Strength, balance and swallowing deficits and outcome after trauma brain injury

1291

Relationship Between Strength, Balance, and Swallowing
Deficits and Outcome After Traumatic Brain Injury:
A Multicenter Analysis
Thao T. Duong, MD, Jeffrey Englander, MD, Jerry Wright, BA, David X. Cifu, MD,
Brian D. Greenwald, MD, Allen W. Brown, MD
ABSTRACT. Duong TT, Englander J, Wright J, Cifu DX, tation discharge. This association remains strong at 1 year after
Greenwald BD, Brown AW. Relationship between strength, TBI. By using this information, clinicians should initiate ther-
balance, and swallowing deficits and outcome after traumatic apeutic interventions that optimize rehabilitation of the identi-
brain injury: a multicenter analysis. Arch Phys Med Rehabil fied impairments and should make necessary arrangement for
2004;85:1291-7. the patient’s anticipated postdischarge needs. Further studies
Objective: To examine the relation among strength, balance, are necessary to delineate the amount of unique variance that
and swallowing deficits, as measured on rehabilitation admis- these early physical examination findings contribute to out-
sion, and functional outcome at discharge and 1 year after come prediction.
traumatic brain injury (TBI). Key Words: Balance; Brain injuries; Multicenter studies;
Design: Multicenter analysis of consecutive admissions to Rehabilitation; Swallowing; Treatment outcome.
designated Traumatic Brain Injury Model Systems (TBIMS) © 2004 by the American Congress of Rehabilitation Medi-
facilities. cine and the American Academy of Physical Medicine and
Setting: Seventeen TBIMS centers. Rehabilitation
Participants: Adults and children older than 16 years of age
with TBI (N 2363) enrolled in the national database from
January 1989 to November 2000.
Interventions: Not applicable.
P EOPLE WITHcaregivers striveBRAIN injuryoutcomesand
their potential
TRAUMATIC
for optimal
(TBI)
in
mobility and self-care. Persons are admitted to rehabilitation
Main Outcome Measures: Transfers, locomotion, stairs, programs to achieve goals in these specific areas, so that they
lower-body dressing, grooming, bathing, upper-body dressing, can regain their functional ability and can safely leave the
toileting, and eating as measured by the FIM instrument at hospital. Rehabilitation professionals are frequently asked by
acute rehabilitation discharge and at 1 year after TBI. family members, acute trauma teams, and third-party payers to
Results: Lower-extremity strength less than 3/5 on admis- predict the likelihood of walking and performing self-care
sion to acute rehabilitation was associated with increased need activities for a given person with TBI. Transfers, ambulation,
for assistance in locomotion, transfers, and lower-body dress- dressing, grooming, and toileting skills are critical in reestab-
ing and less than 3/5 upper-extremity strength was associated lishing routines in the home environment. Gauging the need for
with the need for assistance in self-care at rehabilitation dis- another person to help perform these tasks is important for
charge and at 1 year postinjury. Similar relations were found planning current and future care needs. Knowing the best
between impaired swallowing and assistance with eating, prognostic factors associated with achieving these mobility and
grossly impaired dynamic sitting, or standing balance and self-care goals is important to all parties involved.
assistance with locomotion, transfers, eating, and self-care at Acute trauma factors and impairment levels impact func-
rehabilitation discharge and at 1 year after TBI. tional outcomes in multiple areas, including self-care, mobility,
Conclusions: Assessments of physical strength, swallowing and community reintegration. In 1996, Englander et al1 showed
ability, and dynamic balance on acute rehabilitation admission that, on admission to rehabilitation, subjects with motor
are helpful as screening tests in predicting the need for assis- strength less than 3/5 on manual muscle testing (MMT) and
tance of another person for mobility and self-care at rehabili- moderate to severe incoordination were more likely to need
physical assistance for mobility and self-care at rehabilitation
admission, discharge, and, to some extent, at 1 year postinjury.
Individuals with pelvic fractures, lower-extremity fractures,
From the Department of Physical Medicine and Rehabilitation (Duong, Englander); and upper-extremity fractures had longer lengths of stay
and Rehabilitation Research Center for TBI and SCI (Wright), Santa Clara Valley
Medical Center, San Jose, CA; Department of PM&R, Virginia Commonwealth
(LOSs) in acute care and in rehabilitation. Interestingly, only
University/Medical College of Virginia, Richmond, VA (Cifu); Department of Re- those with lower-extremity fractures had a higher likelihood of
habilitation Medicine, Mount Sinai School of Medicine, New York, NY (Greenwald); requiring physical assistance for self-care and mobility at ad-
and Department of Physical Medicine and Rehabilitation, Mayo Clinic Rochester, mission and discharge from rehabilitation.
Rochester, MN (Brown).
Presented in part at the International Association for the Study of Brain Injury, May
Greenwald et al2 found that people who were below the age
2001, Turin, Italy. of 50 years were more likely to have normal sitting and
Supported in part by the National Institute on Disability and Rehabilitation Re- standing balance on rehabilitation admission. Indications of
search, Office of Special Education and Rehabilitative Services, US Department of higher severity TBI, such as lower Glasgow Coma Scale (GCS)
Education (grant no. H133A70018).
No commercial party having a direct financial interest in the results of the research
score, longer length of coma, longer length of posttraumatic
supporting this article has or will confer a benefit upon the authors(s) or upon any amnesia (PTA), and longer acute care LOS were each associ-
organization with which the author(s) is/are associated. ated with impaired sitting and standing balance ratings at
Reprint requests to Thao T. Duong, MD, Dept of Physical Medicine and Reha- rehabilitation admission. The presence of any midline shift or
bilitation, 751 S Bascom Ave, San Jose, CA 95128, e-mail: thao.duong@hhs.co.
santa-clara.ca.us. brainstem compression on computed tomography (CT) scan
0003-9993/04/8508-8345$30.00/0 was also associated with impairment in standing balance, but
doi:10.1016/j.apmr.2003.11.032 not sitting balance, at rehabilitation admission. Postural control

Arch Phys Med Rehabil Vol 85, August 2004

1292 STRENGTH, BALANCE, AND SWALLOWING AFTER TBI, Duong

is mediated by the complex integration of the sensory and Table 1: Demographics
motor systems. Impairments in any one of these areas can
affect balance reactions. Attainment of sitting and standing Variable n (%)

balance is a critical element in the rehabilitation process. Im- Sex
paired sitting balance affects a large number of activities of Men 1771 (75)
daily living (ADLs) including feeding, dressing, bathing, trans- Women 592 (25)
fers, and wheelchair mobility.3,4 Adequate standing balance is Race/ethnicity
essential for unassisted standing, walking, and negotiation of White 1420 (60)
stairs.5,6 Better scores on the FIM instrument and Disability African American 655 (28)
Rating Scale during inpatient rehabilitation have been found by Hispanic 183 (8)
Cifu et al7 to be predictive of return to work at 1 year postin- Other 103 (4)
jury. In a more recent study,8 high correlations were found Age (y)
between the impairment and disability measures of the leg and 16–25 775 (33)
handicap situations. Interestingly, that study also found disabil- 26–35 507 (21)
ity of the leg to be more strongly associated with handicap than 36–45 486 (21)
was arm disability. 46–55 270 (11)
In a study of a heterogeneous acute inpatient rehabilitation 55 324 (14)
population by Juneja et al,3 the Berg Balance Scale score’s Etiology
sitting-unsupported item was shown to account for 27% of the Motor vehicle crash 1296 (55)
variance in rehabilitation LOS. It was suggested by Juneja that Violence 452 (19)
balance information may be helpful in setting goals in relation Falls 382 (16)
to predicted LOS. Black et al9 also asserted that, after age, the Other 224 (10)
degree of sitting balance impairment at rehabilitation admis- History of TBI
sion was the second most powerful predictor of discharge total Prior TBI 195 (8)
FIM score10,11 and selected elements of the discharge FIM No history 2129 (92)
motor score in patients with TBI.
Wober et al12 found that severe brain injury and deep pa- NOTE. Sample size: N 2363.
renchymal lesions shown by magnetic resonance imaging were
significant indicators of postural imbalance. Englander et al13
studied the association between various brain CT findings and asked by people with TBI, their caregivers, and rehabilitation
the need for assistance with mobility and self-care skills in professionals.
patients with TBI who required acute inpatient rehabilitation. A METHODS
midline shift of greater than 5mm and subcortical contusions
were the lesions most highly correlated with the need for Participants
assistance with ambulation, toileting, lower-body dressing,
continence, and overall supervision needs. A total of 2363 cases were available from the Traumatic
The etiology of balance impairment in the TBI patient has Brain Injury Model Systems (TBIMS) Database. The study
been explored.2,12 Studies to date on resultant disability have population consisted of people with TBI admitted between
focused mostly on rehabilitation discharge.3-6,9 Published re- January 1989 and November 2000 to 1 of the 17 National
search and experiential knowledge suggest that motor strength, Institute on Disability and Rehabilitation Research–funded
balance, and swallowing status on admission to rehabilitation TBIMS facilities. Subjects were enrolled in the study if they
may be useful in predicting the need for assistance at rehabil- met the following criteria: age 16 years or older, presentation to
itation discharge and at 1 year after TBI. Moreover, these a designated system hospital emergency department within 24
measurements are routinely performed by members of the hours of sustaining a TBI, and receiving acute care and inpa-
rehabilitation team. If these routine measurements of impair- tient rehabilitation within the Model Systems hospitals. Sub-
ment can be shown to be associated with or predictive of jects were followed annually. An institutional review board at
disability outcomes, then they can be useful for the rehabilita- each center approved the enrollment of subjects into the
tion program, for patients, and for caregivers in resource plan- TBIMS Database. Informed consent was obtained and signed
ning for individual patients. In this investigation, we examined by the patient or family/legal guardian. Demographic charac-
whether (1) lower-extremity strength as measured on rehabil- teristics of the population are summarized in table 1. Of the
itation admission would be associated with capabilities for subjects followed during inpatient rehabilitation, 1-year postin-
locomotion, transfers, and lower-body dressing at rehabilitation jury data were collected in 1078 subjects at the time of data
discharge and at 1 year after TBI; (2) upper-extremity strength analysis. Subject attrition at 1-year follow-up was examined in
as measured on rehabilitation admission would be associated regard to our variables of interest (swallowing, balance,
with capabilities for eating and self-care at rehabilitation dis- strength). The group followed at 1 year did not differ signifi-
charge and at 1 year after TBI; (3) impaired swallowing as cantly from the group without 1-year data in terms of initial
measured at rehabilitation admission would be associated with assessments in swallowing, balance, or strength. This popula-
eating capabilities at rehabilitation discharge and at 1 year tion was comparable to populations studied in previously pub-
post-TBI; and (4) impaired dynamic sitting balance and stand- lished TBIMS literature and reflected the population served by
ing balance as measured at rehabilitation admission would be the participating regional trauma and brain injury rehabilitation
associated with capabilities for transfers, locomotion, and self- programs.
care at rehabilitation discharge and at 1 year post-TBI.
Although some of these questions have been partially ad- Instruments and Assessment
dressed in the studies cited above, their inclusion here with a Data were collected prospectively. MMT to assess strength
larger database and the end points of both discharge to home was performed by a physiatrist within 72 hours of rehabilitation
and 1 year postinjury may be more applicable to the questions admission. The voluntary muscle strength was recorded for


STRENGTH, BALANCE, AND SWALLOWING AFTER TBI, Duong 1293

each extremity, using a scale from 0 to 5 (0, no movement; 1, a minimum of 45m (150ft) or is able to walk independently
trace activation without joint movement; 2, movement with shorter distances of a minimum of 15m (50ft) with or without
gravity eliminated; 3, movement against gravity; 4, movement a device (household exception). Similarly, for stair climbing, a
against gravity with some resistance; 5, normal strength). The score of 5 may either mean that the person requires supervision
strength recorded for each extremity was the average strength to go up and down 1 flight of stairs or is able to go up and down
for all muscles tested in that extremity. The upper- or lower- 4 to 6 stairs independently. In our data analysis, for ambulation
extremity strength measurement used in the data analysis was and stair climbing, scores of 5 to 7 were considered indepen-
considered less than 3/5 if either the right or left side’s strength dent, and scores of 4 or fewer were considered as definitely
was less than 3/5. Swallowing was evaluated and rated as requiring another person. The FIM raters at different centers
normal, impaired, absent, or nontestable. Dynamic sitting and had either passed the Uniform Data System FIM certification
standing balance were rated as normal, mild impairment (able testing or a similar vignette exercise administered by Santa
to maintain balance with some difficulty), or gross impairment Clara Valley Medical Center.
(unable to maintain balance). Further description of these rat-
ings is available in the TBMIS National Database Syllabus.14
In the present study, examinations of strength, swallowing Statistical Analyses
ability, and dynamic balance were used to document the se- Because of the categoric nature of the variables being ex-
verity of impairments on admission to acute rehabilitation. amined (ie, normal, mild, gross impairment) and the dichoto-
Rating muscle strength is a skill that takes time to learn and mous nature of the outcome variables (need assist, no assist),
perform with reliability because it encompasses both subjective chi-square analyses were performed. Because the chi-square
and objective factors. Reliability and validity of this test also test is influenced by sample size, a measure of effect size is also
depend on the subject’s ability to cooperate, fatigue factor, and reported. Depending on the number of categories analyzed,
adherence to the same protocol in terms of patient positioning, Cramer’s V or was used to measure effect size, with 0.2
joint stabilization, and avoiding substitutions. In a review of the considered a moderate effect and 0.4 a large effect. For each
literature pertaining to the reliability and validity of physical
examination tests for the upper extremity, Marx et al15 found analysis group, only those subjects with complete data for
that MMT had been described as semiquantitative because of indicated items were included.
its ordinal nature and that this method of quantifying muscle
strength was reliable both within and among observers, partic- RESULTS
ularly when the scale is expanded. To increase reliability, the
rating method used in our study only delineated between those Lower-Extremity Strength
who showed antigravity strength and those who did not. This is Significant differences in the percentage of subjects needing
a more easily observed and consistent rating than attempting to assistance existed between those with a lower-extremity
discriminate between various degree of effort (eg, 4 vs 5 strength of less than 3/5 on MMT at rehabilitation admission
ratings). and those with a lower-extremity strength of 3/5 or greater. For
Bedside swallowing evaluation is an important early screen- locomotion, 35.2% of those with less than antigravity strength
ing tool for dysphagia and aspiration risk. Although it is safe required assistance at acute rehabilitation discharge versus
and easily repeated, it has been variable in its sensitivity 11.2% in the group with at least antigravity lower-extremity
(42%–92%), specificity (59%–91%), and interrater reliability strength (P .0001). Differences were 70.1% versus 24.5% for
( range, 0 –1.0), and is poor at detecting silent aspiration.16 stair climbing, 63.1% versus 27.0% for bed to wheelchair
Quantification of dynamic balance has been attempted by using transfers, 64.9% versus 28.7% for toilet transfers, 73.3% versus
various techniques. Reliability estimates for 2 testing days have 42.6% for tub/shower transfers, and 68.6% versus 34.8% for
ranged from .67 to .87 by using the star-excursion test (subject lower-body dressing (all P .0001; table 2). At 1 year, both
to balance on 1 leg while reaching with the other leg) to .58 to groups showed continued improvement, with fewer subjects
.87 for 75% Limits of Stability Test.17,18 Given the limited requiring assistance. However, the difference between the
reliability and validity of the assessment tools used in our groups remained at 1 year (all P .0001; table 2). The effect
investigation, we kept the final subject categorization very sizes, as measured by , were moderate except for locomotion,
simple and qualitative, with only 2 categories used for muscle which was small at 1 year. Between 3 to 5 times as many
strength, 2 for swallowing, and 3 for dynamic balance. individuals in the less than antigravity strength group required
Functional items such as eating, grooming, upper-body assistance for these activites at 1 year post-TBI compared with
dressing, lower-body dressing, bathing, toileting, transfers, lo- the greater than antigravity strength group.
comotion, and stair climbing were rated by the rehabilitation
staff at each center at rehabilitation discharge and by the Upper-Extremity Strength
research staff either in person or via the telephone at 1 year
postinjury, using the FIM instrument.10,11 The FIM is an 18- Statistically significant differences were found between the
item, 7-level scale used to assess self-care, functional mobility, group with upper-extremity strength less than 3/5 and those
communication, cognitive, and bowel and bladder management with strength of 3/5 or greater in needing assistance for eating,
status. Subscale scores for need of another person to complete grooming, bathing, and upper-body dressing at acute rehabili-
the functional task were used in a dichotomous fashion. The tation discharge and at year 1 (all P .0001; table 3). The
FIM outcome variables were dichotomized to lessen the ceiling percentage of subjects with weaker upper-extremity strength
effects of the FIM (especially at 1y).19,20 For eating, bathing, who required assistance at discharge and 1 year post-TBI was
dressing, toileting, and bowel and bladder management, a score 1.5 to 2.3 times and 2.8 to 4.3 times higher, respectively, than
of 6 or 7 is considered independent; a score 5 or less indicates the percentage with stronger upper-extremity strength. In both
the need for another person. For ambulation and stair climbing, groups, the percentage of subjects who still needed assistance
a score of 4 or less definitively indicates the need for another after 1 year after TBI did improve when compared with acute
person. However, with ambulation, a score of 5 may either rehabilitation discharge. The effect sizes, as measured by ,
mean that the person requires standby supervision to walk for were moderate.



Table 2: Lower-Extremity Strength on Admission and Function at Discharge and 1 Year

LE Strength LE Strength LE Strength LE Strength
3/5 3/5 3/5 3/5
% Req % Req % Req % Req
Assist at Assist at Assist at Assist at
2 2
FIM Item Rehab DC (n) Rehab DC (n) Test, 1 Year (n) 1 Year (n) Test,

Locomotion 35.2 (310) 11.2 (1562) 115.3, .25 11.8 (153) 3.5 (720) 18.5, .15
Stairs 70.1 (308) 24.5 (1563) 246.1, .36 23.0 (148) 6.5 (711) 39.5, .21
Bed transfer 63.1 (309) 27.0 (1565) 152.0, .29 18.2 (154) 3.7 (722) 44.9, .23
Toilet transfer 64.9 (308) 28.7 (1565) 149.3, .28 19.5 (154) 3.9 (722) 50.0, .24
Tub transfer 73.7 (308) 42.6 (1567) 100.0, .23 27.5 (153) 7.6 (721) 50.3, .24
LB dressing 68.6 (309) 34.8 (1568) 122.9, .26 25.6 (156) 8.3 (723) 38.3, .21

NOTE. All analyses were statistically significant at P .0001.
Abbreviations: Assist, assistance; DC, discharge; LB, lower body; LE, lower extremity; Rehab, rehabilitation; Req, requiring.

Swallowing Capabilities and 64.0%, 8.7% and 41.7%, and 6.5% and 35.7% for the
In subjects with impaired swallowing on rehabilitation ad- grossly impaired, mildly impaired, and normal balance groups,
mission, 44.9% required assistance for eating at acute rehabil- respectively. Again, the more impaired the standing balance at
itation discharge and 15.5% still needed assistance at 1 year admission, the more likely the need for assistance of another
after TBI as compared with 14.0% and 3.3%, respectively, for person at acute rehabilitation discharge. The effect size at acute
those with normal swallowing (both P .0001; table 4). The rehabilitation discharge was equal to or greater than .20 for all
effect sizes were moderate. functional activities measured except for grooming, upper-
body dressing, and toileting between the grossly impaired and
Sitting Balance mildly impaired groups. The effect size was less than .10
Statistically significant differences were found among the 3 between the mildly impaired and normal groups. At 1 year
groups with normal, mildly impaired, or grossly impaired bal- post-TBI, differences were still statistically significant among
ance, as measured at acute rehabilitation admission, and capa- the 3 groups (all P .0001; table 6), but the effect size between
bilities for self-care and mobility at discharge. The more im- the pair-groups was less than .17 for the grossly impaired and
paired the sitting balance, the more likely was the need for mildly impaired groups and less than .11 for the mildly im-
assistance with each ADL and mobility task measured (all paired and normal groups.
P .0001; table 5). The percentage of subjects needing assis-
tance ranged between 8.0% and 38.2% for the normal group, DISCUSSION
between 14.5% and 52.1% for the mildly impaired group, and This investigation shows that, after TBI, persistent weakness
between 37.8% and 79.6% for the grossly impaired group. of the upper and lower extremities, impaired swallowing, and
When the groups were compared in a pairwise fashion, the abnormal sitting and standing balance at the time of admission
effect size at acute rehabilitation discharge ranged from .23 to to inpatient rehabilitation were all associated with a need for
.30 between the grossly impaired and mildly impaired groups increased assistance at rehabilitation discharge and at 1 year
and was less than .20 between the mildly impaired and the postinjury. These findings are consistent with previously pub-
normal groups. At 1 year, the percentage of subjects needing lished data from the TBIMS projects and extend the analyses of
assistance decreased in all 3 groups. The differences among the Englander1 and Greenwald2 and colleagues by incorporating
groups were still significant using chi-squares analyses (all gross sitting and standing balance into the initial evaluation.
P .0001; table 5), however, the effect sizes between the pair- The confirmation that these early physical examination factors
groups were less than .20 for all, except for tub transfer are important early markers for short- and long-term disability
between the grossly impaired and mildly impaired groups. has significant clinical impact. Although much of inpatient and
outpatient brain injury rehabilitation is standardized in an effort
Standing Balance to achieve predetermined functional levels or goals, the pres-
For standing balance, the differences among the 3 groups ence of some or all of these findings should alert clinicians to
were also statistically significant (all P .0001; table 6). The the need for modifying or individualizing rehabilitation efforts.
level of assistance needed at discharge ranged between 23.7% These modifications should emphasize selecting interventions

Table 3: Upper-Extremity Strength on Admission and Function at Discharge and 1 Year

UE Strength UE Strength UE Strength UE Strength
3/5 3/5 3/5 3/5
% Req Assist at % Req Assist at % Req Assist at % Req Assist at
2 2
FIM Item Rehab DC (n) Rehab DC (n) Test, 1 Year (n) 1 Year (n) Test,

Eating 44.0 (327) 19.1 (1587) 93.6, .22 18.8 (160) 4.4 (734) 42.1, .22
Grooming 54.4 (327) 29.3 (1586) 77.1, .20 21.3 (160) 7.9 (734) 25.4, .17
Bathing 70.3 (327) 45.6 (1585) 66.3, .19 26.9 (160) 9.5 (734) 35.8, .20
UB dressing 58.7 (327) 30.6 (1586) 93.9, .22 23.8 (160) 6.7 (734) 43.6, .22

Abbreviations: UB, upper body; UE, upper extremity.



Table 4: Swallowing on Admission and Eating at Discharge strength is a gross measure, although it is commonly used in
and 1 Year
clinical research to quantify the neurologic examination.21-23
Swallowing Swallowing The global clinical descriptors of balance used we used also
Impaired Normal lack the precision of other standardized measures.24,25 Thus,
% Req % Req even with inherent limitations noted, the clinical factors mea-
FIM Item Assist (n) Assist (n) 2
Test sured are useful screening tests that can assist in prescribing
Eating at rehab DC 44.9 (746) 14.0 (1361) 245.2, P .0001 .34 individual rehabilitation programs and predicting future needs.
Eating at 1y 15.5 (362) 3.3 (613) 47.2, P .0001 .22 If measures more specific to people with TBI are developed,
the association between these physical impairments and short-
and long-term disability may increase in predictive value. De-
velopment of a unique, valid, quantifiable, and brief measure of
that optimize rehabilitation of the identified physical limita- the neurologic impairment that occurs after TBI may provide a
tions, making adjustments to LOS, therapy intensities, and better method both to anticipate future needs and to follow
treatment settings to accommodate greater functional limita- clinical progress.
tions, and making arrangements for necessary aftercare and Impaired strength is an indicator of injury severity, most
family supports. likely an injury involving a focal lesion such as a peripheral
Our findings provide an interesting contrast to a recent nerve injury or cortical and/or subcortical contusion. It is no
TBIMS analysis examining the association between head CT surprise that extremity weakness would be associated with
scan findings during the first week post-TBI and functional difficulties in completing the functional activities performed by
outcomes at rehabilitation discharge and 1 year.13 Although we that extremity. Thus, the findings that upper-extremity weak-
found associations between findings of significant brain insults ness correlates with feeding, grooming, and upper-body dress-
(ie, midline shift 5mm or subcortical contusions) and greater ing, and lower-extremity weakness correlates with transferring,
need for assistance at rehabilitation discharge in ambulation walking, and stair climbing are to be expected. The durability
and ADLs, these associations were weak at 1 year postinjury. of these associations for at least 12 months postinjury had not
Thus, although both severe CT scan abnormalities and the been previously shown and is not necessarily true of all early
presence of significant physical limitations are commonly ac- markers of injury severity, such as CT scan findings.13 The
cepted indicators of injury severity, the value of radiographic same can be said of the demonstrated associations between
indicators in “predicting” disability appears to be limited to the initial swallowing and balance deficits and subsequent disabil-
short term. Importantly, people who have had significant TBIs ities. Functional adaptation to physical limitations by people
and who have shown concomitant physical impairments are after TBI is typical, and this may help to explain the decreased
more likely to require assistance as a result of disabilities for strength of associations over time.
the long term as well. Although the present study shows the association between
The evidence that even gross measures of physical function- early physical examination findings and functional outcome
ing such as strength, balance, and swallowing ability are useful measures, a future study, using multiple regression analyses,
indicators of long-term disability after TBI is an additional may be able to show the amount of unique variance that early
contribution of this investigation. Additionally, although con- physical examinations contribute to outcome prediction as
founding factors, such as concomitant fractures, pain, medica- compared with other predictors (eg, GCS score, length of coma
tions side effects, and behavioral disturbances, may limit the or PTA, CT scan findings). To perform this type of study, one
examiner’s ability to evaluate fully all aspects of a physical would have to focus on only 1 or 2 variables of primary interest
examination at the time of rehabilitation admission, the exam- (ie, transfers or ambulation).
ination components we used in the present investigation are With all these “common-sense” relationships demonstrated,
commonplace. Assigning a single value to describe limb the value of the present research is to highlight the importance

Table 5: Dynamic Sitting Balance on Admission and Function at Discharge and 1 Year

Sitting Sitting Sitting Sitting Sitting Sitting
Balance Balance Balance Balance Balance Balance
Gr Imp Mild Imp Normal Gr Imp Mild Imp Normal
% Req % Req % Req % Req % Req % Req
2 2
Assist at Assist at Assist at Test, Assist at Assist at Assist at Test,
FIM Item Rehab DC (n) Rehab DC (n) Rehab DC (n) Cramer’s V 1 Year (n) 1 Year (n) 1 Year (n) Cramer’s V

Bed transfer 67.4 (417) 36.1 (723) 20.2 (885) 275.1, .37 15.0 (214) 6.5 (340) 1.3 (389) 44.0, .22
Toilet transfer 69.8 (417) 38.6 (725) 21.0 (884) 288.4, .38 14.5 (214) 7.4 (340) 1.3 (389) 40.4, .21
Tub transfer 79.6 (417) 52.1 (725) 35.3 (884) 224.0, .33 25.0 (212) 10.3 (339) 2.8 (389) 71.6, .28
Locomotion 37.8 (418) 14.5 (719) 8.0 (885) 186.2, .30 10.3 (213) 5.0 (340) 1.8 (387) 21.4, .15
Stairs 65.2 (417) 35.2 (722) 21.2 (882) 240.4, .35 20.6 (204) 8.4 (334) 2.9 (384) 52.3, .24
Eating 50.4 (419) 25.8 (726) 12.7 (885) 215.1, .33 16.4 (214) 7.9 (342) 1.3 (390) 48.2, .23
Grooming 60.1 (419) 36.9 (726) 22.1 (884) 182.1, .30 20.1 (214) 12.0 (342) 4.1 (390) 38.5, .20
Bathing 78.8 (419) 52.0 (725) 38.2 (884) 187.0, .30 27.6 (214) 14.3 (342) 4.9 (390) 61.6, .26
UB dressing 65.6 (419) 38.0 (726) 22.6 (884) 218.0, .33 21.0 (214) 11.4 (342) 3.1 (390) 49.8, .23
LB dressing 73.0 (419) 45.2 (726) 27.1 (885) 246.6, .35 22.0 (214) 12.3 (342) 4.1 (390) 45.4, .22
Toileting 61.4 (414) 32.6 (721) 20.2 (871) 214.5, .33 17.8 (214) 6.7 (342) 1.5 (390) 55.3, .24

NOTE. All analyses statistically significant at P .0001.
Abbreviations: Gr, grossly; Imp, impaired; Mild, mildly.



Table 6: Dynamic Standing Balance on Admission and Function at Discharge and 1 Year

Standing Standing Standing Standing Standing Standing
Balance Balance Balance Balance Balance Balance
Gr Imp Mild Imp Normal Gr Imp Mild Imp Normal
% Req % Req % Req % Req % Req % Req
2 2
Assist at Assist at Assist at Test, Assist at Assist at Assist at Test,
FIM Item Rehab DC (n) Rehab DC (n) Rehab DC (n) Cramer’s V 1 Year (n) 1 Year (n) 1 Year (n) Cramer’s V

Bed transfer 49.1 (678) 23.6 (762) 16.5 (340) 153.6, .29 11.0 (318) 3.3 (364) 0.7 (149) 27.2, .18
Toilet transfer 50.7 (681) 24.8 (761) 17.1 (340) 157.5, .29 10.7 (318) 3.8 (364) 0.7 (149) 23.3, .17
Tub transfer 64.0 (681) 38.4 (762) 30.9 (340) 137.1, .28 17.7 (316) 6.6 (363) 2.0 (149) 36.0, .21
Locomotion 23.7 (680) 8.7 (761) 6.5 (387) 86.8, .22 8.2 (319) 2.8 (361) 1.3 (149) 15.6, .14
Stairs 48.0 (680) 20.4 (761) 16.5 (339) 166.3, .31 14.0 (308) 4.8 (357) 2.7 (149) 26.1, .18
Eating 36.1 (684) 18.5 (761) 10.3 (340) 103.3, .24 12.5 (321) 3.6 (364) 0.7 (149) 32.4, .20
Grooming 46.1 (683) 28.2 (762) 20.3 (339) 84.6, .22 17.4 (321) 6.6 (364) 4.0 (149) 29.5, .19
Bathing 61.6 (683) 41.7 (761) 35.7 (339) 83.7, .22 19.6 (321) 9.6 (364) 3.3 (149) 29.3, .19
UB dressing 47.8 (683) 29.7 (762) 19.8 (339) 94.2, .23 15.6 (321) 6.9 (364) 1.3 (149) 28.9, .19
LB dressing 55.5 (683) 31.9 (762) 24.1 (340) 125.3, .27 17.4 (321) 7.7 (364) 2.0 (149) 31.1, .19
Toileting 42.0 (676) 24.3 (750) 17.8 (338) 82.6, .22 12.5 (321) 3.6 (364) 1.3 (149) 30.0, .19


of actively modifying a person’s rehabilitation program when a patients must have been entered into the Model System acute
specific disabling neurologic impairment is identified. What care facility within 24 hours after injury. Similarly, they all
focused interventions are activated based on the initial physical must have been admitted directly to inpatient rehabilitation at
findings by the physicians and rehabilitation team? Are the a Model System center after acute care discharge (ie, no
findings just “assumed” to be present in most patients and the patients may have returned home first or been transitioned to a
management of them already incorporated into rehabilitation subacute program). Additionally, individuals without some sig-
strategies, or are the rehabilitation programs specifically tai- nificant physical impairment who do not require or qualify for
lored to the established needs of these individuals? inpatient rehabilitation services are not included in the Model
This study identifies an important future area of investigation Systems database. The vast majority of the TBIMS are situated
in TBI rehabilitation: namely, the efficacy of specific rehabil- in large, urban, trauma centers. These inherent biases of the
itation interventions in improving functional capabilities. Al- research system provide a greater likelihood that the subjects
though it may be quite interesting for patients, families, and studied had more severe injuries, had more secondary injuries,
clinicians to know at the time of rehabilitation admission that, and lived closer to a major city than patients seen in suburban
based on screening physical examination, a person with TBI
can be identified as being more likely or less likely to have or rural hospitals not associated with major trauma centers.
increased short- and long-term disabilities, it is important to Further, in studies with long-term follow-up, selective attrition
know the specific strategies to overcome these initial impair- may exist in populations with TBI because of substance abuse
ments. The belief that more of a certain therapy type is the issues, socioeconomic status, violent etiologies, and severity of
appropriate intervention for a focal motor deficit—for example, motor deficits.27,28 Lost to follow-up rates of 40% at 1 year are
strengthening exercises in physical therapy (PT) for weak- not unusual in this population, and thus these data need to be
ness—is fairly naive, considering the tremendous complexity viewed carefully in light of all the confounding factors in-
involved in an injury to the brain and the great variability of volved in persons with TBI. Overall, however, the present
patient response. In fact, recent TBIMS research26 has shown investigation strongly suggests that simple physical examina-
that the intensity of speech and language pathology, as well as tion variables at rehabilitation admission are associated with
PT services, is associated with improved ambulation skills. relatively long-term functional outcome in self-care and mo-
More research identifying the value of therapy specificity, bility.
timing, and intensity in different subpopulations of people with With regard to the FIM ratings, no provision exists for
TBI is necessary to provide these answers. Unfortunately, coding the reason for needing assistance. Individuals with TBI
existing limitations in available research methods and the fund- may require supervision, such as a helper, to ensure their safety
ing mechanisms required for that research prevent a clearer either because they lack the physical capability to perform
understanding of these factors. Future studies are needed to test tasks independently or because they have cognitive or behav-
which methods are best for helping patients overcome their ioral deficits. Those needing supervision for cognitive reasons
swallowing, balance, and/or strength impairments in order to do not necessarily require physical assistance, but they may not
achieve self-care and mobility goals and to assess the impacts be safe being left alone. Further, an individual’s functional
these impairments have on handicap after TBI. status at the time of rehabilitation discharge may be signifi-
cantly influenced by his/her level of home supports, type of
Limitations reimbursement (with associated influence on LOS), and pres-
Although the present investigation used more than a decade ence of postacute therapy services, and thus be a “soft” end
of multicenter, prospectively collected, longitudinal data on point to the study. A more relevant end point would be the
more than 2300 persons with moderate to severe TBI, it has functional status at a specific time postinjury, which is captured
some inherent limitations. The generalizability of these find- in this study only at 1 year postinjury. Unfortunately, uniform
ings to non–TBIMS-based TBI patients is unclear. Because of time-interval data before 1 year after TBI are not available with
inclusion and exclusion criteria of the TBIMS program, all the current TBIMS database.



CONCLUSIONS 11. Corrigan JD, Smith-Knapp K, Granger CV. Validity of the func-
tional independence measure for persons with traumatic brain
This study contributes the largest existing dataset supporting injury. Arch Phys Med Rehabil 1997;78:828-34.
the association of physical examination characteristics on ad- 12. Wober C, Oder W, Kolleger H, et al. Posturographic measurement
mission to rehabilitation and the outcome at rehabilitation of body sway in survivors of severe closed head injury. Arch Phys
discharge and 1-year follow-up. Upper-extremity strength is Med Rehabil 1993;74:1151-6.
more predictive of FIM self-care items and lower-extremity 13. Englander J, Cifu DX, Wright JM, Black K. The association of
strength of FIM mobility items. Sitting and standing balance early computed tomography scan findings and ambulation, self-
care, and supervision needs at rehabilitation discharge and at 1
ratings on admission to rehabilitation had a more global effect year after traumatic brain injury. Arch Phys Med Rehabil 2002;
on discharge and 1-year FIM scores. Impaired swallowing on 84:214-20.
admission to rehabilitation was also an important predictor of 14. Traumatic Brain Injury Model Systems National Database Sylla-
ongoing need for assistance with eating on discharge from bus. E Orange (NJ): Kessler Institute of Rehabilitation; 1999-
rehabilitation and 1-year follow-up. By using this information, 2000.
clinicians should initiate interventions that optimize rehabilita- 15. Marx RG, Bombardier C, Wright JG. What do we know about the
reliability and validity of physical examination tests used to ex-
tion of the physical limitations identified, make adjustments to amine the upper extremity? J Hand Surg [Am] 1999;24:185-93.
LOS, provide appropriate therapy intensities and treatment 16. Ramsey DJ, Smithard DG, Kalra L. Early assessments of dyspha-
settings to accommodate greater disability, and arrange for gia and aspiration risk in acute stroke patients. Stroke 2003;34:
necessary aftercare and family supports. This growing body of 1252-7.
literature on outcome predictors may assist clinicians in fore- 17. Kinzey SJ, Armstrong CW. The reliability of the star-excursion
casting early the level of assistance in specific functional areas test in assessing dynamic balance. J Orthop Sports Phys Ther
1998;27:356-60.
that a person with TBI will need over time. 18. Clark S, Rose DJ. Evaluation of dynamic balance among commu-
nity-dwelling older adults fallers: a generalizability study of the
References limits of stability test. Arch Phys Med Rehabil 2001;82:468-74.
1. Englander JS, Cifu DX, Wright J, et al. The impact of acute 19. Hall KM, Cohen ME, Wright J, Call M, Werner P. Characteristics
complications, fractures and motor deficits on functional out- of the Functional Independence Measure in traumatic spinal cord
comes and length of stay after traumatic brain injury: a multicenter injury. Arch Phys Med Rehabil 1999;80:1471-6.
analysis. J Head Trauma Rehabil 1996;11(5):15-26. 20. Hall KM, Bushnik T, Lakisic-Kazazic B, Wright J, Cantagallo A.
2. Greenwald BD, Cifu DX, Marwitz JH, et al. Factors associated Assessing traumatic brain injury outcome measures for long-term
with balance deficits on admission to rehabilitation after traumatic follow-up of community-based individuals. Arch Phys Med Re-
brain injury: a multicenter analysis. J Head Trauma Rehabil 2001; habil 2001;82:367-74.
16:238-52. 21. Albanese MA, Clark WR, Adams HP Jr, Woolson RF. Ensuring
3. Juneja G, Czyrny JJ, Linn RT. Admission balance and outcomes reliability of outcome measures in multicenter clinical trials of
of patients admitted for acute inpatient rehabilitation. Am J Phys treatments of acute ischemic stroke. The program developed for
Med Rehabil 1998;77:388-93. the Trial of Org 10172 in Acute Stroke Treatment (TOAST).
4. Nichols DS, Miller L, Colby LA, Pease WS. Sitting balance: its Stroke 1994;25:1746-51.
relation to function in individuals with hemiparesis. Arch Phys 22. Goldstein LB, Bertels C, Davis JN. Interrater reliability of the
Med Rehabil 1996;77:865-9. NIH stroke scale. Arch Neurol 1989;46:660-2.
5. Wade LD, Canning CG, Fowler V, Felmingham KL, Baguley IJ. 23. Wityk RJ, Pessin MS, Kaplan RF, Caplan LR. Serial assessment
Changes in postural sway and performance of functional tasks of acute stroke using the NIH stroke scale. Stroke 1994;25:362-5.
during rehabilitation after traumatic brain injury. Arch Phys Med 24. Berg KO, Wood-Dauphinee SL, Williams JI, Maki B. Measuring
Rehabil 1997;78:1107-11. balance in the elderly: validation of an instrument. Can J Public
6. Keenan MA, Perry J, Jordan C. Factors affecting balance and Health 1992;83(Suppl 2):S7-11.
ambulation following stroke. Clin Orthop 1984;Jan-Feb(182):165- 25. Berg K, Wood-Dauphinee S, Williams JI. The balance scale:
71. reliability assessment with elderly residents and patients with an
7. Cifu DX, Keyser-Marcus L, Lopez E, et al. Acute predictors of acute stroke. Scand J Rehabil Med 1995;27:27-36.
successful return to work 1 year after traumatic brain injury: a 26. Cifu D, Kreutzer J, Kolakowsky-Hayner S, Marwitz J, Englander
multicenter analysis. Arch Phys Med Rehabil 1997;78:125-31. J. The relationship between therapy intensity and rehabilitative
8. Desrosiers J, Malouin F, Bourbonnais D, Richards CL, Rochette outcomes after traumatic brain injury: a multicenter analysis. Arch
A, Bravo G. Arm and leg impairments and disabilities after stroke Phys Med Rehabil 2003;84:1433-40.
rehabilitation: relation to handicap. Clin Rehabil 2003;17:666-73. 27. Corrigan JD, Harrison-Felix C, Bogner J, et al. Systematic bias in
9. Black K, Zafonte R, Millis S, et al. Sitting balance following brain traumatic brain injury outcome studies because of loss to follow-
injury: does it predict outcome? Brain Inj 2000;14:141-52. up. Arch Phys Med Rehabil 2003;84:153-60.
10. Guide for the Uniform Data Set for Medical Rehabilitation (in- 28. Corrigan JD, Bogner J, Mysiw WJ, Clinchot D, Fugate L. Sys-
cluding the FIM instrument), version 5.1. Buffalo: State Univ tematic bias in outcome studies of persons with traumatic brain
New York; 1997. injury. Arch Phys Med Rehabil 1997;78:132-7.


Strength, balance and swallowing deficits and outcome after trauma brain injury

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Viewers also liked

Viewers also liked (9)

Similar to Strength, balance and swallowing deficits and outcome after trauma brain injury

Similar to Strength, balance and swallowing deficits and outcome after trauma brain injury (20)

More from Connie Dello Buono

More from Connie Dello Buono (20)

Recently uploaded

Recently uploaded (20)

Strength, balance and swallowing deficits and outcome after trauma brain injury