Skip main navigation

Very Early Mobilization After Stroke Fast-Tracks Return to Walking

Further Results From the Phase II AVERT Randomized Controlled Trial
Originally published 2011;42:153–158


Background and Purpose—

Regaining functional independence is an important goal for people who have experienced stroke. We hypothesized that introducing earlier and more intensive out-of-bed activity after stroke would reduce time to unassisted walking and improve independence in activities of daily living.


A Very Early Rehabilitation Trial (AVERT) was a phase II randomized controlled trial. Patients with confirmed stroke (infarct or hemorrhage) admitted <24 hours after stroke and who met physiological safety criteria were eligible. Patients randomized to the very early and intensive mobilization group were mobilized within 24 hours of stroke and at regular intervals thereafter. Control patients received standard stroke unit care. The primary outcome for this analysis was the number of days required to return to walking 50 m unassisted. Secondary outcomes were the Barthel Index and Rivermead Motor Assessment at 3 and 12 months after stroke.


Seventy-one stroke patients with a mean age of 74.7 years were recruited from 2 hospitals. Adjusted Cox regression indicated that very early and intensive mobilization group patients returned to walking significantly faster than did standard stroke unit care controls (P=0.032; median 3.5 vs 7.0 days). Multivariable regression revealed that exposure to very early and intensive mobilization was independently associated with good functional outcome on the Barthel Index at 3 months (P=0.008) and on the Rivermead Motor Assessment at 3 (P=0.050) and 12 (P=0.024) months.


Earlier and more intensive mobilization after stroke may fast-track return to unassisted walking and improve functional recovery.

Clinical Trial Registration—

This trial was not registered because enrollment began before July 2005.

As a consequence of focal brain ischemia or hemorrhage, stroke can have both immediate and ongoing physical effects. Within 12 months of stroke, one third of stroke patients will die and another third are left with disability, restricted in performing simple activities of daily living (ADLs) and requiring some kind of assistance.1 Of all recovery goals, improved walking function is the one most often expressed by patients with stroke.2

The rate of physical recovery is not linear; the most rapid improvement occurs within the first 6 months after stroke.3,4 In a study of 976 acute stroke patients, Wade and Hewer5 reported some recovery between 3 weeks and 6 months in almost all patients, and by 6 months, approximately half of the survivors had regained functional independence. In another large study, approximately two thirds of patients were unable to walk without assistance in the first week after stroke.6 Among those initially unable to walk, 95% of patients recovered within the first 3 months, with leg paresis strongly associated with the rate of recovery. In addition to stroke severity,7,8 increasing age8,9 and diabetes8,10 are factors that have been associated with a lower rate and extent of functional recovery after stroke.

Both the speed and extent of recovery are important to survivors of stroke, yet research examining the efficacy of interventions on recovery milestones is limited. A meta-analysis of studies that included augmented exercise intensity revealed a positive effect on ADLs in the first 6 months after stroke.3 Treatment in a dedicated stroke unit has been associated with more rapid and greater recovery of independence in ADLs when compared with patients treated on a general medical ward.11 Walking recovery was not specifically measured in this study, but a post hoc analysis indicated that a shorter time to start of mobilization after stroke onset was the most important factor associated with discharge to home.12 Given the general lack of physical activity in the first weeks after stroke,13 we believed that there was an opportunity to introduce earlier and more frequent mobilization. The safety and feasibility of this intervention were the primary outcomes tested as part of a phase II clinical trial comparing a very early and intense mobilization protocol (VEM) with standard stroke unit care (SC).14 In the current article, we report the outcomes associated with recovery of walking and independence in ADLs. We hypothesized that patients in the VEM group would walk unassisted sooner than SC controls. We also hypothesized that VEM patients would achieve better functional independence (as measured by the Barthel Index and Rivermead Motor Assessment) than SC controls at 3 and 12 months after stroke.



A Very Early Rehabilitation Trial (AVERT) phase II was a prospective randomized controlled trial with concealed allocation, blinded assessment of outcomes, and intention-to-treat analysis. The setting was the acute stroke units of 2 large hospitals in Melbourne, Australia. Ethics approval was obtained. The methods have been reported in detail elsewhere14 and are summarized in the subsequent paragraphs.


Patients were included if they (1) were ≥18 years; (2) satisfied physiological limits (systolic blood pressure 120 to 220 mm Hg, heart rate 40 to 100 bpm, oxygen saturation >92%, and temperature <38.5°); and (3) could be randomized within 24 hours of symptom onset of a first or recurrent stroke. Patients were excluded if they had a premorbid modified Rankin Scale (mRS)15 score >3 (indicating disability), deterioration within the first hour of admission to the stroke unit or direct admission to intensive care, concurrent progressive neurological disorder, acute coronary syndrome, severe heart failure, lower-limb fracture that prevented mobilization, or required palliative care.


When discussing informed consent, patients were told that they would be given 1 of 2 different types of rehabilitation. Computer-generated, blocked randomization procedures and concealment with opaque envelopes were used to allocate patients to either the VEM or SC group. Randomization was stratified by hospital site and stroke severity on the National Institutes of Health Stroke Scale (NIHSS)16 (mild=0 to 7, moderate=8 to 16, and severe >16) to reduce the likelihood of severity imbalance between groups.


Both VEM and SC groups received standard care from ward therapists and nursing staff in the stroke units. Patients randomized to the VEM group began mobilizing as soon as practical after randomization, with the goal of first mobilization within 24 hours of stroke onset. The VEM group also received additional interventions, with the aim of assisting patients to be upright and out of bed at least twice per day, thereby doubling the standard care “mobilization dose” previously identified. VEM was delivered by a trained nurse and physiotherapist team for the first 14 days after stroke or until discharge from the acute stroke unit (whichever was sooner). The type and dose of therapy for both groups were recorded on personal digital assistants. Occupational health and safety procedures for manual handling of patients were maintained. Blood pressure, heart rate, oxygen saturation, and temperature were monitored before the first 3 mobilizations of VEM patients. “Contamination” of standard care was monitored throughout the study.

Baseline Assessment

Baseline assessment included age, sex, stroke type according to the Oxfordshire stroke classification,17 stroke severity on the NIHSS, diabetes status as well as other stroke risk factors, prestroke living arrangements, and prestroke functional level on the mRS. The mobility scale for acute stroke18 was used to establish baseline restrictions in a patient's ability to move in bed, sit up, and walk 10 m, and information from this scale was used to guide the VEM intervention level.

Outcome Measures

Assessments by the blinded assessor took place at 7 and 14 days and 3, 6, and 12 months after stroke. For this data analysis, the main outcome of interest was time to walking 50 m. This was defined as the number of days from stroke onset until the patient could first walk 50 m without human assistance, a distance used in the Functional Independence Measure19 walking item and commonly marked out within hospital departments. To remove the subjectivity associated with both the need for supervision (Functional Independence Measure score 5) and walking aid (Functional Independence Measure score 6), we defined walking as “walking unassisted by human help (gait aid allowed) for a continuous distance of 50 meters.” Physiotherapists were responsible for completion of the 50-m walk.

The other outcome measures were the Barthel Index5 and Rivermead Motor Assessment,20 both assessed at 3 and 12 months after stroke. The Barthel Index, a valid and reliable measure of ADLs in stroke research,21,22 was used to assess independence in 10 everyday activities. The total score sums to 20, and higher scores reflect better performance. As per common practice with the Barthel Index, patients were classified as either independent (score=20) or dependent (score=0 to 19) in ADLs. The Rivermead Motor Assessment gross function scale of 13 items was used to examine motor activity and has established reliability in stroke.23,24 It includes easy and difficult items, with the first item testing unsupported sitting, and other items testing ability to transfer, walk indoors, walk outdoors, run, and hop on the spot. The best performance of 3 trials on each item was scored, with a score of 0 indicating that the patient was unable to perform any of the activities independently and a score of 13 indicating independence with all activities. Patients were classified as either impaired (score=0 to 9) or not impaired (score=10 to 13) in motor activity, a cutoff chosen to align with the clinically meaningful ability to walk outside the home (community ambulation).

Statistical Analysis

A sample of 70 patients was deemed sufficient to allow examination of the primary safety outcome of the study (death) at 3 months after stroke.14 Although the study was not powered to detect differences in physical outcomes reported in this article, an important part of feasibility testing includes evaluation of performance and utility of the outcome measures used.

Because one third of stroke patients die in the first 12 months,1 analyses must deal with missing data due to deaths. Furthermore, when inclusion criteria are broad, as in this trial, stroke symptoms can vary markedly, with the result that some patients will fail to achieve any score on a given test, particularly those related to higher levels of activity. Excluding patients who died or who failed to achieve a score on a test would bias results, and imputing data values for these subjects requires guesswork. In this study, all patients who died were assigned a score of 0 on the Barthel and Rivermead scales for all subsequent time points. A score of 0 was deemed likely to reflect their status had they still been alive, given that people who die in the first year after stroke are more likely to have had a severe stroke (and would be expected to have a very low score on these measures). Patients alive at the time of assessment with missing data were not included in analyses. Multivariable analyses were used to account for demographic and stroke-related factors known to influence walking and functional independence (age, stroke severity, and diabetes) and other factors that may also impact these outcomes (sex and premorbid disability).

Days to Walking

The number of days between stroke onset and a 50-m unassisted walk was recorded. A Cox regression was performed with adjustment for important prognostic variables (age, sex, NIHSS, premorbid mRS and diabetes).

Barthel Index

Mann-Whitney U tests were used to assess between-group differences in Barthel score. Barthel data were then used to classify patients as either independent or dependent in ADLs, and Fisher's exact tests were used to detect group differences. Multivariable logistic-regression analyses, with the Barthel Index as the binary outcome measure, were then performed to determine the effect of group (VEM or SC), age, sex, NIHSS, premorbid mRS and diabetes on independence in ADLs.

Rivermead Motor Assessment

Between-group differences in Rivermead scores were analyzed by Mann-Whitney U tests. Rivermead data were then used to classify patients as either impaired or not impaired in motor activity, and Fisher's exact tests were used to detect group differences. Multivariable logistic-regression analyses were then performed on the binary Rivermead data in the same way as described for the Barthel Index.


Seventy-one patients were recruited from 2004 to 2006: 60 from the Austin Hospital and 11 from St. Vincent's Hospital (Figure 1). Baseline characteristics between groups were similar (Table 1). Notably, 58% of the sample had moderate or severe stroke. Median time to mobilization was shorter in VEM (18 hours) than in SC (31 hours), and the median total dose of mobilization was greater for VEM (167 minutes) than for SC (69 minutes).14 By 3 months, 11 patients (15%) had died, and at 12 months, 17 patients (24%) had died. The proportion of deaths between groups was not different and was low compared with population-based studies.14 Three surviving patients had missing data, but inspection of their characteristics showed that any systematic bias was unlikely. They were not all from one group (2 VEM, 1 SC), they had a range of ages (58, 82, and 86 years), and they did not have severe stroke (NIHSS of 1, 5 and 7).

Figure 1.

Figure 1. Participant flowchart.

Table 1. Baseline Characteristics of Recruited Patients

SC n=33 No. (%)*VEM n=38 No. (%)*Total N=71 No. (%)*
Patient details
    Age, mean (SD), y74.9 (9.8)74.6 (14.6)74.7 (12.5)
    Female17 (53)16 (42)33 (46)
Stroke details
    Left-sided lesion14 (42)22 (58)36 (51)
    NIHSS score
        Mild (1–7)15 (46)15 (39)30 (42)
        Moderate (8–16)11 (33)13 (34)24 (34)
        Severe (>16)7 (21)10 (26)17 (24)
    Oxfordshire Stroke Classification
        TACI6 (18)10 (26)16 (23)
        PACI10 (30)13 (34)23 (32)
        POCI5 (15)7 (18)12 (17)
        LACI6 (18)5 (13)11 (15)
        ICH6 (18)3 (8)9 (13)
    Prior history of stroke7 (21)11 (29)18 (25)
Stroke risk factors
    Hypertension25 (76)25 (66)50 (70)
    Ischemic heart disease13 (39)7 (18)20 (28)
    Angina9 (27)5 (13)14 (20)
    Hypercholesterolemia11 (33)8 (21)19 (27)
    Diabetes4 (12)11 (29)15 (21)
        Never smoked15 (45)15 (47)30 (46)
        Smoker6 (18)7 (22)13 (20)
        Ex-smoker12 (36)10 (31)22 (34)
        Unknown0 (0)6 (16)6 (8)
Factors limiting mobilization
    Respiratory5 (15)4 (11)9 (13)
    Lower limb§7 (21)11 (29)18 (25)
Premorbid history
    Premorbid mRS
        020 (61)18 (47)38 (54)
        18 (24)6 (16)14 (20)
        22 (6)8 (21)10 (14)
        33 (9)6 (16)9 (13)
    Living arrangement on admission
        Home, alone7 (21)4 (11)11 (15)
        Home, with someone25 (76)30 (79)55 (78)
        Hostel1 (3)4 (11)5 (7)

*No. (percent) unless otherwise indicated.

TACI indicates total anterior circulation infarct; PACI, partial anterior circulation infarct; POCI, posterior circulation infarct; LACI, lacunar infarct; and ICH, intracerebral hemorrhage.

Smoker=current smoker or who quit within the last 2 years. Ex-smoker=quit >2 years ago.

Respiratory limiting factors included emphysema and chronic obstructive airway disease.

§Lower-limb limiting factors included lower-limb arthritis and lower-limb joint replacement.

Days to Walking

On admission, 86% of patients could not walk or required hands-on assistance to walk short distances (mobility scale for acute stroke gait score 1 to 4; VEM n=34, SC n=27). Only 14% of patients were rated as able to walk with supervision (VEM n=4, SC n=6). Median days taken to return to walking 50 m was 3.5 (interquartile range [IQR]=1.5 to 14.0) in the VEM group and 7.0 (IQR=2.0 to 20.0) in the SC group. At 2 weeks after stroke (the end of the intervention period) among surviving patients, 67% (22 of 33) of the VEM group had returned to unassisted walking compared with 50% (16 of 32) of the SC group. Results of the Cox regression analysis, with adjustment for potentially confounding variables, demonstrated a significant impact of the intervention on time to walking (P=0.032; Table 2). The log-minus-log function indicated that the assumption of proportionality was met. Patients in the VEM group returned to independent walking sooner than did SC patients, as shown in the adjusted survival curves (Figure 2). Earlier return to walking after stroke was also independently associated with less severe stroke (NIHSS), younger age, and absence of diabetes (Table 2).

Table 2. Cox Regression for No. of Days to Walking 50 m Unassisted (N=71)

Hazard RatioLower CIUpper CIP
Premorbid mRS0.8670.6451.1650.344
Figure 2.

Figure 2. Number of days to walking 50 m unassisted in VEM and SC groups, adjusted for age, sex, stroke severity (NIHSS), premorbid mRS, and diabetes (N=71). Note: Of those patients who returned to walk in the 12 months after stroke, 142 days was the longest time taken.

Barthel Index

Proxy respondents provided answers for 21% of 3-month survivors and 17% of 12-month survivors (primarily due to the patient having communication difficulties). The distribution of Barthel scores was bimodal, with most patients clustered at either the low or high end of the scale. Scores on the Barthel Index at 3 months were higher for the VEM group (median=18.5, IQR=2.0 to 20.0) than the SC group (median=16.5, IQR=9.0 to 20.0), but this difference was not significant (P=0.713). In VEM, 47% (17 of 36) of patients had a good outcome on the Barthel Index at 3 months compared with 28% (9/32) of SC patients (P=0.136). Scores on the Barthel Index at 12 months were not significantly different between groups, with 39% of patients having a good outcome in both VEM (median=18.0, IQR=0.0 to 20.0) and SC (median=18.0, IQR=7.0 to 20.0) groups. In the multiple regression analysis, having a good outcome on the Barthel Index at 3 months was independently associated with exposure to VEM, less severe stroke on admission (NIHSS), and younger age. Having a good Barthel outcome at 12 months was independently associated with less severe stroke, younger age, male sex, and absence of diabetes, but the effect of exposure to VEM was no longer apparent (Table 3). Results were similar when the analyses were restricted to survivors only (data not shown).

Table 3. Multivariable Binary Logistic Regressions, With Independence in ADLs (Barthel Index=20 vs Barthel Index=0–19) at 3 and 12 Months as the Outcome Variable

3 Months (N=68)*
12 Months (N=69)
Premorbid mRS0.880.332.380.8080.510.161.630.257

OR indicates odds ratio.

*Eleven patients who died were included by assigning a Barthel score of 0; 3 survivors had missing data.

Seventeen patients who died were included by assigning a Barthel score of 0; 2 survivors had missing data.

Rivermead Motor Assessment

Scores on the Rivermead assessment at 3 months were similar in VEM (median=10.0, IQR=0.5 to 11.0) and SC (median=10.0, IQR=3.0 to 11.0) groups (P=0.883). In VEM, 62% (23 of 37) of patients had a good Rivermead outcome at 3 months compared with 56% (18 of 32) of SC patients (P=0.633). Scores on the Rivermead at 12 months were not significantly different between groups, with 53% of VEM patients having a good outcome (median=10.0, IQR=0.0 to 11.0) compared with 45% of SC patients (median=9.0, IQR=1.0 to 11.0). In multiple regression, having a good Rivermead outcome was independently associated with exposure to VEM, less severe stroke (NIHSS), and younger age, at both 3 and 12 months (Table 4). Results were similar when the analyses were restricted to survivors only, although age was no longer associated with good outcome (data not shown).

Table 4. Multivariable Binary Logistic Regressions, With Independence in Motor Function (Rivermead=10–13 vs Rivermead=0–9) at 3 and 12 Months as the Outcome Variable

3 Months (N=69)*
12 Months (N=69)
Premorbid mRS0.660.271.620.3610.440.181.050.065

OR indicates odds ratio.

*Eleven patients who died were included by assigning a Rivermead score of 0; 2 survivors had missing data.

Seventeen patients who died were included by assigning a Rivermead score of 0; 2 survivors had missing data.


The most important finding was that stroke patients who received VEM in addition to standard stroke unit care were able to walk unassisted sooner than patients who received standard stroke unit care alone. One major clinical implication of this result is that walking sooner increases the likelihood of milder stroke patients being discharged from acute care directly to home rather than to a rehabilitation facility. Our data support this possibility: despite the VEM group having more patients with moderate to severe strokes than the SC group, length of stay in the acute care hospital was shorter for VEM than for SC patients (median 6 vs 7 days), and they were more likely to be discharged directly to home (32% vs 24%). The study by Indredavik et al,12 in which a shorter time to start of mobilization was the most important factor associated with discharge to home, also supports this idea. A potential explanation for the faster return to walking is that motor performance was improved by earlier and more frequent out-of-bed experiences. It is also possible that the increased physical activity served to minimize the muscle loss and deterioration in cardiorespiratory function associated with bed rest25 and thus made walking 50 m unassisted less demanding. A further possibility is that self-efficacy was enhanced; that is, the early sessions of mobilization constituted “mastery experiences” that gave the patient increased confidence in his or her own ability to walk.26,27

Our hypothesis that patients exposed to the intervention would have better functional independence than controls at 3 months was supported by the findings from the multiple regression analyses, whereby more VEM patients were independent in ADLs (Barthel Index) and motor function (Rivermead Motor Assessment) than SC patients. Taken together with the walking findings, this indicates that providing earlier and more intensive mobilization after stroke can accelerate the recovery of meaningful physical outcomes. At 12 months, VEM patients remained more independent in motor function than controls, but there was no significant group difference for independence in ADLs. Previous studies have demonstrated that in the longer term, stroke patients who have not received augmented training eventually achieve the same level of function as those who have.3 It should be noted that our study was not powered to detect changes in functional outcomes in response to the addition of an early mobilization program. As Tables 3 and 4 show, the confidence intervals around the effect of intervention group were wide, and whether these promising findings are confirmed in the larger phase III study (currently underway)28 remains to be seen.

Age and stroke severity strongly influenced recovery as measured by the Barthel and Rivermead at both 3 and 12 months after stroke, with younger patients and those with less severe stroke achieving greater functional independence. This is consistent with previous findings.79 The negative influence of diabetes on functional independence found with walking recovery and the Barthel Index at 12 months also has precedent.8,10

As outcome measures in a research context, the Barthel Index and Rivermead Motor Assessment are not perfect. They are brief, ordinal scales that produce data that are not normally distributed, with ceiling effects a particular issue for the Barthel. It is possible that differences in independence in ADLs between groups were present at 12 months in the current study but that a ceiling effect in the Barthel data precluded their detection. More than half of the surviving patients were at ceiling on the Barthel scale (20/20) at 12 months after stroke. The weakness of the Barthel in discriminating between people at moderate versus high levels of function has been well documented.29 Furthermore, neither the Barthel nor the Rivermead measure incorporates death in its scale (as the mRS does). In contrast to many published rehabilitation studies, deaths were not excluded from our analyses of physical outcomes. Rather, we assigned patients who had died a score of 0 for both the Barthel and Rivermead measures to prevent overestimation of recovery in this group. In light of these considerations, it was necessary to dichotomize the Barthel and Rivermead data for analyses so that statistical assumptions were not violated. The consequence of the dichotomization approach is a reduction in sensitivity of the measures. These assessments, however, are widely used and readily understood in rehabilitation settings (both clinical and research), and this explains their choice as outcome measures here.

Any intervention that accelerates functional recovery is important both to stroke survivors and clinicians. This study provides novel evidence that earlier and more intensive mobilization in the acute phase of stroke can accelerate recovery of walking and functional independence. The findings from AVERT phase III will provide greater certainty regarding the efficacy of VEM in the recovery of physical independence after stroke.


We thank all of the stroke survivors and their caregivers who participated in this study. We gratefully acknowledge the support of the clinical staff at the Austin Hospital and St. Vincent's Hospital and all those involved at the National Stroke Research Institute.

Sources of Funding

This trial was supported by grants from the National Heart Foundation of Australia (grant number G 04M 1571), Affinity Health, and an equipment grant from the Austin Health Medical Research Fund. Dr Bernhardt was supported by a National Health and Medical Research Council (Australia) fellowship (157305).




Correspondence to Julie Bernhardt,
Level 2, Neurosciences Building, 300 Waterdale Road, Heidelberg Heights, Victoria, Australia, 3081
. E-mail


  • 1. Thrift AG, Dewey HM, Macdonell RAL, McNeil JJ, Donnan GA. Stroke incidence on the east coast of Australia: the North East Melbourne Stroke Incidence Study (NEMESIS). Stroke. 2000; 31:2087–2092.CrossrefMedlineGoogle Scholar
  • 2. Pound P, Gompertz P, Ebrahim S. A patient-centred study of the consequences of stroke. Clin Rehabil. 1998; 12:338–347.CrossrefMedlineGoogle Scholar
  • 3. Kwakkel G, van Peppen R, Wagenaar RC, Wood Dauphinee S, Richards C, Ashburn A, Miller K, Lincoln N, Partridge C, Wellwood I, Langhorne P. Effects of augmented exercise therapy time after stroke: a meta-analysis. Stroke. 2004; 35:2529–2539.LinkGoogle Scholar
  • 4. Mayo NE, Korner-Bitensky NA, Becker R. Recovery time of independent function post-stroke. Am J Phys Med Rehabil. 1991; 70:5–12.CrossrefMedlineGoogle Scholar
  • 5. Wade DT, Hewer RL. Functional abilities after stroke: measurement, natural history and prognosis. J Neurol Neurosurg Psychiatry. 1987; 50:177–182.CrossrefMedlineGoogle Scholar
  • 6. Jorgensen HS, Nakayama H, Raaschou H, Olsen TS. Recovery of walking function in stroke patients: the Copenhagen stroke study. Arch Phys Med Rehabil. 1995; 76:27–32.CrossrefMedlineGoogle Scholar
  • 7. Glymour MM, Berkman LF, Ertel KA, Fay ME, Glass TA, Furie KL. Lesion characteristics, NIH stroke scale, and functional recovery after stroke. Am J Phys Med Rehabil. 2007; 86:725–733.CrossrefMedlineGoogle Scholar
  • 8. Weimar C, Ziegler A, Konig IR, Diener H-C. Predicting functional outcome and survival after acute ischemic stroke. J Neurol. 2002; 249:888–895.CrossrefMedlineGoogle Scholar
  • 9. Dhamoon MS, Moon YP, Paik MC, Boden-Albala B, Rundek T, Sacco RL, Elkind MSV. Long-term functional recovery after first ischemic stroke: the northern Manhattan study. Stroke. 2009; 40:2805–2811.LinkGoogle Scholar
  • 10. Hankey GJ, Spiesser J, Hakimi Z, Bego G, Carita P, Gabriel S. Rate, degree, and predictors of recovery from disability following ischemic stroke. Neurology. 2007; 68:1583–1587.CrossrefMedlineGoogle Scholar
  • 11. Indredavik B, Bakke F, Solberg R, Rokseth R, Haaheim LL, Holme I. Benefit of a stroke unit: a randomized controlled trial. Stroke. 1991; 22:1026–1031.LinkGoogle Scholar
  • 12. Indredavik B, Bakke F, Slordahl SA, Rokseth R, Haheim LL. Treatment in a combined acute and rehabilitation stroke unit: which aspects are most important?Stroke. 1999; 30:917–923.CrossrefMedlineGoogle Scholar
  • 13. Bernhardt J, Dewey H, Thrift A, Donnan G. Inactive and alone: physical activity within the first 14 days of acute stroke unit care. Stroke. 2004; 35:1005–1009.LinkGoogle Scholar
  • 14. Bernhardt J, Dewey H, Thrift A, Collier J, Donnan G. A very early rehabilitation trial for stroke (AVERT): phase II safety and feasibility. Stroke. 2008; 39:390–396.LinkGoogle Scholar
  • 15. de Haan R, Limburg M, Bossuyt P, van der Meulen J, Aaronson N. The clinical meaning of Rankin ‘handicap’ grades after stroke. Stroke. 1995; 26:2027–2030.CrossrefMedlineGoogle Scholar
  • 16. Brott T, Adams HP, Olinger CP, Marler JR, Barson WG, Biller J, Spilker J, Holleran R, Eberle R, Hertzberg V, Rorick M, Moomaw CJ, Walker M. Measurements of acute cerebral infarction: a clinical examination scale. Stroke. 1989; 20:864–870.LinkGoogle Scholar
  • 17. Bamford J, Sandercock P, Dennis M, Burn J, Warlow C. A prospective study of acute cerebrovascular disease in the community: the Oxfordshire community stroke project–1981–86, 2: incidence, case fatality rates and overall outcome at one year of cerebral infarction, primary intracerebral and subarachnoid haemorrhage. J Neurol Neurosurg Psychiatry. 1990; 53:16–22.CrossrefMedlineGoogle Scholar
  • 18. Simondson JA, Goldie P, Greenwood KM. The mobility scale for acute stroke patients: concurrent validity. Clin Rehabil. 2003; 17:558–564.CrossrefMedlineGoogle Scholar
  • 19. Heinemann AW, Linacre JM, Wright BD, Hamilton BB, Granger C. Prediction of rehabilitation outcomes with disability measures. Arch Phys Med Rehabil. 1994; 75:133–143.MedlineGoogle Scholar
  • 20. Lincoln N, Leadbitter D. Assessment of motor function in stroke patients. Physiotherapy. 1979; 65:48–51.MedlineGoogle Scholar
  • 21. Collin C, Wade DT, Davies S, Horne V. The Barthel ADL index: a reliability study. Int Disabil Studies. 1988; 10:61–63.CrossrefMedlineGoogle Scholar
  • 22. Loewen SC, Anderson BA. Predictors of stroke outcome using objective measurement scales. Stroke. 1990; 21:78–81.CrossrefMedlineGoogle Scholar
  • 23. Collen FM, Wade DT, Bradshaw CM. Mobility after stroke: reliability of measures of impairment and disability. Int Disabil Studies. 1990; 12:6–9.CrossrefMedlineGoogle Scholar
  • 24. Sackley CM, Lincoln NB. The verbal administration of the gross function scale of the Rivermead motor assessment. Clin Rehabil. 1990; 4:301–303.CrossrefGoogle Scholar
  • 25. Allen C, Glasziou P, Del Mar C. Bed rest: a potentially harmful treatment needing more careful evaluation. Lancet. 1999; 354:1229–1233.CrossrefMedlineGoogle Scholar
  • 26. Hellstrom K, Lindmark B, Wahlberg B, Fugl-Meyer AR. Self-efficacy in relation to impairments and activities of daily living disability in elderly patients with stroke: a prospective investigation. J Rehabil Med. 2003; 35:202–207.CrossrefMedlineGoogle Scholar
  • 27. Bonetti D, Johnston M. Perceived control predicting the recovery of individual-specific walking behaviours following stroke: testing psychological models and constructs. Br J Health Psychol. 2008; 13:463–478.CrossrefMedlineGoogle Scholar
  • 28. Bernhardt J, Dewey HM, Thrift AG, Collier JM, Lindley RI, Moodie M, Donnan GA. Lancet protocol review: protocol 06PRT/5424: A Very Early Rehabilitation Trial (AVERT): phase III. Published on-line Lancet. July2007. Scholar
  • 29. O'Connor RJ, Cano SJ, Thompson AJ, Hobart JC. Exploring rating scale responsiveness: does the total score reflect the sum of its parts?Neurology. 2004; 62:1842–1844.CrossrefMedlineGoogle Scholar