predict employability at rehabilitation discharge.
Design: Inception cohort.
Setting: Two inpatient rehabilitation programs.
Participants: A total of 129 patients with TBI seen for
inpatient rehabilitation at 1 of 2 rehabilitation centers. All
subjects had emerged from posttraumatic amnesia before being
assessed for this study.
Interventions: Not applicable.
Main OutcomeMeasures: Impaired self-awareness as mea-
sured by the Awareness Questionnaire (patient self-ratings,
clinician ratings) and employability (rated on the Disability
Rating Scale) at discharge from inpatient rehabilitation.
Results: Regression analysis revealed that early impaired
self-awareness was predicted by age and functional status
(FIM™ instrument total score) at admission to inpatient reha-
bilitation. Spearman correlation coefficients revealed that cli-
nician, family, and significant other ratings of patient function-
ing were related (r
s

.42, P.001), but were not related to
patient self-ratings. Multiple logistic regression analysis re-
vealed that early impaired self-awareness was predictive of
employability at discharge from inpatient rehabilitation. Clini-
cian ratings of patient functioning showed a positive relation to
employability (P
.05), whereas patient self-ratings showed a
trend toward a negative relation to employability (P
.09).
Conclusions: Our results support the importance of early
impaired self-awareness assessment, its predictive value for
complex functional activities, and the need for further research
to determine if treatment programs for impaired self-awareness
enhance functional outcomes.
Key Words: Awareness; Brain injuries; Rehabilitation.
© 2003 by the American Congress of Rehabilitation Medi-
cine and the American Academy of Physical Medicine and
Rehabilitation
T
RAUMATIC BRAIN INJURY (TBI) can result in a vari-
ety of neurobehavioral impairments, including irritability,
restlessness, depression, and impaired self-awareness.
1
Im-
paired self-awareness has received particular attention from
rehabilitation workers because of its association with motiva-
tion for treatment
2,3
and long-term functional outcome.
4-6
Although impaired self-awareness can be evident early after
TBI, most published studies have investigated such impairment
in postacute TBI patients, and have found that it has some
general characteristics. Patients consistently show greater im-
paired self-awareness for cognitive and behavioral deficits than
for physical impairments,
7,8
and in their responses to general
questions about their functioning, rather than in responses to
specific questions.
9
Patients with more acute self-awareness are
more likely to complain of symptoms of depression.
10,11
Al-
though there are some negative findings,
12-14
limited evidence
exists that impaired self-awareness is related to severity of TBI,
with more severely injured patients showing greater impaired
self-awareness.
15,16
Finally, impaired self-awareness appears to
be common after TBI. In a study of 66 postacute TBI patients,
Sherer et al
5
found that, depending on the method of measure-
ment, 76% to 97% of patients showed some degree of impaired
self-awareness.
Although this impaired self-awareness is assumed to be a
neurobehavioral deficit, the neural substrate is poorly under-
stood. Inaccurate perception of one’s abilities is not unique to
TBI or other neurologic disorders. Allen and Ruff
12
found that
unimpaired controls overestimated their abilities on some cog-
nitive tasks and underestimated them on others; however, pa-
tients with TBI overestimated their cognitive abilities on vir-
tually all tasks, and the magnitude of their overestimation was
much greater than it was among the controls.
In contrast to the relatively large number of studies of
impaired self-awareness in postacute TBI patients, our review
found only 9 studies of early impaired self-awareness. We
considered studies to be “early” if patients were studied within
6 months postinjury or during initial inpatient rehabilitation.
McKinlay and Brooks,
17
in a sample of 55 patients with TBI,
noted that at 3 months, 6 months, and 1 year postinjury,
patients and family showed good agreement in reports of
sensory impairments, but poorer agreement in reports of cog-
nitive impairments and even poorer agreement for reports of
emotional and behavioral changes. Patients consistently re-
ported fewer emotional and behavioral problems than did their
family members.
Anderson and Tranel
13
studied 100 neurologic patients, in-
cluding 19 with TBI. Time from injury to evaluation was not
reported, but the description suggests that the TBI patients
were seen in their acute stage. Greater impaired self-awareness
was associated with lower postinjury verbal intelligence and
greater temporal disorientation. Severity of TBI as indicated by
Glasgow Coma Scale (GCS) scores was not associated with
degree of impaired self-awareness.
Ranseen et al
18
investigated impaired self-awareness in 32
patients with TBI who were undergoing inpatient rehabilita-
tion. Patients consistently rated themselves as having higher
levels of functioning than did rehabilitation staff members who
were treating them. Patients with predominantly right hemi-
From the Methodist Rehabilitation Center (Sherer, Thompson, Yablon) and Uni-
versity of Mississippi Medical Center (Sherer, Nick, Thompson, Yablon), Jackson,
MS; Moss Rehabilitation Research Institute, Philadelphia, PA (Hart, Whyte); and
Department of Rehabilitation Medicine, Jefferson Medical College, Thomas Jefferson
University, Philadelphia, PA (Hart, Whyte).
Supported by the National Institute on Disability and Rehabilitation Research
(grant nos. H133A980067, H133A980035, H133A97033).
No commercial party having a direct financial interest in the results of the research
supporting this article has or will confer a benefit upon the author(s) or upon any
organization with which the author(s) is/are associated.
Reprint requests to Mark Sherer, PhD, ABPP/Cn, Methodist Rehabilitation Center,
1350 E Woodrow Wilson, Jackson, MS 39216, e-mail: marks@mmrcrehab.org.
0003-9993/03/8402-7423$35.00/0
doi:10.1053/apmr.2003.50045
168
Arch Phys Med Rehabil Vol 84, February 2003

sphere injuries showed greater impaired self-awareness than
those with left hemisphere or diffuse injuries. The degree of
impaired self-awareness was not significantly related to the
degree of cognitive impairment. Patients with more accurate
self-awareness reported higher levels of depression.
Godfrey et al,
19
by using a cross-sectional design, reported
on the course of recovery of impaired self-awareness in 66
patients with TBI, 24 patients were 6 months postinjury, 19
were 1 year postinjury, and 23 were 2 to 3 years postinjury. The
latter 2 groups were similar in injury severity, whereas the
1-year group had sustained less severe injuries, as indicated by
GCS scores and duration of posttraumatic amnesia (PTA).
These researchers found that patients who were 6 months
postinjury rated themselves as less impaired than did their
significant others, whereas patients who were 1 or more years
postinjury rated themselves similarly to ratings by their signif-
icant others. Patients with more accurate self-awareness re-
ported greater emotional distress.
In another longitudinal investigation with similar results,
Fleming and Strong
20
studied 55 patients with severe TBI at 3
and 12 months postinjury. At 3 months, patients rated them-
selves as less impaired than did their therapists or significant
others, especially in cognitive and emotional functions. Ratings
among different informants were more similar at 12 months
postinjury, although patients persisted in scoring themselves
better than did their collateral raters on a few items.
Hart et al
21
investigated awareness of errors during perfor-
mance of functional tasks in 18 patients with TBI who were
undergoing inpatient rehabilitation and 18 control subjects who
were matched for the total number of errors made. The patients
showed poorer awareness of errors and less frequent self-
correction of errors made while working on the tasks.
Sherer et al
16
studied 64 patients with TBI during their
inpatient rehabilitation and found that they reported fewer
impairments than did their family members. Patients com-
plained more of physical symptoms while family members
reported more cognitive and behavioral impairments. Both
patients and family members showed greater agreement when
asked specific questions about functioning, as opposed to gen-
eral questions about overall functioning.
Newman et al
22
reported on impaired self-awareness in a
mixed sample of 37 patients with TBI or anterior communi-
cating artery aneurysm. The numbers of patients with each
diagnosis were not given. For initial ratings, 36 of 37 patients
(97%) described themselves as being less impaired than did the
rehabilitation staff. A subset of 18 patients was rated a second
time, and these ratings were more similar to staff ratings than
were the initial ones, suggesting a decrease in the degree of
impaired self-awareness. The possible significance of these
findings for understanding impaired self-awareness after TBI is
limited by the failure to report the number of patients with TBI
(or a separate analysis for the TBI patients), or the time period
between the 2 assessments. The reasons for not providing
second ratings for 19 subjects (51%) were not reported.
Finally, Abreu et al
23
studied a group of persons with TBI of
mixed acuity (9d to 119mo). The study also used a ratings-
comparison approach, but unlike other studies in which sub-
jects rate their overall abilities, Abreu asked subjects to judge
their performances on functional tasks immediately after com-
pleting the tasks. Self-ratings were compared with ratings of a
therapist who had observed the task. On most tasks, patients
rated their performances as better than how they were rated by
experienced clinicians. Interestingly, patients were also asked
to rate the ability of other persons to perform the tasks; these
ratings were more accurate, suggesting that generalized “judg-
ment” difficulties did not account for the impaired self-assess-
ment.
Interpretation of studies of impaired self-awareness is com-
plicated by at least 2 important measurement issues. One is that
different methods are used to measure impaired self-awareness
in different studies, consequently, it is difficult to make com-
parisons across studies. All methods for estimating impaired
self-awareness involve estimating the disparity between the
patient’s self-perception or self-report and some external cri-
terion of his/her status. However, differences in the format of
the self-report and/or the external criterion can affect the mea-
surement of impaired self-awareness. For example, the assess-
ment commonly involves comparison of patients’ self-ratings
of function to those of a collateral, usually a family member, a
significant other, or a clinician who is treating the patient. In
studies with postacute patients, the use of these different col-
laterals has produced dissimilar findings because of differences
between family members or significant others and clinicians in
their perceptions of the patients.
24,25
Another important measurement issue is that the patient
versus collateral rating is often expressed as a subtracted score
(ie, a discrepancy score), with the implication being that the
size of the score estimates the magnitude, or severity, of
impaired self-awareness. But the magnitude of such a discrep-
ancy score is actually constrained by the score used as the
external criterion, which is typically selected for its presumed
relation to the “true” level of function. Thus, assuming that a
criterion such as a clinician rating is correlated with the actual
ability level, then as that level increases (ie, toward the more
competent end of the scale), the size of any possible discrep-
ancy decreases. Comparing the discrepancy scores of more
versus less competent subjects could thus make it appear as
though the latter have worse impaired self-awareness (ie, larger
discrepancy scores), whereas some of this effect could be
because of their simply having “more room to disagree” on the
rating scale (this issue is also discussed by Hart et al
26
).
Similarly, in longitudinal studies, this constraint could make it
appear that impaired self-awareness is decreasing over time,
when the change could be a result of functional recovery, that
is, improved scores on the external criterion.
As noted previously, several methods have been used in
impaired self-awareness research, not all of which involve the
calculation of patient versus collateral discrepancy scores. One
method involves comparing patient self-ratings of cognitive
abilities with performances on tests of those abilities. Such
studies have found that patients with TBI overestimate their
abilities,
13
and they overestimate their abilities to a greater
extent than do persons without TBI.
12
Other investigators
27
have shown that persons with TBI often overestimate their
abilities to perform simple functional tasks such as repeating
words or solving arithmetic problems. As noted earlier, Hart
21
showed that patients with TBI made more errors than did
controls on naturalistic tasks such as making toast, wrapping a
gift, and packing a child’s lunchbox. Patients were also less
likely to correct their errors, which indicates a decreased
awareness of the errors. Finally, in an investigation that com-
pared self and other discrepancies between patients with TBI
and unimpaired controls, Prigatano et al
15
found that patients
were more likely than were controls to rate themselves as more
competent than the informant. Taken together, these findings
indicate that impaired self-awareness after TBI is not simply
the result of a tendency of persons to view themselves more
positively than they are viewed by others.
Based on our literature review, we sought to contribute to an
improved understanding of early impaired self-awareness by
studying a sample of patients with TBI during initial inpatient
169EARLY IMPAIRED SELF-AWARENESS AFTER TBI, Sherer
Arch Phys Med Rehabil Vol 84, February 2003