Comments, Opinions, and Reviews

Outcomes Validity and Reliability of the Modified Rankin

Scale: Implications for Stroke Clinical Trials
A Literature Review and Synthesis
Jamie L. Banks, MSc, PhD; Charles A. Marotta, MD, PhD

Background and Purpose—The modified Rankin scale (mRS), a clinician-reported measure of global disability, is widely
applied for evaluating stroke patient outcomes and as an end point in randomized clinical trials. Extensive evidence on
the validity of the mRS exists across a large but fragmented literature. As new treatments for acute ischemic stroke are
submitted for agency approval, an appreciation of the mRS’s attributes, specifically its relationship to other stroke
evaluation scales, would be valuable for decision-makers to properly assess the impact of a new drug on treatment
paradigms. The purpose of this report is to assemble and systematically assess the properties of the mRS to provide
decision-makers with pertinent evaluative information.
Methods—A Medline search was conducted to identify reports in the peer-reviewed medical literature (1957–2006) that
provide information on the structure, validation, scoring, and psychometric properties of the mRS and its use in clinical
trials. The selection of articles was based on defined criteria that included relevance, study design and use of appropriate
statistical methods.
Results—Of 224 articles identified by the literature search, 50 were selected for detailed assessment. Inter-rater reliability
with the mRS is moderate and improves with structured interviews (␬ 0.56 versus 0.78); strong test-re-test reliability
(␬⫽0.81 to 0.95) has been reported. Numerous studies demonstrate the construct validity of the mRS by its relationships
to physiological indicators such as stroke type, lesion size, perfusion and neurological impairment. Convergent validity
between the mRS and other disability scales is well documented. Patient comorbidities and socioeconomic factors
should be considered in properly applying and interpreting the mRS. Recent analyses suggest that randomized clinical
trials of acute stroke treatments may require a smaller sample size if the mRS is used as a primary end point rather than
the Barthel Index.
Conclusions—Multiple types of evidence attest to the validity and reliability of the mRS. The reported data support the
view that the mRS is a valuable instrument for assessing the impact of new stroke treatments. (Stroke. 2007;38:1091-
1096.)
Key Words: cerebrovascular accident 䡲 disability evaluation 䡲 randomized controlled trials 䡲 rankin scale
䡲 reproducibility of results

T he modified Rankin Scale (mRS) is a clinician-reported

measure of global disability that has been widely applied
for evaluating recovery from stroke1,2 and as a primary end
be useful for decision-makers to properly assess their impact.
The purpose of this review is to assemble and systematically
assess the properties of the mRS to provide decision-makers
point in randomized clinical trials (RCTs) of emerging acute with pertinent evaluative information they need for
stroke treatments.3 The value of the mRS as a RCT end point decision-making.
has been examined in several investigations wherein propo-
nents emphasize the importance of the scale’s brevity, sim- Methods
plicity of use and interpretability in the context of stroke An initial literature search was conducted in MEDLINE from 1957
trials.4 Extensive evidence on the validity, reliability and (date of original Rankin Scale (RS) publication to December 2004,
sensitivity of the mRS exists across a broad but fragmented with updates through April 2006) to identify articles on (1) the
literature.5– 8 structure, scoring, and psychometric properties of the mRS and (2)
the effectiveness of the mRS as an end point in RCTs of acute stroke
As new stroke drugs are submitted for agency approval, an treatment. Search criteria included the term Rankin AND one or
in-depth understanding of the mRS in terms of its relationship more of each of the following MeSH terms: psychometrics, reliabil-
to other stroke evaluation scales and clinical outcomes would ity and validity, predictive value of tests, sensitivity and specificity,

Received August 29, 2006; final revision received October 5, 2006; accepted October 6, 2006.
From the Health Economics Research and Quality of Life Evaluation Services (J.L.B.), Abt Associates, Inc, Lexington, Mass; and Health Economics
and Outcomes Research (C.A.M.), Division of Medical Sciences, AstraZeneca Pharmaceuticals LP, Wilmington, Del.
Correspondence to Charles A. Marotta, MD, PhD, AstraZeneca LP, 1800 Concord Pike, Wilmington, DE 19850. E-mail
charles.marotta@astrazeneca.com
© 2007 American Heart Association, Inc.
Stroke is available at http://www.strokeaha.org DOI: 10.1161/01.STR.0000258355.23810.c6
1091 at University of Calgary on May 24, 2015
Downloaded from http://stroke.ahajournals.org/
1092 Stroke March 2007

TABLE 1. Original and Modified RSs

Grade Original RS1 mRS2
0 NA No symptoms at all
1 No significant disability: able to carry out all usual duties No significant disability: despite symptoms, able to carry out all usual
duties and activities
2 Slight disability: unable to carry out some of previous activities but able to Slight disability: unable to perform all previous activities but able to look
look after own affairs without assistance after own affairs without assistance
3 Moderate disability: requiring some help but able to walk Moderate disability: requiring some help but able to walk
without assistance without assistance
4 Moderately severe disability: unable to walk without assistance, and Moderately severe disability: unable to walk without assistance and unable
unable to attend to own bodily needs without assistance to attend to own bodily needs without assistance
5 Severe disability: bedridden, incontinent and requiring constant nursing Severe disability: bedridden, incontinent and requiring constant nursing
care and attention care and attention
6 NA Death*
NA indicates not applicable,
*Used commonly in clinical trials9

disability evaluation, statistics and randomized controlled trials. (IADL; eg, meal preparation, shopping, handling money)4
Other search terms including lesion volume, stroke type, end point and basic ADLs (BADL; eg, walking, dressing, grooming)
and design, were used to search for additional articles. Two review-
with emphasis on compromised motor function.10 The global
ers selected articles for assessment based on relevance to the study
objectives and (1) the use of appropriate statistical methods in nature of the mRS thereby allows the clinician to consider
assessing validity, and (2) analysis of the mRS as a primary end point nonphysical attributes essential to a person’s self-mainte-
in RCTs of acute ischemic stroke treatments. The initial search nance and well-being, such as cognition and language,4 social
identified 224 articles of which 50 were selected for a detailed functioning,11 and poststroke mood disturbances, particularly
assessment (40 relating mainly to validity/reliability, 2 to the original
Rankin and related scales, 8 related to RCTs). Additional articles
depression, that may contribute to perceived disability.12 This
providing information on the limitations of the mRS, and others were allowance for consideration of IADLs and other nonphysical
identified from bibliographies of the retrieved articles. characteristics distinguishes the mRS from BADL-specific
measures, such as the Barthel Index (BI).13†
Results
Historical Development Test-Retest Reliability
The history of the mRS dates from publication of the original Reliability refers to the extent to which a scale consistently and
Rankin Scale in 1957.1 Developed by Dr John Rankin in reproducibly measures the attributes it was intended to measure.
Glasgow, Scotland, the RS comprises 5 grades of stroke severity Test-retest reliability evaluates the consistency of results over
ranging from “no significant disability” to “severe disability” time in the absence of changes in the subject population and the
(Table 1), and was intended as a descriptive categorization of raters.14 The ␬ statistic indicates the extent of agreement among
functional recovery of cerebrovascular disease patients (⬎60 different sets of results not occurring by chance; a weighted ␬
years) at the time of discharge or transfer from the index adjusts for the extent of disagreement, eg, differences of 1 grade
hospitalization.* The mRS was published in 1988 and consists versus 2 grades of the scale.‡
of 6 categories (grades 0 to 5) rather than 5 for the RS; an Strong test-re-test reliability of the mRS was reported in 2
additional category, grade “6” denoting death, is usually incor- independent studies (Table 2). In 1 investigation, 2 raters
porated into the mRS for RCT purposes9 (Table 1). In the mRS: each graded 48 patients on 2 separate occasions with excel-
grade 1 of the original RS (“no significant disability”) is replaced lent consistency.15
by 2 grades, 0 and 1, with grade 0 describing patients without
symptoms and grade 1 describing patients without significant
Inter-Rater Reliability
Inter-rater reliability evaluates the consistency of results among
disability “despite symptoms.” This finer discrimination of mild
raters. Because mRS categories are broad and the assessments
strokes increases the usefulness of the mRS in evaluating RCTs
are subjective, variability may occur across clinical raters. In 3
of acute stroke interventions. Additionally, grade 2 in the mRS
separate studies, inter-rater reliability of the mRS ranged from
(“unable to perform all previous activities”) is more definitive
moderate to nearly perfect, indicated by the weighted ␬ (Table
compared with that grade of the RS (“unable to carry out some
2). In addition to those studies, strong inter-rater reliability has
of previous activities”).
also been reported for a German version of the mRS (␬⫽0.76).16
The mRS is heavily weighted toward global disability (in
Structured interviews have been shown to improve inter-rater
particular, physical disability) and the need for assistance.2,6,7
reliability of the mRS.5,15,17 When used in a RCT setting,
As a global disability measure, the broad categories of the
mRS (Table 1) subsume instrumental activities of daily living
†The Barthel Index ADL items are: feeding, bathing, grooming,
*The mRS was not developed for the acute assessment of stroke. For dressing, bowels, bladder, toilet use, transfers, mobility, stairs.13
example, the scale would have limited value if it were used in the acute ‡Strength of agreement for the ␬ statistic has been categorized as
setting as illustrated by assigning a patient with a mild stroke mRS grade follows: ⬍0⫽poor; 0 – 0.20⫽slight; 0.21– 0.40⫽fair; 0.41– 0.60⫽mod-
4 or 5 on the basis of the need for nursing care and supervision. erate; 0.61– 0.80⫽substantial; 0.81–1.00⫽almost perfect.67

Downloaded from http://stroke.ahajournals.org/ at University of Calgary on May 24, 2015

 Banks and Marotta Validity of the Modified Rankin Scale 1093

TABLE 2. Reliability of the mRS With and Without Structured Interviews

␬

Structured Interview # Raters N Interval Unweighted Weighted References

Test-Re-Test Reliability

No 2 48 1–2 weeks 0.81, 0.95 0.94, 0.99 15

Yes 0.84, 0.97 0.96, 0.99
Inter-Rater Reliability

No 34 100 NA 0.56 0.91 2

No 2 63 NA 0.44 0.78 5
Yes 58 0.70 0.93
No 15 117 NA 0.25 0.71 15
Yes 0.74 0.91
NA indicates not applicable

training and certification of raters should be considered to reduce location, type and extent of stroke injury are closely related to
inter-rater variability. short and longer-term disability (Table 3; detail in supplemental
Table I, available online at http://stroke.ahajournals.org). Nu-
Validity merous investigations have reported an increased risk of poor
Validity is the degree to which an instrument measures the outcome (defined as mRS ⬎2 or ⬎3) from discharge to 6
concept it was intended to measure. Construct validity is applied months for more severe types of stroke.19 –28 For example, a
when a gold standard does not exist. This type of validity study of 198 younger ischemic stroke patients showed that total
assessment uses multiple sources of comparison to test how anterior circulation infarction was an independent predictor of
accurately a measure captures the outcome it claims to measure mRS grade ⱖ2 or death at 3 months (P⫽0.011).22 Similar
in different contexts. Convergent (criterion) validity, a funda- results have been reported for older stroke populations.27,28
mental aspect of construct validity, measures the degree of Studies in small patient series have consistently shown
correlation between different measures of the same construct. significant relationships between lesion volume (measured by
Other forms of validity include predictive validity (ability to diffusion-weighted and other imaging methods) and mRS
predict future events)6,14 and theoretical validity (degree to grades (Table 3; detail in supplemental Table I), with larger
which results are consistent with a priori expectations).18 In this lesions predicting more severe disability.20,29 –34 As would be
report we focus on construct and convergent validity, clinical reasonably expected, improved brain perfusion and recanali-
sensitivity and limitations, and then consider the application of zation after thrombolytic therapy are also associated with
the mRS in clinical trials of acute ischemic stroke treatments. improved mRS disability outcomes.21,33,35,36 For example, in
a study of 177 acute ischemic stroke patients, recanalization
Construct Validity: Relationship to Stroke Severity within 5 hours post–recombinant tissue plasminogen activa-
Construct validity of the mRS has been affirmed by multiple tor treatment (in addition to National Institutes of Health
studies in which it has been consistently observed that the Stroke Scale [NIHSS] baseline scores and other factors)

TABLE 3. Summary of mRS Construct Validity

Predictors/Correlators of mRS Grades*†

Parameter Less severe More severe References

Stroke type Proximal MCA occlusion TACI, POCI, cortical and cardioembolic infarction, tandem 20, 23, 22, 24–26, 19, 27, 28, 46
ICA, MCA, and other arterial infarction
Lesion volume/degree Recanalization within 300 minutes Larger lesion volume, infarct size, continued volume 21, 29, 35, 32, 33, 3420, 29, 31–34, 36
of perfusion post–rt-PA, infarct size, improved brain expansion, lack of recanalization, absence of MCA
perfusion, smaller lesion volume blood flow
Impairment Lower NIHSS score at baseline, lower More severe NIHSS (with or without rt-PA), LAMS, CNS, 11, 20–23, 26, 28, 31, 33, 37–41, 43–45
baseline LAMS score Mathew Scale, Orgogozo Scale, and SSS scores; and,
impairment items: impaired consciousness at admission,
limb weakness, hemianopia, apraxia, perceptual deficits,
aphasia, consciousness commands, leg motor
impairment, left arm palsy
TACI indicates total anterior circulation infarction; ICA, internal carotid artery; MCA, middle cerebral artery; POCI, posterior circulation infarction; PACI, partial anterior
circulation infarction; rt-PA, recombinant tissue plasminogen activator; CNS, Canadian Neurological Scale; LAMS, Los Angeles Motor Scale; ESS, European Stroke
Scale; SSS, Scandanavian Stroke Scale.
*Significance of multivariate regression results expressed by P⬍0.05 and odds ratios with 95% CIs; results include a small number of correlation studies.
†mRS measured at 2 months to 1 year poststroke

Downloaded from http://stroke.ahajournals.org/ at University of Calgary on May 24, 2015

1094 Stroke March 2007

independently predicted mRS grade ⱕ2 at 3 months post- scoring methods provide distinctive information. As de-
stroke (odds ratio⫽4.11, 95% CI⫽2.42 to 6.95; P⬍0.001).21 scribed previously, the BI measures dependence in 10
Multivariate regression analyses have demonstrated that BADLs, whereas the mRS captures higher functioning in
acute impairment score (measured with the NIHSS,20 –22,37 addition to aspects of self-care. Eight of the ten BI domains
Canadian Neurological Scale,26,38 Mathew scale,31 Los An- reflect voluntary motor functions (see footnote†). However,
geles Motor Scale,39 or European Stroke Scale40) indepen- poststroke disability can also affect speech, language or
dently predict mRS grade at 2 months to 1 year poststroke cognitive function.49 For example, a patient with substantial
(detail in supplemental Table I). For example, Demchuk and communication problems may still score ⱖ90 on the BI. This
coworkers observed that patients with the mildest strokes “ceiling effect” is a major disadvantage of the BI relative to
(NIHSS score 1 to 5) compared with those with more severe a global disability instrument.9,45,46,49 –51 Data from a large
strokes (NIHSS scores 11 to 15, 16 to 20, ⬎20) had improved prospective cohort study of stroke patients demonstrated the
chances of achieving a favorable outcome defined by mRS limitations of the BI in mild stroke patients.45 The mRS was
grade 0 to 1 (95% CI⫽0.02 to 0.16, P⬍0.001; 95% CI⫽0.13 also reported to be more sensitive for distinguishing between
to 0.56, P⬍0.001; 95% CI⫽0.20 to 0.79, P⫽0.008, respec- mild and moderate disability, which suggests it may also be
tively).41 Transitions in mRS grades have also been shown to more sensitive to acute stroke treatment effects.45
correspond to transitions in NIHSS scores during stroke
recovery.11 Consistent with these observations, moderate to Clinical Sensitivity
strong correlations (Pearson r⫽0.60 to 0.86) between acute The clinical sensitivity or responsiveness of an instrument refers
impairment and mRS grades at discharge or follow-up have to its ability to detect a clinically important change.52 Limited
also been reported.38,40,42,43 Specific impairments that appear information is available regarding the sensitivity of the mRS to
to contribute significantly to the more severe mRS grades changes in disability levels after a stroke. In a rehabilitation
include those involving alertness, orientation, leg motor setting the sensitivity of 2 global disability measures, the mRS
function, following commands, and arm and hand function and the International Stroke Trial Measure (ISTM),53 and 2 ADL
(detail in supplemental Table I).23,44,45 measures, the BI and FIM,54 were tested and compared in a
nonrandom sample of 95 moderately disabled stroke patients.52
Convergent Validity: Relationship to Other A change of one mRS grade was considered to be clinically
Disability Scales significant based on the range of severity covered by the scale
Convergent validity of the mRS has been demonstrated by grades. Although the mRS was shown to be more sensitive than
comparisons with other disability scales used to evaluate stroke the ISTM (P⬍0.001), it was less sensitive than either the BI
patients, including the American Heart Association’s Stroke (P⬍0.002) or the FIM (P⬍0.005), with the latter being the most
Outcomes Classification (AHA.SOC), the BI, the motor com- sensitive of the 4 instruments under the conditions of the study.
ponent of the Functional Independence Measure (m-FIM), the On this basis, the authors recommended the use of ADL scales
Short Form-36 (SF-36), and the Stroke Impact Scale (SIS; Table for stroke intervention trials. However, other analyses (see
4; detail in supplemental Table II, available online at http:// below) based on treatment effects anticipated in the acute stroke
stroke.ahajournals.org).8,11,43,46 – 48 For example, using stroke setting suggest that those effects may be better detected using the
registry data, Kwon and coworkers quantified the frequency mRS.
distribution of BI scores relative to mRS grades and showed that
the highest BI scores (95 to 100, indicating excellent to complete Limitations
recovery) correspond to mRS grades 0, 1 and 2.46 A number of limitations apply to the mRS when used to measure
The particular relationship between the BI and the mRS disability outcome after stroke. A substantial literature docu-
has been explored in detailed investigations,7,8,46,48 reflecting ments the negative effect of patient comorbidities (including
the common use of these scales as end points in acute stroke cardiovascular disease, diabetes, and arthritis),55–58 surgery,59
treatment RCTs. Post hoc and quantitative translations be- and socioeconomic factors60 on physical functioning, cognitive
tween mRS grades and BI scores have been derived to abilities,58 and overall health status, factors that may have a
facilitate comparisons among trial results.7,48 Whereas the direct impact on the mRS.59 This is particularly important
correlation between trial end points of both scales is strong because comorbidities are common in stroke patients and the
(r⫽0.89, P⬍0.001),43 their different structures, domains and incidence of stroke in socioeconomically disadvantaged popula-

TABLE 4. Summary of Convergent Validity

Parameter Disability Scale*† References
Ability to discriminate mRS grades SIS-Participation: 3 grades 47
BI, mFIM: 3–4 grades 46
SIS-16, SF-36 PF/SF: 4 grades 11, 43, 47
Association/correlation/concordance with mRS grades and grade shifts BI, AHA.SOC (for moderate stroke), Lawton IADL 11, 43
Cross-calibration of scores BI 7
PF indicates Physical Functioning; SF, Social Functioning.
*Statistical significance for all results reported at P⬍0.05.
†mRS measured at 1– 6 months poststroke

Downloaded from http://stroke.ahajournals.org/ at University of Calgary on May 24, 2015

 Banks and Marotta Validity of the Modified Rankin Scale 1095

tions is especially high.61 It is essential for the clinician to take statistical analysis that is most likely to accurately reflect the
these various attributes into account to avoid misapplication and effect of treatment.
misinterpretation of the mRS.
Acknowledgements
End Point in Clinical Trials The authors gratefully acknowledge the financial sponsorship of
The mRS has been used often as an end point in RCTs of acute AstraZeneca LP. We greatly appreciate the contributions of Chris
ischemic stroke treatments based on its straightforward applica- Pashos, PhD for his review and comments and of Yoko Komatsu-
zaki, MPH, Er Chen, MPP, Yu-Chen Yeh, MS and Cassie Cai, MS,
tion, acceptable inter-rater reliability, and ability to discriminate MS for their administrative and editorial assistance.
levels of stroke disability.45,46 Studies have found that the sample
size requirements of trials using mRS-based end points are Disclosures
smaller than BI-based end points without the loss of statistical None.
power.50,62 For example, the sample required for a neuropro-
tectant RCT using a mRS end point (dichotomized at grade ⱕ1) References
was estimated to be 38% of that required for a BI end point 1. Rankin L. Cerebral vascular accidents in patients over the age of 60. II.
(dichotomized at ⱖ60).50 Prognosis. Scott Med J. 1957;2:200 –215.
2. van Swieten JC, Koudstaal PJ, Visser MC, Schouten HJ, van GJ. Inter-
Optimizing mRS end points for acute stroke treatment observer agreement for the assessment of handicap in stroke patients.
RCTs involves careful definition and appropriate statistical Stroke. 1988;19:604 – 607.
analyses. A “favorable” outcome defined as mRS grade ⱕ1 3. Sulter G, Steen C, De KJ. Use of the Barthel index and modified Rankin
scale in acute stroke trials. Stroke. 1999;30:1538 –1541.
or ⱕ2 was estimated to be more powerful than dichotomiza-
4. De Haan R, Limburg M, Bossuyt P, Van Der MJ, Aaronson N. The
tion at higher grades.3,63,64 The importance of the cut-point for clinical meaning of Rankin ‘handicap’ grades after stroke. Stroke. 1995;
dichotomization was illustrated in a post hoc analysis of the 26:2027–2030.
ECASS II trial of alteplase. This study found that expanding 5. Wilson JT, Hareendran A, Grant M, Baird T, Schulz UG, Muir KW, Bone
I. Improving the assessment of outcomes in stroke: use of a structured
the definition of “favorable” outcome from mRS grade ⱕ1 to interview to assign grades on the modified Rankin Scale. Stroke. 2002;
grade ⱕ2 changed a statistically insignificant result to a 33:2243–2246.
significant one.3,64 Concerns over dichotomized end points 6. Wade DT. Measurement in Neurological Rehabilitation. New York:
center on the risk of failure to detect the impact of treatment. Oxford University Press; 1992.
7. Wolfe CD, Taub NA, Woodrow EJ, Burney PG. Assessment of scales of
In a prospective study of 459 stroke patients, 116 subjects had disability and handicap for stroke patients. Stroke. 1991;22:1242–1244.
transitioned from a baseline mRS score of 5 to a 3-month 8. Duncan PW, Wallace D, Lai SM, Johnson D, Embretson S, Laster LJ.
score of 4 or 3; or, from a baseline score of 4 to a 3-month The stroke impact scale version 2.0. Evaluation of reliability, validity, and
score of 3.11 If those observations apply to RCTs, the reported sensitivity to change. Stroke. 1999;30:2131–2140.
9. Uyttenboogaart M, Stewart RE, Vroomen PC, De KJ, Luijckx GJ. Opti-
shifts in mRS grades of more severe stroke patients toward mizing cutoff scores for the Barthel index and the modified Rankin scale
reduced disability may not be captured by an end point for defining outcome in acute stroke trials. Stroke. 2005;36:1984 –1987.
defined as mRS ⱕ2,11 even though all mRS grade transitions 10. Bonita R, Beaglehole R. Recovery of motor function after stroke. Stroke.
1988;19:1497–1500.
are considered to be clinically meaningful.65
11. Lai SM, Duncan PW. Stroke recovery profile and the Modified Rankin
Transition across the entire mRS grade spectrum has been assessment. Neuroepidemiology. 2001;20:26 –30.
proposed as a more comprehensive measure of impact for 12. Dennis M, O’Rourke S, Lewis S, Sharpe M, Warlow C. Emotional
acute stroke interventions.11,66 This type of analysis uses the outcomes after stroke: factors associated with poor outcome. J Neurol
Neurosurg Psychiatry. 2000;68:47–52.
entire data set and has the advantage of reflecting simulta-
13. Mahoney F, Barthel D. Functional evaluation: the Barthel Index. Md State
neously the risk and benefit of an intervention compared with Med J. 1965;14:61– 65.
an analysis based on a dichotomized end point. The SAINT I 14. McDowell I, Newell C. Measuring Health: A Guide to Rating Scales and
trial of the neuroprotectant NXY-059 for acute ischemic Questionnaires (2nd edition). New York: Oxford University Press; 1996.
15. Wilson JT, Hareendran A, Hendry A, Potter J, Bone I, Muir KW.
stroke exemplifies this approach.65 In this trial, the primary Reliability of the modified Rankin Scale across multiple raters: benefits
end point was prespecified as the mRS grade at 90 days (or of a structured interview. Stroke. 2005;36:777–781.
last rating) with the analysis based on the overall difference in 16. Berger K, Weltermann B, Kolominsky-Rabas P, Meves S, Heuschmann
grade distribution of the 2 treatment groups. P, Bohner J, Neundorfer B, Hense HW, Buttner T. The reliability of
stroke scales: the German version of NIHSS, ESS and Rankin scales.
Fortschr Neurol Psychiatr. 1999;67:81–93.
Summary and Conclusions 17. Newcommon NJ, Green TL, Haley E, Cooke T, Hill MD. Improving the
A large and diverse literature supports the view that the mRS is assessment of outcomes in stroke: use of a structured interview to assign
a valid and clinically relevant instrument for assessing recovery grades on the modified Rankin Scale. Stroke. 2003;34:377–378.
18. Ryan M, Scott DA, Reeves C, Bate A, van Teijlingen ER, Russell EM,
from stroke and is a valuable end point for RCTs of acute Napper M, Robb CM. Eliciting public preferences for healthcare: a
ischemic stroke treatments. The scale consists of well-defined systematic review of techniques. Health Technol Assess. 2001;5:1–186.
and easily understood grades that describe the range of global 19. Dawson SL, Manktelow BN, Robinson TG, Panerai RB, Potter JF. Which
disability. When properly administered, the mRS exhibits a parameters of beat-to-beat blood pressure and variability best predict
early outcome after acute ischemic stroke? Stroke. 2000;31:463– 468.
strong relationship with clinical measurements of stroke severity 20. Derex L, Nighoghossian N, Hermier M, Adeleine P, Berthezene Y,
in addition to other disability and outcomes end points. How- Philippeau F, Honnorat J, Froment JC, Trouillas P. Influence of pre-
ever, clinicians should be aware that mRS grading may be treatment MRI parameters on clinical outcome, recanalization and infarct
size in 49 stroke patients treated by intravenous tissue plasminogen
affected by a variety of factors, including patient comorbidities
activator. J Neurol Sci. 2004;225:3–9.
and socioeconomic status. Use of the mRS in RCTs may benefit 21. Molina CA, Alexandrov AV, Demchuk AM, Saqqur M, Uchino K,
from a carefully defined end point as well as from a form of varez-Sabin J. Improving the predictive accuracy of recanalization on

Downloaded from http://stroke.ahajournals.org/ at University of Calgary on May 24, 2015

1096 Stroke March 2007

stroke outcome in patients treated with tissue plasminogen activator. 43. Lai SM, Duncan PW. Evaluation of the American Heart Association
Stroke. 2004;35:151–156. Stroke Outcome Classification. Stroke. 1999;30:1840 –1843.
22. Nedeltchev K, der Maur TA, Georgiadis D, Arnold M, Caso V, Mattle 44. Appelros P, Nydevik I, Viitanen M. Poor outcome after first-ever stroke:
HP, Schroth G, Remonda L, Sturzenegger M, Fischer U, Baumgartner predictors for death, dependency, and recurrent stroke within the first
RW. Ischaemic stroke in young adults: predictors of outcome and recur- year. Stroke. 2003;34:122–126.
rence. J Neurol Neurosurg Psychiatry. 2005;76:191–195. 45. Weimar C, Kurth T, Kraywinkel K, Wagner M, Busse O, Haberl RL,
23. Paciaroni M, Arnold P, Van MG, Bogousslavsky J. Severe disability at Diener HC. Assessment of functioning and disability after ischemic
hospital discharge in ischemic stroke survivors. Eur Neurol. 2000;43: stroke. Stroke. 2002;33:2053–2059.
30 –34. 46. Kwon S, Hartzema AG, Duncan PW, Min-Lai S. Disability measures in
24. Petty GW, Brown RD Jr, Whisnant JP, Sicks JD, O’Fallon WM, Wiebers stroke: relationship among the Barthel Index, the Functional Independence
DO. Ischemic stroke subtypes: a population-based study of functional Measure, and the Modified Rankin Scale. Stroke. 2004;35:918–923.
outcome, survival, and recurrence. Stroke. 2000;31:1062–1068. 47. Lai SM, Perera S, Duncan PW, Bode R. Physical and social functioning
25. Pinto AN, Melo TP, Lourenco ME, Leandro MJ, Brazio A, Carvalho L, after stroke: comparison of the Stroke Impact Scale and Short Form-36.
Franco AS, Ferro JM. Can a clinical classification of stroke predict Stroke. 2003;34:488 – 493.
complications and treatments during hospitalization? Cerebrovasc Dis. 48. Parmigiani G, Ashih HW, Samsa GP, Duncan PW, Lai SM, Matchar DB.
1998;8:204 –209. Cross-calibration of stroke disability measures: Bayesian analysis of
26. Sumer MM, Ozdemir I, Tascilar N. Predictors of outcome after acute longitudinal ordinal categorical data using negative dependence. J Am
ischemic stroke. Acta Neurol Scand. 2003;107:276 –280. Stat Assoc. 2003;98:273–281.
27. Tei H, Uchiyama S, Ohara K, Kobayashi M, Uchiyama Y, Fukuzawa M. 49. Kelly-Hayes M, Robertson JT, Broderick JP, Duncan PW, Hershey LA,
Deteriorating ischemic stroke in 4 clinical categories classified by the Roth EJ, Thies WH, Trombly CA. The American Heart Association
Oxfordshire Community Stroke Project. Stroke. 2000;31:2049 –2054. Stroke Outcome Classification. Stroke. 1998;29:1274 –1280.
28. Ueda T, Sakaki S, Kumon Y, Ohta S. Multivariable analysis of predictive 50. Young FB, Lees KR, Weir CJ. Strengthening acute stroke trials through
factors related to outcome at 6 months after intra-arterial thrombolysis for optimal use of disability end points. Stroke. 2003;34:2676 –2680.
acute ischemic stroke. Stroke. 1999;30:2360 –2365. 51. Duncan PW, Lai SM, Bode RK, Perera S, DeRosa J. Stroke Impact Scale-16:
29. Kluytmans M, van Everdingen KJ, Kappelle LJ, Ramos LM, Viergever A brief assessment of physical function. Neurology. 2003;60:291–296.
MA, van der GJ. Prognostic value of perfusion- and diffusion-weighted 52. Dromerick AW, Edwards DF, Diringer MN. Sensitivity to changes in
MR imaging in first 3 days of stroke. Eur Radiol. 2000;10:1434 –1441. disability after stroke: a comparison of four scales useful in clinical trials.
30. Barber PA, Parsons MW, Desmond PM, Bennett DA, Donnan GA, Tress J Rehabil Res Dev. 2003;40:1– 8.
53. Lindley R. Very simple methods of assessing outcome after stroke can be
BM, Davis SM. The use of PWI and DWI measures in the design of
valid. Cerebrovasc Dis. 1992;2:209.
“proof-of-concept” stroke trials. J Neuroimaging. 2004;14:123–132.
54. Guide for the Uniform Data Set for Medical Rehabilitation. Uniform Data
31. Vila N, Filella X, Deulofeu R, Ascaso C, Abellana R, Chamorro A.
System for Medical Rehabilitation. 1993.
Cytokine-induced inflammation and long-term stroke functional
55. Lai S, Studenski S, Duncan PW, Perera S. Persisting consequences of stroke
outcome. J Neurol Sci. 1999;162:185–188.
measured by the Stroke Impact Scale. Stroke. 2002;249:1160–1167.
32. Parsons MW, Li T, Barber PA, Yang Q, Darby DG, Desmond PM,
56. Lalonde L, O’Connor A, Joseph L, Growver SA, Group CCCA. Health-
Gerraty RP, Tress BM, Davis SM. Combined (1)H MR spectroscopy and
related quality of life in cardiac patients with dyslipidemia and hyper-
diffusion-weighted MRI improves the prediction of stroke outcome. Neu-
tension. Qual Life Res. 2004;13:793– 804.
rology. 2000;55:498 –505.
57. MacKenzie AE, Chang AM. Predictors of quality of life following stroke.
33. Lev MH, Segal AZ, Farkas J, Hossain ST, Putman C, Hunter GJ, Budzik
Disabil Rehabil. 2002;24:259 –265.
R, Harris GJ, Buonanno FS, Ezzeddine MA, Chang Y, Koroshetz WJ,
58. Nichols-Larsen DS, Clark PC, Zeriongue A, Greenspan A, Blanton.
Gonzalez RG, Schwamm LH. Utility of perfusion-weighted CT imaging
Factors influencing stroke survivors’ quality of life during subacute
in acute middle cerebral artery stroke treated with intra-arterial
recovery. Stroke. 2005;36:1480 –1484.
thrombolysis: prediction of final infarct volume and clinical outcome. 59. Rojas M, Silver A, Llewellyn C, Rances L. A study of adverse occur-
Stroke. 2001;32:2021–2028. rences and major functional impairment following surgery. Agency for
34. van Everdingen KJ, van der GJ, Kappelle LJ, Ramos LM, Mali WP. Healthcare Research and Quality, Rockville, MD, US Department of
Diffusion-weighted magnetic resonance imaging in acute stroke. Stroke. Health and Human Services, May 2005.
1998;29:1783–1790. 60. Horner RD, Swanson JW, Bosworth HB, Matchar DB. VA Acute Stroke
35. Chalela JA, Kang DW, Luby M, Ezzeddine M, Latour LL, Todd JW, Study Team (VAST). Effects of race and poverty on the process and
Dunn B, Warach S. Early magnetic resonance imaging findings in outcome of inpatient rehabilitation services among stroke patients. Stroke.
patients receiving tissue plasminogen activator predict outcome: insights 2003;34:1027–1031.
into the pathophysiology of acute stroke in the thrombolysis era. Ann 61. Gresham GE, Phillips TF, Wolf PA, McNamara PM, Kannel WB,
Neurol. 2004;55:105–112. Dawber TR. Epidemiologic profile of long-term stroke disability: the
36. Liu Y, Karonen JO, Vanninen RL, Nuutinen J, Koskela A, Soimakallio S, Framingham Study. Arch Phys Med Rehabil. 1979;60:487– 491.
Aronen HJ. Acute ischemic stroke: predictive value of 2D phase-contrast 62. Sacco RL, DeRosa JT, Haley EC Jr, Levin B, Ordronneau P, Phillips SJ,
MR angiography–serial study with combined diffusion and perfusion MR Rundek T, Snipes RG, Thompson JL. Glycine antagonist in neuropro-
imaging. Radiology. 2004;231:517–527. tection for patients with acute stroke: GAIN Americas: a randomized
37. Appelros P, Terent A. Characteristics of the National Institute of Health controlled trial. JAMA. 2001;285:1719 –1728.
Stroke Scale: results from a population-based stroke cohort at baseline 63. Broderick JP, Lu M, Kothari R, Levine SR, Lyden PD, Haley EC, Brott
and after one year. Cerebrovasc Dis. 2004;17:21–27. TG, Grotta J, Tilley BC, Marler JR, Frankel M. Finding the most
38. Fiorelli M, Alperovitch A, Argentino C, Sacchetti ML, Toni D, Sette G, powerful measures of the effectiveness of tissue plasminogen activator in
Cavalletti C, Gori MC, Fieschi C; Italian Acute Stroke Study Group. the NINDS tPA stroke trial. Stroke. 2000;31:2335–2341.
Prediction of long-term outcome in the early hours following acute 64. Hacke W, Kaste M, Fieschi C, von KR, Davalos A, Meier D, Larrue V,
ischemic stroke. Arch Neurol. 1995;52:250 –255. Bluhmki E, Davis S, Donnan G, Schneider D, ez-Tejedor E, Trouillas P.
39. Llanes JN, Kidwell CS, Starkman S, Leary MC, Eckstein M, Saver JL. Randomised double-blind placebo-controlled trial of thrombolytic
The Los Angeles Motor Scale (LAMS): a new measure to characterize therapy with intravenous alteplase in acute ischaemic stroke (ECASS II).
stroke severity in the field. Prehosp Emerg Care. 2004;8:46 –50. Second European-Australasian Acute Stroke Study Investigators. Lancet.
40. Hantson L, De Weerdt W, De Keyser J, Diener HC, Franke C, Palm R, 1998;352:1245–1251.
Van Orshoven M, Schoonderwalt H, De Klippel N, Herroelen L. The 65. Lees KR, Zivin JA, Ashwood T, Davalos A, Davis SM, Diener HC,
European Stroke Scale. Stroke. 1994;25:2215–2219. Grotta J, Lyden P, Shuaib A, Hardemark HG, Wasiewski WW. NXY-059
41. Demchuk AM, Tanne D, Hill MD, Kasner SE, Hanson S, Grond M, for acute ischemic stroke. N Engl J Med. 2006;354:588 – 600.
Levine SR. Predictors of good outcome after intravenous tPA for acute 66. Saver JL. Number needed to treat estimates incorporating effects over the
ischemic stroke. Neurology. 2001;57:474 – 480. entire range of clinical outcomes: novel derivation method and application to
42. De Haan R, Horn J, Limburg M, Van Der MJ, Bossuyt P. A comparison thrombolytic therapy for acute stroke. Arch Neurol. 2004;61:1066–1070.
of five stroke scales with measures of disability, handicap, and quality of 67. Feinstein AR. Clinimetrics. New Haven, Connecticut: Yale University
life. Stroke. 1993;24:1178 –1181. Press; 1987.

Downloaded from http://stroke.ahajournals.org/ at University of Calgary on May 24, 2015

Outcomes Validity and Reliability of the Modified Rankin Scale: Implications for Stroke
Clinical Trials: A Literature Review and Synthesis
Jamie L. Banks and Charles A. Marotta

Stroke. 2007;38:1091-1096; originally published online February 1, 2007;

doi: 10.1161/01.STR.0000258355.23810.c6
Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 2007 American Heart Association, Inc. All rights reserved.
Print ISSN: 0039-2499. Online ISSN: 1524-4628

The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://stroke.ahajournals.org/content/38/3/1091

Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Stroke can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.
Once the online version of the published article for which permission is being requested is located, click
Request Permissions in the middle column of the Web page under Services. Further information about this
process is available in the Permissions and Rights Question and Answer document.

Reprints: Information about reprints can be found online at:

http://www.lww.com/reprints

Subscriptions: Information about subscribing to Stroke is online at:

http://stroke.ahajournals.org//subscriptions/

Downloaded from http://stroke.ahajournals.org/ at University of Calgary on May 24, 2015