Relationships between long-term stroke outcomes27323. .docx
1. Relationships between long-term stroke outcomes
273
23. Jarman B. Identification of underprivileged areas. Br Med J
1983; 286: 1705–08.
24. Office of Population Censuses and Surveys. Standard
Occupational
Classification: Volume 2. Coding Index. London: HSMO, 1990.
25. Siegel S, Castellan NJ. Nonparametric Statistics for the
Behav-
ioural Sciences, 2nd edn. Singapore: McGraw-Hill, 1988.
26. Hosmer DH, Lemeshow S. Applied Logistic Regression, 2nd
edn. New York: Wiley, 2000.
27. Clipp EC, George LK. Dementia and cancer: a comparison
of
spouse caregivers. Gerontologist 1993; 33: 534–41.
28. Meyers JL, Gray LN. The relationships between family
primary
caregiver characteristics and satisfaction with hospice care,
quality of life and burden. Oncol Nurs Forum 2001; 28: 73–82.
29. Chan CWH, Chang AM. Experience of palliative home care
according to caregivers’ and patients’ ages in Hong Kong
Chinese people. Oncol Nurs Forum 2001; 27: 1601–5.
30. Folkman S, Lazarus RS, Pimley S, Novacek J. Age
3. was categorised as severe, moderate, mild and independent (0–
9, 10–14, 15–19 and 20); FAI was categorised as inactive,
moderately active and very active (0–15, 16–30 and 31–45). SF-
36 domains include: Physical Functioning (PF), Role Physical
(RP), Bodily Pain (BP), General Health (GH), Vitality (VT),
Social Functioning (SF), Role Emotional (RE) and Mental
Health (MH). Physical (PHSS) and Mental Health (MHSS)
Summary Scores were computed.
Results: at 1 and 3 years, 26.1 and 26.3%, respectively, were
disabled (BI <15); 55 and 51%, respectively, were handicapped
(FAI = 0–15); and survivors had low mean PHSS (37.1 and
37.9), but satisfactory mean MHSS (46.6 and 47.7). There was a
graded positive relationship between all SF-36 domains and the
categories of BI and FAI. Spearman rank correlations were
significant between BI and all SF-36 domains at both time
points: strong (r>0.70) with PF, moderate (r = 0.31–0.70) with
RP, SF and PHSS, but weak (r <0.30) with other domains.
Correlations between FAI and SF-36 domains were strong with
PF, weak with BP, RE and MHSS, and moderate with other
domains.
Conclusions: disability and handicap remain highly prevalent up
to 3 years after stroke. Patients’ perception of physical
health is persistently low, but mental health perception is
satisfactory up to 3 years. Due to variable correlations between
dif-
ferent HRQOL domains with disability and handicap, it is
suggested that disability, handicap and HRQOL should all be
assessed to acquire a broader measure of stroke outcome.
Keywords: stroke, long-term outcomes, disability, handicap,
quality of life, elderly
D
ow
nloaded from
4. https://academ
ic.oup.com
/ageing/article-abstract/35/3/273/40353 by guest on 27 M
arch 2019
mailto:[email protected]
mailto:[email protected]
M. D. Patel et al.
274
Introduction
Stroke remains one of the major chronic illnesses worldwide
that health care organisations will need to address for the next
several decades. This is because stroke can impact on virtually
all human functions [1] and, unlike other disabling conditions,
the onset of stroke is sudden, leaving the individual and the
family ill prepared to deal with its sequelae[1]. Stroke is a
chronic disease, yet studies have focused on measuring short-
term outcomes mainly in the domain of impairments and disa-
bility. The concept of broader outcomes was globally recog-
nised in 1980 with the development of the international
classification of impairments, disabilities and handicaps by the
World Health Organisation (WHO) [2]. The few community-
based reports on the long-term prognosis of an unselected
cohort have mainly looked at stroke survival, recurrence,
impairment and disability [3–5]. A review of stroke outcome
measures in 174 acute stroke trials showed that death was
recorded in 76%, impairment in 76%, disability in 42% and
handicap in only 2% [6]. Another distinct paradigm that is also
relevant in assessing the impact of stroke, from a patient’s per-
5. spective, is health-related quality of life (HRQOL) [7]. Meas-
urement of HRQOL in stroke patients can potentially provide
researchers with a more holistic picture of stroke recovery and
assist in the evaluation of medical interventions [8].
Many clinical trials have used scales assessing more than
one domain of outcome, yet there is a paucity of long-term
stroke studies exploring the relationships between the dif-
ferent domains [6]. Exploring the relationships of HRQOL
to disability and handicap would enable one to establish
whether these ‘objective’ tools of assessment of stroke out-
come are also relevant to patients themselves, examine the
correlations of outcomes with one another and allow one to
focus on specific measures [e.g. the Barthel index (BI)]
rather than the numerous measures currently used in the
hope of capturing stroke outcomes.
There are limitations to studies that have examined stroke
outcomes beyond a year, including restriction to a selective
groups such as hospital admissions [9–13], stroke units
[14,15], specific age groups [5,13,15] or specific subtypes of
stroke [9,11]. The design of the studies was predominantly
cross-sectional so subjects were assessed at variable times
after their stroke, and thus were at differing stages of the nat-
ural recovery from stroke. Earlier studies [9,14] used non-
standardised instruments to assess HRQOL, which made it
difficult to interpret their results. Finally, associations
between disability, handicap and HRQOL are not at all well
established, even in short-term stroke studies [13].
The aims of this longitudinal study were to describe
long-term disability, handicap and HRQOL up to 3 years
after stroke, and to examine the relationships between these
domains of stroke outcomes in an unbiased community-
based sample of stroke subjects.
6. Methodology
Stroke subjects were recruited from the South London
Stroke Register (SLSR) which is a longitudinal, population-
based register of first-ever strokes in subjects of all age
groups covering a multiethnic, inner-city population of
234,533. Detailed methodology of the compilation of this
register has been reported earlier [16]. Consent was obtained
from Guy’s & St Thomas’ Ethics Committee according to
the revised Declaration of Helsinki of 1983. Stroke was
defined according to the WHO criteria [17]. ‘Hot pursuit’ of
cases and multiple sources of notification were used to
achieve a high level of case ascertainment. Data collected on
initial assessment included patient demography, risk factors,
and pre-morbid disability using the BI. The Oxfordshire
Community Stroke Project Classification (OCSP) was used
to distinguish between total and partial anterior, lacunar and
posterior circulatory infarctions [18].
Subjects were followed up at 1 and 3 years after stroke.
For this study, subjects who sustained their strokes between
1 January 1995 and 31 December 1997 were included.
Their 3-year follow-up was completed by 31 December
2000. Follow-up data were collected by personal inter-
views with the subjects and/or their carers. The outcome
measures collected were survival rates, disability using the
BI [19], handicap using the Frenchay activities index (FAI)
[20], and HRQOL which was evaluated by the UK version
of the Medical Outcomes Study 36 item Short-Form
Health Survey (SF-36) [21]. The HRQOL assessments
were done mainly by asking the subjects themselves,
except in those who were too confused or dysphasic to
undergo these assessments themselves. This ensured that
the SF-36 represented the subjects’ own subjective views
on their health.
7. The BI was used because it is simple, and has a well-
established validity, communicability and efficiency [19].
The BI was interpreted as follows [22]: 0–9, severely disa-
bled; 10–14, moderately disabled; 15–19, mildly disabled;
and 20, functionally independent. Defined specifically for
stroke subjects, the FAI was used as a measure of social
function and of handicap as it is simple, has good validity
and is widely used [20]. The FAI was interpreted as follows:
0–15, inactive; 16–30, moderately active; and 31–45, very
active [23]. The SF-36 was selected to measure HRQOL
because of its strong psychometric properties, wide use, reli-
ability, validity and responsiveness [21,24]. It assesses eight
domains of health status: Physical Functioning (PF); Role
Physical (RP); Bodily Pain (BP); General Health (GH);
Vitality (VT); Social Functioning (SF); Role Emotional
(RE); and Mental Health (MH). Each domain is scored
between 0 and 100. Absence of problems is indicated by
scores of 100 for PF, RP, BP, SF and RE, and scores of 50
in GH, VT and MH. These domains were then computed to
produce two summary scales representing physical and
mental health [25]. Domains for the physical health sum-
mary scale (PHSS) include PF, RP, BP and GH. The mental
health summary scale (MHSS) includes VT, SF, RE and
MH. The summary scales are based on norms with a mean
of 50 and an SD of 10.
Due to unavoidable practical problems with the register,
SF-36 assessments at the 3-year follow-up were adminis-
tered only for subjects registered between 1 January 1995
and 29 February 1996. Thereafter, for subjects registered
between 1 January 1997 and 31 December 1997, HRQOL
D
ow
8. nloaded from
https://academ
ic.oup.com
/ageing/article-abstract/35/3/273/40353 by guest on 27 M
arch 2019
Relationships between long-term stroke outcomes
275
was assessed using the shorter version of SF-36, SF-12 [26].
Thus, there were fewer SF-36 assessments (n = 99) done at
3 years. As the 12 items in the SF-12 have been adopted
from the SF-36 verbatim, and SF-12 and SF-36 summary
scores are replicable and reproducible [8], the responses
given to the specific 12 items in the SF-36 version, by the
earlier group of subjects (n = 99), were combined with those
obtained from the SF-12 in the later group (n = 51) to
obtain a larger cohort with SF-12 assessments at 3 years
after stroke (n = 150 out of 294 survivors, 51%). This
cohort was used to calculate the mean PHSS and MHSS at
3 years.
Statistical analyses were carried out using χ2 tests and
t-tests as appropriate. The SF-36 scores were presented
according to accepted guidelines [21], and non-parametric
tests were used for analyses involving SF-36: Wilcoxon
signed-rank tests for analyses between continuous variables,
Kruskal–Wallis rank tests between categorical and continu-
ous variables, and Spearman rank correlation coefficients
for correlations with other outcomes.
9. Results
There were 946 subjects who had strokes between 1 January
1995 and 31 December 1997. At 1 and 3 years respectively,
case fatality rates were 42% (397) and 54% (507), and data
were unavailable for 6.2% (59) and 10.2% (97) of subjects
because they were either lost to follow-up or refused to be
assessed. Data were thus available for 490 subjects at 1 year
and 342 at 3 years after stroke. Appendix 1 in the
supplementary
data on the journal website (www.ageing.oxfordjournals.
org) describes their demographic characteristics.
Potential selection biases
Univariate analyses showed no differences between subjects
included (n = 490) and excluded (n = 59) from the 1-year
analyses. This is discussed in the supplementary data on the
journal website (Appendix 2, available on the journal web-
site: www.ageing.oxfordjournals.org).
Long-term disability, handicap and HRQOL
(Table 1 and Appendix 3)
At 1 and 3 years after stroke, 40 and 34% of subjects,
respectively, were independent (BI = 20). Twenty-six per
cent of subjects were moderately or severely disabled (BI 0–14)
at 1 and 3 years after stroke. In terms of FAI, 16% were very
active at 1 year and 14% at 3 years. At both time points,
more than half the survivors were inactive. SF-36 summary
scales revealed low mean PHSS scores but satisfactory
MHSS scores. There were no significant changes observed
in any domains between 1 and 3 years, except for a small
improvement in RE. The Spearman correlation between BI
and FAI was strong at 0.83 and 0.79 at 1 and 3 years, respec-
10. tively. It is important to note that analyses for Table 1 were
conducted for only those subjects with data available at
both time points. Hence, BI was compared in 342 subjects,
FAI in 294, HRQOL summary scales in 150, and individual
HRQOL domains in 90.
Table 1. Disability, handicap and health-related quality of life
(HRQOL) at 1 and 3 years after stroke
aStatistical analyses were only conducted in subjects with data
available at both times.
bSubjects registered between 1 March and 31 December 1996
did not have their HRQOL assessed at 3 years after their stroke.
See text for further details.
Subjects, n (%) at 1 year Subjects, n (%) at 3 years Mean
changea (95% CI)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Case fatality rates 397 (42%) 507 (54%) –
Alive and followed-up 490 342 –
Institutionalisation 76 (15.5%) 51 (15.0%) –
Barthel index
Severe (0–9) 60 (12.2%) 49 (14.3%) –
Moderate (10–14) 68 (13.9%) 41 (12.0%)
Mild (15–19) 168 (34.3%) 136 (39.8%)
Independent (20) 194 (39.6%) 116 (33.9%)
Mean 16.3 15.9
Median (range) 19.0 (0–20) 18.0 (0–20) –0.83 (–1.20 to 0.46)
Frenchay activity index n = 462 n = 323
Inactive (0–15) 254 (55.0%) 166 (51.4%) –
Moderately active (16–30) 133 (28.8%) 111 (34.4%)
11. Very active (31–45) 75 (16.2%) 46 (14.2%)
Mean 15.4 15.8
Median (range) 13.0 (0–45) 15.0 (0–44) –1.05 (–2.07 to 0.04)
SF-36 scales n = 397 n = 150b
Physical health summary scales
Mean, SD 37.1 (12.4) 37.9 (11.5) –0.46 (–2.39 to 1.48)
Range 11.7–65.4 19.0–59.8
Mental health summary scales
Mean, SD 46.6 (12.1) 47.7 (10.2) –0.95 (–3.04 to 1.15)
Range 15.4–71.9 20.3–68.0
D
ow
nloaded from
https://academ
ic.oup.com
/ageing/article-abstract/35/3/273/40353 by guest on 27 M
arch 2019
http://www.ageing.oxfordjournals.org
http://www.ageing.oxfordjournals.org
http://www.ageing.oxfordjournals.org
M. D. Patel et al.
276
Relationships between disability and HRQOL
Correlations between disability and all HRQOL domains
12. were significant at both 1 and 3 years after stroke, but there
was variation in the strength of these correlations; it was
strong [26] (r>0.70) with PF, moderate (r = 0.31–0.70) with
RP, SF and PHSS, and weak (r < 0.30) with the other
domains (Table 2). One year after stroke, there was a graded
positive relationship between all the domains of HRQOL
and categorical levels of disability (Figure 1). At 3 years, sim-
ilar analyses showed that this graded relationship was evi-
dent only for PF, RP and SF. However, caution must be
exercised when interpreting the 3-year results as only 14
subjects had severe disability.
Relationships between handicap and HRQOL
Correlations between handicap and HRQOL domains were
also all significant, but their strength was variable (Table 2);
cor-
relation was strong with PF, weak with BP, RE and MHSS, and
moderate with the other domains. There was a graded positive
relationship between all the domains of HRQOL and categori-
cal levels of handicap at 1 year after stroke (Figure 1). This
relationship was also observed at 3 years after stroke.
Discussion
This large, population-based longitudinal study of first-ever
strokes has shown that the long-term impact of stroke is
considerable. Three years after stroke, 26% of subjects were
moderately or severely disabled (BI <15), 51% were handi-
capped (inactive, FAI <15), mean scores for physical HRQOL
(PHSS) were poor, but the mean scores for mental HRQOL
(MHSS) were satisfactory.
Disability remains prevalent for years following a
stroke, with studies reporting 13–66% of their sample hav-
ing some disability [5,14,15]. The prevalence of handicap in
13. previous stroke studies ranges from 12 to 64% [15,27] (fur-
ther discussed in Appendix 4 on the journal website:
www.ageing.oxfordjournals.org). Apart from individual
domains of HRQOL, PHSS and MHSS were also com-
puted in this study. These scales were particularly useful as
they allowed meaningful and simple comparisons to be
made between subjects’ perceived physical and mental
health and objective measures used in research and clinical
practice. Subjects perceived their physical health to be low
up to 3 years after a stroke, whereas their perception of
their mental health was satisfactory. The acceptable MHSS
found in this study could be due to higher initial mortality
in stroke, meaning this ‘natural selection’ resulted in assess-
ments being undertaken in a ‘survivor cohort’ who were in
better spirits, younger and living at home [29]. The physical
scores were also low in subjects with minimal or no disabil-
ity, which confirms earlier reports that in many instances low
HRQOL is reported, yet the individual is independent in
activities of daily living [13]. In contrast, Hackett [29] meas-
ured HRQOL using the SF-36 in 639 subjects from a popu-
lation-based study at 6 years after stroke, and concluded that
HRQOL was acceptable for most stroke survivors com-
pared with controls, despite ongoing physical limitations.
This study showed that most domains of HRQOL had a
graded relationship with the categorical levels of disability
and handicap. As previously recommended [6], categorical
analysis was done as it is more appropriate than continuous
analysis in terms of clinical relevance that can usually be
applied to different categories, such as describing disability
as mild, moderate or severe. The graded relationship was
more consistent with FAI than with BI at both time points,
which may be either due to the small number of subjects
with severe disability at 3 years or due to the slightly
stronger correlation between most domains of HRQOL
14. and FAI (Table 2). These results are comparable with previ-
ous studies. Wilkinson [5] examined 106 subjects in a cross-
sectional study, and reported a high correlation between BI
and FAI (r = 0.83) at a mean follow-up of 5 years. Another
hospital-based study of 81 patients with lacunar infarctions
reported a strong correlation between disability (Katz activ-
ities of daily living) and handicap (Oxford Handicap Scale)
[27]. Duncan [30] indicated that standardised assessment of
stroke subjects must evaluate across the entire continuum
of health-related functions, and recommended that mea-
sures such as SF-36 be used in addition to the BI, which has
a ceiling effect and captures only physical functions.
Roberts [6] also concluded that as there is no simple
Table 2. Relationship between disability (Barthel index),
handicap (Frenchay activities index) and HRQOL (SF-36) at 1
and
3 years after stroke, using Spearman’s rank correlation
coefficient
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
Disability
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . .
Handicap
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
One year, n = 401
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Three years, n = 99
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
16. https://academ
ic.oup.com
/ageing/article-abstract/35/3/273/40353 by guest on 27 M
arch 2019
http://www.ageing.oxfordjournals.org
Relationships between long-term stroke outcomes
277
Figure 1. Comparison of SF- 36 according to disability (Barthel
index) and handicap (Frenchay activities index) 1 year after
stroke
using Kruskal–Wallis rank tests.
0
10
20
30
40
50
60
70
80
19. bi 0-14, n=73 bi 15-19, n=145 bi 20, n=183
SF-36 & Barthel Index
0
10
20
30
40
50
60
70
80
90
100
Physical
Functioning
p<0.001
Role Physical
p<0.001
20. Bodily Pain
p<0.001
General Health
p<0.001
Vitality p<0.001 Social functioning
p<0.001
Role Emotional
p=0.031
Mental Health
p<0.001
Physical health
summary scale
p<0.001
Mental health
summary scale
p<0.001
domains of HRQOL
m
21. e
a
n
p
e
rc
e
n
ta
g
e
s
c
o
re
s
fai 0-15, n=179 fai 16-30, n=129 fai 31-45, n=71
SF-36 & Frenchay Activity Index
D
ow
nloaded from
https://academ
ic.oup.com
/ageing/article-abstract/35/3/273/40353 by guest on 27 M
22. arch 2019
M. D. Patel et al.
278
relationship between disability, handicap and HRQOL, out-
come measures should include items relating only to a single
level because a mixture of levels is conceptually confusing
and difficult to interpret clinically. In view of the variable
correlations of disability and handicap with different
domains of HRQOL observed in this study, we suggest that
a multitude of standardised assessments should be con-
ducted to obtain a more holistic measure of stroke out-
come. This would provide useful information for planners
of rehabilitation services and long-term care, and scope for
new interventions to improve these outcomes.
There are some limitations to this study. First, it is
acknowledged that although it is widely used, SF-36 is not a
stroke-specific measure, but instead a generic measure that
was first developed to assess HRQOL outcomes that are
closely affected by disease and treatment. Thus, it may not
be sensitive or specific enough to detect the psychological
domains of mental health that are relevant to stroke [29].
Secondly, data were missing for several subjects, particularly
for subjects who were very dysphasic or confused and they
had no carers to act as proxies, and for subjects registered
during part of the study period due to reasons already men-
tioned. These ‘missing’ data may have introduced a bias into
the study so that the frequency of poor outcomes may actu-
ally have been underestimated. Thirdly, SF-12 was used for
some subjects instead of SF-36, though the summary scores
produced by either of them have been shown to be replica-
23. ble [8]. Fourthly, this study did not consider the effects that
stroke recurrence or other major co-morbidities such as
vascular cognitive impairment may have had on the sub-
jects’ disability or handicap during the 3 years after stroke.
Finally, the strong correlation between BI and FAI with the
physical domains of SF-36, particularly PF, may be due to
common factors being assessed, i.e. questions relating to
this domain assess mobility and physical activities which are
also assessed by similar questions in BI and FAI.
We conclude that the prevalence of disability and handi-
cap following stroke remains high up to 3 years after a
stroke. By embracing a wider perspective than merely sur-
vival, this study provides a broader and more representative
description of long-term stroke outcome than has been pre-
viously reported. Patients’ perception of physical HRQOL
is persistently low, while their perception of mental
HRQOL is satisfactory up to 3 years post-stroke. There is a
graded relationship of HRQOL to disability and handicap.
The correlation of disability and handicap is variable with
different domains of HRQOL (SF-36); it is strong with
PHSS, but weak with MHSS. In view of these variable cor-
relations, it is suggested that future studies should assess
disability, handicap and HRQOL, using standardised assess-
ments, as primary outcome measures in order to acquire a
broader measure of stroke outcome.
Key points
• This report provides a broader and more representative
description of long-term stroke outcome than has been
previously reported.
• It provides population-based estimates of impact of
stroke in terms of disability, handicap and HRQOL up to
3 years after stroke.
24. • The correlation of disability and handicap is variable with
different domains of HRQOL (SF-36); it is strong with
PHSS, but weak with MHSS.
• Future studies on stroke should assess disability, handi-
cap and HRQOL as primary outcome measures in order
to acquire a broader measure of stroke outcome.
References
1. Mayo NE, Wood-Dauphinee S, Ahmed S et al. Disablement
following stroke. Disabil Rehabil 1999; 21: 258–68.
2. Wood P. International Classification of Impairments,
Disabilities
and Handicaps: A Manual of Classification Relating to the
Conse-
quences of Disease. Geneva: World Health Organisation, 1980.
3. Sacco RL, Wolf PA, Kannel WB, McNamara PM. Survival
and
recurrence following stroke. The Framingham study. Stroke
1982; 13: 290–5.
4. Dennis MS, Burn JPS, Sandercock PAG, Bamford JM, Wade
DT,
Warlow CP. Long-term survival after first-ever stroke: the
Oxford-
shire Community Stroke Project. Stroke 1993; 24: 796–800.
5. Wilkinson PR, Wolfe CD, Warburton FG et al. Longer term
quality of life and outcome in stroke patients: is the Barthel
index alone an adequate measure of outcome? Qual Health
Care 1997; 6: 125–30.
25. 6. Roberts L, Counsell C. Assessment of clinical outcomes in
acute stroke trials. Stroke 1998; 29: 986–91.
7. Bowling A. Measuring Disease: A Review of Disease-specific
Quality of Life Measurement Scales. Buckingham: Open Uni-
versity Press, 1995.
8. Pickard AS, Johnson JA, Penn A, Lau F, Noseworthy T.
Repli-
cability of SF-36 summary scores by the SF-12 in stroke
patients. Stroke 1999; 30: 1213–7.
9. Kwa VI, Limburg M, de Haan RJ. The role of cognitive
impairment in the quality of life after ischaemic stroke. J Neu-
rol 1996; 243: 599–604.
10. Harwood RH, Gompertz P, Pound P, Ebrahim S. Determi-
nants of handicap 1 and 3 years after a stroke. Disabil Rehabil
1997; 19: 205–11.
11. Greveson GC, Gray CS, French JM, James OF. Long-term
outcome for patients and carers following hospital admission
for stroke. Age and Ageing 1991; 20: 337–44.
12. King R. Quality of life after stroke. Stroke 1996; 27: 1467–
72.
13. McEwen S, Mayo N, Wood-Dauphinee S. Inferring quality
of
life from performance-based assessments. Disabil Rehabil
2000; 22: 456–63.
14. Astrom M. Asplund K, Astrom T. Psychosocial function and
life satisfaction after stroke. Stroke 1992; 23: 527–31.
15. Dijkerman HC, Wood VA, Hewer RL. Long-term outcome
26. after discharge from a stroke rehabilitation unit. J R Coll Phy-
sicians Lond 1996; 30: 538–46.
16. Stewart J, Dundas R, Howard RS, Rudd AG, Wolfe CDA.
Ethnic differences in incidence of stroke: prospective study
with stroke register. Br Med J 1999; 318: 967–71.
17. Aho K, Harmsen P, Hatano S, Marquardsen J, Smirnov VE,
Strasser T. Cerebrovascular disease in the community: results
of a WHO collaborative study. Bull WHO 1980; 58: 113–30.
18. Bamford J, Sandercock P, Dennis M, Burn J, Warlow C.
Clas-
sification and natural history of clinically identifiable subtypes
of cerebral infarction. Lancet 1991; 337: 1521–6.
D
ow
nloaded from
https://academ
ic.oup.com
/ageing/article-abstract/35/3/273/40353 by guest on 27 M
arch 2019
Relationships between long-term stroke outcomes
279
19. Wade DT, Collin C. The Barthel ADL Index: a standard
meas-
ure of physical disability? Int Disabil Stud 1988; 10: 64–7.
27. 20. Wade DT, Legh-Smith J, Langton Hewer J. Social activities
after stroke: measurement and natural history using Frenchay
Activities Index. Int Rehabil Med 1985; 7: 176–81.
21. Ware JE, Snow KK, Kosinski M, Gandek B. SF-36® Health
Survey: Manual and Interpretation Guide. Lincoln, RI: Quality
Metric Incorporated, 1993.
22. Wolfe CDA, Taub NA, Woodrow J, Burney PGJ.
Assessment
of scales of disability and handicap for stroke patients. Stroke
1991; 22: 1242–4.
23. Anderson CS, Jamrozik KD, Broadhurst RJ, Stewart-Wynne
EG. Predicting survival for 1 year among different subtypes of
stroke. Stroke 1994; 25: 1935–44.
24. Anderson C, Laubscher S, Burns R. Validation of the Short
Form 36 (SF-36) health survey questionnaire among stroke
patients. Stroke 1996; 27: 1812–6.
25. Ware JE, Kosinski M, Keller SD. SF-36 Physical and Mental
Health Summary Scales: A User’s Manual. Boston, MA: The
Health Assessment Laboratory, 1994.
26. Ware JE, Kosinski M, Keller SD. SF-12®: How to score the
SF-12® Physical and Mental Health Summary Scales, 3rd edn.
Lincoln, RI: Quality Metric Incorporated, 1998.
27. Samuelsson M, Soderfeldt B, Olsson GB. Functional
outcome in patients with lacunar infarction. Stroke 1996;
27: 842–6.
28. Wade DT. Measurement in Neurological Rehabilitation.
Oxford: Oxford University Press, 1992.
28. 29. Hackett ML, Duncan JR, Anderson CS, Broad JB, Bonita R.
Health-related quality of life among long-term survivors of
stroke: results from the Auckland Stroke Study, 1991–1992.
Stroke 2000;31(2):440–7.
30. Duncan PW, Samsa GP, Weinberger M et al. Health status
of
individuals with mild stroke. Stroke 1997; 28: 740–7.
Received 10 November 2005; accepted in revised form 6
February
2006
D
ow
nloaded from
https://academ
ic.oup.com
/ageing/article-abstract/35/3/273/40353 by guest on 27 M
arch 2019