INTRODUCTION
Rehabilitation professionals have recently been encouraged to use time-limited walk tests, such as the 6-minute walk test (6MWT), to evaluate functional walking capacity across the care continuum poststroke owing to the quality of the tests' measurement properties and perceived utility.1–3 Use of standardized assessments as a cornerstone of evidence-informed physical therapy practice has been well described.2 , 4 The existence of assessment recommendations, however, does not ensure their implementation in clinical practice. Based on self-report data, an estimated 11% of physical therapists use the 6MWT and 26% use the 2-minute walk test (2MWT) to evaluate walking poststroke.5 Physical therapists who have the primary responsibility for the rehabilitation of walking poststroke6 have described how knowledge of psychometric properties, and, more importantly, perceived clinical relevance of these tests influences their decisions to adopt them into routine clinical practice.1 , 7 , 8 To date, research evidence of the speed and distance requirements for community ambulation9 and walk test normative values10 has been synthesized to improve the interpretability of walk test results. Evidence of reliability, construct validity, minimal detectable change, and minimal clinically important difference is needed to provide convincing evidence of the quality and clinical relevance of these tests.
The knowledge-to-action (KTA) framework is a conceptual model of knowledge translation developed by Graham and colleagues11 to guide thinking about the process of knowledge creation and application. The knowledge creation funnel in the KTA model is used to illustrate the filtering process required to develop knowledge products or tools tailored to the user. First-generation knowledge at the base of the funnel refers to the numerous individual sources of information (eg, research articles, and reports) on a topic that are of variable quality and time-consuming to acquire. Second-generation knowledge, or knowledge synthesis, is described as an essential precursor to the development of user-friendly knowledge tools, such as evidence-based algorithms, guides, and guidelines, that will help persuade and prepare physical therapists to adopt time-limited walk tests into clinical practice. Based on the KTA process, synthesis and critical appraisal of the literature is necessary to inform the development of a guideline on the use of time-limited walk tests in people poststroke.
The objective of the study was to appraise and synthesize the research literature describing (1) reliability, measurement error, construct validity, and sensitivity to change; (2) walk test protocol; and (3) the effect of walk test protocol elements (eg, walkway length and encouragement) on test performance for time-limited walk tests in adults poststroke. In addition, our review aimed to identify gaps in the evaluation of measurement properties of time-limited walk tests, and to identify considerations for the administration and interpretation of performance on time-limited walk tests poststroke to enhance acceptance, utility, and value for practicing clinicians.
METHODS
Overview
A systematic review was conducted according to a review protocol that was developed by the research team. The PRISMA checklist was used to guide reporting.12
Search Strategy
We searched 7 electronic databases (MEDLINE (Ovid), EMBASE, PubMed, CINAHL, Scopus, PEDro, and The Cochrane Library ) from 1946 to July 2013. Search strategies for each database were developed with input from an information specialist. Search terms included cerebrovascular accident, stroke, 6-minute walk test, and a wide variety of terms associated with walk tests (see PubMed search strategy in the Appendix). No limitations were applied during the search. The principal investigator's personal collection of research literature and reference lists of included studies were reviewed for potentially relevant articles.
Selection Criteria
Studies were considered eligible if (1) participants were adults 18 years or older poststroke; (2) the study reported on reliability, measurement error, construct validity, and sensitivity to change, or the effect of walkway length, encouragement, walking aids, and practice trials on performance of time-limited walk tests (for construct validity, studies reporting associations between walk test performance and other variables, regardless of whether this was framed as validity testing, were included); (3) test duration and track distance were reported to enable test replication; and (4) the report was written in English, French, or Spanish. Because the literature was extensive, we only included studies reporting unadjusted correlations and associated P -values or confidence intervals (CIs) between walk test performance and measures of motor function, aerobic capacity, balance, balance self-efficacy, strength, walking, stairs, mobility, physical activity, participation, health-related quality of life, and discharge destination for construct validity; and studies reporting minimal detectable change (MDC), standard error of measurement (SEM), and SEM% for measurement error. We excluded studies in which (1) less than 80% of the study sample comprised people poststroke; (2) the walk test was completed on a treadmill or embedded within another test; (3) the study objective was to validate the comparator measure or where studies assessed the correlation of change scores (for construct validity); and (4) the study was a conference proceeding, dissertation, case report/series, or limited to abstract form.
Study Selection and Data Extraction
Two authors independently screened titles and abstracts. A single author (NMS or PT) determined the inclusion of potentially relevant studies, and extracted data on general study information, study and participant characteristics, walk test protocol, and results from included studies. To ensure data accuracy, a single author (PT) randomly selected and verified data extracted from 30% of included studies.
Method of Quality Assessment
The methodological quality of included studies was assessed using the COnsensus-based Standards for the selection of health Measurements INstruments (COSMIN) critical appraisal tool.13 The first version of COSMIN comprises checklists that each relate to a specific measurement property (reliability, measurement error, hypothesis testing) or the interpretability and generalizability of studying findings. Operational definitions were developed to optimize scoring consistency. For example, for reliability and measurement error, we defined a retest time interval (item 6) over which patient stability would be assumed for 3 recovery phases poststroke as 1 day or less (acute), 5 days or less (subacute), and 3 weeks or less (chronic) based on results from longitudinal studies of walking14 , 15 and consensus among authors. Adequate sample sizes were defined as 25 or more and 30 or more for assessing reliability/measurement error and construct validity, respectively, based on minimum sample sizes required to reproduce reliability and validity estimates from larger datasets.16 Type, side, severity, and time poststroke were identified as disease characteristics relevant to the generalizability of study findings. For stroke severity, reporting of scores on a measure of stroke severity, motor function, or functional burden (ie, Functional Independence Measure17 ) was considered acceptable. Irrelevant checklist items were removed. The research team developed a checklist assessing sensitivity to change based on the format of the COSMIN checklists.13 Checklist response options were “yes/no” for 32 items and 27 items had a third response option of “can't tell.” A single author (NMS or PT) assessed the methodological quality of included studies.
During the review, COSMIN developers published a new scoring system that classified each measurement property as excellent, good, fair, or poor based on the lowest score reported on the corresponding checklist.18 Thus, a single author additionally applied the new rating system to studies examining reliability and measurement error. Throughout the review, an author not involved in the quality assessment was consulted to resolve uncertainty.
Data Synthesis and Analysis
The intraclass correlation coefficients (ICCs) used to evaluate reliability were interpreted as excellent (ICC ≥ 0.75), acceptable (ICC >0.40 to <0.75) or poor (ICC ≤ 0.40).19 To enable comparison, SEM and minimal detectable change at the 90% confidence level (MDC90 ) were computed across studies reporting test-retest reliability estimates and standard deviation of baseline score using the following equations: 1 20 and .20 For construct validity, constructs were classified using the International Classification of Functioning, Disability and Health.21 Correlation coefficients were interpreted as strong (≥0.70), moderate (0.50-0.69), weak (0.30-0.49), or negligible (<0.30).22 Simple correlations were derived from R 2 values where reported.
Time poststroke was determined by the range/interquartile range or by the mean/median values if the range was not reported. Study participants were considered to be in the acute, subacute, or chronic phase poststroke if they were less than 1 month, 1 to 6 months, or more than 6 months poststroke, respectively. To enable comparison across studies, results were converted to a common metric unit, frequency data were converted to percentages, and values were rounded to a consistent decimal place.
RESULTS
Study Selection
The search and article selection results are presented in Figure 1 . Of the 12 180 records identified, 43 articles23–65 representing 42 studies were eligible and included in the review. Of the 13 authors contacted to clarify and/or obtain select data not reported in published studies, 10 authors (77%) responded with requested information.
Figure 1.: Process of study selection (PRISMA).
Study Characteristics
All included articles were written in English. Five time-limited walk tests, including the 2-, 3-, 5-, 6-, and 12-minute walk test (2MWT, 3MWT, 5MWT, 6MWT, and 12MWT, respectively) were identified. The 6MWT was most commonly evaluated (n = 36, 82%). Table 1 provides a cross-tabulation of the number of evaluations of each measurement property, and the effect of turning direction, walkway length, practice trials, and walking aids on test performance by walk test.
Table 1. -
Frequency of Evaluations of Measurement Properties by Walk Test
Measurement Property or Protocol Element Examined
Number of Evaluations
2MWT
3MWT
5MWT
6MWT
12MWT
Total
Cross-sectional construct validity
–
1
2
32
1
36
Reliability
2
1
1
6
1
11
Measurement error
1
–
1
4
–
6
Predictive validity
–
–
–
3
–
3
Sensitivity to change
–
–
–
–
1
1
Effect of turning direction
–
–
–
1
–
1
Effect of walkway length
–
–
–
1
–
1
Effect of practice trials
–
–
–
1
–
1
Effect of walking aids
–
–
–
1
–
1
Number of articles
2a
1a
3a
36a
2a
Abbreviations: 2MWT, 2-minute walk test; 3MWT, 3-minute walk test; 5MWT, 5-minute walk test; 6MWT, 6-minute walk test; 12MWT, 12-minute walk test.
a Select articles reported on more than one measurement property and more than one test.
Appraisal of Study Methodology
Figures 2, 3, and 4 present the item-level COSMIN scores for articles assessing reliability, measurement error, and construct validity, respectively. Across evaluations of reliability and measurement error, common methodological issues were uncertainty of whether administrations were independent (ie, performance on the first test was not provided to the evaluator or participant at retest) (reliability [R] 100%; measurement error [ME] 100%), inadequate sample size (R 73%; ME 67%), and uncertainty whether participants were stable in the interim period (R 55%; ME 50%). The 3 most common methodological issues among evaluations of construct validity were failure to report a hypothesis (83%), inadequate sample size (67%), and failure to report the measurement properties of comparator instruments (50%). Important flaws were reported across all included studies with respect to insufficient description of the walk test protocol, thus limiting replication. For the generalizability checklist, the common methodological issues were failure to report the method used to select participants (60%), important disease characteristics, and/or description of treatment (47%).
Figure 2.: Quality appraisal (COSMIN) for studies examining reliability (n = 11). Gen, generalizability checklist; Rel, reliability checklist; +, yes; −, no; ?, cannot tell. Important flaws across included studies commonly related to insufficient description of the walk test protocol.
Figure 3.: Quality appraisal (COSMIN) for studies examining measurement error (n = 6). Gen, generalizability checklist; ME, measurement error checklist; +, yes; −, no; ?, cannot tell. Important flaws across included studies commonly related to insufficient description of the walk test protocol.
Figure 4.: Quality appraisal (COSMIN) for studies examining construct validity (n = 36). Gen, generalizability checklist; HT, hypothesis testing; +, yes; −, no; ?, cannot tell. Important flaws across included studies commonly related to insufficient description of the walk test protocol.
Participant Characteristics and Walk Test Protocols
The table (Supplemental Digital Content 2, https://links.lww.com/JNPT/A151 ) provides details of participant characteristics across articles. The number of articles describing people in the acute, subacute and chronic phase poststroke was 2 (5%), 4 (9%), and 31 (72%), respectively. In 6 studies, the sample consisted of a mix of participants in the acute and subacute (7%, n = 3) and subacute and chronic phases poststroke (7%, n = 3). Usual assistive devices used were reported across 32 (74%) articles.
This table (Supplemental Digital Content 3, https://links.lww.com/JNPT/A152 ) describes the walk test protocols used in the 42 studies described in 43 articles. All studies reporting the 2MWT, 3MWT, and 5MWT evaluated walking using a straight walkway with test distances ranging from 5 to 30 m.23–28 The 12MWT was examined using a rectangular walkway 42 m63 and 122 m62 long. Of the 35 studies describing 36 protocols for the 6MWT, 76% (n = 28) of the walk tests were performed along a straight walkway ranging from 10 to 85 m in length. The 3 most frequently reported straight walkway distances were 30 m (36%), 30.5 m (11%), and 33 m (11%). Rectangular,30 , 32 , 38 , 63 oval,41 , 49 and circular44 walkways were also used. Of the 36 6MWT protocols reported, 14% were the American Thoracic Society 6MWT protocol.66
Across the 44 walk test protocols, participants were instructed to walk at either a comfortable (n = 13) or fast (n = 4) pace where pace was described. The position of the evaluator was reported in 13 walk test protocols (30%). Evaluators walked behind (n = 10),23 , 24 , 30 , 35 , 42 , 45 , 53 , 56 , 59 , 65 or beside (n = 2)27 , 41 participants, or remained at the starting line (n = 1).55 Authors described that the evaluator either guarded the participant for safety concerns during the walk test48 or provided assistance for balance, weight-shifting, or leg advancement as necessary.41 , 56 , 62
Influence of Walk Test Protocol Elements on Test Performance
Effect of Turning Direction and Walkway Distance
In one study,53 26 people with chronic stroke walked significantly further and took significantly fewer turns, on average, as walkway length increased from 10, to 20 to 30 m. For example, walk distance improved by an average of 38.2 and 40.6 m after increasing the walking length from 10 to 30 m, respectively.53 Turning direction (to the paretic/nonparetic side) did not influence 6MWT performance.53
Effect of a Practice Trial
In a study42 of 83 people within the first-year poststroke, the mean difference ± standard deviation in 6MWT performance across 2 trials performed approximately 30 minutes apart was 0 ± 35 m. Across trials, 58% improved their performance (median 13 m), 40% deteriorated (median −17 m), and 2% did not change.
Effect of Walking Aids
Twenty-five people undergoing inpatient rehabilitation poststroke completed 3 6MWTs on 3 consecutive days using 3 walking aids (4-point cane, simple cane with ergonomic handgrip, Nordic stick) in random order.44 On average, participants walked significantly further when using the simple cane (115.5 ± 55.0 m), compared with the 4-point cane (101.4 ± 54.1 m) and the Nordic stick (98.0 ± 51.3 m).
Reliability and Measurement Error
Across the 11 articles reporting reliability, interrater, intrarater, and test-retest reliability was reported in 1, 2, and 9 articles, respectively (Table 2 ). Test-retest reliability and measurement error were examined in participants in the subacute (n = 2), subacute/chronic (n = 1), and chronic (n = 6; note: one study did not report measurement error) phases poststroke. Inter- and intrarater reliability was examined in a group of participants in the acute/subacute (n = 2) phase poststroke. Across walk tests, point estimates for ICCs and lower 95% CI limits were 0.90 or greater (range 0.90-1.00) with 2 exceptions. The ICC for test-retest reliability of the 6MWT in people requiring physical assistance to walk was 0.80 (95% CI 0.44-0.94),41 and the ICC for intra- and interrater reliability of the 12MWT was 0.71 and 0.68, respectively.62 Table 2 presents the SEM and computed or reported MDC values for the 2MWT (n = 1), 5MWT, (n = 1) and 6MWT (n = 6).
Table 2. -
Reliability and Measurement Error (11 Studies)
Author
Walk Test
Level of Gait Deficit
Walk Test Protocol
Re-test Time Interval
N
Test-Retest Reliabilitya ICC (95% CI)
Measurement Error, m
COSMIN Quality Score
Path Distance, Shape, Location
#Practice/Test Trials, Position of Rater, Timing Tool
Walk Aids Allowed
Encouragement (Timing), Assistance Allowed
Acute (<1 mo) and subacute (1-6 mo)
English et al23
2MWT
NR
10 m, straight, indoor
0/1b , behind, stopwatch
Yesb
Yesb (NR), NR
4 wk, video-taped trials rescored
10
Intra-R: 1.00 (NR)
Poor
Kosak and Smith62
12MWT
FIM walking subscore: 3 ± 1
122 m, rectangle, indoor
NR, NR, NR
Yes
NR, yes
1 d
18
Inter-R: 0.68 (P < 0.0007)
Poor
Day 1, 3
18
Intra-R: 0.71 (P < 0.0003)
Subacute (1-6 mo)
Da Cunha-Filho et al27
5MWT
NR
5 m, straight, indoor
1/1; behindb , stopwatch
Yesb
Nob , NR
≥10 min
9
0.97 (NR)
SEM = 6.9 MDC95 = 19.2 MDC90 = 24.4c
Poor
Fulk et al41
6MWT
NR
46/76 m, oval, indoor
0/1b , beside, NR
Yesb
No, yes
1-3 d
37 24, walk on owng 13, walk with assistg
0.97 (0.92-0.99) 0.97 (0.91-0.99) 0.80 (0.44-0.94)
SEM = 23.2 MDC90 = 54.1 MDC90 = 52.1c SEM = 26.1 MDC90 = 61.0 MDC90 = 59.6c SEM = 16.7 MDC90 = 39.0 MDC90 = 33.9c
Fair
Subacute (1-6 mo) and chronic (>6 mo)
Liu et al42
6MWT
CGS ≥0.7, 0.3-0.7, <0.3 m/s: 38%, 42%, 21%
20 m, straight, indoor
0/1, slightly behindb , stopwatchb
Yes
Yes (30 s), nob ,d
30 min
83
0.98 (0.97-0.99)
MDC90 = 39.2c
Fair
Chronic (>6 mo)
Hiengkaew et al24
2MWT
NR
20 m, straight, indoor
0/1b , behindb , stopwatch
Yes
Nob , NRd
5-10 d
61 12, MAS = 0e 32, MAS = 1-1+e 17, MAS ≥2e
0.98 (0.97-0.99) 0.98 (0.92-0.99) 0.98 (0.96-0.99) 0.96 (0.90-0.99)
SEM = 4.8 MDC90 = 11.4c SEM = 5.1 MDC90 = 11.5c SEM = 4.8 MDC90 = 11.2c SEM = 4.7 MDC90 = 11.6c
Fair
Sakai et al25
3MWT
NR
30 m, straight, indoor
NR, NR, NR
Yes
NR, NRd
2-5 d
14
0.90 (NR)
Poor
Eng et al30
6MWT
CGS: 1.0 ± 0.3 m/s
42 m, rectangle, indoor
0/1b , behindb , stopwatchb
Yes
Yesb (as per ATS), nob
Additional days
12
0.99 (NR)
SEM = 12.4 MDC90 = 28.6c
Poor
Ng and Hui-Chan65
6MWT
NR
33 m, straight, indoor
0/1b , behindb , NR
Yesb
Yesb (as per ATS), NRd
Different days in 1 wk, same time
10
0.98 (0.94-0.99)
MDC90 = 29.0c
Poor
Flansbjer et al31
6MWT
CGS: 0.9 ± 0.3 m/s
30 m, straight, indoor
0/1b , at mid-walkwayb , stopwatch
Yes
No (informed when 3 min left), nob
7 d, same timef
50
0.99 (0.98-0.99)
SEM = 18.6 SEM% = 4.8 MDC90 = 43.3 MDC90 = 30.7c
Fair
Wevers et al56
6MWT
FAC score of 3/4/5, n: 1/3/23
30 m, straight, outdoors
0/1b , behind, GPS/MW
Yesb
Yes (1 min), yesd
1.5 ± 1.3 d (maximum 5 d)
27, GPS
0.96 (0.96-0.98)
SEM = 18.1 MDC95 = 50.2 MDC90 = 42.1c
Poor
Zalewski and Dvorak57
27, MW
0.98 (0.98-1.00)
SEM = 11.9 MDC95 = 33.0 MDC90 = 27.7c
Abbreviations: A, acute (<1 month); ATS, American Thoracic Society; C, chronic (≥ 6 months); CI, confidence interval; CGS, comfortable gait speed; GPS, global positioning system; FAC, Functional Ambulation Classification; FIM, Functional Independence Measure; ICC, intraclass correlation coefficient; Inter-R, interrater reliability; Intra-R, intrarater; MAS, Motor Assessment Scale; MDC90 /MDC95 , minimal detectable change at 90%/95% confidence level; MW, measuring wheel; NR, not reported; SA, subacute (1-6 months); SEM, standard error of measurement; 2MWT, 2-minute walk test; 3MWT, 3-minute walk test; 5MWT, 5-minute walk test; 6MWT, 6-minute walk test; 12MWT, 12-minute walk test.
a Test-retest reliability presented unless otherwise indicated.
b Data obtained from author.
c Computed by authors using published data.
d Participants were required to walk independently to be eligible.
e Participants categorized based on Modified Ashworth Scale score for ankle plantarflexor tone, where 0 = no increase in tone; 1-1+ = slight increase in tone; ≥2 = marked/considerable increase in tone or showing rigid ankle plantarflexors.
f For 2 participants, the interval was 10 and 13 days; for 3 participants, the test session was not performed at the same time of day.
g Participants categorized based on FIM locomotion score, where ≥5 = walk without assistance; <5 = required physical assistance to walk.
Construct Validity
Table 3 presents 89 correlation coefficients for relationships between measures of targeted constructs and performance on the 3MWT (1 correlation), 5MWT (3 correlations), 6MWT (81 correlations), and 12MWT (4 correlations). Across the acuity levels, 67% (n = 60) of the correlation coefficients were evaluated among participants in the chronic phase poststroke. No studies examined relationships with discharge destination. Of the 89 correlations, 6 were predictive in nature examining the ability of the 6MWT to predict physical activity45 , 49 and health-related quality of life.35 Table 4 summarizes the reliability, measurement error, and construct validity findings by walk test and recovery phase poststroke.
Table 3. -
Construct Validity (N = 36)
Walk Test
ICF Classification
Construct
Measure
Results Pearson r (P-value , n) Spearman ρ (P-value , n)
Acute (<1 mo)
5MWT
Activity
Walking independence
Functional ambulation classification
r = 0.55 (P < 0.05, n = 20)26
Walking speed
5-m fast walk test
r = 0.80 (P < 0.01, n = 20)26
Acute (<1 mo) and subacute (1-6 mo)
6MWT
Body function
Aerobic capacity
o
2
peak /age-predicted o
2
max
r = 0.84 (P < 0.001, n = 17)29
Activity
Mobility
Timed Up and Go
ρ = −0.73 (P < 0.001, n = 41, discharge)34
ρ = −0.80 (P < 0.001, n = 41, admission)34
Walking speed
10-m comfortable walk test
r = 0.91 (P < 0.05, n = 17)29
10-m fast walk test
r = 0.89 (P < 0.05, n = 17)29
Subacute (1-6 mo)
6MWT
Body function
Aerobic capacity
Maximal exercise test duration
r = 0.60 (P < 0.001, n = 36)36
Activity
Walking speed
5-m comfortable walk test
r = 0.79 (P < 0.001, n = 34)36
r = 0.89 (P < 0.000, n = 37)41
5-m fast walk test
r = 0.82 (P < 0.001, n = 30)36
Subacute (1-6 mo) and chronic (>6 mo)
6MWT
Body function
Aerobic capacity
Peak oxygen consumption (o
2
peak )
ρ = 0.34 (NS, n = 30, indoor 6MWT)40
ρ = 0.39 (P ≤ 0.05, n = 30, outdoor 6MWT)40
r = 0.39 (P > 0.05, n = 34)30
r = 0.40 (P < 0.005, n = 63)32
r = 0.45 (P < 0.05, n = 48)55
r = 0.56 (P < 0.001, n = 33)36
r = 0.58a (P < 0.05, n = 48)55
r = 0.64 (P < 0.001, n = 74)39
r = 0.68 (P < 0.05, n = 14)52
r = 0.71 (p = 0.04, n = 8)54
r = 0.73 (p = 0.027, n = 9)48
Body function
Balance self-efficacy
Activities-specific balance confidence scale
ρ = 0.43 (95% CI 0.24-0.59, n = 86)35
Activity
Walking speed
10-m comfortable walk test
r = −0.74 (correlation with time, P ≤ 0.01, n = 21)45
r = 0.84 (P < 0.001, n = 50)31
Activity
Walking speed
10-m fast walk test
r = 0.94 (P < 0.001, n = 50)31
Non-ICF
Health-related quality of life
SF-36-PF Scale
r = 0.64 (P < 0.05, n = 62)35 predictive
SF-36-PCS
r = 0.39 (P < 0.05, n = 61)35 predictive
EQ-5D VAS
r = 0.22 (P ≥ 0.05, n = 64)35 predictive
Chronic (>6 mo)
3MWT
Body function
Motor function-lower extremity
Brunnstrom recovery stage
r = 0.49 (P < 0.05, n = 153)25
5MWT
Body function
Strength-knee extensorb
Dynamometer
r = 0.41 (P = 0.24, n = 10)28
6MWT
Body function
Aerobic capacity
Relative o
2
r = 0.66 (P < 0.05, n = 12)30
Motor function-Lower extremity
Fugl-Meyer lower extremity score
ρ = 0.72 (P ≤ 0.001, n = 34, indoor 6MWT)40
ρ = 0.74 (P ≤ 0.001, n = 34, outdoor 6MWT)40
ρ = 0.80 (P < 0.01, n = 31)58
ρ = 0.80 (P < 0.01, n = 31)69
Chedoke-McMaster stroke assessment
r = 0.75 (P < 0.01, n = 25)63
Strength-hip flexorb
Dynamometer
r = 0.40 (P = 0.02, n = 48)37
Strength-hip extensorb
r = 0.40 (P = 0.002, n = 48)37
Strength-knee flexorb
r = 0.20 (P = 0.2, n = 48)37
r = 0.71 (P < 0.01, n = 50)33
Strength-knee extensorb
r = 0.39 (P < 0.01, n = 63)38
r = 0.40 (P = 0.01, n = 48)37
r = 0.41 (P < 0.005, n = 63)32
r = 0.57 (P < 0.001, n = 62)39
r = 0.70 (P < 0.01, n = 50)33
Strength-ankle dorsiflexorb
Dynamometer
r = 0.50 (P = 0.001, n = 48)37
Peak torque
r = 0.79 (P ≤ 0.000, n = 62)59
Strength-ankle plantarflexorb
Dynamometer
r = 0.43 (P < 0.05, n = 25)63
r = 0.50 (P = 0.001, n = 48)37
Peak torque
r = 0.35 (P = 0.005, n = 62)59
Strength (arm, leg, hand, ankle/foot)b
Stroke impact scale-strength
r = 0.52 (P = 0.003, n = 30)43
Activity
Balance
Berg balance scale
r = 0.67 (P ≤ 0.01, n = 21)46
r = 0.69 (P < 0.001, n = 74)39
r = 0.78 (P < 0.01, n = 25)63
ρ = 0.67 (P ≤ 0.001, n = 34, indoor 6MWT)40
ρ = 0.69 (P ≤ 0.001, n = 34, outdoor 6MWT)40
ρ = 0.85 (P < 0.005, n = 63)32
Mobility
Timed Up and Go
ρ = −0.96 (P < 0.01, n = 10)65
r = −0.89 (P < 0.001, n = 50)31
Physical activity
Physical activity scale for individuals with physical disabilities
ρ = 0.31 (P = 0.06, n = 40)51 r = 0.32 (P < 0.05, n = 63)38
ICF measure-activity
r = −0.32 (P < 0.05, n = 77)60
Accelerometer activity kilocounts
ρ = 0.67 (P = 0.001, n = 40)51
SAM-steps/d
r = 0.55 (P < 0.05, n = 17)57
SAM-# steps at low ratec
r = 0.58 (P < 0.01, n = 49)47
SAM-# steps at high ratec
r = 0.60 (P < 0.01, n = 49)47
SAM-mean steps/d
r = 0.67 (P < 0.01, n = 49)47
SAM peak activity index (steps/min)
r = 0.72 (P < 0.01, n = 49)47
activPAL–mean minutes spent walking/standing/d
ρ = 0.48 (P = 0.051, n = 17)64
IDEEA-time spent on ft/dd
r = 0.55 (P < 0.001, n = 42)45 predictive
IDEEA-activity counts
r = 0.60 (P < 0.001, n = 42)45 predictive
SAM-mean steps/d
r = 0.68 (P = 0.001, n = 19)49 predictive
Stair function
Stair climbing-descent
r = −0.80 (P < 0.001, n = 50)31
Stair climbing-ascent
r = −0.82 (P < 0.001, n = 50)31
Strength-lower extremity
Five times sit-to-stand test
r = −0.60 (P < 0.001, n = 68)50
Walking ability
Walk-12
ρ = 0.59 (P ≤ 0.01, n = 50)61
Walking distance
12-min walk test
r = 0.97 (P < 0.01, n = 25)63
Walking speed
8-m comfortable walk test
r = 0.92 (P < 0.01, n = 25)63
Activity and participation
Activity and participation
ICF measure-total
r = −0.34 (P < 0.05, n = 77)60
Participation
Mobility Participation
Stroke impact scale-mobility Stroke impact scale-participation
r = 0.72 (P < 0.001, n = 30)43 r = 0.56 (P = 0.001, n = 30)43 r = 0.53 (P < 0.01, n = 50)33 ρ = 0.43 (P = 0.082, n = 17)64
ICF measure-participation
r = −0.31 (P < 0.05, n = 77)60
12MWT
Body function
Motor function-lower extremity
Chedoke-McMaster stroke assessment
r = 0.69 (P < 0.01, n = 25)63
Strength-ankle plantarflexorb
KinCom
r = 0.35 (P ≥ 0.05, n = 25)63
Activity
Balance
Berg balance scale
r = 0.80 (P < 0.01, n = 25)63
Walking speed
8-m comfortable walk test
r = 0.91 (P < 0.01, n = 25)63
Abbreviations: ICF, International Classification of Functioning, Disability and Health; IDEEA, Intelligent Device for Energy Expenditure and Activity; NS, not significant; PCS, physical component summary; PF, physical function; SAM, StepWatch activity monitor; SSWS, self-selected walking speed; UE, upper extremity; VAS, visual analog scale; 3MWT, 3-minute walk test; 5MWT, 5-minute walk test; 6MWT, 6-minute walk test; 6MWD, 6-minute walk distance; 12MWT, 12-minute walk test.
a Normalized values corrected for the influence of age, sex, height, and weight.
b Strength of the paretic-side examined.
c High rate defined as less than 60 steps per minute, low rate defined as less than 30 steps per minute.
d Sum of time spent walking, going up and down stairs, standing, and in sit-to-stand transition per day averaged over 2 days.
Table 4. -
Summary of Reliability, Measurement Error, and Construct Validity Findings by Walk Test and Recovery Phase
a Poststroke
Walk Test
Reliability Coefficient (# Studies)
MDC90 , m (Rounded) (# Studies)
Constructs Correlated With Walk Test Performanceb (# Correlations with P < 0.05)
Acute
Subacute
Chronic
Acute
Subacute
Chronic
Acute
Subacute
Chronic
2MWT
1.00 (1)
1.00 (1)
0.98 (1)
11 (1)
3MWT
0.90 (1)
Motor functionc (1)
5MWT
0.97 (1)
24 (1)
Walk independence (1)
Walk speed (1)
6MWT Walk with assist
0.97-0.98 (2) 0.80 (1)
0.96-0.99 (5)
39-52 (2) 34 (1)
28-42 (5)
Aerobic capacity (1)
Aerobic capacity (11)
Aerobic capacity (10)
Mobility (2)
Mobility (2)
Mobility (2)
Walk speed (2)
Walk speed (8)
Walk speed (4)
Balance SE (1)
Balance SE (1)
Motor functionc (5)
Strengthc (15)
Balance (6)
Physical activity (11)
Stairs (2)
Walk distance (1)
Walk ability-SR (1)
Participation (5)
12MWT
0.68-0.71 (1)
0.68-0.71 (1)
Motor functionc (1) Balance (1)
Walk speed (1)
Abbreviations: MDC90 , minimal detectable change at the 90% confidence level; SE, self-efficacy; SR, self-reported; 2-, 3-, 5-, 6-, 12MWT, 2-, 3-, 5-, 6-, 12-minute walk test.
a Results from studies combining people in 2 phases of stroke recovery were listed under both phases.
b Constructs with statistically significant correlations (P < 0.05) in 80% or more of correlations performed were listed.
c The lower extremities.
Sensitivity to Change
Sensitivity to change was reported in one study62 examining the 12MWT among participants in the acute and subacute phase poststroke. The standardized response mean for the 12MWT was 1.90 (ie, large67 , 68 ).
DISCUSSION
This is one of the first comprehensive systematic reviews to synthesize research evidence describing the measurement properties and walk test protocols of time-limited walk tests in people with stroke. Findings support the excellent test-retest reliability of various 6MWT protocols in the subacute41 , 42 and chronic30 , 31 , 42 , 56 , 65 phases of stroke recovery and the construct validity of 6MWT performance as a measure of functional walking capacity in people with acute,29 , 34 subacute,29 , 31 , 32 , 34–36 , 39 , 41 , 46 , 48 , 52 , 54 , 55 and chronic30–33 , 35–40 , 43 , 45 , 46–52 , 54 , 55 , 57–61 , 63 , 69 stroke. Estimates of MDC90 for the 6MWT range from 3942 to 52 m41 and from 2856 to 42 m30 , 31 , 42 , 65 in the subacute and chronic phases of stroke recovery, respectively. Few reports evaluating the 2-, 3-, 5-, and 12MWT poststroke were available. Preliminary evidence indicates that for the 2MWT, intrarater reliability is excellent in acute and subacute recovery phases,23 test-retest reliability is excellent in the chronic phase,24 and the estimated MDC90 is 11 m in people with chronic stroke.24 No studies examining the construct validity of the 2MWT poststroke were found. In people with chronic stroke, the 3MWT demonstrates excellent test-retest reliability with evidence of construct validity limited to one correlation with motor function.25 The 5MWT has excellent test-retest reliability in people with subacute stroke,27 an estimated MDC90 of 24 m27 (subacute stroke), and preliminary evidence of construct validity in the acute26 recovery phase. Intra- and interrater reliability of the 12MWT is acceptable in people with acute stroke,62 and preliminary evidence supports construct validity in the chronic phase.63 Across walk tests, only the 12MWT was evaluated for sensitivity to change and it was shown to be a responsive indicator in an acute and subacute stroke population.62
In alignment with the KTA framework, synthesis and analysis of the extensive research evidence on the 6MWT in this review has yielded the following considerations that would inform the development of a clinical practice guide. First, a walk test protocol selected for use in a particular practice setting should have evidence of test-retest reliability obtained in patients with characteristics similar to those seen in the setting of interest. For example, although evidence of reliability of the 6MWT in the acute phase of stroke recovery is lacking, excellent reliability of the 6MWT has been observed in studies where evaluators provided physical assistance to walk as necessary.41 , 56 These findings suggest that reliability of the 6MWT protocol used would be acceptable in the acute setting where 24% of patients who can walk require assistance.15
Second, it is essential to use a standard written protocol and documentation procedure if comparison of walk test performances over time or between patients within and across settings is desired. Although excellent reliability of the 6MWT has been reported for the subacute and chronic phases of stroke recovery, 6MWT protocols used in each phase varied in terms of walkway length and shape, location (indoor/outdoor), and encouragement, which can influence the distance walked and limit comparisons.53 , 70–72 For example if a patient completed the 6MWT on a 10-m walkway in an acute care setting, and later walked 25 m further during the test on a 30-m walkway in a rehabilitation hospital, the improvement could have resulted from using a longer walkway53 instead of an increase in walking capacity.
Third, review findings support inclusion of specific elements in a standardized protocol for 6MWT administration poststroke. An excellent basis for adapting a protocol to use poststroke is the recently updated 6MWT protocol for chronic respiratory disease73 that is widely used in research and clinical practice.10 This protocol involves screening the patient for relative and absolute contraindications, and using a standardized set of equipment, instructions and encouragement statements delivered each minute, and a straight, 30-m walkway.73 Patients are asked to wear comfortable clothing, supportive shoes, and their usual walking aids. Documentation of the distance walked, and use of any mobility devices, is recommended. Although 2 trials of the 6MWT is advised for people with respiratory disease, results from the current review show reliability is excellent conducting only 1 6MWT trial,30 , 31 , 41 , 42 , 56 , 65 and no practice trial31 , 41 , 42 , 56 , 65 ; thus, a single 6MWT administration is recommended poststroke. We also advise recording the level of physical assistance provided, and the walkway length in communications of patient status to help colleagues who may readminister the test to better interpret the influence of these factors on observed change in performance. Once consensus on a standardized 6MWT protocol to use poststroke is reached, evaluating the reliability and measurement error of the protocol across the care continuum is advised to support its use.
The MDC90 estimates obtained for the 6MWT in this review varied widely for subacute41 , 42 (39-52 m) and chronic30 , 31 , 42 , 56 , 65 (28-42 m) phases of stroke recovery. This variability was likely due to differences in 6MWT protocols and stroke populations across studies that influenced reliability estimates (ICC values) and standard deviations used to calculate the MDC90 . A methodological issue that emerged from this review is that MDC90 values derived using the standard deviation of performances and test-retest reliability estimates were consistently lower than values computed using within-subject error variance.20 This finding highlights the importance of adequately reporting the statistical methods used to compute MDC values, and comparing values derived using the same method.
An MDC value is used to interpret whether the change in a patient's performance is sufficiently large to represent true change in ability.20 The MDC at the 90% confidence level (MDC90 ) means that 90% of truly unchanged patients will display random fluctuations in performance within the range of the MDC value.20 Change must therefore exceed the MDC90 value to be considered as “true change” in ability.20 For example, the 6MWT MDC90 value of 39 m for people with subacute and chronic stroke42 indicates that 90% of truly unchanged patients may display random fluctuations (improvements or deteriorations) in performance as large as 39 m. Therefore, a patient has to improve by greater than 39 m on the 6MWT to interpret the change as true improvement in walking capacity. Clinicians should consider whether sample size was adequate, defined as n ≥ 25 in this study, prior to selecting a 6MWT protocol based on reliability results, or an MDC value to interpret performance, and examine whether characteristics of study participants are similar to patients seen in clinical practice.
Review findings revealed gaps in the literature. Despite extensive evaluation of the 6MWT, no studies examining reliability and measurement error of the 6MWT in the acute phase and few studies in the subacute phase of stroke recovery were identified. The vast majority of the literature reporting associations between 6MWT performance and measures of physical capacity, physical activity, and participation involved people with chronic stroke. Evidence of reliability and validity of the 6MWT poststroke in acute and subacute care settings would help influence physical therapists to implement the test in these settings.1 Finally, there is a general lack of research investigating the measurement properties of alternate time-limited walk tests. Consensus as to which time-limited walk tests should be used poststroke would guide the focus of future research.
Limitations of this review include the inability to include all constructs in validity studies or a more current review due to the extensiveness of the literature in this area. To optimize the comprehensiveness of our review of published validity evidence, we included any study examining associations between walk test performance and measures of targeted constructs.
CONCLUSIONS
Various 6MWT protocols demonstrate excellent test-retest reliability and yield estimates of measurement error in the subacute and chronic phases of stroke recovery. Evidence supports the construct validity of using the 6MWT, with a standardized administration protocol, in people with acute, subacute, and chronic stroke. Investigation of the measurement properties of the 2-, 3-, 5-, and 12MWT is limited. Methodological weaknesses in the literature highlighted in this review will inform future research. Considerations for advancing implementation of the 6MWT as a recommended measure of functional walking capacity poststroke are provided.
REFERENCES
1. Pattison KM, Brooks D, Cameron JI, Salbach NM. Factors influencing physical therapists' use of standardized measures of walking capacity post-stroke across the care continuum. Phys Ther. 2015;95:1507–1517.
2. Sullivan JE, Crowner BE, Kluding PM, et al. Outcome measures for individuals with stroke: process and recommendations from the American Physical Therapy Association Neurology Section Task Force. Phys Ther. 2013;93:1383–1396.
3. Lindsay MP, Gubitz G, Bayley M, Hill MD, Phillips S, Smith EE. Canadian Stroke Best Practice Recommendations Overview and Methodology. 5th ed. On behalf of the Canadian Stroke Best Practices Advisory Committee and Writing Groups. Ottawa, Canada: Heart and Stroke Foundation; 2014.
www.strokebestpractices.ca . Accessed July 2, 2015.
4. McCulloch KL, de Joya AL, Hays K, et al. Outcome measures for persons with moderate to severe traumatic brain injury: recommendations from the American Physical Therapy Association Academy of Neurologic Physical Therapy TBI EDGE Task Force. J Neurol Phys Ther. 2016;40:269–280.
5. Salbach NM, Guilcher SJ, Jaglal SB. Physical therapists' perceptions and use of standardized assessments of walking ability post-stroke. J Rehabil Med. 2011;43:543–549.
6. Latham NK, Jette DU, Slavin M, et al. Physical therapy during stroke rehabilitation for people with different walking abilities. Arch Phys Med Rehabil. 2005;86:S41–S50.
7. Jette DU, Halbert J, Iverson C, Miceli E, Shah P. Use of standardized outcome measures in physical therapist practice: perceptions and applications. Phys Ther. 2009;89:125–135.
8. McGlynn M, Cott CA. Weighing the evidence: clinical decision making in neurological physical therapy. Physiother Can. 2007;59:241–254.
9. Salbach NM, O'Brien K, Brooks D, et al. Speed and distance requirements for community ambulation: a systematic review. Arch Phys Med Rehabil. 2014;95:117–128.
10. Salbach NM, O'Brien KK, Brooks D, et al. Reference values for standardized tests of walking speed and distance: a systematic review. Gait Posture. 2014;41:341–360.
11. Graham ID, Logan J, Harrison MB, Straus SE, Tetroe J, Caswell W, Robinson N. Lost in knowledge translation: time for a map? J Contin Educ Health Prof. 2006;26:13–24.
12. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. J Clin Epidemiol. 2009;62:e1–34 beginning with Liberati et al, references need to be renumbered from 12 onward.
13. Mokkink LB, Terwee CB, Patrick DL, et al. The COSMIN checklist for assessing the methodological quality of studies on measurement properties of health status measurement instruments: an international Delphi study. Qual Life Res. 2010;19:539–549.
14. Richards CL, Olney SJ. Hemiparetic gait following stroke. Part II: recovery and physical therapy. Gait Posture. 1996;4:149–162.
15. Jorgensen HS, Nakayama H, Raaschou HO, Olsen TS. Recovery of walking function in stroke patients—the Copenhagen Stroke Study. Arch Phys Med Rehabil. 1995;76:27–32.
16. Hobart JC, Cano SJ, Warner TT, Thompson AJ. What sample sizes for reliability and validity studies in neurology? J Neurol. 2012;259:2681–2694.
17. Granger CV, Hamilton BB, Linacre JM, Heinemann AW, Wright BD. Performance profiles of the functional independence measure. Am J Phys Med Rehabil. 1993;72:84–89.
18. Terwee C, Mokkink L, Knol D, Ostelo R, Bouter L, de Vet H. Rating the methodological quality in systematic reviews of studies on measurement properties: a scoring system for the COSMIN checklist. Qual Life Res. 2012;21:651–657.
19. Andresen E. Criteria for assessing the tools of disability outcomes research. Arch Phys Med Rehabil. 2000;81:S15–S20.
20. Stratford P. Getting more from the literature: estimating the standard error of measurement from reliability studies. Physiother Can. 2004;56:27–30.
21. World Health Organization. International Classification of Functioning, Disability and Health: ICF. 1st ed. Geneva, Switzerland; World Health Organization; 2001.
22. Landis J, Koch G. The measurement of observer agreement for categorical data. Biometrics. 1977;33:159–174.
23. English CK, Hillier SL, Stiller KR, Warden-Flood A. Circuit class therapy versus individual physiotherapy sessions during inpatient stroke rehabilitation: a controlled trial. Arch Phys Med Rehabil. 2007;88:955–963.
24. Hiengkaew V, Jitaree K, Chaiyawat P. Minimal detectable changes of the Berg Balance Scale, Fugl-Meyer Assessment Scale, Timed “Up & Go” Test, gait speeds, and 2-minute walk test in individuals with chronic stroke with different degrees of ankle plantarflexor tone. Arch Phys Med Rehabil. 2012;93:1201–1208.
25. Sakai T, Tanaka K, Holland G. Functional and locomotive characteristics of stroke survivors in Japanese community-based rehabilitation. Am J Phys Med Rehabil. 2002;81:675–683.
26. Da Cunha-Filho IT, Lim PA, Henson H, Monga T, Qureshy H, Protas EJ. Performance-based gait tests for acute stroke patients. Am J Phys Med Rehabil. 2002;81:848–856.
27. Da Cunha-Filho IT, Henson H, Wankadia S, Protas EJ. Reliability of measures of gait performance and oxygen consumption with stroke survivors. J Rehabil Res Devt. 2003;40:19–25.
28. Kobayashi T, Leung AK, Hutchins SW. Correlations between knee extensor strength measured by a hand-held dynamometer and functional performance in patients with chronic stroke. Isokinet Exerc Sci. 2011;19:33–37.
29. Kelly JO, Kilbreath SL, Davis GM, Zeman B, Raymond J. Cardiorespiratory fitness and walking ability in subacute stroke patients. Arch Phys Med Rehabil. 2003;84:1780–1785.
30. Eng JJ, Dawson AS, Chu KS. Submaximal exercise in persons with stroke: test-retest reliability and concurrent validity with maximal oxygen consumption. Arch Phys Med Rehabil. 2004;85:113–118.
31. Flansbjer UB, Holmback AM, Downham D, Patten C, Lexell J. Reliability of gait performance tests in men and women with hemiparesis after stroke. J Rehabil Med. 2005;37:75–82.
32. Pang MY, Eng JJ, Dawson AS. Relationship between ambulatory capacity and cardiorespiratory fitness in chronic stroke: influence of stroke-specific impairments. Chest. 2005;127:495–501.
33. Flansbjer UB, Downham D, Lexell J. Knee muscle strength, gait performance, and perceived participation after stroke. Arch Phys Med Rehabil. 2006;87:974–980.
34. Langhammer B, Lindmark B, Stanghelle J. The relation between gait velocity and static and dynamic balance in the early rehabilitation of patients with acute stroke. Adv Physiother. 2006;8:60–65.
35. Salbach NM, Mayo NE, Robichaud-Ekstrand S, Hanley JA, Richards CL, Wood-Dauphinee S. Balance self-efficacy and its relevance to physical function and perceived health status after stroke. Arch Phys Med Rehabil. 2006;87:364–370.
36. Tang A, Sibley KM, Bayley MT, McIlroy WE, Brooks D. Do functional walk tests reflect cardiorespiratory fitness in sub-acute stroke? J Neuroeng Rehabil. 2006;3:23.
37. Yang YR, Wang RY, Lin KH, Chu MY, Chan RC. Task-oriented progressive resistance strength training improves muscle strength and functional performance in individuals with stroke. Clin Rehabil. 2006;20:860–870.
38. Liu-Ambrose T, Pang MY, Eng JJ. Executive function is independently associated with performances of balance and mobility in community-dwelling older adults after mild stroke: implications for falls prevention. Cerebrovasc Dis. 2007;23:203–210.
39. Patterson SL, Forrester LW, Rodgers MM, et al. Determinants of walking function after stroke: differences by deficit severity. Arch Phys Med Rehabil. 2007;88:115–119.
40. Carvalho C, Willen C, Sunnerhagen KS. Relationship between walking function and 1-legged bicycling test in subjects in the later stage post-stroke. J Rehabil Med. 2008;40:721–726.
41. Fulk GD, Echternach JL, Nof L, O'Sullivan S. Clinometric properties of the six-minute walk test in individuals undergoing rehabilitation poststroke. Physiother Theory Pract. 2008;24:195–204.
42. Liu J, Drutz C, Kumar R, et al. Use of the six-minute walk test poststroke: is there a practice effect? Arch Phys Med Rehabil. 2008;89:1686–1692.
43. Muren MA, Hutler M, Hooper J. Functional capacity and health-related quality of life in individuals post stroke. Top Stroke Rehabil. 2008;15:51–58.
44. Allet L, Leemann B, Guyen E, et al. Effect of different walking aids on walking capacity of patients with poststroke hemiparesis. Arch Phys Med Rehabil. 2009;90:1408–1413.
45. Alzahrani MA, Dean CM, Ada L. Ability to negotiate stairs predicts free-living physical activity in community-dwelling people with stroke: an observational study. Aust J Physio. 2009;55:277–281.
46. Kluding P, Gajewski B. Lower-extremity strength differences predict activity limitations in people with chronic stroke. Phys Ther. 2009;89:73–81.
47. Mudge S, Stott NS. Timed walking tests correlate with daily step activity in persons with stroke. Arch Phys Med Rehabil. 2009;90:296–301.
48. Tseng BY, Kluding P. The relationship between fatigue, aerobic fitness, and motor control in people with chronic stroke: a pilot study. J Geriatr Phys Ther. 2009;32:97–102.
49. Fulk GD, Reynolds C, Mondal S, Deutsch JE. Predicting home and community walking activity in people with stroke. Arch Phys Med Rehabil. 2010;91:1582–1586.
50. Ng S. Balance ability, not muscle strength and exercise endurance, determines the performance of hemiparetic subjects on the timed-sit-to-stand test. Am J Phys Med Rehabil. 2010;89:497–504.
51. Rand D, Eng J, Tang R, Hung C, Jeng J. Daily physical activity and its contribution to the health-related quality of life of ambulatory individuals with chronic stroke. Health Qual Life Outcomes. 2010;8:80.
52. Calmels P, Degache F, Courbon A, et al. The feasibility and the effects of cycloergometer interval-training on aerobic capacity and walking performance after stroke. Preliminary study. Ann Phys Rehabil Med. 2010;54:3–15.
53. Ng SS, Tsang WW, Cheung TH, Chung JS, To FP, Yu PC. Walkway length, but not turning direction, determines the six-minute walk test distance in individuals with stroke. Arch Phys Med Rehabil. 2011;92:806–811.
54. Ovando AC, Michaelsen SM, de Carvalho T, Herber V. Evaluation of cardiopulmonary fitness in individuals with hemiparesis after cerebrovascular accident. Arq Bras Cardiol. 2011;96:140–146.
55. Severinsen K, Jakobsen JK, Overgaard K, Andersen H. Normalized muscle strength, aerobic capacity, and walking performance in chronic stroke: a population-based study on the potential for endurance and resistance training. Arch Phys Med Rehabil. 2011;92:1663–1668.
56. Wevers LE, Kwakkel G, van de Port IG. Is outdoor use of the six-minute walk test with a global positioning system in stroke patients' own neighbourhoods reproducible and valid? J Rehabil Med. 2011;43:1027–1031.
57. Zalewski KR, Dvorak L. Barriers to physical activity between adults with stroke and their care partners. Top Stroke Rehabil. 2011;18:666–675.
58. Danielsson A, Willen C, Sunnerhagen KS. Physical activity, ambulation, and motor impairment late after stroke. Stroke Res Treat. 2012;2012:818513.
59. Ng SS, Hui-Chan CW. Contribution of ankle dorsiflexor strength to walking endurance in people with spastic hemiplegia after stroke. Arch Phys Med Rehabil. 2012;93:1046–1051.
60. Schmid AA, Van PM, Altenburger PA, et al. Balance and balance self-efficacy are associated with activity and participation after stroke: a cross-sectional study in people with chronic stroke. Arch Phys Med Rehabil. 2012;93:1101–1107.
61. Brogardh C, Flansbjer UB, Lexell J. Self-reported walking ability in persons with chronic stroke and the relationship with gait performance tests. PM R. 2012;4:734–738.
62. Kosak M, Smith T. Comparison of the 2-, 6-, and 12-minute walk tests in patients with stroke. J Rehabil Res Dev. 2005;42:103–108.
63. Eng JJ, Chu KS, Dawson AS, Kim CM, Hepburn KE. Functional walk tests in individuals with stroke: relation to perceived exertion and myocardial exertion. Stroke. 2002;33:756–761.
64. Salbach N, Brooks D, Romano J, Woon L. The relationship between clinical measures and daily physical activity and participation in ambulatory, community-dwelling people with stroke. J Nov Physiother. 2013;3:182.
65. Ng SS, Hui-Chan CW. The timed up & go test: its reliability and association with lower-limb impairments and locomotor capacities in people with chronic stroke. Arch Phys Med Rehabil. 2005;86:1641–1647.
66. ATS Committee on Proficiency Standards for Clinical Pulmonary Function Laboratories. ATS Statement: guidelines for the six-minute walk test. Am J Respir Crit Care Med. 2002;166:111–117.
67. Salter K, Jutai J, Teasell R, Foley N, Bitensky J, Bayley M. Issues for selection of outcome measures in stroke rehabilitation: ICF participation. Disabil Rehabil. 2005;27:507–528.
68. Cohen J. Statistical Power Analysis for the Behavioral Sciences. Rev. ed. New York, NY: Academic Press; 1977.
69. Danielsson A, Willen C, Sunnerhagen KS. Is walking endurance associated with activity and participation late after stroke? Disabil Rehabil. 2011;33:2053–2057.
70. Bansal V, Hill K, Dolmage TE, Brooks D, Woon LJ, Goldstein RS. Modifying track layout from straight to circular has a modest effect on the 6-min walk distance. Chest. 2008;133:1155–1160.
71. Brooks D, Solway S, Weinacht K, Wang D, Thomas S. Comparison between an indoor and an outdoor 6-minute walk test among individuals with chronic obstructive pulmonary disease. Arch Phys Med Rehabil. 2003;84:873–876.
72. Guyatt GH, Pugsley SO, Sullivan MJ, et al. Effect of encouragement on walking test performance. Thorax. 1984;39:818–822.
73. Holland AE, Spruit MA, Troosters T, et al. An official European Respiratory Society/American Thoracic Society technical standard: field walking tests in chronic respiratory disease. Eur Respir. 2014;44:1428–1446.