From the Department of Clinical Neurosciences, Western General Hospital,
Edinburgh, Scotland.
Correspondence to Dr Carl Counsell, Department of Neurology, Austin & Repatriation Medical Centre, Repatriation Campus, Banksia St, West Heidelberg 3081, Victoria, Australia.
Methods—Acute stroke trials from the Cochrane Stroke Group's
database were included from 1955 to 1995 if they were published in full
text in English. For each trial we collected year of publication,
number of patients randomized, blinding of outcome assessment, the
specific outcome instruments used, the statistical methods used for
analysis, and the significance of the results. The validity and
reliability of each outcome measure were assessed by review of the
literature.
Results—Our study included 174 trials. Outcomes were assessed
blindly in 69%. Death was recorded in only 76% of trials,
impairment in 76%, disability in 42%, and handicap or quality of life
in only 2%. Of the trials that measured impairment, 35% used a
measure of established validity or reliability. For disability and
handicap, the proportions with valid or reliable measures were 70% and
25%, respectively. Impairment and handicap measures were primarily
analyzed as continuous variables, while disability was
mainly analyzed as a dichotomous variable. Continuous data
were usually analyzed with inappropriate parametric
statistics. There was no relationship between the method of
analysis, the type of outcome, and the statistical significance
of results.
Conclusions—Most acute stroke trials up to 1995 have used
clinical outcome measures that were inadequate in terms of their
content, reliability, validity, blinded assessment, and statistical
analysis. This has important implications for future stroke
research.
To influence clinical practice, acute stroke trials should use measures
of outcome that are relevant to patients. Thus, phase III trials ought
to concentrate on measuring disability, handicap, or health-related
quality of life. In contrast, small phase II trials are generally not
designed to influence practice but aim to establish whether a treatment
influences the disease process and therefore might have an important
clinical benefit. In these trials it is often appropriate to
concentrate on measuring impairments. However, phase II trials should
also measure more clinically relevant outcomes because these results
may help in designing later phase III trials (eg, by providing
estimates of realistic treatment effects) and can also be included in
ongoing systematic reviews.
Other types of outcomes also need to be considered in acute stroke
trials. The numbers of deaths should always be reported because both
doctors and patients need to know whether a treatment is associated
with more deaths even if it decreases disability in survivors. Length
of hospital stay measures the economic implications of an
intervention,11 while outcomes in the caregiver
may also be important after the patient has been discharged from the
hospital. Given the variety of outcomes that can be measured in trials
of treatments for acute stroke, the outcomes should be prespecified and
classified as primary (assessing the main hypothesis) or secondary
(assessing the secondary hypotheses) to avoid inappropriate "data
dredging" or multiple subgroup analyses to find a spurious
positive result.
Outcome measures should also be valid,5
reliable,5 sensitive to important clinical
changes, assessed blindly to avoid measurement bias, and
analyzed appropriately. Measures of impairment, disability, and
handicap almost always consist of a mixture of categories designated by
numbers to form an ordinal scale in which the gaps between the
different points are not necessarily equal.12
Strictly speaking, it is therefore statistically incorrect to treat
these numbers as though they represent values on a continuous
scale. The statistical methods used should generally be
nonparametric and suitable for ranked data.
Analysis that categorizes results into two or more groups of
patients with similar outcomes is perhaps even more appropriate.
Given the above considerations, we wished to assess how appropriate the
measurement of outcomes has been in acute stroke trials in terms of the
types of outcomes recorded, whether they were assessed blindly,
their validity and reliability, and the method of analysis.
Data Collection
Definitions for the Classification of Outcomes
Impairment, disability, and handicap scales were further classified
into those that had been shown to have some validity or reliability and
those that were neither valid nor reliable. A rigorous appraisal of the
validity and reliability of each measure was beyond the scope of this
study. Instead, we identified previous
reviews5 15 16 17 that discussed the properties of
outcomes measures used in stroke research and followed their
recommendations about which measures were valid and reliable. For those
outcome measures not mentioned in these reviews, we tried to identify
primary research studies that documented their validity or reliability.
Unnamed measures that were not referenced in the original trial report
were assumed to be untested unless otherwise stated. The classification
of impairment, disability, and handicap scales is shown in Table 1
All data were collected by the two authors independently and were
cross-checked. Disagreements were resolved by consensus. We did not
formally measure interobserver variability.
Analysis
Assessment of the Type of Outcomes
The relationship between the type of outcome measured and the
size of the trial is demonstrated in Figure 1
Statistical Analysis of Outcomes
Twelve trials assessed disability with the Barthel Index and
analyzed it as a dichotomous variable, but there was no
standardization in the cutoff point used to define a good outcome (ie,
which patients were independent). Five trials defined a good outcome as
a Barthel Index of 12 or more of 20 (or 60 of 100); 4 trials as a
Barthel Index of more than 14 (70 of 100); 1 trial as a Barthel Index
of 18 or more (90 of 100); 1 trial as a Barthel Index of 19 or more (95
of 100); and 1 trial as an improvement in the Barthel Index from the
baseline score.
Measures analyzed as continuous data were not more likely to
report statistically significant results than those analyzed as
dichotomous data (impairment:
Historically, impairment and disability measures were introduced into
stroke trials about the same time (the early 1960s), but whereas the
use of impairment rapidly increased, the measurement of disability has
increased only gradually. This may have been because there were so many
measures of impairment or because impairment measures were perceived to
be easier to use. However, several simple disability measures such as
the Rankin or Barthel scales have existed for many years. Many of the
trials may have been phase II studies and therefore concentrated on
impairment to demonstrate some effect on the disease process. This is
partly supported by the fact that larger trials were more likely to
measure disability. However, few trials were formally called phase II
studies, and even phase II studies should report deaths, but 25% of
trials did not even do this. Moreover, only 60% of trials with more
than 300 patients reported disability. Detailed measurement of
impairment may also be more likely to show small but statistically
significant effects that may be important to detect in exploratory
phase II trials. However, we did not find that trials measuring
impairment were more likely to show statistically significant results
than those measuring disability. There was little evidence to suggest
that reporting of important outcomes had improved over time.
Any measure used to assess outcome, whether impairment or disability,
should be shown to be valid and reliable, and yet many of those used in
stroke trials were not. Many trialists used their own measure of
impairment or disability rather than an established and tested one. The
statistical analysis of outcomes recorded on categorical or
ordinal scales was also extremely poor. Most were analyzed as
continuous data, but this has inherent problems. A difference of a few
points in a continuous score between the treated and control groups has
little clinical meaning. If such a scoring system is used, the scoring
of death is also problematic (should death merely be
assigned the worst possible score, or is it worse than that?). In
addition, most trials analyzed continuous data with
parametric statistics, which is invalid if the trial is small
and the distribution is skewed. This was the case in many of the
trials, particularly those in which death was scored as zero or
negative on a scale.
Categorical analysis is much more appropriate because clinical
meaning can usually be applied to different categories, eg, disability
scores can be divided into "independent in activities of daily
living" or "dependent." Death can be added as a separate category
or combined with another category (eg, "dead or dependent").
However, problems also existed in the reporting of results of measures
analyzed appropriately as two categories. Very few of these
trials reported a measure of relative treatment effect such as relative
risk or odds ratio. In addition, there was no standardization in the
point at which disability scales, such as the Barthel Index, were split
into different categories. Comparison of these outcomes between
different trials, for example in meta-analysis, therefore
becomes difficult. Reassuringly, the method of analysis
(continuous versus categorical, parametric versus
nonparametric) was not associated with the proportion of
statistically significant results. If it had been, it might have
suggested that trialists were using inappropriate statistics to
increase the significance of their findings.
We found other problems with the process and reporting of outcome
assessment. Only two thirds of the trials stated that outcomes were
blindly assessed, yet unblinded assessments may be
biased.36 37 Even in so-called double-blind
trials, the outcome assessor may in fact not have been blind,
particularly if the assessor was involved in the care of the patient
and the treatment was associated with particular side
effects.38 It was usually impossible to determine
from the reports which were the primary and secondary outcomes. More
emphasis should be given to the results of the prespecified primary
outcomes because these are the ones the trial was specifically designed
to test. Secondary outcomes may be less reliable, especially if
multiple outcomes are measured and only the statistically significant
ones are reported. Finally, the period of follow-up was often too short
to assess the full impact of treatment on the final status of the
patient.
Our study has some weaknesses. First, we excluded non–English language
trials, and therefore our results may not be generalizable. However,
there is little evidence to suggest that non–English language trials
are any better or worse than English language
trials.39 We also did not include trials from
1996 onward, and therefore it is possible that the assessment of
outcomes has improved recently. However, given the scale of the
problems we found, it is unlikely that they will have been resolved
since 1995. Certainly, the trends shown in Figure 2
This study highlights several improvements that should be made in
future acute stroke trials. Outcome measures should be relevant to the
patient, concentrating on disability or handicap, and deaths must be
reported. Even small exploratory phase II trials should assess these
outcomes because the results can then be included in
meta-analyses. Other important factors such as side effects,
length of hospital stay, and assessment of the caregiver should also be
considered. The outcome measures must be well established with proven
validity and reliability and should be blindly assessed. Statistical
analyses should be appropriate for ranked data or, better
still, involve division of patients into clinically meaningful groups,
eg, those alive and independent compared with those dead or dependent.
The definitions of dependency in existing disability scales need to be
standardized to facilitate comparison between trials in
meta-analyses.
Received December 17, 1997;
revision received February 6, 1998;
accepted February 9, 1998.
© 1998 American Heart Association, Inc.
Original Contributions
Assessment of Clinical Outcomes in Acute Stroke Trials
Abstract
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
Background and Purpose—Adequate
outcome assessment is crucial to randomized trials. We wished to assess
the types of outcomes used in acute stroke trials and the
appropriateness of these outcomes and their analyses.
Key Words: clinical trials • outcome assessment • stroke, acute
Introduction
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
A randomized
controlled trial aims to assess the effects of treatment, but if the
outcomes are measured inappropriately, the trial cannot provide
reliable results.1 2 3 Clinical outcome can be
classified into impairment (signs of underlying pathology), disability
(the functional results of impairment), and handicap (the social impact
of the disease).4 In this system, impairment is
the least clinically relevant to the patient and handicap the
most.4 5 Health-related quality of life may be
even more relevant than handicap, but it has proved difficult to
define.6 7 Since there is no simple relationship
between impairment, disability, handicap, or health-related quality of
life,8 9 outcome measures should only include
items relating to a single level because a mixture of levels is
conceptually confusing and difficult to interpret
clinically.10 In general, the more relevant the
outcome measure is to the patient, the more difficult it is to define
and assess.10
Subjects and Methods
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
Trial Selection
We included all completed randomized trials of
interventions in the acute phase of stroke (ie, started within 30 days
of onset) provided they were published in full in English before 1996.
Trials were identified from the Cochrane Stroke Group's comprehensive
register of trials, which is continually updated by means of multiple
overlapping search strategies (including electronic searching of
MEDLINE and other databases and hand searching of journals, books, and
reference lists).13 We limited ourselves to
studies published before 1996 because, at the time of our study, we
could not be sure that the database was complete for 1996.
The following data were collected on each trial: the year of
publication (for trials with more than one report, the year of the main
report was taken); the number of patients originally randomized; the
type of intervention tested; the reported clinical outcome measures and
whether they were categorized as primary or secondary (nonclinical
outcomes such as biochemical measurements were excluded); the method of
analysis of each outcome measure (continuous or
categorical/dichotomous, parametric or
nonparametric); the statistical significance of each trial
(results with P<0.05 were interpreted as significant); the
duration of follow-up; and whether outcomes were assessed blinded to
treatment allocation (if the report simply stated that the trial was
"double-blind," we assumed that the assessor was blind). Data were
only collected on these criteria in relation to the whole study
population at the end of follow-up. Subgroup analyses were
ignored.
After data collection, each outcome was classified into one of
the following categories: impairment, disability, handicap, death,
place of residence at follow-up, length of hospital stay, side effects
(recorded as "side effects," "adverse events," or
"complications" in the original trial reports), recurrent vascular
events (including deep venous thrombosis, myocardial infarction, and
recurrent stroke), and assessments of the patient's caregiver.
Measures were classified as impairment, disability, and handicap scales
according to the definitions of Wade,5 with the
exception of the Rankin/Modified Rankin Scale, which was classified as
a measure of disability rather than handicap.14
Quality of life measures were classified as measures of handicap
because there were too few of them to consider separately and, in terms
of importance to the patient, they are closer to handicap than
impairment. 
along with the references to support
our decisions on validity and reliability.
View this table:
[in a new window]
Table 1. Classification of Impairment, Disability, and
Handicap/Quality of Life Scales
The data were entered onto a computer database (DBase IV).
Percentages and the corresponding 95% confidence intervals were
calculated with the Confidence Interval Analysis Program, and
differences in proportions were assessed with the
2 test in Epi Info (version 5.01b).
Results
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
A total of 174 of 374 acute stroke trials on the database
fulfilled the selection criteria, most of which were published from
1975 onward. Trial size ranged from 8 to 1267 patients (median, 68
patients; interquartile range, 37 to 159 patients), and the duration of
follow-up ranged from 1 or 2 days to more than a year (median, 12
weeks; interquartile range, 3 to 24 weeks). The most frequently tested
interventions were hemodilution (18 trials), anticoagulants (18
trials), calcium antagonists (18 trials),
corticosteroids (17 trials), and
thrombolytic agents (15 trials). In most studies it was
impossible to tell which was the primary and which were the secondary
outcomes, and therefore all outcomes were considered together. In 69%
of trials the outcome was assessed blindly, 6% were not blinded, and
in 25% the blinding of outcome assessment was not clearly described in
the published report.
Table 2 demonstrates the number of
trials assessing each of the clinical outcomes considered. A
substantial minority of trials did not report data on death (24%) or
impairment (24%), while the majority of trials did not report data on
disability or handicap. Ten trials (6%) measured only biochemical
markers and did not assess death, impairment, disability, or handicap.
The proportion of trials measuring impairment was significantly higher
than that measuring disability (76% versus 42%;
P
0.00001). The most common impairment measures used were a
trialist's own scale (eg, signs better, worse, or unchanged) (47/174
trials, 27%), the Mathew/Modified Mathew Scale (28/174, 16%), the
National Institutes of Health Stroke Scale (11/174, 6%), the
Scandinavian Stroke Scale (9/174, 5%), and the Toronto Stroke
Score (7/174, 4%). For disability, the most common measures were the
Barthel Index (37/174, 21%), the trialist's own scale (20/174, 11%),
and the Rankin or Modified Rankin Scale (15/174, 9%). Several trials
used more than one measure of impairment or disability. Many of the
measures were inadequate (Table 2): of the trials that measured
impairment, only 35% (47/133) used a measure with proven reliability
or validity, while only 70% of trials (51/73) measuring disability
used an adequate measure. Most of the untested measures appeared to
have been developed by the trialists themselves.
View this table:
[in a new window]
Table 2. Number of Trials (n=174) Assessing Each Type of
Outcome . Impairment was assessed in
approximately 70% to 80% of trials regardless of size. Trials with
more than 100 patients were significantly more likely to assess
disability than smaller ones (60% versus 25%;
2=21.5, df=3,
P=0.00008), while larger trials were also more likely to
report deaths (2=18.15, df=3,
P=0.0004); approximately 80% of trials with more than 50
patients reported deaths compared with 60% of smaller trials. Figure 2 shows the changes in the proportions of
trials measuring each type of outcome over time. There was a trend for
more trials to record disability over time (P=0.16) and
for fewer to report death (P=0.1), but neither of these was
significant. However, the proportion of trials measuring impairment did
increase over time (2=21.42, df=7,
P=0.003), with fewer trials measuring it before 1964
compared with after (20% versus 80%).
View larger version (43K):
[in a new window]
Figure 1. Relationship between type of outcome measures used
and size of trial.
View larger version (38K):
[in a new window]
Figure 2. Relationship between type of outcome measures used
and year of study.
Table 3 shows how the main outcome
measures were analyzed. Two trials did not analyze the
results for impairment despite having recorded it, and one did not
analyze disability. Impairment and handicap measures were
primarily analyzed only as continuous data, while a greater
proportion of disability measures were analyzed as dichotomous
data (ie, patients were separated into those who were dead or dependent
and those alive and independent). Continuous data were mainly
analyzed with parametric statistical methods. Most of
the trials that analyzed dichotomous data simply reported the
percentages with a good or bad outcome in each group rather than an
odds ratio or relative risk. Of the 10 trials that did report a
relative risk, seven were published in 1995.
View this table:
[in a new window]
Table 3. Methods of Analysis Used for Impairment,
Disability, and Handicap Measures 2=3.3,
P=0.07; disability: 2=0.01,
P=0.9). Similarly, for continuous variables there was no
significant difference between the proportion of significant results
obtained with parametric and nonparametric methods
(impairment: 2=1.27, P=0.53;
disability: 2=0.54, P=0.76). There
was also no significant difference between the proportion of trials
with significant results when impairment was the outcome of interest
compared with when disability was analyzed.
Discussion
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
This study has shown that few trials in acute stroke have
measured the outcomes of most relevance to the patient. Death was
measured in most trials, but given the importance and robustness of
this outcome it should have been reported in all trials. Most trials
also measured impairment, which has the advantage of being easy to
measure and is relatively objective10 but has
limited clinical relevance. The patient's state is more than merely
the sum of his or her signs.12 Handicap and
quality of life are much more important to the patient, but such
measures are more difficult to define, more subjective, and more
difficult to validate.10 It is therefore not
surprising that only four trials recorded these and, of these, two
used measures of handicap that could be regarded as measuring mainly
disability (the Frenchay and Viitanen scales5 ).
Assessment of disability has been suggested to be the most feasible
compromise,35 but only 40% of trials assessed
this. 
do not suggest a
dramatic improvement in the type of outcome assessed over time. Our
classification of reliability and validity of outcome measures was
usually only defined by previous authors' conclusions. This was the
only feasible way to assess these, however, since it is difficult to
set uniform, rigorous criteria for validity and reliability. We also
defined adequate outcome measures as those with proven validity or
reliability and not both. In addition, one of the disability measures
defined as reliable (the Glasgow Outcome Scale, which was used in three
trials) has been validated in head injury but not in
stroke.27 We may therefore have been
overoptimistic in our assessment of reliability and validity. Finally,
other important characteristics of outcome measures were not
considered, such as sensitivity to change and feasibility. Measures
that take several hours to complete are not feasible for large
trials.
Acknowledgments
This study was supported by a Wellcome Trust research training
fellowship in clinical epidemiology (to Dr
Counsell). We thank Dr Peter Sandercock and the two anonymous reviewers
for their comments on the manuscript.
References
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
This article has been cited by other articles:
 |
|
 |
|
  T. J. Quinn, J. Dawson, J. S. Lees, T.-P. Chang, M. R. Walters, K. R. Lees, and for the GAIN and VISTA Investigators Time Spent at Home Poststroke: "Home-Time" a Meaningful and Robust Outcome Measure for Stroke Trials Stroke, January 1, 2008; 39(1): 231 - 233. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. J. Quinn, K. R. Lees, H.-G. Hardemark, J. Dawson, and M. R. Walters Initial Experience of a Digital Training Resource for Modified Rankin Scale Assessment in Clinical Trials Stroke, August 1, 2007; 38(8): 2257 - 2261. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. D. Patel, C. McKevitt, E. Lawrence, A. G. Rudd, and C. D. A. Wolfe Clinical determinants of long-term quality of life after stroke Age Ageing, May 1, 2007; 36(3): 316 - 322. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. M. Boyd, C. O. Weiss, J. Halter, K. C. Han, W. B. Ershler, and L. P. Fried Framework for Evaluating Disease Severity Measures in Older Adults With Comorbidity J. Gerontol. A Biol. Sci. Med. Sci., March 1, 2007; 62(3): 286 - 295. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G Gromadzka, I Sarzynska-Dlugosz, and A Czlonkowska IL1RN intron 2 polymorphism caused by variable number tandem repeats is associated with 1-year outcome in patients with ischaemic stroke J. Neurol. Neurosurg. Psychiatry, February 1, 2007; 78(2): 183 - 186. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. D. Patel, K. Tilling, E. Lawrence, A. G. Rudd, C. D. A. Wolfe, and C. McKevitt Relationships between long-term stroke disability, handicap and health-related quality of life. Age Ageing, May 1, 2006; 35(3): 273 - 279. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Song, C. Jerosch-Herold, R. Holland, M. d. L. Drachler, K. Mares, and I. Harvey Statistical methods for analysing Barthel scores in trials of poststroke interventions: a review and computer simulations Clinical Rehabilitation, April 1, 2006; 20(4): 347 - 356. [Abstract] [PDF]
|
|
|
|
 |
|
 C. A. Anderson, D. B. Arciniegas, and C. M. Filley Treatment of Acute Ischemic Stroke: Does It Impact Neuropsychiatric Outcome? J Neuropsychiatry Clin Neurosci, November 1, 2005; 17(4): 486 - 488. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. B. Young, C. J. Weir, K. R. Lees, and for the GAIN International Trial Steering Committe Comparison of the National Institutes of Health Stroke Scale With Disability Outcome Measures in Acute Stroke Trials Stroke, October 1, 2005; 36(10): 2187 - 2192. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Marti-Fabregas, M. Gomis, A. Arboix, A. Aleu, J. Pagonabarraga, R. Belvis, D. Cocho, J. Roquer, A. Rodriguez, M. D. Garcia, et al. Favorable Outcome of Ischemic Stroke in Patients Pretreated with Statins Stroke, May 1, 2004; 35(5): 1117 - 1121. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. T. Engelter, S. G. Wetzel, E. W. Radue, M. Rausch, A. J. Steck, and P. A. Lyrer The clinical significance of diffusion-weighted MR imaging in infratentorial strokes Neurology, February 24, 2004; 62(4): 574 - 580. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
German Stroke Study Collaboration Predicting outcome after acute ischemic stroke: An external validation of prognostic models Neurology, February 24, 2004; 62(4): 581 - 585. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T Hillen, S Davies, A G Rudd, T Kieselbach, and C D Wolfe Self ratings of health predict functional outcome and recurrence free survival after stroke J. Epidemiol. Community Health, December 1, 2003; 57(12): 960 - 966. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Als-Nielsen, W. Chen, C. Gluud, and L. L. Kjaergard Association of Funding and Conclusions in Randomized Drug Trials: A Reflection of Treatment Effect or Adverse Events? JAMA, August 20, 2003; 290(7): 921 - 928. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Di Carlo, M. Lamassa, M. Baldereschi, G. Pracucci, A. M. Basile, C. D.A. Wolfe, M. Giroud, A. Rudd, A. Ghetti, and D. Inzitari Sex Differences in the Clinical Presentation, Resource Use, and 3-Month Outcome of Acute Stroke in Europe: Data From a Multicenter Multinational Hospital-Based Registry Stroke, May 1, 2003; 34(5): 1114 - 1119. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. A. Mayer, K. T. Kreiter, D. Copeland, G. L. Bernardini, J. E. Bates, S. Peery, J. Claassen, Y. E. Du, and E. S. Connolly Jr. Global and domain-specific cognitive impairment and outcome after subarachnoid hemorrhage Neurology, December 10, 2002; 59(11): 1750 - 1758. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Weimar, T. Kurth, K. Kraywinkel, M. Wagner, O. Busse, R. L. Haberl, and H.-C. Diener Assessment of Functioning and Disability After Ischemic Stroke Stroke, August 1, 2002; 33(8): 2053 - 2059. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. J. Gladstone, S. E. Black, and A. M. Hakim Toward Wisdom From Failure: Lessons From Neuroprotective Stroke Trials and New Therapeutic Directions Stroke, August 1, 2002; 33(8): 2123 - 2136. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. G. Celani, T. A. Cantisani, E. Righetti, L. Spizzichino, and S. Ricci Different Measures for Assessing Stroke Outcome: An Analysis From the International Stroke Trial in Italy Stroke, January 1, 2002; 33(1): 218 - 223. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. G. Hamedani, C. K. Wells, L. M. Brass, W. N. Kernan, C. M. Viscoli, J. N. Maraire, I. A. Awad, and R. I. Horwitz A Quality-of-Life Instrument for Young Hemorrhagic Stroke Patients Stroke, March 1, 2001; 32(3): 687 - 695. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Lamassa, A. Di Carlo, G. Pracucci, A. M. Basile, G. Trefoloni, P. Vanni, S. Spolveri, M. C. Baruffi, G. Landini, A. Ghetti, et al. Characteristics, Outcome, and Care of Stroke Associated With Atrial Fibrillation in Europe : Data From a Multicenter Multinational Hospital-Based Registry (The European Community Stroke Project) Stroke, February 1, 2001; 32(2): 392 - 398. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. van Straten, R.J. de Haan, M. Limburg, and G.A.M. van den Bos Clinical Meaning of the Stroke-Adapted Sickness Impact Profile-30 and the Sickness Impact Profile-136 Stroke, November 1, 2000; 31(11): 2610 - 2615. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Buck, A. Jacoby, A. Massey, and G. Ford Evaluation of Measures Used to Assess Quality of Life After Stroke Stroke, August 1, 2000; 31(8): 2004 - 2010. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. W. Duncan, H. S. Jorgensen, and D. T. Wade Outcome Measures in Acute Stroke Trials : A Systematic Review and Some Recommendations to Improve Practice Stroke, June 1, 2000; 31(6): 1429 - 1438. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Di Carlo, M. Lamassa, G. Pracucci, A. M. Basile, G. Trefoloni, P. Vanni, C. D. A. Wolfe, K. Tilling, S. Ebrahim, and D. Inzitari Stroke in the Very Old : Clinical Presentation and Determinants of 3-Month Functional Outcome: A European Perspective Stroke, November 1, 1999; 30(11): 2313 - 2319. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. W. Duncan, D. Wallace, S. M. Lai, D. Johnson, S. Embretson, and L. J. Laster The Stroke Impact Scale Version 2.0 : Evaluation of Reliability, Validity, and Sensitivity to Change Stroke, October 1, 1999; 30 (10): 2131k - 2140. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||