http://informahealthcare.

com/dre
ISSN 0963-8288 print/ISSN 1464-5165 online

Disabil Rehabil, 2014; 36(5): 353358

! 2014 Informa UK Ltd. DOI: 10.3109/09638288.2013.793411

RESEARCH PAPER

Long-term prediction of functional outcome after stroke using single

items of the Barthel Index at discharge from rehabilitation centre
Liesbet De Wit1,2, Koen Putman2,3, Hannes Devos1, Nadine Brinkmann4, Eddy Dejaeger5, Willy De Weerdt1,
Walter Jenni6, Nadina Lincoln7, Birgit Schuback6, and Wilfried Schupp4
1
Department of Rehabilitation Sciences, Faculty of Kinesiology and Rehabilitation Sciences, Katholieke Universiteit Leuven, Leuven, Belgium,
2
Department of Medical Sociology and Health Sciences, Faculty of Medicine and Pharmacy, Vrije Universiteit Brussel, Brussel, Belgium, 3School of
Health, University of Central Lancashire, Preston, Lancashire, United Kingdom, 4Fachklinik Herzogenaurach, Herzogenaurach, Germany, 5University
Hospital Pellenberg, Pellenberg, Belgium, 6Rehaclinic Zurzach, Zurzach, Switzerland, and 7Institute of Work, Health and Organisations, University of
Nottingham, Nottingham, United Kingdom

Abstract Keywords
Purpose: To determine the prognostic value of single items of the Barthel Index (BI) at discharge Activities of daily living, prognosis, stroke
from rehabilitation, in predicting independence in personal activities of daily living
(ADL) (BI score 95/100) at five years after stroke. Method: People with stroke were History
recruited consecutively from four European rehabilitation centres. BI was assessed on discharge
and at five years after stroke. Stepwise multivariate logistic regression analysis was used Received 4 December 2012
to determine independent predictors of BI score 95/100 at five years after stroke. Revised 28 March 2013
Thereupon, percentage chance of reaching BI  95/100 at five years after stroke was calculated. Accepted 3 April 2013
Results: Data were available for 153 patients. Independence in dressing (odds ratio (OR) 5.22, Published online 21 May 2013
95% confidence interval (CI) 1.8514.76, p 0.002) and bathing (OR 8.10, 95%
CI 3.4019.32, p50.0001) were independent predictors. Independence in both items resulted
in 74.1% (57.685.8) chance of reaching BI  95/100 at five years after stroke. Dependence in
both items resulted in 6.3% (5.17.9) chance. Independence in bathing, but dependence in
dressing resulted in 35.4% (30.740.4) chance whereas the opposite resulted in 26.1%
(20.732.3) chance. Conclusion: Simple assessment of dressing and bathing on discharge from
rehabilitation enables therapeutic staff to predict prognosis for long-term independence in
personal ADL. This method can be used for early identification of persons with stroke who need
intensive follow-up.

Implications for Rehabilitation

 (In)dependence for dressing and bathing at discharge from a rehabilitation centre are
significant factors in the prediction of (in)dependence in personal ADL at five years after
stroke.
 This predictive tool can be used for targeting inpatient stroke rehabilitation and early
identification of those patients who need intensive follow-up. 14
20

Introduction of others is of concern to patients, their relatives and to those

Discharge from rehabilitation is one of stroke patients main
desires but at the same time, patients experience ambivalent
feelings [1]. They feel relieved being able to return home, but are
also uncertain about the new life situation [2]. For health care
providers, discharge is an important time as it involves transfer to
another setting either with or without another type of care. The
possibility of being able to survive in the long term, independently
Address for correspondence: Liesbet De Wit, Department of Medical
Sociology and Health Sciences, Faculty of Medicine and Pharmacy, Vrije
Universiteit Brussel, Laarbeeklaan 103, 1090 Brussel, Belgium.
Tel: 0032/ (0)2 477 47 67. Fax: 0032/ (0)2 477 47 11. E-mail:
Liesbet.De.Wit@vub.ac.be
providing or planning health care [3].
Few studies have attempted to predict dependency beyond the
point of six months after stroke onset. Pettersen et al. found that a
Barthel Index (BI) score 515/20 on admission to a rehabilitation
centre was the strongest predictor of functional dependence
(BI  19/20) at three years after stroke [4]. Lin et al. reported that,
amongst other predictors, a poorer score on the functional
independence measure (FIM) at discharge from the rehabilitation
centre was independently associated with severe disability one
year after stroke [5]. In a community-based study, the initial
BI was an independent predictor of death or disability five years
after stroke [6]. In a recent study, Alexander et al. found that,
amongst others, a score 45 on the Alberta Stroke Program Early
Computed Tomography Score (ASPECTS) had significant
354 L. De Wit et al.
predictive value for greater functional independence at one year
after stroke [7].
The total score on both BI [8] and FIM [9] indicate patients
global, overall need for assistance in carrying out activities of
daily living (ADL). In a recent study, Gialanella et al. investigated
the influence of single ADL activities as possible predictors of
functional outcome [10]. The authors found that social inter-
action, problem solving, grooming and upper body dressing on
admission were the most important predictors of functional
outcome at discharge from the centre. No previous studies were
identified which have investigated whether single ADL activities
predict long-term functional outcomes after stroke. Identification
of single ADL items that are predictive for long-term functional
independence at discharge from the rehabilitation centre may help
in the discharge planning by early identification of persons who
are likely in need for intensive follow-up.
The aim of the study was to determine the prognostic value of
single BI items assessed at discharge from rehabilitation, in
predicting independence in personal ADL at five years after
stroke.
Methods
Patients and settings
This prospective rehabilitation cohort study was a follow-up of the
CERISE project (Collaborative Evaluation of Rehabilitation in
Stroke across Europe) [11]. This project compared stroke care and
recovery patterns between four European rehabilitation centres:
University Hospital, Leuven, Belgium (BE); Nottingham
University Hospitals, Nottingham, United Kingdom (UK);
RehaClinic Zurzach, Zurzach, Switzerland (CH) and Fachklinik
Herzogenaurach, Herzogenaurach, Germany (DE). In each centre,
inpatient multidisciplinary care was provided in a stroke rehabili-
tation unit. Between March 2002 and September 2004, patients
were recruited consecutively using the following inclusion
criteria: (1) first-ever stroke as defined by WHO [12]; (2) age
4085 years; and (3) Rivermead Motor Assessment scores [13]:
Gross Function (RMA-GF) 11 and/or Leg and Trunk function
(RMA-LT) 8 and/or Arm function (RMA-A) 12 on admission
to the centre. The exclusion criteria were: (1) other neurological
impairments with permanent damage; (2) stroke-like symptoms
attributable to subdural haematoma, tumour, encephalitis
or trauma; (3) admission to the centre 4six weeks after stroke;
(4) no informed consent; and (5) pre-stroke BI550.
Baseline assessment and follow-up
In the CERISE project, a trained researcher in each centre
collected the following data of participants on admission to the
rehabilitation centre: age, gender, centre (BE, UK, CH, DE),
stroke type (ischaemic infarct, haemorrhagic infarct, unknown)
and BI on admission (0100). The BI measures functional
disability by quantifying patients performance on 10 activities of
daily life. These activities are feeding (score: 0510), grooming
(score: 05), bathing (score: 05), dressing (score: 0510),
bowels (score: 0510), bladder care (score: 0510), toilet use
(score: 0510), ambulation (score: 051015), transfers (score: at five years after stroke was calculated. Using a cut-off of 0.50, the
051015) and stair climbing (score: 0510) [8]. The scores sensitivity, specificity and predictive accuracy of the regression
assigned to each activity are based on time and amount of actual
physical assistance required if a patient is unable to perform the
activity. Maximum score is not given for an activity if the patient
needs even minimal help and/or supervision, i.e. if he cannot
safely perform the activity without someone present in the room
with him [8].
At discharge, the length of stay (LOS) was recorded
and patients functional status was re-assessed using the BI. The
Disabil Rehabil, 2014; 36(5): 353358
assessments for the present five-year follow-up study were carried
out in the period 20082009. In this follow-up study, the Belgian
and Swiss researchers contacted the patients by telephone to
explain study protocol and invite them to participate. In line with
ethical requirements, the British and German researchers con-
tacted patients by letter and asked them to indicate whether they
were willing to take part. If patients agreed, an appointment was
made for a home visit, at which informed consent was obtained
and the BI was administered by interview. Before the study began,
all researchers attended a workshop to familiarise them with the
study protocol. The project manager (L.D.W.) visited each
researcher to ensure standardized implementation of the research
protocol. The study was approved by the relevant ethics commit-
tee in each country.
Statistical analysis
Age, gender and BI on admission and at discharge from the
rehabilitation centre were compared between those who dropped
out and those who completed the study. Patients characteristics
are presented in terms of means with SDs, medians with
interquartile ranges (IQR) or frequencies with percentages,
depending on the type of data. Each of the 10 BI items, assessed
at discharge from the rehabilitation centre, were dichotomized,
i.e. independent (maximum score) or dependent (5maximum
score). Patients were classified into two categories based on their
total BI score at five years after stroke. Patients who scored 95/
100 on the BI were considered independent in personal ADL;
those with a BI score 595/100 were considered dependent in
personal ADL. This cut-off is widely accepted and has been used
in previous studies [4,6].
First, the relationships between gender, centre, type of stroke
and independence at discharge for each BI item and BI  95/100
at five years after stroke were investigated using chi square tests.
A T test was used to determine whether a difference existed
between the mean age at stroke onset for the independent
(BI  95/100) and dependent (BI595/100) groups at five years
after stroke. Variables with p values 50.10 were subjected to
multivariate logistic regression analysis. A forward stepwise
selection procedure was used whereby p50.15 was used as the
entry and p50.05 was used as the retain criterion. Variables
retained in the final model were considered as independent
predictors of BI  95/100 at five years after stroke. From the
logistic regression equation, the percentage chance that patients
have a BI score 95/100 at five years after stroke can be
computed according to the following formula:
expb0 bn  x
p  100
1 expb0 bn  x
p probability to have a BI score 95/100 at five years
after stroke.
b0 intercept.
bn parameter coefficient.
Based on this formula, the chance of reaching BI score 95/100
equation was calculated [14]. All statistical procedures were
carried out using SAS version 9.2 (SAS Institute Inc., Cary, NC).
Results
Five hundred and thirty-two consecutive stroke patients were
included in the CERISE study. At the five year follow-up,
247 patients (46%) were assessed. Reasons for drop-out are
DOI: 10.3109/09638288.2013.793411 Predicting functional outcome after stroke 355
Figure 1. Flow diagram.
CERISE-study
n=532

no contact data: n=28 (5%)

died: n=167 (31%)
moved > 100km away: n=5 (0.9%)
refused: n=84 (16%)
medically unstable: n=1

five year after stroke onset

n=247 (46%)

German centre: n=55 Swiss centre: n=80

British centre: n=38 Belgian centre: n=74

presented in Figure 1. Of these, 80 patients had received Table 1. Demographic and clinical characteristics of
rehabilitation in the Swiss, 74 in the Belgian, 55 in the German patients (n 153).
and 38 in the British centre. This represented 59%, 58%, 41%
and 28% of the initial sample recruited in each rehabilitation Variables n (%)
centre, respectively. In the Belgian and Swiss centres, Female gender 76 (49.7)
we found no significant differences for age, gender, BI
Centre
on admission and at discharge between those who dropped BE 73 (47.7)
out and those who took part in the study. In the British and CH 80 (52.3)
German centres, no significant differences were found for Type of stroke
age, gender and initial BI. BI at discharge from the British and Haemorrhagic 30 (19.6)
German centres was significantly (p50.05) lower for those Ischaemic 119 (77.8)
who dropped out compared to those who completed the study. Unknown 4 (2.6)
Because of the high drop-out (550% of the initial German BI 5 years: n (%)
and British patient samples were assessed at five years 595/100 91 (59.5)
after stroke), we based all further analyses on data from the 95/100 62 (40.5)
Swiss and Belgian patients only. Data from the British and Variables Mean (SD)
German patients were used for external validation of the LOS (days) 61.0 (32.9)
prediction model. TSOA (days) 21.4 (9.1)
Variables Median (IQR)
From the initial 262 patients recruited (135 Swiss 127 BI admission 55 (3085)
Belgian patients), 79 (30%) died, 19 (7%) refused to participate,
5 (2%) could not be contacted, 5 (2%) moved more than 100 km TSOA indicates time between stroke onset and admission
away from the rehabilitation centre and 1 was medically unstable to the centre.
at time of five year follow-up. Hence, follow-up data were
available for 153 patients (Table 1). Mean age at stroke onset
was 67.8 years (SD 10.5). There were equal numbers of men
noticed at five years after stroke. The percentage of
(52.3%) and women (49.7%) and a slightly higher percentage of
patients independent for transfer, feeding, bowels, toilet
Swiss (52.3%) compared to Belgian (47.7%) patients. Thirty
use and bathing decreased by 510% between discharge and
patients (20%) suffered from a haemorrhagic stroke, 119 (78%) had
five years after stroke. A decrease of 410% was noted for the
an ischaemic stroke and for 4 (2%) exact diagnosis was unknown.
items dressing, bladder and grooming.
At five years after stroke, 62 patients (40.5%) scored 95/100
on the BI.
Predictors of BI score 95/100 at five years after stroke
Figure 2 presents the percentage of patients that were
independent on each BI item at discharge from the rehabilitation Significant univariate associations were found between centre and
centre and at five years after stroke onset. At discharge from independence on each BI item at discharge from the rehabilitation
the centre, the highest percentages of independent patients centre and BI score  95/100 at five years after stroke (Table 2).
were found for bowels (96.1% continent), grooming (90.8%) These 11 parameters were included in a multivariate logistic
and bladder (84.6% continent). The lowest percentages of regression analysis. Only independence in dressing (odds ratio
independent patients were found for feeding (58.2%), climbing (OR) 5.22, 95% confidence interval (CI) 1.8514.76,
stairs (52.3%) and bathing (47.7%). The only item in which more p 0.002) and bathing (OR 8.10, 95% CI 3.4019.32,
patients were dependent (52.3%) than independent (47.7%) p50.0001) at discharge from the rehabilitation centre were
was bathing. Compared to the situation at discharge, an retained in the multivariate model, which accounted for 34.0% of
increase in patients independent for mobility and stairs was the variance.
356 L. De Wit et al. Disabil Rehabil, 2014; 36(5): 353358

Table 2. Univariate analysis (n 153).

BI  95/100 at
five years after stroke
Variables
Yes (n 62) No (n 91) p Value*
Mean (SD) Mean (SD)
Age at stroke onset 65.88 (10.47) 69.06 (10.45) 0.06y
n (%) n (%)
Gender
Female 26 (41.94) 50 (54.95) 0.10
Male 36 (58.06) 41 (45.05)
Centre
BE 19 (30.65) 54 (59.34) 0.0004
CH 43 (69.35) 37 (40.66)
Type of stroke
Haemorrhagic 13 (20.97) 17 (18.68) 0.74
stroke
Ischaemic infarct 48 (77.42) 71 (78.02)
Unknown 1 (1.61) 3 (3.30)
Individual BI items at discharge from the rehabilitation centre
Bowels
Figure 2. Percentage of patients independent for individual items of the Continent 62 (100) 85 (93.41) 0.04
Barthel Index on discharge from the rehabilitation centre and at five years Incontinent 0 (0) 6 (6.59)
after stroke onset (n 153).
Bladder
Continent 62 (100) 72 (79.12) 0.0001
Logistic regression equation Incontinent 0 (0) 19 (20.88)
Grooming
Given the results of the multivariate regression analysis, the Independent 62 (100) 77 (84.62) 0.001
logistic regression equation contained the two variables dressing Dependent 0 (0) 14 (15.38)
and bathing and read as following: Toilet use
Percentage chance to reach BI score 95/100 at five years Independent 62 (100) 59 (64.84) 50.0001
after stroke Dependent 0 (0) 32 (35.16)
  Feeding
exp0:8210 dressing
Independent 53 (85.48) 36 (39.56) 50.0001
0:8265 bathing  1:0459 Dependent 9 (14.52) 55 (60.44)
   100
1 exp0:8210 dressing Transfer (bed to chair)
0:8265 bathing  1:0459 Independent 60 (96.77) 55 (60.44) 50.0001
Dependent 2 (3.23) 36 (39.56)
Mobility
Dressing 1, if patient scores 10 on BI item dressing at Independent 59 (95.16) 44 (48.35) 50.0001
discharge from the rehabilitation centre; 1, Dependent 3 (4.84) 47 (51.65)
if patient scores 0 or 5 on BI item dressing Dressing
Independent 56 (90.32) 39 (42.86) 50.0001
at discharge from the rehabilitation centre; Dependent 6 (9.68) 52 (57.65)
Bathing 1, if patient scores 5 on BI item bathing at Stairs
discharge from the rehabilitation centre; 1, if Independent 52 (83.87) 28 (30.77) 50.0001
patient scores 0 on BI item bathing at Dependent 10 (16.13) 63 (69.23)
discharge from the rehabilitation centre. Bathing
Independent 51 (82.26) 22 (24.18) 50.0001
Based on this equation, the percentage chance of reaching a BI Dependent 11 (17.74) 69 (75.82)
score 95/100 at five years after stroke was calculated (Table 3)
for the following four potential combinations: (1) independence in *p Value is derived from chi square tests unless otherwise indicated.
y
p Value is derived from T test.
both dressing and bathing at discharge from the rehabilitation
centre resulted in 74.1% (57.685.8) chance, (2) dependence on
both items resulted in 6.3% (5.17.9) chance, (3) independence
for bathing, but dependence for dressing resulted in 35.4% (30.7 score 95/100 at five years after stroke. External validation of the
40.4) chance whereas the opposite resulted in 26.1% (20.732.3) regression formula using data from the British and German
chance of reaching a BI score 95/100 at five years after stroke. patients resulted in following figures: sensitivity 80.4%, speci-
If the chance of reaching a BI score 95/100 at five years after ficity 76.6%, positive predictive accuracy 77.1% and negative
stroke is 50%, the participant is predicted to reach a BI score predictive accuracy 80.0% (Table 4b).
95/100, if the chance is550%, the participant is predicted not to
reach a BI score 95/100. As seen in Table 4(a), 122 (80%) of the
Discussion
153 participants were correctly classified. The sensitivity was
77% and specificity 81%. Any patient predicted to reach a BI The aim of this study was to determine the role of single items of
score 95/100 (positive predictive accuracy) had 74% chance of the BI, at discharge from the rehabilitation centre, in predicting
actually reaching a BI score 95/100 at five years after stroke. independence in personal ADL at five years after stroke. We
Any patient predicted not to reach a BI score 95/100 (negative aimed to translate these findings into a user-friendly method of
predictive accuracy) had 84% chance of actually not reaching a BI prediction for individual patients.
DOI: 10.3109/09638288.2013.793411 Predicting functional outcome after stroke 357
Table 3. Percentage chance of reaching a BI score 95/100 at five years after stroke based on the logistic regression equation.

(In) dependent for BI* item dressing (In) dependent for BI* item bathing % chance (95% CI) of reaching BI* 95/100
At discharge from the rehabilitation centre At five years after stroke
1 2
Independent Independent 74.10 (57.6385.76)
Dependent3 Independent2 35.40 (30.6840.42)
Independent1 Dependent4 26.11 (20.7432.30)
Dependent3 Dependent4 6.34 (5.107.85)

*BI indicates Barthel Index; 1 score 10/10; 2 score 5/5; 3 score 05/10; 4 score 0/5.

Table 4. Comparisons of participants predicted and actual functional
status at five years after stroke based on the logistic regression equation.
Actual functional status
Predicted functional status BI  95/100, n (%) BI595/100, n (%)
(a) Patients from Belgian and Swiss centres (n 153)
BI  95/100, n (%) a 48 (31.4) b 17 (11.1)
BI595/100, n (%) c 14 (9.2) d 74 (48.3)
(b) Patients from British and German centres (external validation, n 93)
BI  95/100, n (%) a 37 (39.8) b 11 (11.8)
BI595/100, n (%) c 9 (9.7) d 36 (38.7)
Sensitivity (a/a c)  100; specificity (d/b d)  100; positive
predictive accuracy (a/a b)  100; negative predictive
accuracy (d/c d)  100.
Although each BI item was significantly associated with
independence in personal ADL at five years after stroke in
univariate analysis, only dressing and bathing at discharge from
the rehabilitation centre were retained in the multivariate model
that accounted for 34% of the variance. A straightforward
comparison of our results with the literature is not possible as
we have not identified any previous studies investigating the
predictive value of single ADL items for long-term independence
in personal ADL after stroke. Similar ADL items (grooming and
upper body dressing on admission to the centre) were retained in a
multivariate model predicting functional outcome at discharge
from a centre [10]. Independence in dressing (score of 10/10)
implies that patients are able to dress their upper and lower body
including buttons, zips, laces, etc. [8]. Walker and Lincoln found
that difficulty in dressing was associated with physical impair-
ment, visual inattention and sensory disturbance [15]. The
complexity of this task might explain why it is retained in the
multivariate model. Independence in bathing (score of 5/5)
implies that patients are able to take a bath or shower without any
help [8]. Again, this is a complex activity requiring bimanual
activity, coordination, balance control and good cognitive func-
tioning. In our sample, 60% of patients were dependent for
bathing which was the highest percentage of dependent patients
for any BI item. Being independent in bathing may reflect a good
overall recovery, explaining its predictive value in the model.
Based on the regression formula, patients who can dress and
bath themselves independently at discharge from the rehabilitation
centre have more than 10 times higher chance (74% versus 6%
chance) of reaching independence in self-care and mobility at five
years after stroke than those who are dependent for both BI items.
The specificity (81%) of the prediction is higher than the
sensitivity (77%). This indicates that the two BI items are more
accurate in identifying those stroke patients who will not reach
independence in self-care and mobility at five years after stroke
than those who will. Positive and negative predictive accuracy
were 74% and 84%, respectively. External validation resulted in
comparable percentages indicating that the prediction formula
was not specific for this sample, underlining the robustness of our
findings. However, additional validation, using data from a
separate data set is required before widespread generalization and
implementation in clinical practice can be justified [16].
Some further reflections are required. Only data from the
Swiss and Belgian centres were included in the analyses. Because
of the high dropout rates in the British (72%) and German (59%)
centres and the significant difference in BI score at discharge
from the centre, we believe that excluding data from these centres
was justified, given the realistic risk of bias. In the present study,
the dependent and independent variable(s) were represented by
the dichotomized BI total score and its individual items,
respectively. It is not surprisingly that the BI total score was
predicted by its individual items, but the aim of the study was to
determine which items were more predictive than others. Long-
term independence in personal ADL was predicted from patients
performance on single BI items assessed at discharge from the
rehabilitation centres. Discharge varies in time, so the prediction
is not made at a consistent time after stroke onset. Starting the
prediction at discharge from the centre has several advantages: (1)
at discharge, all patients are still in the rehabilitation centre. As
the BI is a measure of what the patient actually does, rather than
ability, scoring may be location dependent [8]. By choosing a
similar care setting, the influence of location on patients
performance on individual BI items was excluded; (2) at
discharge, many questions arise from patients, their relatives
and caregivers about future prognosis. The method developed
enables therapists and physicians to provide answers to these
questions at the moment questions are asked; (3) discharge is an
important time point for health care providers as it defines a
change in care management. Patients that are dependent for self-
care and mobility require more intensive therapy, care and support
than those who are functionally independent. The method
developed enables early identification of these vulnerable
patients. Survival analysis on the same sample of patients
showed that BI at six months was a powerful predictor of long-
term mortality [17]. It was concluded that maximum functional
independence at six months post stroke should be promoted
through medical interventions and rehabilitation. Our present
study confirms the important role of independence in personal
ADL in targeting the rehabilitation process after stroke. A recent
systematic review, investigated the methodological quality of
prognostic studies identifying variables predicting ADL outcome
after stroke [18]. The authors used a 27-item checklist which
considered six major sources of bias: study participation, study
attrition, predictor measurement, outcome measurement, statis-
tical analysis and clinical performance/validity. Based on this
quality assessment, the present study would receive a score of 23/
27, indicating that 23 out of the 27 scoring items were scored
positively. This classifies the study as of high quality (20/27).
In conclusion, the studys results suggest a method by which
therapists and physicians could predict patients prognosis for
independence in personal ADL in the long term after stroke,
based on the assessment of dressing and bathing abilities at
discharge from the rehabilitation centre. Additionally, this method
358 L. De Wit et al.
could be used for targeting inpatient stroke rehabilitation
and early identification of those patients who need intensive
follow-up.
Acknowledgements
The authors wish to thank all patients for their participation in the
CERISE study and its follow-up and R. Buyl, PhD (Department of
Biostatistics and Medical Informatics, Vrije Universiteit Brussel)
for the statistical advice.
Declaration of interest
L.D.W. was funded by the Funds for Scientific Research Flanders
(FWO) and by the Vrije Universiteit Brussel (VUB) to carry out
this follow-up study. H.D. is a postdoctoral fellow of the Funds for
Scientific Research Flanders (FWO).
References
1. Maclean N, Pound P, Wolfe C, Rudd A. Qualitative analysis of
stroke patients motivation for rehabilitation. BMJ 2000;321:
10514.
2. Cartier C. From home to hospital and back again: economic
restructuring, end of life, and the gendered problems of place-
switching health services. Soc Sci Med 2003;56:2289301.
3. Olofsson A, Andersson SO, Carlberg B. If only I manage to get
home Ill get better Interviews with stroke patients after
emergency stay in hospital on their experiences and needs.
Clin Rehabil 2005;19:43340.
4. Pettersen R, Dahl T, Wyller TB. Prediction of long-term functional
outcome after stroke rehabilitation. Clin Rehabil 2002;16:14959.
5. Lin J, Tsai AY, Lo SK, et al. Predicting the grade of disability 1 year
after stroke following rehabilitation. Kaohsiung J Med Sci 2005;21:
21219.
Disabil Rehabil, 2014; 36(5): 353358
6. Hankey GJ, Jamrozik K, Broadhurst RJ, et al. Long-term disability
after first-ever stroke and related prognostic factors in the Perth
community stroke study. Stroke 2002;33:103440.
7. Alexander LD, Pettersen JA, Hopyan JJ, et al. Long-term prediction
of functional outcome after stroke using the Alberta Stroke Program
Early Computed Tomography Score in the subacute stage. J Stoke
Cerebrovasc Dis 2012;21:73744.
8. Mahoney FI, Barthel DW. Functional evaluation: the Barthel Index.
Md State Med J 1965;14:615.
9. Granger CV, Cotter AC, Hamilton BB, Fiedler RC. Functional
assessment scales: a study of persons after stroke. Arch Phys Med
Rehabil 1993;74:1338.
10. Gialanella B, Santoro R, Ferlucci C. Predicting outcome after stroke:
the role of basis activities of daily living. Eur J Phys Rehabil Med
2012;48:19.
11. De Wit L, Putman K, Schuback B, et al. Motor and functional
recovery patterns after stroke: a comparison of four European
rehabilitation centers. Stroke 2007;38:21017.
12. WHO MONICA project principal investigators. The world health
organization MONICA project (Monitoring trends and determinants
in cardiovascular disease): a major international collaboration.
J Clin Epidemiol 1988;41:10514.
13. Lincoln N, Leadbitter D. Assessment of motor function in stroke
patients. Physiotherapy 1979;65:4851.
14. Altman DG, Bland JM. Diagnostic tests: sensitivity and specificity.
BMJ 1994;308:1552.
15. Walker MF, Lincoln NB. Factors influencing dressing performance
after stroke. J Neurol Neurosurg Psychiatry 1991;54:699701.
16. Adams ST, Leveson SH. Clinical prediction rules. BMJ 2012;344:
d8312.
17. De Wit L, Putman K, Devos H, et al. Five-year mortality and related
prognostic factors after inpatient stroke rehabilitation: a European
multi-centre study. J Rehabil Med 2012;44:54752.
18. Veerbeek JM, Kwakkel G, van Wegen EEH, et al. Early prediction
of outcome of activities of daily living after stroke: a systematic
review. Stroke 2011;42:14828.
Copyright of Disability & Rehabilitation is the property of Taylor & Francis Ltd and its
content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission. However, users may print, download, or email
articles for individual use.