. 2020 Jun 26;10(8):e01742. doi: 10.1002/brb3.1742

TABLE 1.

Characteristics of the included studies

Study (Country)	Design and assessment time points	Participants	Characteristics of the participants	Intervention	Instruments	Key finding(s)	Attrition rate
Barker 2017 (Australia)	3‐arm RCT in repeated measures design, at baseline, post‐treatment, 26 and 52 weeks post‐treatment	Consecutively admitted patients at one acute stroke unit (n = 50) were randomized into 3 groups using random permuted blocks	Mean age = 53.6 ± 15 Male = 68% Stroke intervals = 5.9 ± 3 weeks	Treatment groups SMART Arm ^b training with stimulation for elbow, forearm, wrist, and hand plus conventional therapy (n = 17) SMART Arm ^b training without stimulation plus conventional therapy (n = 16) Both groups received 1‐hr session on weekdays for 4 weeks Content: perform reaching task in a straight‐line path. Control group ( n = 17) 1‐hr conventional therapy, including physiotherapy, occupational therapy, and therapy assistant time, and involved a mix of one‐on‐one and group therapy sessions on weekdays for 4 weeks	MAS –item 6 ^a MRC MAS RAI MAS –item 6,7,8 SIS MAL kinematics and kinetics of reaching.	The RT group achieved better upper arm function when compared to the control group at post‐training (OR = 1.47, 95% CI = 1.23–1.71) and at 26 weeks (OR = 1.31, 95% CI = 1.05–1.57), respectively. There was no significant between‐group difference in arm function (p = .37). All groups showed significant improvements in arm function and quality of life (all ps < 0.001) over time. No significant improvements in muscle tone among all groups at any assessment time points.	Post‐treatment = 6%; 26 weeks post‐treatment = 22%; 52 weeks post‐treatment = 30%
Daunoraviciene 2018 (Lithuania)	2‐arm RCT in pre‐post design	Subacute stroke patients (n = 34) were randomly assigned into 2 groups (Sampling strategies not reported)	Mean age = 65.7 ± 4.48 Male = 65% Stroke intervals = 9.1 ± 5 weeks	Treatment group ( n = 17) RT with Armeo Spring ^c for elbow, forearm, wrist, and hand for 1‐hr session on weekdays for 2 weeks Content: perform a sequence of motor tasks in 5–7 exercise cycles Control group ( n = 17) Occupational therapy sessions, including exercising, physical activities, active table games etc.	FIM‐self‐care ^a FMA HAM‐D HAM‐A ACE‐R MAS Active ROM	The RT group showed significantly better functional independence when compared with the UT group at postintervention (p < .03). No significant between‐group difference in motor control at postintervention (p = .287)	No attrition was found at postintervention
Dehem 2019 (Belgium)	2‐arm RCT in repeated measures design at baseline, postintervention, and 6‐month poststroke	Subacute stroke patients (n = 45) with < 1 month poststroke from inpatient rehabilitation centers were randomized into two groups using computer‐generated sequence	Mean age = 67.9 ± 15.4 Male % = 46.7% Stroke intervals = 27.8 ± 5.5 days	Treatment group ( n = 23) Four 45‐min RT sessions for wrist and hand with REAplan robot ^d (25%) and twelve UT sessions (75%) per week for 9 weeks Content: exercises with game involving moving the paretic hand along a reference trajectory while passing through checkpoints Control group ( n = 22) Sixteen 45‐min UT session per week for 9 weeks Content: motor rehabilitation that matched with patients’ personal needs	FMA‐UE ^a BBT WMFT ABILHANDACTIVLIM SIS	RT showed significantly greater improvement in gross manual dexterity (p = .02), upper‐limb ability (p =.02) and patient social participation (p = .01) compared with control group at six months poststroke Both group show similar improvement in abilities to perform manual activities and activities of daily living	Postintervention = 28.9%, six months poststroke = 37.8%
Hesse 2014 (Germany)	2‐arm RCT in repeated measures design at baseline, post‐treatment and three‐month post‐treatment	Subacute patients from two inpatient stroke rehabilitation units (n = 50) were assigned into 2 groups by web‐based randomization tool	Mean age = 70.6 ± 16.1 Male = 56% Stroke intervals = 4.5 ± 2 weeks	Treatment group ( n = 25) 30‐min of RT in arm studio plus 30‐min individual arm therapy on weekdays for 4 weeks Content: 1) repetitive practice of finger, wrist, forearm, and shoulder movement; 2) task‐oriented motor relearning program and impairment‐oriented arm ability training Control group ( n = 25) 1‐hr individual arm therapy on weekdays for 4 weeks, consisting of the task‐oriented motor relearning program and the impairment‐oriented arm ability training (repetitions of movements and shaping)	FMA ^a ARAT BBT MRC MAS BI	No significant between‐group differences were found in motor control, upper extremity performance, muscle tone, and functional independence at all measurement points. Both groups showed significant improvements in motor control and upper extremity performance at post‐training and at three‐month follow‐up (all ps <0.001)	Post‐treatment = 2%, 3 months post‐treatment = 8%
Masiero 2014 (Italy)	2‐arm RCT in repeated measures design at baseline, post‐treatment, 3 months and 7 months post‐treatment	Subacute subjects from Stroke Unit (n = 34) were randomly allocated to 2 groups by a computer program	Mean age = 66.3 ± 8.55 Male = 66.7% Stroke intervals = 1.3 ± 0.4 week	Treatment group ( n = 16) ~75‐min of standard therapy plus ~ 45‐min of RT for elbow, foreman, wrist, and hand on weekday for 5 weeks Content: 1) NeReBot training ^e with exercises (flexion and extension, adduction and abduction, pronation and supination, circumduction); 2) Conventional functional rehabilitation including proprioceptive exercises, functional reduction, gait training, occupational therapy, passive, and active assisted mobilization of the hand, wrist, and upper paretic arm Control group ( n = 18) 2‐hr of daily rehabilitation treatment for 5 weeks including proprioceptive exercises, functional re‐education, gait training, occupational therapy, and passive and active assisted mobilization of the hand and wrist.	FMA ^a FIM‐motor ^a FAT ^a MRC MAS BBT tolerability and acceptability of treatment	No significant between‐group differences were found in motor function, muscle tone, and functional independence at all assessment time‐points.	At post‐treatment = 11.8%; 7 months post‐treatment = 17.6%
Orihulela Espina 2016 (Mexico)	2‐arm RCT in pre‐post design	Subacute stroke patients from Neurologic Rehabilitation Unit (n = 17) were randomly assigned into 2 groups by block randomization	Mean age = 55.6 ± 20.3 Male = 64.7% Stroke intervals = 1 week to 4 months	Treatment group ( n = 9) 1‐hr of RT for wrist and hand on weekdays for around 8 weeks Content: 1) passive activities; 2) partial assistance/ resistance activities; 3) active movement Control group ( n = 8) 1‐hr of classical occupational therapy on weekdays for around 8 weeks Content: massage and conventional occupational exercises, including passive movements, strengthening exercises and active grasps movement and personalized activities for fine pinching control	FMA‐hand ^a MI ^a	RT showed significantly greater improvement in hand motor function compared with UT at post‐treatment (p < .01). Both groups showed significant improvement in motor control over time (Nonparametric Cliff's delta‐within effect sizes: dwOT‐FMA = 0.5, dwRT‐FMA = 1)	No attrition was found at postintervention
Sale 2014 (Italy)	2‐arm RCT in repeated measures design at baseline, after 15th and 30th treatment sessions	Subacute stroke patients from the rehabilitation center (n = 53) were randomly allocated into 2 groups by dedicated software	Mean age = 67.7 ± 14.2 Male = 58.5% Stroke intervals = 4.3 ± 1week	Treatment group ( n = 26) 45‐min of RT with MIT‐MANUS ^h for shoulder and elbow plus 3‐hr physiotherapy on weekdays for 6 weeks Content: 1) dexterity and gait training, 2) goal‐directed, planar reaching tasks, including both unassisted and assisted repetitions Control group ( n = 27) 45‐min of conventional therapy plus 3‐hr of physiotherapy on weekdays for 6 weeks Content: 1) dexterity and gait training; 2) assisted stretching, shoulder and arm exercises and functional reaching tasks exercising, physical activities, and active table games	FMA ^a MAS‐S ^a MAS‐E ^a pROM MI	Both groups showed significant improvements in motor control after 15th and 30th session, with significant greater improvement found in the RT group after the first 15th sessions (p <.0001). Significant improvement in muscle tone for shoulder and elbow (p <.05) was only found in the RT group.	No attrition was found after 30th treatment sessions
Stinear 2014 (New Zealand)	2‐arm RCT in repeated measures design at baseline, 6, 12 and 26 weeks post‐treatment	Consecutive subacute stroke patients from a stroke unit (n = 57) were randomized by customized software (www.rando.la)	Mean age = 68 ± 25 Male = 45.6% Stroke intervals < 26 days	Treatment group ( n = 29) 15‐min Bilateral priming for wrists and hands plus 30‐min physiotherapy and occupational therapy on weekdays for 4 weeks Control group ( n = 28) 15‐min Intermittent cutaneous electric stimulation plus 30‐min physiotherapy and occupational therapy on weekdays for 4 weeks	ARAT ^a SIS MRS	At 12th weeks, greater proportion of the participants in the treatment group achieved their recovery plateaus in upper extremity performance than the control group (χ² = 4.25; p = .039). At 26th weeks, no between‐group difference was found in quality of life⁵	At post‐treatment = 7.0% At 12 weeks post‐treatment = 10.5% At 26 weeks post‐treatment = 15.8%
Villafane 2018 (Italy)	2‐arm RCT in pre‐post design	Acute stroke patients (n = 32) with hand paralysis from rehabilitation hospitals were randomized into two groups using simple randomization	Mean age = 68.9 ± 11.6 Male% = 65.6% Stroke intervals: < 3 moths	Treatment group ( n = 16) 1hr physical and occupational therapy on weekdays + 30‐min RT on 3 days per week for three weeks Content: passive mobilization of hand through robotic device Gloreha ^f Control group ( n = 16) 1hr physical and occupation therapy + 30 min standard rehabilitation on 3 days per week for three weeks Content: assisted stretching, shoulder, and arm exercises and functional reaching tasks	NIHSS ^a MAS BI MI QuickDASH VAS	RT showed greater reduction in pain compared with UT at postintervention (Cohen's d = 1.73) Except MAS, NIHSS, BI, MI, and QuickDASH showed improvement in both group at post‐treatment ( p <.001)	No attrition was found at post‐treatment
Volpe 2000 (USA)	2‐arm RCT in pre‐post design	Subacute patients from inpatient rehabilitation stroke unit (n = 56) was randomly assigned into 2 groups (Sampling strategies not reported)	Mean age = 64.3 ± 3.20 Male = 53.4% Stroke intervals = 2.1 ± 0.2 weeks Baseline FMA‐ shoulder and elbow M = 5.54 SD = 2.01	Treatment group ( n = 30) Standard physical and occupational poststroke therapy plus 1‐hr RT per day with MIT‐MANUS ^h on weekdays for 5 weeks Content: >1,500 repetitions of goal‐directed shoulder, elbow, wrist, and hand movement to a target Control group ( n = 26) Standard physical and occupational poststroke therapy plus 1‐hr per week of exposure to the robot without training	FMA‐SEC ^a FMA‐WH ^a MS‐SE MS‐WH MP FIM‐Motor FIM‐Cognition	The RT group showed significantly better functional independence compared to the UT group at post‐treatment (p < .01). No significant between‐group difference was found in motor control.	No attrition was found at post‐treatment
Wolf 2015 (USA)	2‐arm RCT in pre‐post design	Subacute stroke patients (n = 99) were randomly assigned into 2 groups using a stratified, computer‐driven randomization procedure	Mean age = 57.0 ± 13.4 Male = 56.6% Stroke intervals = 17.1 ± 7 weeks	Treatment group ( n = 51) 3‐hr session including RT with the Hand Mentor Pro (HMP) ^g and home exercise program on weekdays for 8–12 weeks Content: 1) Wrist and fingers exercises; 2) functional activity Control group ( n = 48) 3‐hr of home exercise program on weekdays for 8–12 weeks Content: 1) Traditional impairment‐based activities, for example, weight‐bearing activates, active assisted exercises, shoulder exercises etc.; 2) functional activities	ARAT ^a WFMT FMA	No significant between‐group difference was found in motor control and upper extremity performance. Both groups showed improvement in motor control and upper extremity performance over time (all ps < 0.001)	7.1%

Study (Country)

Design and assessment time points

Participants

Characteristics of the participants

Intervention

Instruments

Key finding(s)

Attrition rate

Barker 2017

(Australia)

3‐arm RCT in repeated measures design, at baseline, post‐treatment, 26 and 52 weeks post‐treatment

Consecutively admitted patients at one acute stroke unit (n = 50) were randomized into 3 groups using random permuted blocks

Mean age = 53.6 ± 15

Male = 68%

Stroke intervals = 5.9 ± 3 weeks

Treatment groups

SMART Arm ^b training with stimulation for elbow, forearm, wrist, and hand plus conventional therapy (n = 17)
SMART Arm ^b training without stimulation plus conventional therapy (n = 16)

Both groups received 1‐hr session on weekdays for 4 weeks

Content: perform reaching task in a straight‐line path.

Control group ( n = 17)

1‐hr conventional therapy, including physiotherapy, occupational therapy, and therapy assistant time, and involved a mix of one‐on‐one and group therapy sessions on weekdays for 4 weeks

MAS –item 6 ^a

MRC

MAS

RAI

MAS –item 6,7,8

SIS

MAL

kinematics and kinetics of reaching.

The RT group achieved better upper arm function when compared to the control group at post‐training (OR = 1.47, 95% CI = 1.23–1.71) and at 26 weeks (OR = 1.31, 95% CI = 1.05–1.57), respectively. There was no significant between‐group difference in arm function (p = .37).
All groups showed significant improvements in arm function and quality of life (all ps < 0.001) over time.
No significant improvements in muscle tone among all groups at any assessment time points.

Post‐treatment = 6%; 26 weeks post‐treatment = 22%; 52 weeks post‐treatment = 30%

Daunoraviciene 2018

(Lithuania)

2‐arm RCT in pre‐post design

Subacute stroke patients

(n = 34) were randomly assigned into 2 groups

(Sampling strategies not reported)

Mean age = 65.7 ± 4.48

Male = 65%

Stroke intervals = 9.1 ± 5 weeks

Treatment group ( n = 17)

RT with Armeo Spring ^c

for elbow, forearm, wrist, and hand for 1‐hr session on weekdays for 2 weeks

Content: perform a sequence of motor tasks in 5–7 exercise cycles

Control group ( n = 17)

Occupational therapy sessions, including exercising, physical activities, active table games etc.

FIM‐self‐care ^a

FMA

HAM‐D

HAM‐A

ACE‐R

MAS

Active ROM

The RT group showed significantly better functional independence when compared with the UT group at postintervention (p < .03).
No significant between‐group difference in motor control at postintervention (p = .287)

No attrition was found at postintervention

Dehem 2019

(Belgium)

2‐arm RCT in repeated measures design at baseline, postintervention, and 6‐month poststroke

Subacute stroke patients (n = 45) with < 1 month poststroke from inpatient rehabilitation centers were randomized into two groups using computer‐generated sequence

Mean age = 67.9 ± 15.4

Male % = 46.7%

Stroke intervals = 27.8 ± 5.5 days

Treatment group ( n = 23)

Four 45‐min RT sessions for wrist and hand with REAplan robot ^d (25%) and twelve UT sessions (75%) per week for 9 weeks

Content: exercises with game involving moving the paretic hand along a reference trajectory while passing through checkpoints

Control group ( n = 22)

Sixteen 45‐min UT session per week for 9 weeks

Content: motor rehabilitation that matched with patients’ personal needs

FMA‐UE ^a

BBT

WMFT

ABILHANDACTIVLIM

SIS

RT showed significantly greater improvement in gross manual dexterity (p = .02), upper‐limb ability (p =.02) and patient social participation (p = .01) compared with control group at six months poststroke
Both group show similar improvement in abilities to perform manual activities and activities of daily living

Postintervention = 28.9%, six months poststroke = 37.8%

Hesse 2014

(Germany)

2‐arm RCT in repeated measures design at baseline, post‐treatment and three‐month post‐treatment

Subacute patients from two inpatient stroke rehabilitation units

(n = 50) were assigned into 2 groups by web‐based randomization tool

Mean age = 70.6 ± 16.1

Male = 56%

Stroke intervals = 4.5 ± 2 weeks

Treatment group ( n = 25)

30‐min of RT in arm studio plus 30‐min individual arm therapy on weekdays for 4 weeks

Content: 1) repetitive practice of finger, wrist, forearm, and shoulder movement; 2) task‐oriented motor relearning program and impairment‐oriented arm ability training

Control group ( n = 25)

1‐hr individual arm therapy on weekdays for 4 weeks, consisting of the task‐oriented motor relearning program and the impairment‐oriented arm ability training (repetitions of movements and shaping)

FMA ^a

ARAT

BBT

MRC

MAS

No significant between‐group differences were found in motor control, upper extremity performance, muscle tone, and functional independence at all measurement points.
Both groups showed significant improvements in motor control and upper extremity performance at post‐training and at three‐month follow‐up (all ps <0.001)

Post‐treatment = 2%, 3 months post‐treatment = 8%

Masiero

2014

(Italy)

2‐arm RCT in repeated measures design at baseline, post‐treatment, 3 months and 7 months post‐treatment

Subacute subjects from Stroke Unit

(n = 34) were randomly allocated to 2 groups by a computer program

Mean age = 66.3 ± 8.55

Male = 66.7%

Stroke intervals = 1.3 ± 0.4 week

Treatment group ( n = 16)

~75‐min of standard therapy plus ~ 45‐min of RT for elbow, foreman, wrist, and hand on weekday for 5 weeks

Content: 1) NeReBot training ^e with exercises (flexion and extension, adduction and abduction, pronation and supination, circumduction); 2) Conventional functional rehabilitation including proprioceptive exercises, functional reduction, gait training, occupational therapy, passive, and active assisted mobilization of the hand, wrist, and upper paretic arm

Control group ( n = 18)

2‐hr of daily rehabilitation treatment for 5 weeks including proprioceptive exercises, functional

re‐education, gait training, occupational therapy, and passive and active assisted mobilization of the hand and wrist.

FMA ^a

FIM‐motor ^a

FAT ^a

MRC

MAS

BBT

tolerability and acceptability of treatment

No significant between‐group differences were found in motor function, muscle tone, and functional independence at all assessment time‐points.

At post‐treatment = 11.8%;

7 months post‐treatment = 17.6%

Orihulela Espina 2016

(Mexico)

2‐arm RCT in pre‐post design

Subacute stroke patients from Neurologic Rehabilitation Unit

(n = 17) were randomly assigned into 2 groups by block randomization

Mean age = 55.6 ± 20.3

Male = 64.7%

Stroke intervals = 1 week to 4 months

Treatment group ( n = 9)

1‐hr of RT for wrist and hand on weekdays for around 8 weeks

Content: 1) passive activities; 2) partial assistance/ resistance activities; 3) active movement

Control group ( n = 8)

1‐hr of classical occupational therapy on weekdays for around 8 weeks

Content: massage and conventional occupational exercises, including passive movements, strengthening exercises and active grasps movement and personalized activities for fine pinching control

FMA‐hand ^a

MI ^a

RT showed significantly greater improvement in hand motor function compared with UT at post‐treatment (p < .01).
Both groups showed significant improvement in motor control over time (Nonparametric Cliff's delta‐within effect sizes: dwOT‐FMA = 0.5, dwRT‐FMA = 1)

No attrition was found at postintervention

Sale 2014

(Italy)

2‐arm RCT in repeated measures design at baseline, after 15th and 30th treatment sessions

Subacute stroke patients from the rehabilitation center

(n = 53) were randomly allocated into 2 groups by dedicated software

Mean age = 67.7 ± 14.2

Male = 58.5%

Stroke intervals = 4.3 ± 1week

Treatment group ( n = 26)

45‐min of RT with MIT‐MANUS ^h for shoulder and elbow plus 3‐hr physiotherapy on weekdays for 6 weeks

Content: 1) dexterity and gait training, 2) goal‐directed, planar reaching tasks, including both unassisted and assisted repetitions

Control group ( n = 27)

45‐min of conventional therapy plus 3‐hr of physiotherapy on weekdays for 6 weeks

Content: 1) dexterity and gait training; 2) assisted stretching, shoulder and arm exercises and functional reaching tasks exercising, physical activities, and active table games

FMA ^a

MAS‐S ^a

MAS‐E ^a

pROM

Both groups showed significant improvements in motor control after 15th and 30th session, with significant greater improvement found in the RT group after the first 15th sessions (p <.0001).

Significant improvement in muscle tone for shoulder and elbow (p <.05) was only found in the RT group.

No attrition was found after 30th treatment sessions

Stinear 2014

(New Zealand)

2‐arm RCT in repeated measures design at baseline, 6, 12 and 26 weeks post‐treatment

Consecutive subacute stroke patients from a stroke unit

(n = 57) were randomized by customized software (www.rando.la)

Mean age = 68 ± 25

Male = 45.6%

Stroke intervals < 26 days

Treatment group ( n = 29)

15‐min Bilateral priming for wrists and hands plus 30‐min physiotherapy and occupational therapy on weekdays for 4 weeks

Control group ( n = 28)

15‐min Intermittent cutaneous electric stimulation plus 30‐min physiotherapy and occupational therapy on weekdays for 4 weeks

ARAT ^a

SIS

MRS

At 12th weeks, greater proportion of the participants in the treatment group achieved their recovery plateaus in upper extremity performance than the control group (χ² = 4.25; p = .039).
At 26th weeks, no between‐group difference was found in quality of life⁵

At post‐treatment = 7.0%

At 12 weeks post‐treatment = 10.5%

At 26 weeks post‐treatment = 15.8%

Villafane 2018

(Italy)

2‐arm RCT in pre‐post design

Acute stroke patients (n = 32) with hand paralysis from rehabilitation hospitals were randomized into two groups using simple randomization

Mean age = 68.9 ± 11.6

Male% = 65.6%

Stroke intervals: < 3 moths

Treatment group ( n = 16)

1hr physical and occupational therapy on weekdays + 30‐min RT on 3 days per week for three weeks

Content: passive mobilization of hand through robotic device Gloreha ^f

Control group ( n = 16)

1hr physical and occupation therapy + 30 min standard rehabilitation on 3 days per week for three weeks

Content: assisted stretching, shoulder, and arm exercises and functional reaching tasks

NIHSS ^a

MAS

QuickDASH

VAS

RT showed greater reduction in pain compared with UT at postintervention (Cohen's d = 1.73)
Except MAS, NIHSS, BI, MI, and QuickDASH showed improvement in both group at post‐treatment ( p <.001)

No attrition was found at post‐treatment

Volpe 2000

(USA)

2‐arm RCT in pre‐post design

Subacute patients from inpatient rehabilitation stroke unit

(n = 56) was randomly assigned into 2 groups

(Sampling strategies not reported)

Mean age = 64.3 ± 3.20

Male = 53.4%

Stroke intervals = 2.1 ± 0.2 weeks

Baseline FMA‐ shoulder and elbow

M = 5.54

SD = 2.01

Treatment group ( n = 30)

Standard physical and occupational poststroke therapy plus 1‐hr RT per day with MIT‐MANUS ^h on weekdays for 5 weeks

Content: >1,500 repetitions of goal‐directed shoulder, elbow, wrist, and hand movement to a target

Control group ( n = 26)

Standard physical and occupational poststroke therapy plus 1‐hr per week of exposure to the robot without training

FMA‐SEC ^a

FMA‐WH ^a

MS‐SE

MS‐WH

FIM‐Motor

FIM‐Cognition

The RT group showed significantly better functional independence compared to the UT group at post‐treatment (p < .01).

No significant between‐group difference was found in motor control.

No attrition was found at post‐treatment

Wolf 2015

(USA)

2‐arm RCT in pre‐post design

Subacute stroke patients

(n = 99) were randomly assigned into 2 groups using a stratified, computer‐driven randomization procedure

Mean age = 57.0 ± 13.4

Male = 56.6%

Stroke intervals = 17.1 ± 7 weeks

Treatment group ( n = 51)

3‐hr session including RT with the Hand Mentor Pro (HMP) ^g and home exercise program on weekdays for 8–12 weeks

Content: 1) Wrist and fingers exercises; 2) functional activity

Control group ( n = 48)

3‐hr of home exercise program on weekdays for 8–12 weeks

Content: 1) Traditional impairment‐based activities, for example, weight‐bearing activates, active assisted exercises, shoulder exercises etc.; 2) functional activities

ARAT ^a

WFMT

FMA

No significant between‐group difference was found in motor control and upper extremity performance.
Both groups showed improvement in motor control and upper extremity performance over time (all ps < 0.001)

7.1%

Abbreviations: ACE‐R, Addenbrooke Cognitive Examination‐Revised; Active ROM, Active Range of Motion; ARAT, Action Research Arm Test; BBT, Box and Block Test; BI, Barthel Index; FAT, Frenchay Arm Test; FIM, Functional Independence Measurement; FMA, Fugl‐Meyer Assessment; HAM‐A, Hamilton Rating Scale for Anxiety; HAM‐D, Hamilton Rating Scale for Depression; MAL, Motor Activity Log‐28; MAS, Modified Ashworth Scale; MAS, Motor Assessment Scale; MI, Motricity Index; MP, Motor Power Scale; MRC, Medical Research Council, MRS, Modified Rankin Scale; MS, Motor Status Score; NIHSS, the National Institutes of Health Stroke Scale; pROM, passive Range of motion; QuickDASH, short version of the Disabilities of the Arm, Shoulder and Hand; RAI, Ritchie Articular Index; RCT, randomized controlled trial; RT, Robot‐assisted therapy; SIS, Stroke Impact Scale; VAS, Visual Analog Scale; WFMT, Wolf Motor Function Test.

^{^a}

Primary outcome(s) of the included study.

^{^b}

SMART ArmTM http://smartarm.com.au/development/.

^{^c}

Armeo Spring: https://www.hocoma.com/solutions/armeo‐spring/.

^{^d}

REAplan robot https://www.axinesis.com/en/.

^{^e}

Masiero, S., Celia, A., Armani, M., & Rosati, G. (2006). A novel robot device in rehabilitation of post‐stroke hemiplegic upper limbs. Aging clinical and experimental research, 18(6), 531–535.

^{^f}

Gloreha https://www.gloreha.com.

^{^g}

The Hand Mentor Pro (HMP) https://motusnova.com/products/hand‐mentor‐pr.

^{^h}

MIT‐MANUS/ InMotion2, (Interactive Motion Technologies, Inc., Watertown., MA, USA).