1357

The Relation Between Impairments and Functional

Outcomes Poststroke
Atul T. Patel, MD, Pamela W. Duncan, PhD, PT, Sue-Min Lai, PhD, Stephanie Studenski, MD, MPH
ABSTRACT. Patel AT, Duncan PW, Lai S-M, Studenski S.
Relation between impairments and functional outcomes post-
stroke. Arch Phys Med Rehabil 2000;81:1357-63.
T HE TIME COURSE and extent of recovery of neurologic
impairment and functional abilities poststroke is critical for
treatment planning and prognosis but is not well understood.
Several investigators studied the outcome and/or time of
Objective: To assess the influence of initial stroke impair- recovery poststroke1-10 but either did not study the entire time
ments on the severity of basic and higher level functional course of recovery, failed to assess the breadth of relevant
deficits over time and to determine the cumulative impact on functional abilities, or did not account for the presence of
functional deficits beyond severity of motor deficits alone. multiple neurologic impairments.
Design: Observational study. Impairments accumulating across systems worsen prognosis
Setting: Twelve participating hospitals in the Greater Kansas for many generic conditions and affect functional outcomes in
City area, as part of the Kansas City Stroke Study (October stroke, measured by basic activities of daily living (ADLs).11
1995–March 1998). Initial stroke severity clearly influences recovery and initial
Participants: Individuals (n ⫽ 459) who sustained an functional status and is a strong predictor of functional status at
eligible stroke were evaluated prospectively using standardized discharge. A model of accumulating impairments in stroke may
assessments at enrollment (within 14 days of stroke onset, help predict a broader range of functional outcomes. This
8.8 ⫾ 3.5 days). predictive ability is best examined in the context of the full-time
Main Outcome Measures: Mobility and activities of daily course of recovery from stroke.
living (ADLs) were assessed at 1, 3, and 6 months poststroke The purpose of the present study is to assess the influence of
using the Functional Independence MeasureTM, Barthel index, accumulated stroke impairments on the basic ADL, instrumen-
Lawton Instrumental Activities of Daily Living (IADL), and the tal activities of daily living (IADL), and higher level functional
Medical Outcomes Study Short-Form Health Survey instru- mobility (Medical Outcomes Study Short-Form Health Survey
ments. [SF-36]). We modified the classification scheme of Reding and
Results: The cumulative probability of achieving indepen- Potes11 (motor deficit only, motor deficit plus somatosensory
dence with walking, a Barthel index of 60 or greater or 90 or deficit, motor deficit plus somatosensory deficit plus homony-
greater, and independence in 3 or more IADL was significantly mous visual deficit). We used consistent times to assess
different for the following 4 impairment groups in descending recovery: within 14 days and 1, 3, and 6 months poststroke.
order: motor; motor and somatosensory; motor and hemiano-
pia; and motor, sensory, and hemianopia. Although motor
SUBJECTS AND METHODS
severity was a strong predictor of outcome (p ⬍ .0001), the
additional somatosensory and hemianopia deficits significantly
Subjects
(p ⬍ .05) affected time and likelihood of achieving these levels
of function. The participants were 459 individuals who sustained an
Conclusion: Cumulative deficits poststroke affect patients’ eligible stroke and were recruited for the Kansas City Stroke
functional outcome in the first 6 months poststroke beyond the Study.12 Case ascertainment for the Kansas City Stroke Study
effect of motor severity alone. started in October 1995 and ended in March 1998. The eligible
Key Words: Activities of daily living; Cerebrovascular study participants, recruited from 12 participating hospitals in
accident; Motor activity; Rehabilitation. the Greater Kansas City area, were identified by (1) review of
r 2000 by the American Congress of Rehabilitation Medi- daily admission records; (2) referrals from physicians, clinical
cine and the American Academy of Physical Medicine and nurse specialists, or therapists on medical, neurology, and
Rehabilitation rehabilitation units; and (3) review of discharge codes.
To be accepted, patients had to have a confirmed eligible
stroke, defined by the World Health Organization13 as ‘‘rapid
onset and of vascular origin reflecting a focal disturbance of
cerebral function, excluding isolated impairments of higher
From the Departments of Rehabilitation Medicine (Patel), Preventive Medicine function and persisting longer than 24 hours.’’ The stroke was
(Lai), and Internal Medicine (Studenski), and the Center on Aging (Patel, Duncan, confirmed by clinical assessment and/or computed tomography
Studenski), University of Kansas Medical Center; and the Veterans Affairs Medical or magnetic resonance imaging. Trained nurses or physical
Center (Patel, Duncan), Kansas City, MO.
Accepted in revised form January 14, 2000.
therapists reviewed medical records and interviewed both
Supported in part by the Department of Veterans Affairs Rehabilitative Research and patients and physicians to determine eligibility before obtaining
Development (grant no. E879RC0) and Glaxo-Wellcome Pharmaceuticals. consent for enrollment. Patients were excluded if they were (1)
Preliminary data presented in a poster session at the American Geriatrics Society younger than 18 years of age; (2) had stroke onset since more
meeting, Seattle, WA, May 8, 1998.
No commercial party having a direct financial interest in the results of the research
than 14 days; (3) had stroke caused by subarachnoid hemor-
supporting this article has or will confer a benefit upon the authors or upon any rhage; (4) had hepatic failure documented by medical chart; (5)
organization with which the authors are associated. had renal failure documented by medical chart or were undergo-
Reprint requests to Atul T. Patel, MD, Department of Rehabilitation Medicine, ing dialysis; (6) had New York Heart Association functional
University of Kansas Medical Center, 3901 Rainbow Blvd, Kansas City, Kansas
66160, e-mail: [email protected].
grade III or IV heart failure (ie, patients with cardiac disease
0003-9993/00/8110-5805$3.00/0 resulting in inability or marked limitation to perform physical
doi:10.1053/apmr.2000.9397 activity without discomfort); (7) were not expected to live more

Arch Phys Med Rehabil Vol 81, October 2000

1358 IMPAIRMENTS FUNCTIONAL OUTCOME AND POSTSTROKE, Patel

than 6 months; (8) lived in a nursing home before the stroke; (9) Sensory and visual function were assessed using components
were unable to take care of their own affairs prestroke; (10) of the 14-item NIHSS index. We used 1 item to assess visual
were lethargic, obtunded, or comatose; and (11) lived more than function and 1 item to assess sensory function; each was
70 miles from the participating hospital. dichotomized to report the presence or absence of deficit.
Accumulated deficits were used to form 4 impairment
Procedures groups: motor deficits only (M); motor and somatosensory
The patients were evaluated by standardized assessments at deficits (MS); motor and hemianopia (MH); and motor, somato-
enrollment and followed up at 1, 3, and 6 months after stroke by sensory, and hemianopia (MSH).
health care professionals who were nurses or physical thera- Outcome variables: functional abilities. These variables
pists. Follow-up assessment was conducted at home or at an were measured within 14 days and at 1, 3, and 6 months
extended care facility. Each study nurse or physical therapist poststroke. The main outcomes measured were mobility and
received at least 2 weeks of training in the administration of the ADL using the FIM walking score, Barthel index,17 IADL,18
measures and were certified in the administration of the US and SF-36.19 The FIM walking score (subset of the FIM) was
National Institutes of Health Stroke Scale (NIHSS)14 and used to assess independent locomotion with or without an
Functional Independence MeasureTM (FIM).15 Assessments in- assistive device (score ⱖ6). The Barthel score was used to
cluded baseline demographics and self-reported poststroke assess ADL. We used cutoff scores of 60 (dependence) and 90
physical function, stroke characteristics, and the administration points (independence).24 To assess such higher level activities
of other stroke-related assessment instruments. Measurements as using the phone, shopping for groceries, or preparing meals,
were performed at baseline (within 14 days of stroke onset) and we used the IADL. For the purpose of the present study, we
1, 3, and 6 months poststroke. For the present study, the
observed whether the patient was independent in 3 or more of
following validated assessment tools and instruments were
the 9 IADL. Higher levels of functional mobility were assessed
used: Fugl-Meyer,16 NIHSS,14 FIM,15 Barthel index,17 IADL,18
and the SF-36 Physical Function Index (PFI)19 (question 3). by means of the SF-36 PFI (question 3), which addresses such
vigorous activities as running and such moderate activities as
Measures moving a table or pushing a vacuum cleaner; lifting or carrying
Predictor variables: criteria for individual and accumu- groceries; climbing several flights of stairs; climbing 1 flight of
lated impairments. Individual deficits in motor, sensory, and stairs; bending, kneeling, or stooping; walking more than a
visual (hemianopia) function were assessed. All measures were mile; walking several blocks; walking 1 block; and bathing or
obtained at baseline. Motor function was assessed using the dressing. We assessed the probability of individuals achieving
Fugl-Meyer motor score,16 which is a 100-point scoring system 80% of their prestroke physical function.
that includes both upper and lower extremities. For the purpose Adjustment variables: stroke severity and hemisphere in-
of this study, we only considered the lower extremity portions volved. We adjusted for stroke severity by using the total
of the Fugl-Meyer index (total, 34 points). This instrument Fugl-Meyer motor score. Analysis was performed to assess for
provides a reliable and valid measure of specific motor function the effects of cumulative deficits versus stroke severity in
that is also sensitive to change.20 Motor recovery of upper and predicting outcomes. We adjusted for missing Fugl-Meyer
lower extremity weakness after stroke has been reported to be scores. If more than 8 items were missing, the subject was
similar21; however, functional improvement (ie, level of disabil- excluded from this analysis. If a subject had fewer than 8 items
ity) is most dependent on lower extremity motor function.22 We missing, we assigned an average score based on the items
used a cutoff score of 28 points to differentiate patients with available. Ninety percent (38 of 42) of the subjects with missing
significant motor involvement, which identified the subjects information had 5 or fewer items missing. We also adjusted for
with less than normal power for hip flexion, measured by the location of stroke; ie, right versus left hemisphere. This analysis
Motricity index.23 excluded those patients with a brainstem or cerebellar stroke.

Table 1: Patient Demographics and Stroke Characteristics

All Patients M MS MH MSH
Variables (n ⫽ 360) (n ⫽ 122) (n ⫽ 111) (n ⫽ 35) (n ⫽ 92)

Age (yr) 70.4 ⫾ 11.4 68.4 ⫾ 10.8 69.2 ⫾ 10.7 70.7 ⫾ 11.2 74.3 ⫾ 12.2
Men (%) 45.6 45.1 56.8 40 34.8
Ethnicity
White (%) 78 78 75 77 84
Black (%) 18 17 22 17 13
Other (%) 4 5 3 6 3
Time since onset (days) 8.8 ⫾ 3.5 8.8 ⫾ 3.8 8.3 ⫾ 3.4 9.4 ⫾ 3.2 9.3 ⫾ 3.4
Prestroke functional status*
Prior Barthel score 97 ⫾ 7 98 ⫾ 5.7 97 ⫾ 7.9 96 ⫾ 6 96 ⫾ 7.5
SF-36 PFI 68 ⫾ 28.5 72 ⫾ 27.4 70 ⫾ 27.9 57 ⫾ 27.5 65 ⫾ 30
Ischemic stroke (%) 93.9 97.5 90.1 94.3 93.5
Stroke location
Right (%) 45.8 37.7 46.8 40 57.6
Left (%) 47.2 51.6 46.8 48.6 41.3
Brainstem (%) 6.7 8.2 6.3 11.4 3.3
Cerebellar (%) 3.1 4.1 1.8 2.9 3.3
Values expressed as mean ⫾ SD.
* Self-reported.

Arch Phys Med Rehabil Vol 81, October 2000

 IMPAIRMENTS FUNCTIONAL OUTCOME AND POSTSTROKE, Patel 1359

Table 2: Stroke Outcomes by Impairment Groups

M MS MH MSH
Outcome n Mean ⫾ SD n Mean ⫾ SD n Mean ⫾ SD n Mean ⫾ SD

Barthel
1 Month 118 76 ⫾ 23.4 108 71 ⫾ 28.3 32 61 ⫾ 29 86 44 ⫾ 33.2
3 Month 110 86 ⫾ 18.5 99 82 ⫾ 22.5 31 70 ⫾ 29.2 72 55 ⫾ 35.4
6 Month 104 87 ⫾ 20.3 95 84 ⫾ 21.8 27 74 ⫾ 29.3 64 61 ⫾ 35.7
IADL
1 Month 118 17 ⫾ 4.8 107 18 ⫾ 4.5 31 16 ⫾ 4 86 13 ⫾ 4.3
3 Month 109 20 ⫾ 5 99 20 ⫾ 5 31 17 ⫾ 4.5 71 15 ⫾ 5
6 Month 104 21 ⫾ 5 95 20 ⫾ 5.3 27 17 ⫾ 5.2 65 15 ⫾ 5.3
SF-36 PFI
1 Month 113 31 ⫾ 25.7 107 30 ⫾ 26.3 28 24 ⫾ 26.9 60 19 ⫾ 20.7
3 Month 107 45 ⫾ 28.2 97 38 ⫾ 28.3 27 35 ⫾ 33 51 27 ⫾ 26.6
6 Month 100 46 ⫾ 27 92 41 ⫾ 28.7 24 38 ⫾ 28.1 45 32 ⫾ 31.2

Statistical Analysis and were included in the present analysis. All subjects were
Descriptive statistics were used to show demographics, community dwelling before their strokes. Subject demograph-
prestroke functional status, stroke characteristics, severity of ics and associated stroke characteristics are listed in table 1.
impairment caused by stroke, and scores on the Barthel index, Eighty-three patients (23%) had a history of prior stroke or
IADL, and SF-36 PFI measured at 1, 3, and 6 months transient ischemic attack; however, all subjects were living at
poststroke. Kaplan-Meier survival curves using the various home and independent in ADLs before admission to the
thresholds for functional status for the 4 impairment groups hospital.
were compared using the log-rank test. Finally, the likelihood of Descriptive statistics on baseline stroke severity measured
achieving 80% of the premorbid SF-36 PFI score was deter- within 14 days poststroke (mean ⫾ SD, 8.8 ⫾ 3.5 days; range,
mined for the 4 groups by means of Kaplan-Meier estimates of 3–14 days) and subjects’ prestroke functional scores are also
cumulative probabilities. The Cox proportional hazards model listed in table 1. Table 2 lists outcome measures for each of the 4
was used to estimate the risk ratio after controlling for the impairment groups (M, MS, MH, MSH).
Fugl-Meyer lower extremity subscore. Figures 1 and 2 show the cumulative probabilities of
achieving a Barthel index of 60 or greater and 90 or greater for
RESULTS the 4 impairment groups over a period of 6 months. Figure 3
Of the 459 subjects enrolled in the Kansas City Stroke Study, shows the cumulative probability of achieving independence
28 subjects died, 10 moved out of town, 30 refused follow-up with walking (FIM, 6 or 7) for the 4 groups. Figure 4 shows the
assessment, and 1 subject was lost to follow-up for unknown cumulative probability of achieving independence in 3 or more
reasons. Of the remaining 390 patients, 360 met the criteria of at ADL, defined by the IADL, for the 4 groups over the 6-month
least some motor deficit (hemiparesis, Fugl-Meyer score ⬍ 28) period. The cumulative probabilities of achieving these scores

Fig 1. Kaplan-Meier estimates

of cumulative probability of
achieving a Barthel index of 60
or greater according to impair-
ment group.

Arch Phys Med Rehabil Vol 81, October 2000

1360 IMPAIRMENTS FUNCTIONAL OUTCOME AND POSTSTROKE, Patel

Fig 2. Kaplan-Meier estimates

of cumulative probability of
achieving a Barthel index of 90
or greater according to impair-
ment group.

for the 4 impairment groups are significantly different (p ⬍ as likely at 6 months to achieve a Barthel index of at least 60
.0001). compared with the M group.
Table 3 lists the likelihoods of achieving the specified Figure 5 shows the likelihood of reaching a score of 80% or
functional score when controlling for severity of lower extrem- greater of prestroke physical function for the 4 impairment
ity motor function. The motor deficit group (M) was the groups during the 6-month period. Less than 50% of the M
reference group. Although motor severity was a strong predic- group, 40% of the MH and MS groups, and less than 20% of the
tor of outcome (p ⬍ .0001), the additional somatosensory and MSH group achieved an SF-36 PFI score of 80% or greater of
hemianopia deficits significantly affected time and likelihood of their prestroke physical function.
achieving certain functions. For example, the group with motor, When comparing the side of hemispheric involvement and
somatosensory, and hemianopia deficits (MSH) was 0.3 times controlling for motor severity, we found no significant differ-

Fig 3. Kaplan-Meier estimates

of cumulative probability of
achieving independence with
walking (FIM, 6 or 7) according
to impairment group.

Arch Phys Med Rehabil Vol 81, October 2000

 IMPAIRMENTS FUNCTIONAL OUTCOME AND POSTSTROKE, Patel 1361

Fig 4. Kaplan-Meier estimates

of cumulative probability of
achieving independence in 3
or more ADL, defined by the
IADL, according to impairment
group.

ence in the likelihood of achieving the previously mentioned Jongbloed,8 reviewing 33 studies and summarizing factors
functional outcomes. that predict final walking ability and self-care functions,
identified several limitations, such as differences in patient
DISCUSSION samples, timing of assessments, criteria by which outcomes
Predicting the extent and time course of stroke recovery is were measured, and measuring instrument used. More recent
complex; however, this information is extremely important in studies have focused on specific impairments and functional
the rehabilitation of stroke patients. Newer changes in health deficits but are limited by the small numbers of subjects,
care, including pressures to reduce costs, demand an ability to variable timing of entry onto the study, and variable time of
predict outcomes early poststroke to help place patients in follow-up.6,9,10 Jorgensen et al1 found that timing and degree of
appropriate settings. We found that specific impairments help walking function recovery in stroke patients could be predicted
predict outcome beyond the effect of stroke severity alone. based on the initial impairment of walking and leg weakness.
Other studies have assessed stroke severity and outcomes, Subsequent studies suggest that initial status determined by the
but they have several limitations. Severity of stroke,4,25 severity Scandinavian Neurological Stroke Scale and the Barthel index
of hemiparesis, and presence of such deficits as hemianopsia can predict the time course of neurologic and functional
and sensory loss have been associated with less favorable recovery poststroke. Global impairment measures rather than
outcomes1-3,26-29; however, none of these studies addressed specific impairment scores were used as predictors, and informa-
cumulative deficits and controlled for stroke severity. tion about higher level ADL was not provided.
Reding and Potes11 used the approach of categorizing
Table 3: Likelihood of Achieving Specified Functional Scores When unilateral hemispheric stroke patients into subgroups based on
Controlling for Motor Impairment Severity
their motor, somatosensory, and visual field deficits. They
Impairment Risk showed that recovery was significantly different among the
Category Ratio* 95% CI p subgroups; however, their study was limited by variable time of
Barthel ⱖ60 MS 0.6 .40-.95 .028 entry, brief follow-up to rehabilitation discharge only, use of
MH 0.5 .24-.89 .012 limited functional outcomes, and failure to control for stroke
MSH 0.3 .16-.42 .0001 severity.
Barthel ⱖ90 MS 0.7 .47-1.06 .0906 In the present study, we used fixed poststroke time points
MH 0.3 .17-.64 .001 rather than the time around hospitalization or rehabilitation to
MSH 0.4 .24-.63 .0003 assess outcomes and functional status. Compared with other
FIM ⱖ6 MS 0.7 .48-1.05 .089 studies, we assessed a greater range of outcomes, from basic to
MH 0.4 .22-.75 .004 higher levels of physical function, using various validated
MSH 0.3 .21-.59 .0001 instruments (ADL, IADL, SF-36 PFI).
IADL ⱖ3 MS 0.9 .60-1.33 .575 The results of this study showed that cumulative deficits and
MH 0.6 .35-1.11 .108 severity of the motor deficits make a difference in the likelihood
MSH 0.5 .30-.73 .0009 of reaching a particular functional level. It also showed that
Abbreviation: CI, confidence interval.
deficits in somatosensory function and/or hemianopia affect the
* Risk ratio was calculated after adjusting for the Fugl-Meyer lower likelihood of achieving specified functional scores when control-
extremity subscore. ling for the severity of stroke assessed by motor deficits.

Arch Phys Med Rehabil Vol 81, October 2000

1362 IMPAIRMENTS FUNCTIONAL OUTCOME AND POSTSTROKE, Patel

Fig 5. Kaplan-Meier estimates

of cumulative probability of
achieving a score of 80% or
greater of prestroke physical
function during the 6-month
period according to impair-
ment group.

Our analysis did not show a significant difference in func- improvement by 6 months is very low, and appropriate long-
tional outcomes based on the side of hemispheric involvement. term placement plans can be instituted. This information
Previous research is controversial in this regard.4,30-34 Discrep- provides useful evidence for establishing clinical pathways and
ancies between reports may be caused by differences in the allocating and timing rehabilitation services.
timing of assessment, population studied, or methods. The Results in this heterogeneous stroke population suggest that
discriminating factor may be the actual impairments, such as rehabilitation interventions should be based more on impair-
neglect, hemianopsia, or aphasia, rather than the side of ments poststroke than on the diagnosis of stroke itself or the
hemispheric involvement. side of lesion location. Rehabilitation services could also be
An important finding of this study is that poststroke assess- individualized to serve the time course of recovery most
ments must consider a hierarchy of functional outcomes. The appropriate for the expected functional outcomes.
choice of the appropriate instrument may depend not only on
the function being assessed, but also on the timing of the CONCLUSIONS
assessment. For example, when assessing the probability of Beyond the initial stroke, severity of cumulative deficits
achieving a Barthel score of at least 90 (fig 2), little discrimina- affects the patient’s functional outcome at 1, 3, and 6 months.
tion exists at 1 month poststroke between patients with a motor Furthermore, somatosensory deficits and/or hemianopia impact
deficit only (M) and patients with motor and somatosensory on the likelihood of achieving specified functional outcomes, as
deficits (MS), but discrimination exists at 6 months. The other the present study shows by controlling for motor severity. This
problem to be considered is ceiling and floor effects. When we study highlights the need for better assessment of function
assessed the likelihood of reaching 80% of prestroke function across different levels of severity. Such assessment can help
measured by the SF-36 instrument, we found that even in the evaluate the effect of various rehabilitation interventions. Better
group with the best outcomes (M group), less than 50% understanding of impairments and outcomes can help establish
achieved this level at 6 months. Instruments with broad ranges clinical pathways and facilitate better allocation and timing of
of discrimination are needed to detect changes over time. rehabilitation services.
This study also more accurately estimates the likelihood and
timing of recovery based on initial impairment. This informa- Acknowledgments: Participating facilities in the Greater Kansas
tion can help determine the appropriate placement and timing of City area include Baptist Medical Center, Department of Veterans
rehabilitation efforts for different groups of patients. For Affairs Medical Centers at Kansas City and Leavenworth, Liberty
example, this study showed probabilities of achieving a Barthel Hospital, Medical Center of Independence, Mid-American Rehabilita-
score of 60 or greater (able to go home with help) for the group tion Hospital, Rehabilitation Institute, Research Medical Center, St
with motor, somatosensory, and hemianopia deficits (MSH) at Joseph Health Center, St Luke’s Hospital, Trinity Lutheran Hospital,
and the University of Kansas Medical Center.
1, 3, and 6 months of 35%, 50%, and 52%, respectively. It may
be appropriate to place a patient with these deficits in a less
intense rehabilitation setting initially (once medically stabi- References
1. Jorgensen HS, Nakayama H, Raaschou HO, Olsen TS. Recovery
lized) to monitor recovery and prevent complications and of walking function in stroke patients: The Copenhagen Stroke
conserve resources. If significant recovery occurs by 1 to 3 Study. Arch Phys Med Rehabil 1995;76:27-32.
months poststroke, the patient can be transitioned into a more 2. Jorgensen HS, Nakayama H, Raaschou HO, Olsen TS. Outcome and
intense rehabilitation program. If the patient does not improve time course of recovery in stroke. Part I: outcome. The Copenhagen
much by 3 months poststroke, the probability of significant Stroke Study. Arch Phys Med Rehabil 1995;76:399-405.

Arch Phys Med Rehabil Vol 81, October 2000

 IMPAIRMENTS FUNCTIONAL OUTCOME AND POSTSTROKE, Patel 1363

3. Jorgensen HS, Nakayama H, Raaschou HO, Vive-Larsen J, Stoier 19. Ware JE, Sherbourne CD. The MOS 36-Item short-form health
M, Olsen TS. Outcome and time course of recovery in stroke. Part survey (SF-36): I. Conceptual framework and item selection. Med
II: time course of recovery. The Copenhagen Stroke Study. Arch Care 1992;30:473-83.
Phys Med Rehabil 1995;76:406-12. 20. Duncan PW, Propst M, Nelson SG. Reliability of the Fugl-Meyer
4. Ween JE, Alexander MP, D’Esposito M, Roberts M. Factors assessment of the sensorimotor recovery following cerebrovascu-
predictive of stroke outcome in a rehabilitation setting. Neurology lar accident. Phys Ther 1983;63:1607-10.
1996;47:388-92. 21. Duncan PW, Goldstein LB, Horner RD, Landsman PB, Samsa GP,
5. Partridge CJ, Johnston M, Edwards S. Recovery from physical Matchar DB. Similar motor recovery of upper and lower extremi-
disability after stroke: normal patterns as a basis for evaluation. ties after stroke. Stroke 1994;25:1181-8.
Lancet 1987;1:373-5. 22. Olson TS. Arm and leg paresis as outcome predictors in stroke
6. Loewen SC, Anderson BA. Predictors of stroke outcome using rehabilitation. Stroke 1990;21:247-51.
objective measurement scales. Stroke 1990;21:78-81. 23. Collin C, Wade D. Assessing motor impairment after stroke: a pilot
7. Kalra L, Perez I, Gupta S, Wittink M. The influence of visual reliability study. J Neurol Neurosurg Psychiatry 1990;53:576-9.
neglect on stroke rehabilitation. Stroke 1997;28:1386-91. 24. Granger CV, Dewis LS, Peters NC, Sherwood CC, Barrett JE.
8. Jongbloed L. Prediction of function after stroke: a critical review. Stroke rehabilitation: analysis of repeated Barthel index measures.
Stroke 1986;17:765-76. Arch Phys Med Rehabil 1979;60:14-7.
9. Sandin KJ, Smith BS. The measure of balance in sitting in stroke 25. Dove HG, Schneider KC, Wallace JD. Evaluating and predicting
rehabilitation prognosis. Stroke 1990;21:82-6. outcome of acute cerebral vascular accident. Stroke 1984;15:
10. Franchignoni FP, Tesio L, Ricupero C, Martino MT. Trunk control 858-64.
test as an early predictor of stroke rehabilitation outcome. Stroke 26. Kaplan J, Hier D. Visuospatial deficits after right hemisphere
1997;28:1382-5.
stroke. Am J Occup Ther 1982;36:314-21.
11. Reding MJ, Potes E. Rehabilitation outcome following initial
27. Wade DT, Hewer RL, Wood VA. Stroke: influence of patient’s sex
unilateral hemispheric stroke. Life table analysis approach. Stroke
1988;19:1354-8. and side of weakness on outcome. Arch Phys Med Rehabil
12. Lai SM, Duncan PW, Keighly J. Prediction of functional outcome 1984;65:513-6.
after stroke. Comparison of the Orpington Prognostic Scale and the 28. Feigenson JS, McDowell FH, Meese P, McCarthy ML, Greenberg
NIH stroke scale. Stroke 1998 29:1838-42. SD. Factors influencing outcome and length of stay in a stroke
13. Report on the WHO Task Force on stroke and other cerebrovascu- rehabilitation unit. Stroke 1977;8:651-6.
lar disorders. Stroke-1989: recommendations on stroke prevention, 29. Stern PH, McDowell F, Miller SM, Robinson M. Factors influenc-
diagnosis and therapy. Stroke 1989;20:1407-31. ing stroke rehabilitation. Stroke 1971;3:213-8.
14. Brott T, Adams HP Jr, Olinger CP, Marler JR, Barsen WG, Biller J, 30. Ring H, Feder M, Schwartz J, Samuels G. Functional measures of
et al. Measurements of acute cerebral infarction: a clinical first-stroke rehabilitation patients: usefulness of the Functional
examination scale. Stroke 1989;20:864-70. Independence Measure total score with a clinical rationale. Arch
15. Keith RA, Granger CV, Hamilton BB, Sherwin FS. The functional Phys Med Rehabil 1997;78:630-5.
independence measure: a new tool for rehabilitation. In: Eisenberg 31. Macciocchi SN, Diamond PT, Alves WM, Mertz T. Ischemic
MG, Grzesiak RC, editors. Advances in clinical rehabilitation, stroke: relation of age, lesion location, and initial neurologic deficit
volume 1. New York: Springer-Verlag; 1987. p. 6-18. to functional outcome. Arch Phys Med Rehabil 1998;79:1255-7.
16. Fugl-Meyer AR, Jaasko L, Leyman I, Olsson S, Steglind S. The 32. Alexander MP. Stroke rehabilitation outcome: a potential use of
post-stroke hemiplegic patient: a method for evaluation of physical predictive variables to establish levels of care. Stroke 1994;25:
performance. Scand J Rehabil Med 1975;7:13-31. 128-34.
17. Mahoney FI, Barthel DW. The Barthel index. Md State Med J 33. Cifu DX, Lorish TR. Stroke rehabilitation. 5. Stroke outcome.
1965;14: 61-5. Arch Phys Med Rehabil 1994;75(Suppl):56-60.
18. Lawton MP. Assessing the competence of older people. In: Kent D, 34. Kotila M. Four year prognosis of patients under the age of 65
Kastenbaum R, Sherwood S, editors. Research planning and action surviving their first ischemic brain infarction. Ann Clin Res
for the elderly. New York: Behavioral Publications; 1972. 1986;18:76-9.

Arch Phys Med Rehabil Vol 81, October 2000