Repetitive task training for improving functional ability after stroke

Beverley French; Lois H Thomas; Jacqueline Coupe; Naoimh E McMahon; Louise Connell; Joanna Harrison; Christopher J Sutton; Svetlana Tishkovskaya; Caroline L Watkins

doi:10.1002/14651858.CD006073.pub3

. 2016 Nov 14;2016(11):CD006073. doi: 10.1002/14651858.CD006073.pub3

Repetitive task training for improving functional ability after stroke

Beverley French ¹, Lois H Thomas ^2,^✉, Jacqueline Coupe ², Naoimh E McMahon ², Louise Connell ², Joanna Harrison ³, Christopher J Sutton ², Svetlana Tishkovskaya ⁴, Caroline L Watkins ²

Editor: Cochrane Stroke Group

PMCID: PMC6464929 PMID: 27841442

Abstract

Background

Repetitive task training (RTT) involves the active practice of task‐specific motor activities and is a component of current therapy approaches in stroke rehabilitation.

Objectives

Primary objective: To determine if RTT improves upper limb function/reach and lower limb function/balance in adults after stroke.    Secondary objectives: 1) To determine the effect of RTT on secondary outcome measures including activities of daily living, global motor function, quality of life/health status and adverse events. 2) To determine the factors that could influence primary and secondary outcome measures, including the effect of 'dose' of task practice; type of task (whole therapy, mixed or single task); timing of the intervention and type of intervention.

Search methods

We searched the Cochrane Stroke Group Trials Register (4 March 2016); the Cochrane Central Register of Controlled Trials (CENTRAL) (the Cochrane Library 2016, Issue 5: 1 October 2006 to 24 June 2016); MEDLINE (1 October 2006 to 8 March 2016); Embase (1 October 2006 to 8 March 2016); CINAHL (2006 to 23 June 2016); AMED (2006 to 21 June 2016) and SPORTSDiscus (2006 to 21 June 2016).

Selection criteria

Randomised/quasi‐randomised trials in adults after stroke, where the intervention was an active motor sequence performed repetitively within a single training session, aimed towards a clear functional goal.

Data collection and analysis

Two review authors independently screened abstracts, extracted data and appraised trials. We determined the quality of evidence within each study and outcome group using the Cochrane 'Risk of bias' tool and GRADE (Grades of Recommendation, Assessment, Development and Evaluation) criteria. We did not assess follow‐up outcome data using GRADE. We contacted trial authors for additional information.

Main results

We included 33 trials with 36 intervention‐control pairs and 1853 participants. The risk of bias present in many studies was unclear due to poor reporting; the evidence has therefore been rated 'moderate' or 'low' when using the GRADE system.

There is low‐quality evidence that RTT improves arm function (standardised mean difference (SMD) 0.25, 95% confidence interval (CI) 0.01 to 0.49; 11 studies, number of participants analysed = 749), hand function (SMD 0.25, 95% CI 0.00 to 0.51; eight studies, number of participants analysed = 619), and lower limb functional measures (SMD 0.29, 95% CI 0.10 to 0.48; five trials, number of participants analysed = 419).

There is moderate‐quality evidence that RTT improves walking distance (mean difference (MD) 34.80, 95% CI 18.19 to 51.41; nine studies, number of participants analysed = 610) and functional ambulation (SMD 0.35, 95% CI 0.04 to 0.66; eight studies, number of participants analysed = 525). We found significant differences between groups for both upper‐limb (SMD 0.92, 95% CI 0.58 to 1.26; three studies, number of participants analysed = 153) and lower‐limb (SMD 0.34, 95% CI 0.16 to 0.52; eight studies, number of participants analysed = 471) outcomes up to six months post treatment but not after six months. Effects were not modified by intervention type, dosage of task practice or time since stroke for upper or lower limb. There was insufficient evidence to be certain about the risk of adverse events.

Authors' conclusions

There is low‐ to moderate‐quality evidence that RTT improves upper and lower limb function; improvements were sustained up to six months post treatment. Further research should focus on the type and amount of training, including ways of measuring the number of repetitions actually performed by participants. The definition of RTT will need revisiting prior to further updates of this review in order to ensure it remains clinically meaningful and distinguishable from other interventions.

Keywords: Adult, Humans, Activities of Daily Living, Physical Therapy Modalities, Recovery of Function, Extremities, Motor Activity, Randomized Controlled Trials as Topic, Stroke Rehabilitation, Stroke Rehabilitation/methods, Task Performance and Analysis, Walking

Plain language summary

Repetitive task training for improving functional ability after stroke

Review question: What are the effects of repeated practice of functional tasks on recovery after stroke when compared with usual care or placebo treatments?

Background: Stroke can cause problems with movement, often down one side of the body. While some recovery is common over time, about one third of people have continuing problems. Repeated practice of functional tasks (e.g. lifting a cup) is a treatment approach used to help with recovery of movement after stroke. This approach is based on the simple idea that in order to improve our ability to perform tasks we need to practice doing that particular task numerous times, like when we first learned to write. The types of practice that people do, and the time that they spend practicing, may affect how well this treatment works. To explore this further we also looked at different aspects of repetitive practice that may influence how well it works.

Study characteristics: We identified 33 studies with 1853 participants. Studies included a wide range of tasks to practice, including lifting a ball, walking, standing up from sitting and circuit training with a different task at each station. The evidence is current to June 2016.

Key results: In comparison with usual care (standard physiotherapy) or placebo groups, people who practiced functional tasks showed small improvements in arm function, hand function, walking distance and measures of walking ability. Improvements in arm and leg function were maintained up to six months later. There was not enough evidence to be certain about the risk of adverse events, for example falls. Further research is needed to determine the best type of task practice, and whether more sustained practice could show better results.

Quality of the evidence: We classified the quality of the evidence as low for arm function, hand function and lower limb functional measures, and as moderate for walking distance and functional ambulation. The quality of the evidence for each outcome was limited due poor reporting of study details (particularly in earlier studies), inconsistent results across studies and small numbers of study participants in some comparisons.

Summary of findings

for the main comparison.

Repetitive task training compared with usual care or attention control for patients with stroke
Patient or population: people with stroke Settings: hospital, clinic or home Intervention: repetitive task training (RTT) Comparison: usual care, attention control or no treatment
Outcomes	Illustrative comparative risks (95% CI)		Relative effect  (95% CI)	No of participants  (studies)	Quality of the evidence  (GRADE)	Comments
	Assumed risk	Corresponding risk
	Estimated score / value with control	Absolute reduction in score / value with RTT^a
Arm function	Arm function score in the repetitive task training groups was on average 0.25 standard deviations (0.01 to 0.49) higher than in the control groups. SD units, measured using different instruments; higher scores mean better arm function.		SMD 0.25, 95% CI 0.01 to 0.49	11 studies 749 participants	⊕⊕⊝⊝  low	Downgraded by one level for inconsistency (1² 58%). Downgraded by one level for study design (random sequence generation unclear in 4/11 trials and high risk in 1/11 trials in the meta‐analysis; allocation concealment unclear in 7/11 trials and high risk in 1/11 trials).
Hand function	Hand function score in the repetitive task training groups was on average 0.25 standard deviations (0.00 to 0.51) higher than in the control groups. SD units, measured using different instruments; higher scores mean better hand function.		SMD 0.25, 95% CI 0.00 to 0.51	8 studies 619 participants	⊕⊕⊝⊝  low	Downgraded by one level for inconsistency (1² 54%). Downgraded by one level for study design (random sequence generation unclear in 2/8 trials and high risk in 1/8 trials in the meta‐analysis; allocation concealment unclear in 4/8 trials and high risk in 1/8 trials).
Walking distance: change from baseline	The mean change in walking distance (metres walked in six minutes; a higher score means greater walking distance) in the control groups ranged from ‐1.0 to 118.5.	The mean change in walking distance (metres walked in six minutes; a higher score means greater walking distance) in the repetitive training group ranged from 19 to 221.	MD 34.80, 95% CI 18.19 to 51.41	9 studies 610 participants	⊕⊕⊕⊝  moderate	Downgraded by one level for study design (random sequence generation unclear in 6/9 trials in the meta‐analysis; allocation concealment unclear in 6/9 trials and high risk in 3/9 trials).
Walking speed	The mean walking speed in the control groups ranged from  0.29 to 2.47 metres per second. A higher score means faster walking speed.	The mean walking speed in the intervention groups ranged from 0.39 to 2.03 metres per second. A higher score means faster walking speed.	SMD 0.39, 95% CI ‐0.02 to 0.79	12 studies 685 participants	⊕⊕⊝⊝  low	Downgraded by one level for inconsistency (1² 80%). Downgraded by one level for study design (random sequence generation unclear in 7/12 trials in the meta‐analysis; allocation concealment unclear in 9/12 trials and high risk in 3/12 trials).
Functional ambulation	Functional ambulation score in the repetitive task training groups was on average 0.35 standard deviations (0.04 to 0.66) higher than in the control groups. SD units, measured using different instruments; higher scores mean better function.		SMD 0.35, 95% CI 0.04 to 0.66	8 studies 525 participants	⊕⊕⊕⊝  moderate	Downgraded by one level for study design (random sequence generation unclear in 4/8 trials in the meta‐analysis; allocation concealment unclear in 7/8 trials and high risk in 1/8 trials).
Lower limb functional measures	Lower limb functional measures in the repetitive task training groups were on average 0.29 standard deviations (0.10 to 0.48) higher than in the control groups. SD units, measured using different instruments; higher scores mean better function.		SMD 0.29, 95% CI 0.10 to 0.48	5 studies 419 participants	⊕⊕⊝⊝  low	Downgraded by one level for study design (random sequence generation unclear in 3/5 trials in the meta‐analysis; allocation concealment unclear in 3/5 trials and high risk in 1/5 trials). Downgraded by one level for publication bias; 4 out of 5 are small studies (less than 50 participants).
Global motor function scales	Global motor function in the repetitive task training groups was on average 0.38 standard deviations (0.11 to 0.65) higher than in the control groups. SD units, measured using different instruments; higher scores mean better function.		SMD 0.38, 95% CI 0.11 to 0.65	5 studies 222 participants	⊕⊕⊕⊝  moderate	Downgraded by one level for study design (random sequence generation unclear in 4/5 trials in the meta‐analysis; allocation concealment unclear in 4/5 trials and high risk in 1/5 trials).
Adverse events	Barreca 2004: 3/25 (12%) falls in the intervention group versus 4/23 (17.4%) in the control group, OR 0.65, 95% CI 0.13 to 3.27. Holmgren 2010: 11 participants in total fell during study (32%), five in the intervention group and six in the attention control group. van de Port 2012: 29 falls reported in the circuit training group and 26 in the usual physiotherapy group (P = 0.93). Two serious adverse events were reported in the circuit training group: one participant fell and consulted a GP and one patient experienced arrhythmias during one session. Winstein 2016: 168 serious adverse events involving 109 participants. The most common were hospitalisation (n = 143, 25% of randomised participants) and recurrent stroke (n = 42, 9% of randomised participants). Adverse events were not presented by trial arm. Salbach 2004: intervention‐related reasons for withdrawal that could be interpreted as adverse events included one participant out of 47 in a mobility training group who experienced the onset of groin pain. Four participants also fell during the mobility intervention but did not suffer injury and continued to participate in the group. Two falls also occurred during evaluation. Two trials narratively reported no adverse effects (de Sèze 2001; McClellan 2004).
^a As a rule of thumb, 0.2 SD represents a small difference, 0.5 a moderate, and 0.8 a large difference. CI: confidence interval; MD: mean difference; SMD: standardised mean difference; OR: odds ratio; SD: standard deviation
GRADE Working Group grades of evidence  High quality: Further research is very unlikely to change our confidence in the estimate of effect.  Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.  Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.  Very low quality: We are very uncertain about the estimate.

STUDY	Task practice dose	Time since stroke	Type of intervention	Practice intensity	Allocation conceal	Comparison group	Therapy equivalence	Small trials
	1 = 20 hours or less 2 = more than 20 hours	1 = 1 to 14 days 2 = 15 days to 6 months 3 = more than 6 months	1 = whole therapy 2 = mixed task 3 = single task	1 = 1 to 4 weeks or less 2 = more than 4 weeks	A = adequate B = inadequate/unclear	AC = attention control UC = usual care	EQ = equivalent therapy time ADD = additional therapy time	1 = less than 25 participants 2 = 25 or more participants
Arya 2012	Not reported	2	1	1	A	UC	EQ	2
Baer 2007	Not reported	3	2	1	B	UC	ADD	2
Barreca 2004	1	2	1	2	B	AC	ADD	2
Blennerhassett 2004	1	2	2	1	B	AC	EQ	2
Dean 1997	1	3	3	1	B	AC	EQ	1
Dean 2000	1	3	2	1	B	AC	EQ	1
Dean 2007	1	2	3	1	B	AC	EQ	1
de Sèze 2001	1	2	3	1	B	UC	EQ	2
Frimpong 2014	1	2	2	2	B	UC	ADD	1
Gordon 2013	1	3	2	2	B	AC	EQ	2
Holmgren 2010	2	2	2	2	A	UC	ADD	2
Howe 2005	1	2	3	1	A	UC	ADD	2
Kim 2012	1	3	2	1	B	UC	ADD	1
Kim 2014	1	3	2	1	B	UC	ADD	2
Kim 2016	2	2	2	1	B	UC	ADD	1
Kwakkel 1999	2	1	2	2	B	AC	EQ	2
Langhammer 2000	1	1	1	1	B	UC	EQ	2
Lennon 2009	1	1	2	1	B	UC	EQ	2
McClellan 2004	2	3	2	2	B	AC	EQ	2
Mudge 2009	1	3	2	1	B	AC	EQ	2
Olawale 2011	2	3	2	2	B	UC	EQ	2
Park 2011	1	3	2	1	B	UC	ADD	2
Peurala 2009	2	1	2	1	B	UC	ADD	2
Ross 2009	2	3	2	2	A	UC	ADD	2
Salbach 2004	1	3	2	2	B	AC	EQ	2
Song 2015	1	3	2	1	B	UC	ADD	1
Tung 2010	1	3	1	1	B	UC	ADD	2
Turton 1990	2	2	2	2	B	UC	ADD	1
van de Port 2012	2	2	2	2	B	UC	ADD	2
Van Vliet 2005	1	1	1	1	B	UC	EQ	2
Winstein 2004	1	1	2	1	B	UC	ADD	2
Winstein 2016	2	2	2	2	A	UC	EQ	2
Yen 2005	2	3	2	1	B	UC	EQ	2

Author and year	Global function	Lower limb function	Balance/sit‐to‐stand	Upper limb function	Hand function	ADL function	QOL, health status	Adverse events
Arya 2012				Action Research Arm Test ‐ gross arm movement
Barreca 2004			Number of participants able to stand				Dartmouth COOP	Falls
Blennerhassett 2004; Blennerhassett 2004a; Blennerhassett 2004b		6 Minute Walk Test; Step Test	Timed Up & Go Test	Motor Assessment Scale ‐ arm	Motor Assessment Scale ‐ hand
Dean 1997		10 Metre Walk Speed	Reaching distance
Dean 2000		6 Minute Walk Test; 10 Metre Walk Speed; Step Test	Timed Up & Go Test
Dean 2007		10 Metre Walk Test	Reaching distance
de Sèze 2001		Functional Ambulation Classification	Sitting and Standing Equilibrium Index			Functional Independence Measure
Frimpong 2014		10 Metre Walk Test Functional Ambulatory Category
Gordon 2013		6 Minute Walk Test				Barthel Index	SF‐36 physical health component
Holmgren 2010			Berg Balance Scale			Barthel Index
Howe 2005			Lateral reach ‐ time, sit‐to‐stand ‐ time
Kim 2012		10 Metre Walk Speed	Berg Balance Scale; Timed Up & Go Test
Kim 2014	Stroke Impact Scale ‐ social participation subscale	10 Metre Walk Test 6 Minute Walk Test
Kim 2016		6 Minute Walk Test	Berg Balance Scale			Korean version of Modified Barthel Index
Kwakkel 1999; Kwakkel 1999a; Kwakkel 1999b		Functional Ambulation Classification; Walking speed		Action Research Arm Test		Barthel Index	Nottingham Health Profile
Langhammer 2000	Motor Assessment Scale	Motor Assessment Scale ‐ walking; Sødring Motor Evaluation Scale ‐ trunk, balance and gait	Motor Assessment Scale ‐ balanced sitting, Motor Assessment Scale ‐ sit‐to‐stand	Motor Assessment Scale ‐ arm	Motor Assessment Scale ‐ hand	Barthel Index	Nottingham Health Profile
Lennon 2009		5 Metre Walk Speed
McClellan 2004		Motor Assessment Scale ‐ walking	Functional Reach Test
Mudge 2009		6 Minute Walk Test
Olawale 2011		10 Metre Walk Speed
Park 2011		10 Metre Walk Speed; 6 Minute Walk Test; Walking ability questionnaire	Activities‐Specific Balance Confidence Scale
Peurala 2009	Rivermead Mobility Index
Ross 2009					Wolf Motor Function Test (functional score)	Canadian Occupational Performace Measure
Salbach 2004; Salbach 2004a; Salbach 2004b		6 Minute Walk Test; 5 Metre Walk Speed	Timed Up and Go Test; Berg Balance Scale	Box & Block Test	9 Hole Peg Test	Barthel Index
Tung 2010			Berg Balance Scale
Turton 1990				Southern Motor Group's Motor Assessment ‐ upper extremity	10 Hole Peg Test
van de Port 2012		6 Minute Walk Test; 5 Metre Walk Speed; Stroke Impact Scale ‐ mobility domain	Timed Balance Test
Van Vliet 2005	Rivermead Motor Assessment ‐ gross function	Rivermead Motor Assessment ‐ leg and trunk; 6 Minute Walk Test; Motor Assessment Scale ‐ walking, Motor Assessment Scale ‐ leg and truck	Motor Assessment Scale ‐ balanced sitting, Motor Assessment Scale ‐ sit‐to‐stand	Motor Assessment Scale ‐ arm	Motor Assessment Scale ‐ hand	Barthel Index
Winstein 2004				Functional Test of the Hemiparetic Upper Extremity
Winstein 2016				Log Wolf Motor Function Test	Stroke Impact Scale ‐ hand function
Yen 2005				Wolf Motor Function Test

Date	Event	Description
22 September 2016	New citation required and conclusions have changed	The conclusions of the review have changed since the original review was published in 2007; there is now low‐quality evidence for the effectiveness of repetitive task training on upper limb function.
22 September 2016	New search has been performed	We updated selected searches to March 2016. We have added 19 new studies with 1355 participants, bringing the total number of included studies to 33, involving a total of 2014 participants. We have revised the review throughout and included an assessment of the quality of the evidence (presented in a 'Summary of findings' table).

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Arm function	11	749	Std. Mean Difference (IV, Random, 95% CI)	0.25 [0.01, 0.49]
2 Hand function	8	619	Std. Mean Difference (IV, Random, 95% CI)	0.25 [0.00, 0.51]
3 Sitting balance/reach	6	222	Std. Mean Difference (IV, Fixed, 95% CI)	0.28 [0.01, 0.55]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 All outcomes	9		Std. Mean Difference (IV, Fixed, 95% CI)	Subtotals only
1.1 Under 6 months post treatment	3	153	Std. Mean Difference (IV, Fixed, 95% CI)	0.92 [0.58, 1.26]
1.2 6 to 12 months post treatment	6	412	Std. Mean Difference (IV, Fixed, 95% CI)	0.10 [‐0.09, 0.30]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Dosage of task practice	15	833	Std. Mean Difference (IV, Random, 95% CI)	0.33 [0.11, 0.56]
1.1 0 to 20 hours	9	383	Std. Mean Difference (IV, Random, 95% CI)	0.23 [0.00, 0.46]
1.2 More than 20 hours	6	450	Std. Mean Difference (IV, Random, 95% CI)	0.38 [‐0.03, 0.80]
2 Time since stroke	15	833	Std. Mean Difference (IV, Random, 95% CI)	0.33 [0.11, 0.56]
2.1 0 to 15 days	4	239	Std. Mean Difference (IV, Random, 95% CI)	0.21 [‐0.04, 0.47]
2.2 16 days to 6 months	7	421	Std. Mean Difference (IV, Random, 95% CI)	0.48 [0.06, 0.91]
2.3 More than 6 months	4	173	Std. Mean Difference (IV, Random, 95% CI)	0.24 [‐0.23, 0.72]
3 Type of intervention	15	833	Std. Mean Difference (IV, Random, 95% CI)	0.33 [0.11, 0.56]
3.1 Whole therapy	3	240	Std. Mean Difference (IV, Random, 95% CI)	0.51 [‐0.18, 1.20]
3.2 Mixed training	8	509	Std. Mean Difference (IV, Random, 95% CI)	0.14 [‐0.03, 0.32]
3.3 Single task training	4	84	Std. Mean Difference (IV, Random, 95% CI)	0.71 [0.11, 1.30]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Walking distance: change from baseline	9	610	Mean Difference (IV, Random, 95% CI)	34.80 [18.19, 51.41]
2 Walking speed	12	685	Std. Mean Difference (IV, Random, 95% CI)	0.39 [‐0.02, 0.79]
3 Functional ambulation	8	525	Std. Mean Difference (IV, Random, 95% CI)	0.35 [0.04, 0.66]
4 Sit‐to‐stand: post treatment/change from baseline	7	346	Std. Mean Difference (Fixed, 95% CI)	0.35 [0.13, 0.56]
5 Lower limb functional measures	5	419	Std. Mean Difference (IV, Fixed, 95% CI)	0.29 [0.10, 0.48]
6 Standing balance/reach	9	504	Std. Mean Difference (IV, Fixed, 95% CI)	0.24 [0.07, 0.42]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Dosage of task practice	24	1144	Std. Mean Difference (IV, Random, 95% CI)	0.32 [0.12, 0.53]
1.1 0 to 20 hours	16	583	Std. Mean Difference (IV, Random, 95% CI)	0.39 [0.07, 0.71]
1.2 More than 20 hours	8	561	Std. Mean Difference (IV, Random, 95% CI)	0.33 [0.16, 0.50]
2 Time since stroke	24	1144	Std. Mean Difference (IV, Random, 95% CI)	0.32 [0.12, 0.53]
2.1 0 to 15 days	5	288	Std. Mean Difference (IV, Random, 95% CI)	0.16 [‐0.15, 0.46]
2.2 16 days to 6 months	9	428	Std. Mean Difference (IV, Random, 95% CI)	0.52 [‐0.03, 1.07]
2.3 More than 6 months	10	428	Std. Mean Difference (IV, Random, 95% CI)	0.41 [0.21, 0.60]
3 Type of intervention	24	1144	Std. Mean Difference (IV, Random, 95% CI)	0.32 [0.12, 0.53]
3.1 Whole therapy	2	138	Std. Mean Difference (IV, Random, 95% CI)	0.10 [‐0.24, 0.43]
3.2 Mixed training	17	894	Std. Mean Difference (IV, Random, 95% CI)	0.42 [0.17, 0.67]
3.3 Single task training	5	112	Std. Mean Difference (IV, Random, 95% CI)	0.07 [‐0.42, 0.55]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Activities of daily living function	9	527	Std. Mean Difference (IV, Fixed, 95% CI)	0.28 [0.10, 0.45]
2 Global motor function scales	5	222	Std. Mean Difference (IV, Fixed, 95% CI)	0.38 [0.11, 0.65]
3 Quality of life/health status	4	264	Std. Mean Difference (IV, Fixed, 95% CI)	0.28 [0.04, 0.53]

Methods	An assessor‐blinded, multicentre randomised controlled design
Participants	India Participants were recruited from an inpatient neurology ward and occupational therapy unit of a rehabilitation institute. Date of recruitment not reported 103 participants: 51 RTT, 52 control Inclusion criteria: first episode of unilateral stroke with hemiparesis, 4 to 24 weeks post stroke, functional ambulation classification level I and above, ability to understand instructions (Hindi Mental State Examination > 24), National Institute of Health Stroke Scale score < 14, able to cope with intensive training program, Brunnstrom stage of arm recovery of 2 to 5 Exclusion criteria: perceptual deficits such as neglect and apraxia, dementia, depression, impaired vision, impaired conscious level, concomitant medical illness, cardiovascular instability (resting systolic blood pressure > 200 mmHg and resting diastolic blood pressure > 100 mmHg), shoulder subluxation, aphasia, sensory loss Mean age: RTT 51.67 years (SD 7.96), control 50.21 years (SD 7.60) 60.2% male Stroke details: first stroke, ischaemic 66.9%, haemorrhagic 33.0% Timing post stroke: RTT 11.92 weeks (SD 6.49), control 12.37 weeks (SD 6.64) Pre‐intervention functional ability level: Functional ambulation classification level I and above
Interventions	RTT intervention: Meaningful Task Specific Training (MTST) is a training program for upper extremity rehabilitation of post stroke clients based on principles of motor learning, experience dependent neuroplasticity, and shaping techniques MTST mainly comprises the specific number of meaningful tasks, which are common to all the patients The tasks have to be practiced repetitively either with unilateral (the most affected extremity) or bilateral upper limb/s, depending on the task requirement It also has a component of individualised meaningful tasks, which have to be selected from a task bank for repetitive practice Sessions were delivered as one‐to‐one outpatient rehabilitation in day care units 1 hour per day (unclear number of times per week) for 4 weeks Comparison group: the control group was given an intervention of the same duration based on the Brunnstrom movement therapy and Bobath neurodevelopmental technique
Outcomes	Outcome measures were recorded at baseline, 4 weeks (post treatment) and 8 weeks Upper limb functional outcome measures: Fugl‐Meyer assessment, Action Research Arm Test, Graded Wolf Motor Function Test, Motor Activity Log
Notes	No significant differences at baseline 1 experimental dropout due to personal reasons No adverse events reported
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	A random‐number generator program was used
Allocation concealment (selection bias)	Low risk	Intervention assignments were enclosed in sealed envelopes, which were opaque and sequentially numbered
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Blinded assessors were trained to administer the measures properly, and they did not participate in providing the interventions
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Missing data balanced across intervention groups An intention‐to‐treat analysis was used with the last observation carried forward for the missing data
Selective reporting (reporting bias)	Unclear risk	No study protocol

Methods	Prospective, single‐blind randomised controlled pilot trial
Participants	UK Participants were recruited from the community 64 participants: 20 Whole practice, 23 Part practice, 21 control Inclusion criteria: Age over 18 At least 12 months post stroke Residual neurological physical deficit due to stroke Discharged from formal Physiotherapy Mini Mental State Examination (MMSE) score ≥ 22 Functional Reach Test ≥15 cm Able to understand the nature of the study and give informed consent Exclusion criteria: Age under 18 Pre‐existing gross neuropathology – e.g. Multiple Sclerosis, Parkinson's Disease Co‐existing pathology that would prohibit independent exercise – e.g. lower limb fracture Pre‐existing disabilities with grossly limited mobility (e.g. lower limb amputation) History of two falls within the previous six months Mean age: 72.9 ± 9.0 48.4% male Stroke details: 38 right CVA, 26 left CVA Timing post stroke: 30.3 ± 28.8 months Pre‐intervention functional activity level: RTT WP MAS mean 26.5 (SD 8.5), RTT PP 28.1 (7.9), control 26.5 (8.9); (MAS score ≥32 designated as "mild", ≤ 31 designated "moderate to severe")
Interventions	RTT intervention: the exercises consisted of practising standing up from a chair, sitting down, stepping onto a step, stepping off a step, pronation and supination holding a bottle and reaching and grasping. All participants allocated to an exercise "arm" of the trial practised the same functional exercises but in different ways: entirety (whole practice) or component parts (part practice). The clinical research assistant encouraged participants to increase the number of repetitions of exercises practiced if assessed to be appropriate The target number of repetitions of each exercise was documented in the exercise diary and participants were requested to document the actual number undertaken Comparison group: participants did not receive any physical intervention or exercise instruction but received the same number of visits by the research assistant to counteract the possible therapist interaction effect
Outcomes	Outcome measures were recorded at baseline (2 measurements 2 weeks apart), 4 weeks (post treatment) and 72 hours and 3 months post treatment Upper limb functional outcome measure: Frenchay Arm Test Lower limb functional outcome measures: Timed Up and Go 2m, Step Test, Step‐up count of 15 seconds Other outcome measures: Barthel Index, MAS, Frenchay Activity Index, Stroke Impact Scale (activity)
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomised blocks within 4 strata of participants were used, based on side and severity of stroke, due to the small numbers of participants
Allocation concealment (selection bias)	High risk	Method of concealment not described
Blinding of outcome assessment (detection bias)   All outcomes	High risk	Outcome data collected by the principal investigator who was blind to group allocation. The outcome assessor became aware of group allocation for 3 participants
Incomplete outcome data (attrition bias)   All outcomes	Low risk	At the short‐term follow‐up point, data were missing for 5 control, 3 RTT PP and one RTT WP participant. Reasons are given by intervention group; for 2/5 of participants in the control group reasons were not illness related.
Selective reporting (reporting bias)	Unclear risk	No study protocol

Methods	Single centre RCT
Participants	Canada 48 participants: 25 RTT, 23 control Participants were recruited from stroke rehabilitation units between 2000 and 2001 Inclusion criteria: between the ages of 18 to 90 years, medically stable, had a postural control of Stage 3 or greater as measured by the Chedoke‐McMaster Stroke Assessment (CMSA), and failed the third item of the CMSA Stage 4 Postural Control Exclusion criteria: none stated Median age: RTT 67 years (IQR 56 ‐ 72), control 70 years (IQR 64 ‐ 78) 65% male Stroke details: not stated whether first or recurrent stroke, 73% ischaemic, 42% right hemiparesis Timing post stroke: RTT median 30 days (IQR 21 to 48), control median 31 days (IQR 18 to 50) Pre‐intervention functional ability level: lack of postural control
Interventions	RTT intervention: sit‐to‐stand training. Group class practice in attaining standing from sitting from a variety of different heights and surfaces Training was additional to usual care, which included daily strengthening exercise, repetitive training, functional training, electrical stimulation and other exercise Sessions were 45 minutes, 3 times per week until competence or discharge (approximately 6 weeks) = 13.5 hours + practice on ward Each session aimed to involve 3 practice sets of 5 sit‐to‐stand manoeuvres per class Average total repetitions during training = 450 to 500 Classes had 6 to 7 participants, supervised by 2 registered practical nurses, with extra practice delivered by nurses trained on the sit‐to‐stand protocol in a ward setting using videotapes, written instruction and practice Comparison group: usual care and recreation therapy
Outcomes	Outcome measures were recorded at baseline and at competence or discharge (approximately 6 weeks) Balance/sit‐to‐stand outcome measures: ability to stand independently and safely on 2 consecutive occasions Adverse events outcome measures: number of falls QoL/health status outcome measures: satisfaction with ability to stand, Dartmouth Primary Care Cooperative Chart
Notes	No significant differences in baseline characteristics
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Assigned by coin flip
Allocation concealment (selection bias)	Unclear risk	Inadequately reported
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	A research physiotherapist, blind to the study, tested the participants' STS movement once per week
Incomplete outcome data (attrition bias)   All outcomes	Low risk	No loss to follow‐up at end of treatment phase
Selective reporting (reporting bias)	Unclear risk	No study protocol

Methods	Single‐centre RCT
Participants	Australia 30 participants: 15 mobility group, 15 upper limb group Participants were recruited from inpatient admissions, with a primary diagnosis of stroke, to a rehabilitation centre between 2001 and 2003 Inclusion criteria: able to walk 10 metres and provide informed consent Exclusion criteria: deteriorating medical condition, independent community ambulation Mean age: mobility group 53.9 years (SD 19.8), upper limb group 56.3 years (SD 10.5) 56.6% male Stroke details: first or recurrent stroke, 73% ischaemic, 47% right hemiparesis Timing post stroke: mobility group 36 days (SD 25.1), upper limb group 50 days (SD 49.2) Pre‐intervention functional ability level: Six Minute Walk Test: mobility group 183 metres (SD 85), upper limb group 181 metres (SD 85)
Interventions	RTT interventions: Mobility group: circuit training including sit‐to‐stand, step‐ups, obstacle course, plus stretching/strengthening exercise, and some endurance training (stationary bikes/treadmill) Upper limb group: reach and grasp, hand‐eye co‐ordination activities, stretching and strengthening exercises Sessions were during inpatient rehabilitation and additional to usual care of 5 hours per week Sessions were 60 minutes, 5 times per week for4 weeks = 20 hours Each circuit included 10 x 5‐minute workstations Sessions were delivered by a physical therapist in groups of up to 4 participants Comparison group: Blennerhassett 2004a lower limb attention control Blennerhassett 2004b upper limb attention control
Outcomes	Outcome measures were recorded at baseline, 4 weeks (post treatment) and 6 months after training Upper limb functional outcome measures: MAS, Jebsen Taylor Test of Hand Function Lower limb functional outcome measures: Six Minute Walk Test, Step Test Balance/sit‐to‐stand outcome measures: Timed Up and Go Test
Notes	No significant differences reported at baseline  3% lost to follow‐up at end of treatment phase  No likely intervention‐related withdrawals  Average attendance was approximately 80%, with no significant difference between the groups
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Inadequately reported
Allocation concealment (selection bias)	Unclear risk	Randomisation performed by a person independent from the study, drawing a pre‐sealed opaque envelope that specified group allocation, unclear if sequentially numbered
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Outcome assessors blinded to group allocation
Incomplete outcome data (attrition bias)   All outcomes	Low risk	One withdrawal in the upper limb group due to hip fracture from fall post discharge
Selective reporting (reporting bias)	Unclear risk	No study protocol

Methods	See Blennerhassett 2004
Participants
Interventions
Outcomes
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	See Blennerhassett 2004
Allocation concealment (selection bias)	Unclear risk	See Blennerhassett 2004
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	See Blennerhassett 2004
Incomplete outcome data (attrition bias)   All outcomes	Low risk	See Blennerhassett 2004
Selective reporting (reporting bias)	Unclear risk	See Blennerhassett 2004

Methods	Single‐centre pilot RCT
Participants	France 20 participants: 10 RTT, 10 control group Participants recruited from a neurorehabilitation unit in 1998 Inclusion criteria: hemiplegia caused by a single stroke occurring at least 1 month previously, static imbalance of the trunk resulting from the stroke Exclusion criteria: multiple cerebral lesions, disorders of the locomotor system, a severe visual or auditory deficit, a severe deficit of executive functions, or deterioration in the general state of health that might alter postural performances Mean age: RTT 63.5 years (SD 17), control 67.7 years (SD 15) 55% male Stroke details: first stroke, 35% ischaemic, 25% right hemiparesis Time since stroke: RTT 36.8 days (SD 25), control 27.7 days (SD 15) Pre‐intervention functional ability level: lack of postural balance
Interventions	RTT intervention: postural training using the Bon Saint Côme device ‐ a custom‐moulded orthosis that holds a pointing device, used by the participant to point to targets on a vertical panel which are activated to emit light and sound signals Sessions were in addition to 1 hour of usual care and were 60 minutes (unclear whether 5 or 7 days per week), for 4 weeks = 20 to 28 hours Sessions were delivered individually by a physical therapist Comparison group: 2 hours of usual care (Bobath inspired approach and functional therapy plus a session of occupational therapy 5 days per week)
Outcomes	Outcomes were recorded at baseline, 4 weeks (post treatment), and 2 months Balance/sit‐to‐stand outcome measures: Sitting Equilibrium Index, Upright Equilibrium Index Lower limb functional outcome measures: Functional Ambulation Classification Impairment outcome measures: Trunk Control Test, Motricity Index, Ashworth Scale ADL outcome measures: Functional Independence Measure
Notes	Postural deficit and unilateral neglect tended to be more severe in the device group at baseline, although not significant No intervention‐related reasons for withdrawal Attendance: all participants completed training
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	The participants were distributed consecutively into 2 groups of 10 each by using a randomisation table
Allocation concealment (selection bias)	Unclear risk	No details reported
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	The clinician who evaluated the patients did not know to which group they belonged
Incomplete outcome data (attrition bias)   All outcomes	Low risk	No withdrawals or dropouts
Selective reporting (reporting bias)	Unclear risk	No study protocol

Methods	RCT
Participants	Australia 20 participants: 10 RTT, 10 control group Participants were recruited from stroke clubs around Sydney. Date of recruitment not reported Inclusion criteria: diagnosis of stroke resulting in hemiplegia at least 12 months previous, discharged from all rehabilitation services, ability to understand instructions and give informed consent, no orthopaedic problem that would interfere with seated reaching, ability to sit unsupported for 20 minutes Exclusion criteria: none stated Mean age: RTT 68.2 years (SD 8.2), control 66.9 years (SD 8.2) 70% male Stroke details: not stated whether first or recurrent stroke; 40% right‐sided stroke Time since stroke: RTT 6.7 years (SD 5.8), control 5.9 years (SD 2.9) Pre‐intervention functional ability level: walking speed: RTT 0.41 m/s (SD 0.25), control 0.52 m/s (SD 0.28)
Interventions	RTT intervention: training designed to improve sitting balance and involving emphasis on appropriate loading of the affected leg while practicing reaching tasks using the unaffected hand to grasp objects located beyond arm's length Intervention was after discharge from all rehabilitation programmes Sessions were 30 minutes, 5 days per week for 2 weeks = 5 hours Sessions were delivered by a physical therapist in the participant's own home. Comparison group: upper extremity attention control ‐ performance of cognitive manipulative tasks while seated at a table
Outcomes	Outcomes were recorded at baseline and at 2 weeks (post treatment) Lower limb functional outcome measures: 10 Metre Walk Speed Balance/sit‐to‐stand outcome measures: reaching distance, reaching speed
Notes	No significant differences reported at baseline  5% loss to follow‐up at end of treatment phase  No intervention‐related reasons for withdrawal
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Not reported
Allocation concealment (selection bias)	Unclear risk	Participants allocated by drawing a card from a box of 10 experimental and 10 control cards
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Walking speed and cognitive‐manipulative tasks were evaluated by an assessor blinded to the participant's group allocation Biomechanical data collection and analysis for the seated reaching tasks and sit‐to‐stand were computerised, which minimised experimenter bias because group allocation was not evident to the operator
Incomplete outcome data (attrition bias)   All outcomes	Low risk	1 withdrawal in control group due to medical complications
Selective reporting (reporting bias)	Unclear risk	No study protocol

Methods	Pilot RCT
Participants	Canada 12 participants: 6 RTT, 6 control group Participants were recruited from a rehabilitation research group database. Date of recruitment not reported Inclusion criteria: first stroke, at least 3 months post stroke, discharged from all rehabilitation services, able to attend a rehabilitation centre 3 times per week for 4 weeks, able to walk 10 metres Exclusion criteria: any medical condition that would prevent participation Mean age: RTT 66.2 years (SD 7.7), control 62.3 years (SD 6.6) 58% male Stroke details: first stroke, 58% right hemiparesis Timing post stroke: RTT 2.3 years (SD 0.7), control 1.3 years (SD 0.9) Pre‐intervention functional ability level: Walking velocity: RTT 76 cm/s (SD 44), control 76 cm/s (SD 39)
Interventions	RTT intervention: lower limb circuit training of 10 workstations including sitting reach, sit‐to‐stand, stepping, heel lifts, standing balance, leg strengthening, treadmill walking, obstacle walking, slope and stair walking, plus participation in walking races and relays Intervention was after discharge from all rehabilitation programmes Sessions were 60 minutes, 3 times per week for 4 weeks = 12 hours Sessions were delivered to a group of 6 participants by 2 physical therapists, in an rehabilitation centre setting Comparison group: circuit programme designed to improve function of the affected upper limb
Outcomes	Outcomes were recorded at baseline, 4 weeks (post treatment), and 2 months after completion of training Lower limb functional outcome measures: Six Minute Walk Test, 10 Metre Walk Speed (with and without assistive device), Step Test Balance/sit‐to‐stand outcome measures: Timed Up and Go Test
Notes	No significant difference in walking velocity at baseline for total group, but after withdrawals, measures of walking speed and distance favoured the control group 25% loss to follow‐up at end of treatment phase Two participants withdrew before training (one due to transport costs) Nine participants attended at least 9 out of 12 sessions
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Inadequately reported
Allocation concealment (selection bias)	High risk	Drawing cards from a box
Blinding of outcome assessment (detection bias)   All outcomes	High risk	Independent rater blinded to participant allocation for the clinical assessments but may have been unmasked as a result of the observer inadvertently viewing 1 training session
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Missing outcome data balanced in numbers across groups
Selective reporting (reporting bias)	Unclear risk	No study protocol

Methods	Randomised placebo‐controlled trial
Participants	Australia 12 participants: 6 RTT, 6 control group Participants were recruited from a hospital rehabilitation facility between January and June 2000. Inclusion criteria: a diagnosis of first stroke resulting in hemiplegia within the previous three months; no orthopaedic problems which would interfere with the ability to perform seated reaching tasks; no visual problems which would interfere with reaching to pick up objects or reading; a score of at least 3 on Item 3 (sitting balance) of the MAS for Stroke; the ability to reach with intact arm a distance equivalent to 140% of arm's length; no major cognitive or perceptual problems identified using the short portable mental status  questionnaire; no left neglect identified using the Letter Cancellation Test; the ability to give informed consent; and the ability to understand instructions.  Exclusion criteria: none stated Mean age: RTT 60 years (SD 7), control 74 years (SD 12) % male not reported. Ratio of males to females: RTT 5:1, control 4:2 Stroke details: side of hemiplegia RTT 3:3, control 1:5 Timing post stroke: RTT 21 days (SD 8), control 37 days (SD 23) Pre‐intervention functional activity level: not reported
Interventions	RTT intervention: sitting training protocol designed to improve sitting by reaching beyond arm's length using the unaffected hand whilst focusing on: smooth co‐ordinated motion of the trunk and arm to get the hand to the object; appropriate loading of the affected foot; and preventing the use of maladaptive strategies such as widening the base of support. While reaching beyond arm's length, reach distance, direction, thigh support, seat height, and task were varied systematically. Training was progressed over the 2‐week period by increasing the reach distance and the number of repetitions Sessions were delivered to individuals by the first or second author or undergraduate physiotherapy students Comparison group: sham training protocol; participants completed a series of 11 cognitive‐manipulative tasks Participants were seated at a table, well supported in a chair with back and armrests, with their forearms resting on the table. The workspace was confined so that reach distance was less than 50% of arm's length which minimised perturbations to balance. Training was progressed over the 2‐week period by increasing the number of repetitions and cognitive difficulty of the cognitive‐manipulative tasks Both training programmes were 10 sessions of 30 minutes spread over a two week period = 5 hours
Outcomes	Outcomes were recorded at baseline, 2 weeks (post treatment) and 28 weeks Upper limb functional outcome measures: Functional Reach Test (primary outcome); standardised "reach to grasp and drink a glass of water" task; average reach movement time Lower limb functional outcome measures: 10 metre Walk Test
Notes	Potential baseline imbalance in time from stroke to admission to trial: RTT mean 21 days (SD 8), control mean 37 days (SD 23)
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Sequence generated by a person independent of the study using random number tables, blocked to ensure equal numbers of experimental and control participants
Allocation concealment (selection bias)	Unclear risk	Randomisation was concealed from the recruiter and assessor by using sealed opaque envelopes containing the allocation; not clear if envelopes sequentially numbered
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	"The third author remained blinded to group allocation and collected the outcomes measures post training and six months later. The collection of some outcome measures required two persons, one of whom was not blinded. To reduce bias, the blinded assessor (third author) gave all instructions and measured outcomes which were not collected by the computer."
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Missing outcome data balanced across groups 1 participant in the RTT group and 2 participants in the control group lost to follow‐up at 6 months
Selective reporting (reporting bias)	Unclear risk	No protocol available

Methods	RCT
Participants	Ghana 20 participants: 10 RTT, 10 control Participants were recruited from stroke survivors referred for physiotherapy. Date of recruitment not reported Inclusion criteria: first‐episode single stroke, stroke duration of < 3 months, ability to walk 10 metres independently with or without walking aid and Functional Ambulatory Category (FAC) score of 3 or more Exclusion criteria: participants with aphasia, cardiac arrhythmias or any other conditions making exercises contraindicated Mean age: RTT 57.6 ± 0.3 years, control 55.8 ± 6.7 64% male Stroke details: ischaemic stroke 66.6% (6), haemorrhagic stroke 33.3% (3) Time since stroke: RTT 2.2 months (SD 0.8), control 2.4 months (SD 0.9) Pre‐intervention functional ability: 6 Minute Walk Test: RTT 249.5 metres (SD 10.7), control 253.0 (SD 12.5)
Interventions	RTT intervention: circuit training for 105 minutes, 3 times per week for 8 weeks including treadmill walking, push‐ups, squatting, straight leg raise, stairs walking and cycling exercises Comparison group: conventional therapy of passive and active exercises. Participants also performed upper limb strengthening exercises, walking re‐education, as well as standing and balance retraining carried out between parallel bars
Outcomes	Outcomes were recorded at baseline, week 4 and week 8 (post intervention) Lower limb functional outcome measures: 6 Minute Walk Test, 10 Metre Walk Test, Functional Ambulatory Category
Notes	No apparent baseline imbalance
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	"Subjects were randomized into two groups"
Allocation concealment (selection bias)	Unclear risk	Method of concealment not reported
Blinding of outcome assessment (detection bias)   All outcomes	Unclear risk	No mention of blinding
Incomplete outcome data (attrition bias)   All outcomes	Unclear risk	Number of participants not reported in data tables
Selective reporting (reporting bias)	Unclear risk	No protocol available

Methods	RCT
Participants	Jamaica 128 participants: 64 RTT, 64 control Participants were recruited from 3 hospitals. Date of recruitment not reported Inclusion criteria: 40 years of age or older, community dwelling, 6 to 24 months after stroke, able to walk with or without assistive devices, not currently in a rehabilitation or regular exercise programme, not having any disorder that would compromise exercise training, such as unstable cardiovascular diseases, no cognitive deficits Exclusion criteria: none stated Mean age: RTT 63.4 years (SD 9.4), control 64.9 years (SD 11.1) 45.3% male Stroke details: ischaemic 71.1% (91), haemorrhagic 11.7% (15) Time since stroke: RTT 12.8 months (SD 3.6), control 11.8 months (SD 3.6) Pre‐intervention functional ability level: use of walking aid at recruitment: RTT 26.6% (n = 17), control 32.8% (n = 21)
Interventions	RTT intervention: participants were supervised by trained instructors to walk briskly along a prescribed course for 15 minutes, 3 times per week, for 12 weeks initially, progressing by 5 minutes per week up to 30 minutes in their home or community = 9 to 18 hours Comparison group: light massage to the affected limbs for 25 minutes, 3 times per week for 12 weeks at home
Outcomes	Outcomes were recorded at baseline, 6 weeks and 12 weeks (post treatment) Lower limb functional outcome measures: 6 Minute Walk Test Impairment outcome measures: Motricity Index QoL/health status outcome measures: Physical and Mental Component Summary scores of the Medical Outcomes Survey 36‐Item Short‐Form Health Survey (SF‐36) ADL outcome measures: Barthel Index, instrumental ADL dimension of the Older Americans Resources and Services Questionnaire
Notes	No significant differences between groups at baseline 2 intervention‐related withdrawals (programme too difficult (n = 1) and participant not happy with group assignment (n = 1)) No major adverse events during or immediately after the sessions
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Block randomisation used but not clear how the sequence was generated
Allocation concealment (selection bias)	Unclear risk	Not reported
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Assessment by a physical therapist blinded to group assignment
Incomplete outcome data (attrition bias)   All outcomes	Unclear risk	Similar number of dropouts from intervention and control groups (7 and 5 respectively) Time of dropout not reported Reasons per group not reported
Selective reporting (reporting bias)	Unclear risk	No study protocol

PERMALINK

Repetitive task training for improving functional ability after stroke

Beverley French

Lois H Thomas

Jacqueline Coupe

Naoimh E McMahon

Louise Connell

Joanna Harrison

Christopher J Sutton

Svetlana Tishkovskaya

Caroline L Watkins

Abstract

Background

Objectives

Search methods

Selection criteria

Data collection and analysis

Main results

Authors' conclusions

Plain language summary

Summary of findings

for the main comparison.

Background

Description of the condition

Description of the intervention

How the intervention might work

Why it is important to do this review

Objectives

Methods

Criteria for considering studies for this review

Types of studies

Types of participants

Types of interventions

Types of outcome measures

Primary outcomes

Secondary outcomes

Timing of outcome assessment

Search methods for identification of studies

Electronic searches

Searching other resources

Data collection and analysis

Selection of studies

Data extraction and management

Assessment of risk of bias in included studies

Measures of treatment effect

Unit of analysis issues

Studies with multiple treatment groups

Dealing with missing data

Assessment of heterogeneity

Assessment of reporting biases

Data synthesis

Subgroup analysis and investigation of heterogeneity

1. Criteria for subgroup and sensitivity analyses.

Sensitivity analysis

Results

Description of studies

Results of the search

1.

Included studies

Design

Sample size

Country

Participants

Mean time since stroke

Interventions

Setting

Amount of task practice

Duration of training

Intervention delivery

Comparison interventions

Outcomes

2. Outcome measures used from the included trials.

Primary outcomes

Upper limb functional outcome measures

Lower limb functional outcome measures

Secondary outcomes

ADL measures

Quality of life/health status measures

Adverse events

Outcomes used at follow‐up

Methods	Single‐blind RCT
Participants	Sweden 34 participants: 15 RTT, 19 control Participants were recruited from Umeå Stroke Unit. There were 3‐monthly recruitment periods between February 2005 and June 2007 Inclusion criteria: first‐ever or recurrent ischaemic or haemorrhagic stroke 3 to 6 months before enrolment and randomisation, age ≥ 55, the ability to walk 10 metres with or without a walking device, the ability to understand and comply with instructions in Swedish, risk of fall at the time of enrolment according to subjective clinical observations in the assessment situation performed by the experienced physiotherapists in the study Exclusion criteria: the ability to walk outdoors independently, i.e. without personal assistance or walking device, severe aphasia or severe vision or hearing impairment, a medical condition that a physician determined was inconsistent with study participation, e.g. cancer or severe congestive heart failure with expected short remaining life expectancy, recurrent stroke within 3 months before study start, living more than 100 km away from the training facilities Mean age: RTT 77.7 years (SD 7.6), control 79.2 years (SD 7.5) 62% male Stroke details: first or recurrent stroke, 97% ischaemic, 3% haemorrhagic Time since stroke: RTT 139.7 days (SD 37.3), control 126.8 days (SD 28.2) Pre‐intervention functional activity level: Barthel Index: RTT 44.3 (CI 40.0 to 48.7), control 44.2 (CI 39.3 to 49.2)
Interventions	RTT intervention: the intervention was based on the HIFE (High Intensity Functional Exercise) program, to improve the participants lower‐limb strength, balance and gait ability The program includes lower‐limb strength (e.g. chair stand) and balance exercises (e.g. weight shifting outside support surface), standing (e.g. knee bend) and walking (e.g. obstacle crossing course) A home visit was conducted by a physiotherapist and an occupational therapist to determine each participant's ability to perform ADLs and lifestyle activities and to experience the participants daily difficulties in their own environment Sessions were 45 minutes, 6 times per week (twice daily) for 5 weeks = 22.5 hours Comparison group: participants met once per week for a 1 hour of educational session during the 5‐week period The session was led by an occupational therapist, group discussions were about communication difficulties, fatigue, depressive symptoms, mood swings, personality changes and dysphagia, all more or less hidden dysfunctions after stroke and how to cope with these difficulties There was no special focus on the risks of falling in these discussions
Outcomes	Outcomes were recorded at baseline, 5 weeks (post treatment), 3 months and 6 months Balance/sit‐to‐stand outcome measures: Berg Balance Scale, Falls Efficacy Scale, number of falls ADL outcome measures: Barthel Index, Frenchay Activities Index
Notes	No significant differences between groups at baseline 11 participants in total fell during study (32%), RTT (n = 5), control (n = 6)
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Minimisation software program
Allocation concealment (selection bias)	Low risk	The randomisation procedure was conducted by the 2 principal investigators who were involved neither in the assessments, nor in the RTT or control group Both investigators were blinded to allocation at the time of randomisation, which was made possible by using code numbers for each participant
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	All assessments were done by blinded staff, who were instructed that if they had any reason to believe that they had revealed a participant's group they should make an adverse event report. The staff in the intervention did not take part in any of the assessments
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Small number of dropouts, reasons provided
Selective reporting (reporting bias)	Unclear risk	No study protocol

Methods	RCT
Participants	Korea 20 participants: 10 RTT, 10 control Inpatient recruitment, date of recruitment not reported Inclusion criteria: ability to walk 10 metres independently using an aid or orthotic with or without supervision or aid and a minimum score of 20 in the Korean Mini‐Mental State Examination Exclusion criteria: joint contraction, pain or fracture of the musculoskeletal system, hemianopsia Mean age: RTT 52.50 years (SD 11.72), control 53.40 years (SD 12.11) % male: not reported Stroke details: not reported Time since stroke: RTT 7.70 years (SD 6.11), control 13.10 years (SD 10.62) Pre‐intervention functional activity level: Timed Up and Go Test: RTT 29.84 seconds (SD 13.32), control 39.10 seconds (SD 14.97)
Interventions	RTT intervention: the training consists of 10 walking‐related tasks designed to strengthen the lower extremities, and enhance the walking balance, speed and distance in a progressive manner The 10 tasks were: step‐ups; balance beam; kicking a ball; stand up and walk; obstacle course; treadmill; walk and carry; speed walk; walk backwards; and stairs. Before commencing training, the participants warmed up for 5 minutes to improve their range of motion and flexibility. Each item was practiced for 5 minutes, and 1 minute of rest time was allowed between each item Sessions were in addition to conservative physical therapy and were 1 hour, 3 times per week for 4 weeks = 12 hours Comparison group: conservative physical therapy for 1 hour per day, 5 days per week for 4 weeks Conservative physical therapy consisted of joint mobilisation, muscle strengthening, and balance training
Outcomes	Outcomes were recorded at baseline and 4 weeks (post treatment) Lower limb functional outcome measures: 10 Metre Walk Speed, Balance/sit‐to‐stand outcome measures: Trunk Impairment Scale, Berg Balance Scale, Timed Up & Go Test
Notes	Equivalence not reported, but baseline values for Time Since Stroke and the Timed up and Go Test appear different across groups
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Patients 'randomly allocated' but no further information provided
Allocation concealment (selection bias)	Unclear risk	Not reported
Blinding of outcome assessment (detection bias)   All outcomes	Unclear risk	Not reported
Incomplete outcome data (attrition bias)   All outcomes	Low risk	No dropouts
Selective reporting (reporting bias)	Unclear risk	No study protocol

Methods	RCT
Participants	Republic of Korea 26 participants: 13 RTT; 13 control Participants were recruited from inpatients in a rehabilitation hospital. Date of recruitment not reported Inclusion criteria: hemiparesis from a single stroke occurring at least six months before; sufficient cognition to follow simple instructions and understand the purpose of the study (Korean version of the Mini‐Mental State Examination score of ≥ 24 points); gait speed < 0.8 m/s; ability to walk 10 metres independently without an assistive device; absence of a musculoskeletal condition that could potentially affect the ability to walk safely; and absence of hemispatial neglect Exclusion criteria: participation in other studies or rehabilitation programs; or severe heart disease or uncontrolled hypertension and pain Mean age: 50.45 years 50% male Stroke details: 59% right hemiparesis (no baseline data reported for 4 participants who dropped out) Time since stroke: 231.64 days Pre‐intervention functional ability: 10 metre Walk Test (m/s) RTT mean 0.51 ± 0.16, control mean 0.48 ± 0.18
Interventions	RTT intervention: a Community Walking Training Programme comprising various community environments, including walking near the hospital setting, walking outside of the hospital setting on uneven ground, walking outside of the hospital setting on uneven ground with obstacles, and visiting a shopping centre Comparison group: All participants took part in the same standard rehabilitation programme consisting of conventional physical and occupational therapy. Conventional physical therapy, including increased trunk stability, lower‐extremity muscle strength, and gait, was performed for 30 minutes per day, 5 times a week, for 4 weeks. Occupational therapy, consisting of an upper‐extremity training program for ADL, was performed for 30 minutes per day, 5 times a week, for 4 weeks
Outcomes	Outcomes were recorded at baseline and 4 weeks (post treatment) Lower limb functional outcome measures: 10 Metre Walk Test, 6 Minute Walk Test and Community Walk Assessment QoL: Stroke Impact Scale social participation domain
Notes	No apparent baseline imbalance
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	"patients were randomly assigned"
Allocation concealment (selection bias)	High risk	"sealed envelopes were prepared in advance and marked on the inside with an O or X."
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Blinded outcome assessment.
Incomplete outcome data (attrition bias)   All outcomes	Unclear risk	"Two subjects each in the CWTP and control groups dropped out due to health conditions, personal reasons, or discharge." Reasons not given by intervention group
Selective reporting (reporting bias)	Unclear risk	No study protocol

Methods	Randomised controlled single‐blind study
Participants	Korea 20 participants: 10 RTT; 10 control Inpatient recruitment between August 2012 and October 2013 Inclusion criteria: a clinical diagnosis of a first stroke confirmed by neuroimaging (CT or MRI); a hemiparesis; a time interval between stroke and recruitment of 3 months or less; the ability to comprehend the instructions for the testing procedures; and mild to moderate walking deficit, as indicated by Functional Ambulation Category (FAC) between 3 and 4 Exclusion criteria: severe cognitive impairment (K‐MMSE ≤ 10) or aphasia; previous stroke history; not independent 'sit‐to‐stand' activity (Berg Balance Scale score < 18); acute systemic illness or infection; a significant orthopaedic condition or pain that limited participation in exercise; and visual impairment or vestibular system deficit that caused balance impairment Mean age: 65.6 ± 9.2 years 65% male Stroke details: first stroke; ischaemic 80% (16), haemorrhage 20% (4) Time since stroke: RTT 30.1 days (SD 21.8), control 29.9 days (SD 20.3) Pre‐intervention functional activity level: 6 Minute Walk Test RTT 167.5 metres (SD 121.8), control 157.5 metres (SD 64.0)
Interventions	RTT intervention: participants participated in 90‐minute circuit‐training classes, 5 times per week for 4 weeks. Circuit training consisted of a 5‐minute warm‐up period, five classes of 15 minutes duration interspersed with a 1 minute rest and a 5‐minute cool‐down period. There were 5 categories of complex exercises including trunk exercise and active sitting practice, sit‐to‐stand practice, standing and walking practice, aerobic exercise training and strengthening training Comparison group: participants in the control group received conventional individual physiotherapy for 30 minutes twice a day (total 60 minutes), 5 days a week for 4 weeks
Outcomes	Outcomes were recorded at baseline and 4 weeks (post treatment) Lower limb functional outcome measures: Fugl‐Meyer lower limb score, Berg Balance Scale, 6 Minute Walk Test ADL: Korean version of the Modified Barthel Index
Notes	No apparent baseline imbalance
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	"Participants were randomly allocated"
Allocation concealment (selection bias)	Unclear risk	A sealed envelope technique was used; unclear if envelopes were opaque and sequentially numbered
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Blinded outcome assessment
Incomplete outcome data (attrition bias)   All outcomes	Low risk	All participants completed the study
Selective reporting (reporting bias)	Unclear risk	No protocol available

Methods	Multicentre RCT
Participants	The Netherlands 101 participants: 31 leg training group, 33 arm training group, 37 control Participants recruited from 7 hospitals in the Netherlands between1994 and1997 Inclusion criteria: primary first‐ever stroke in the territory of the middle cerebral artery, confirmed by CT or MRI, aged 30 to 80 years, impaired motor function of the arm and leg, inability to walk at first assessment Exclusion criteria: complicating medical history or severe deficits in communication, memory or understanding Mean age: leg training group 64.5 years (SD 9.7), arm training group 69 years (SD 9.8), control group 64.1 years (SD 15) 43% male Stroke details: first‐ever stroke,41% right hemiparesis Timing post stroke: leg training group 7.0 days (SD 2.5), arm training group 7.2 days (SD 2.8), control group 7.5 days (SD 2.9) Pre‐intervention functional ability level: Barthel Index of 9 or lower
Interventions	Leg training group: sitting, standing and weight‐bearing exercise, with an emphasis on achieving stability and improving gait velocity Treadmill training was used if available If treatment at disability level was not possible, strengthening exercises were used Arm training group: functional exercise to facilitate forced arm and hand activity such as leaning, punching a ball, grasping, reaching, dressing, hair‐combing and moving objects If treatment at disability level not possible, strengthening exercises were used Intervention was in addition to basic rehabilitation, which consisted of 15 minutes arm rehabilitation, 15 minutes leg rehabilitation and 1.5 hours per week of ADL training by an occupational therapist Sessions were delivered individually by a physiotherapist and were 30 minutes, 5 days per week for 20 weeks = 50 hours Comparison group: immobilisation of the paretic arm and leg by means of an inflatable pressure splint Kwakkel 1999a: arm training versus splint control Kwakkel 1999b: leg training versus splint control
Outcomes	Outcomes were recorded at baseline, and weekly between weeks 1 to 10, and every 2 weeks between week 11 to 26 Final measurements were at 26 weeks Results are presented for baseline, weeks 6, 12, 20 and 26 Lower limb functional outcome measures: Functional Ambulation Classification, walking speed (comfortable and maximum) Upper limb functional outcome measures: Action Research Arm test ADL outcome measures: Barthel Index QoL/health status outcome measures: Nottingham Health Profile
Notes	No significant differences reported at baseline 12% lost to follow‐up at end of treatment phase No likely intervention‐related reasons for withdrawal, although 2 participants refused the splint control treatment Compliance with delivery of intended amounts of training was monitored, and achieved
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Restricted randomisation (permuted blocks of nine) was applied, using random number tables for each of 3 participating hospitals
Allocation concealment (selection bias)	Unclear risk	Allocation was concealed by use of sealed envelopes
Blinding of outcome assessment (detection bias)   All outcomes	High risk	Assessors were blind to group allocation, Treatment assignment was unintentionally disclosed for 10 participants (1 leg training, 4 arm training, 5 control group)
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Number of dropouts balanced across groups, reasons do not appear to be related to the intervention
Selective reporting (reporting bias)	Unclear risk	No study protocol

Methods	See Kwakkel 1999
Participants
Interventions
Outcomes
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	See Kwakkel 1999
Allocation concealment (selection bias)	Unclear risk	See Kwakkel 1999
Blinding of outcome assessment (detection bias)   All outcomes	High risk	See Kwakkel 1999
Incomplete outcome data (attrition bias)   All outcomes	Low risk	See Kwakkel 1999
Selective reporting (reporting bias)	Unclear risk	See Kwakkel 1999

Methods	Stratified, single‐centre RCT
Participants	Norway 61 participants: 33 RTT, 28 control Participants were recruited from a hospital in Norway between 1996 and1997 Inclusion criteria: first‐ever stroke with hemiparesis verified clinically and by CT Exclusion criteria: more than1 stroke incident, subarachnoid bleeding, tumours of the brain, other severe medical conditions in combination with stroke, 5 or more points on each of the scores on the MAS Mean age: 78 years (SD 9), range 49 to 75 years 59% male Stroke details: first stroke, 56% right hemiparesis Time since stroke: baseline measures taken within 3 days of admission Pre‐intervention functional ability level: Barthel Index: RTT 56 (SD 28), control 46 (SD 36)
Interventions	RTT intervention: Motor Relearning Programme as per Carr 1987 Functional task training in ordinary settings, with ordinary tasks, using the principles of maximal repetition, task and setting variation RTT intervention was instead of usual care Sessions were delivered by hospital and outpatient physiotherapists and were 40 minutes minimum per session, 5 days per week for as long as hospitalised, and continuing into the community, although receipt of physiotherapy in community settings was variable After discharge, some participants received therapy in their own homes, at rehabilitation centres, or private outpatient departments, dependent on need Comparison group: Bobath Programme
Outcomes	Outcomes were recorded at baseline, 2 weeks, 3 months, 1 year and 4 years post stroke Lower limb functional outcome measures: MAS, Sødring Motor Evaluation Scale ‐ subscale for trunk/balance/gait Balance/sit‐to‐stand outcome measures: Berg Balance Scale (1 year only) Impairment outcome measures: Sødring Motor Evaluation Scale ‐ subscales for leg function, arm function ADL outcome measures: Barthel Index QoL/health status outcome measures: Nottingham Health Profile
Notes	Baseline differences: control group slightly more dependent at entry, but no significant difference in MAS, Sødring Motor Evaluation Scale or Barthel Index
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Inadequately reported
Allocation concealment (selection bias)	Unclear risk	Not reported
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	The study was double‐blind, and the code was sealed until the last test was performed at 3 months follow‐up
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Number of dropouts balanced across groups and reasons provided
Selective reporting (reporting bias)	Unclear risk	No study protocol

Methods	RCT
Participants	Ireland 61 participants, 31 RTT, 30 control Participants were recruited from an inpatient stroke unit between October 2004 and July 2007 Inclusion criteria: stroke admissions within 21 days post stroke Exclusion criteria: none stated Mean age: RTT 71.55 (SD 13.31), control 72.13 (SD 9.93) 58.4% male Stroke details: first or recurrent stroke, 55.7% right hemisphere Timing post stroke: RTT 10.81 days (SD 5.22), control 10.73 (SD 5.09) Pre‐intervention functional ability level: Barthel Index: RTT 10.10 (SD 3.75), control 9.93 (SD 3.66); MAS: RTT 18.32 (SD 11.59), control 19.97 (SD 10.98)
Interventions	RTT intervention: 20 minutes of conventional therapy based on Bobath principles and 20 minutes of gait‐specific training administered by a research therapist 5 times per week for 4 weeks = 6.6 hours of RTT Comparison group: 40 minutes of conventional therapy based on Bobath principles administered by the stroke unit therapists
Outcomes	Outcomes were recorded at baseline, 4 weeks (post treatment) and at 3 and 6 months Lower limb functional outcome measures: MAS, walking speed, Modified RMI, Step Test ADL outcome measures: Barthel Index QoL/health status outcome measures: London Handicap Score
Notes	Baseline characteristics were similar in terms of age, gender, side of hemiplegia, time since stroke onset, stroke severity and walking speed There was insufficient contrast in treatment between the groups (i.e. therapists in the Bobath group practiced early ambulation more frequently than therapists in the RTT group) Within the RTT group, there were three times the number of Total Anterior Circulation Infarct strokes (a poor prognostic indicator for recovery of independent mobility), more participants with a previous stroke and more participants requiring the assistance of 2 people to walk
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Inadequately reported
Allocation concealment (selection bias)	Unclear risk	Stratified by age and walking ability, in blocks of 4 using sealed envelopes
Blinding of outcome assessment (detection bias)   All outcomes	High risk	3 physiotherapists assisted with both intervention and outcome assessment following resignation of 1 research associate
Incomplete outcome data (attrition bias)   All outcomes	Low risk	No dropouts
Selective reporting (reporting bias)	High risk	Findings for the London Handicap Score (measured at 6 months) not reported

Methods	RCT
Participants	Australia 26 participants: 15 RTT, 11 control Participants were recruited on discharge from physiotherapy services in 6 hospitals in 1 region. Date of recruitment not reported Inclusion criteria: stroke within the past 18 months, 45 years and older, living in the community, score > 0 and < 6 on MAS, score < 6 on Item 7 or 8 of the MAS Exclusion criteria: unable to consent, uncontrolled cardiac symptoms or other medical conditions that limited exercise, or with a pacemaker Mean age: RTT 69 years (SD 13), control 72 years (SD 9) 50% male Stroke details: unclear whether first or recurrent stroke, 50% right hemiparesis Timing post stroke: RTT median 6.5 months (IQR 5.5), control median 4.5 months (IQR 3) Pre‐intervention functional ability level: all participants could walk, but with difficulty
Interventions	RTT intervention: home‐based exercise programme aimed at improving mobility in standing balance and walking, based on a list of 23 activities arranged hierarchically on their challenge to balance The home programme used video self‐modelling prepared on the baseline visit to the clinic to prescribe the exercise programme, telephone monitoring to encourage compliance, and 2 clinic visits for programme review Sessions were prescribed 60 minutes per day over 6 weeks = 42 hours Participants were required to keep a record of practice Comparison group: home‐based exercise programme of same duration based on improving upper limb function, starting from basic movement through to functional activity, using the same self‐instructional video, self‐ and telephone‐monitoring and clinic visits as the experimental group
Outcomes	Outcomes were recorded at baseline, 6 weeks (post treatment) and 14 weeks Lower limb functional outcome measures: MAS ‐ walking Balance/sit‐to‐stand outcome measures: Functional Reach Test
Notes	No baseline comparisons reported 19% lost to follow‐up by end of treatment phase No likely intervention‐related reasons for withdrawal Participants reported 75% compliance with prescribed exercises
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Inadequately described
Allocation concealment (selection bias)	Unclear risk	Randomisation by numbered, sealed, opaque envelopes
Blinding of outcome assessment (detection bias)   All outcomes	Low risk	Outcome measures were collected by a measurer blinded to group allocation
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Number of dropouts balanced across groups and reasons provided
Selective reporting (reporting bias)	Unclear risk	No study protocol

Methods	Prospective randomised controlled study
Participants	Africa 40 participants: 20 RTT, 20 control Participants were recruited from people referred for outpatient management at the physiotherapy department of a tertiary hospital. Date of recruitment not reported Inclusion criteria: all participants were people whose stroke occurred not less than 3 months, and not more than 24 months, before entering the study. Participants were included if they were able to walk 10 metres independently with or without a walking aid Exclusion criteria: none specifically reported but ability to walk < 10 metres excluded Mean age: RTT 56.8 years (SD 8.3), control 57.2 years (SD 5.9) 55% male Stroke details: 52.5% right hemisphere Timing post stroke: RTT 10.7 months (SD 6.8), control 10.3 (SD 5.9) Pre‐intervention functional ability level: able to walk ≥ 10 metres
Interventions	RTT intervention: on each day of treatment/training, participants observed a pre‐exercise rest period of 10 minutes during which heart rate and blood pressure measurements were made. The RTT group went through a 1‐hour session of conventional physiotherapy including 25 minutes of overground walking exercise training. The overground walking exercise training involved walking over ground at a natural safe speed (i.e. walking at own pace in order to cover as much ground as possible within the training period) on a 15 x 10 metre walk course marked out on the flat floor of a remedial gymnasium. In each case, exercise would be terminated any time the participant reported symptoms of exertional intolerance, i.e. outside the target zone on Borg's rate of perceived exertion (RPE) scale. Participants took part in 3 x 25 minute sessions per week for a 12‐week period = 15 hours Comparison group: the conventional physiotherapy rehabilitation consisted of 1 hour of active and passive range of motion (ROM) exercises, strength training and balance training, as applicable
Outcomes	Outcomes were recorded at baseline, 4 weeks, 8 weeks and 12 weeks (post treatment) Lower limb functional outcome measures: 10 Metre Walk Speed, 6 Minute Walk Test
Notes	Equivalence of groups at baseline was not reported 5 participants were lost to follow‐up
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Inadequately reported
Allocation concealment (selection bias)	Unclear risk	Not reported
Blinding of outcome assessment (detection bias)   All outcomes	Unclear risk	Not reported
Incomplete outcome data (attrition bias)   All outcomes	Unclear risk	Withdrawals not explained
Selective reporting (reporting bias)	Unclear risk	No study protocol