Evidence reviews for robot-assisted arm training

Evidence reviews for robot-assisted arm training

Stroke rehabilitation in adults (update)

Evidence review M

NICE Guideline, No. 236

London: National Institute for Health and Care Excellence (NICE); 2023 Oct.

ISBN-13: 978-1-4731-5462-9

1. Robot-assisted arm training

1.1. Review question

In people after stroke, what is the clinical and cost effectiveness of robot-assisted arm training in improving function and reducing disability?

1.1.1. Introduction

Robot assisted arm training is an intervention which allows people with arm weakness following stroke to perform repetitive functional tasks with the aim of improving strength and function. Repetitive functional task practice is known to help recovery following stroke and robot assisted arm training is a potential mechanism to increase the intensity and frequency of rehabilitation. In previous guidance, robot assisted arm training could only be recommended in the context of a clinical trial and it is important to understand whether recent evidence might support its use as an intervention or adjunct to improve arm recovery.

In current clinical practice, robot assisted arm training is not widely available. New technologies are being developed which are potentially more accessible to both in hospital and community services. It is not yet understood the extent to which robot assisted arm training could benefit arm recovery, or indeed whether use of robots may potentially cause harm to the weaker arm following stroke. In addition, there are discrepancies around the use of this technology regarding whether it can be used independently or requires supervision by health care professionals.

Implementation of robot assisted arm training will require investment in training and equipment in the majority of services and review of evidence is necessary to understand both the effectiveness and cost effectiveness of its implementation within stroke rehabilitation services.

1.1.2. Summary of the protocol

Table 1

PICO characteristics of review question.

For full details see the review protocol in Appendix A.

1.1.3. Methods and process

This evidence review was developed using the methods and process described in Developing NICE guidelines: the manual. Methods specific to this review question are described in the review protocol in Appendix A and the methods document.

Declarations of interest were recorded according to NICE’s conflicts of interest policy.

1.1.4. Effectiveness evidence

1.1.4.1. Included studies

One systematic review⁷⁹ and in total eighty-one randomised controlled trial studies (one hundred and five papers) were included in the review¹^–⁵^, ⁸^–¹¹^, ¹⁴^–²⁰^, ²³^, ²⁴^, ²⁶^, ²⁷^, ³⁰^, ³³^–³⁶^, ³⁸^–⁴¹^, ⁴³^–⁵⁰^, ⁵³^–⁵⁸^, ⁶¹^, ⁶²^, ⁶⁴^–⁶⁸^, ⁷²^–⁷⁴^, ⁷⁷^, ⁸²^, ⁸⁴^–⁸⁸^, ⁹⁰^, ⁹²^–¹⁰⁶^, ¹⁰⁹^–¹¹²; these are summarised in Table 2 below. Evidence from these studies is summarised in the clinical evidence summary below (Table 3).

This review updated a previous Cochrane review, Mehrholz 2018⁷⁹. This review included forty-five trials from up to January 2018. A search from January 2018 was completed and an additional thirty trials were added to the review¹^, ⁵^, ¹⁰^, ¹⁴^–¹⁹^, ²⁶^, ²⁷^, ³⁴^, ³⁶^, ⁴⁵^, ⁴⁶^, ⁴⁹^, ⁵⁰^, ⁵⁴^, ⁵⁵^, ⁵⁸^, ⁸⁴^, ⁸⁶^–⁸⁸^, ⁹²^, ⁹³^, ⁹⁷^, ¹⁰³^, ¹¹⁰^, ¹¹². This included six cross-over trials (of which only the first phase was included in the analysis as per the Cochrane review protocol)³^, ⁹^, ¹⁶^, ⁴⁰^, ⁴⁹^, ⁷⁴.

Trials included comparisons of robot assisted arm therapy to any other intervention (including usual care/conventional rehabilitation, no treatment and other interventions. All comparisons have been pooled for the analysis as in Mehrholz 2018⁷⁹.

Robot assisted arm training was usually offered alongside conventional rehabilitation exercises or in two studies as a combination with other therapies (including botulinum toxin A injection and functional electrical stimulation).

Studies included a range of robotic devices which performed different movement types (including active, active/assisted, passive or a combination) and which targeted different parts of the joint (for example: proximal or distal). The type of movement and the region of the limb trained was poorly reported in many studies, but the majority of the robotic devices provided a combination of passive and active assisted movements and trained both the proximal and distal limb to perform tasks such as reaching and grasping. Nearly all of the trials reported supervised robot assisted arm training and the healthcare professional delivering the therapy was most commonly an occupational therapist or physiotherapist.

The people included in the studies were from a mixture of time periods after stroke, being split between subacute and chronic periods. However, the majority of studies included people in the subacute phase post stroke.

Indirectness

7 outcomes were downgraded for indirectness due to outcome indirectness arising from a short follow up duration. Specifically, any outcomes reported after the post intervention follow up were included in the ≥6 month follow up category and if these were reported at less than 6 months they were downgraded.

Inconsistency

A number of outcomes showed significant heterogeneity. In each case, this was not resolved by sensitivity or subgroup analyses and so random effects models were used, and the outcomes were downgraded for inconsistency.

See also the study selection flow chart in Appendix C, study evidence tables in Appendix D, forest plots in Appendix E and GRADE tables in Appendix F.

1.1.4.2. Excluded studies

See the excluded studies list in Appendix J.

1.1.5. Summary of studies included in the effectiveness evidence

Table 2

Summary of studies included in the evidence review.

See Appendix D for full evidence tables.

1.1.6. Summary of the effectiveness evidence

Table 3

Clinical evidence summary: Robot-assisted arm training compared to any other intervention.

See Appendix F for full GRADE tables.

1.1.7. Economic evidence

1.1.7.1. Included studies

Two health economic studies with the relevant comparison were included in this review.³¹^,⁴⁹ These are summarised in the health economic evidence profile below (Table 4) and the health economic evidence tables in Appendix H. Note that one study,³¹ as well as the RCT⁸⁸ that formed the basis of the analysis are also included as part of the evidence review for this guideline that assessed the clinical and cost-effectiveness of more intensive rehabilitation.

1.1.7.2. Excluded studies

No relevant health economic studies were excluded due to assessment of limited applicability or methodological limitations.

See also the health economic study selection flow chart in Appendix G.

1.1.8. Summary of included economic evidence

Table 4

Health economic evidence profile: Robot-assisted arm training versus usual care.

1.1.9. Economic model

This area was not prioritised for new cost-effectiveness analysis.

1.1.10. Unit costs

Relevant unit costs are provided below to aid consideration of cost effectiveness.

The main additional resource use of robot-assisted arm training is the cost of the robotic device. The studies included in the clinical review used different robots. The RATULS RCT (Rodgers 2019⁸⁸), conducted as part of the Health Technology Assessment (HTA) programme, provided UK costs associated with the MIT-Manus robotic gym. This included the initial capital investment and maintenance fees. Costs associated with a trial centre’s estate and facilities were included in the salary costs of the staff delivering the therapy, and so are not incorporated in the robotic device costing below (staff costs were incorporated in the cost effectiveness analysis above however). No additional storage facilities were identified as the robotic gyms were installed in the therapy room. The allocation of these capital costs was conducted following the ‘equivalent annual cost’ methodology by Drummond 2005²⁸. This method allowed for the initial capital cost to be converted into an annual sum that equals the resources invested plus their opportunity cost.

The equivalent annual cost of each robot session was calculated under the following assumptions:

Robot usage: 35 average number of sessions per week (seven sessions held on an 8-hour day). Weeks per year that the MIT-Manus robotic gym system is in use: 52 weeks.
Useful lifespan of the MIT-Manus robotic gym system is 5 years.
Training costs are not included as they are not considered to drive any differences in costs between randomisation groups.
The capital cost of the robotic gym was spread over its lifespan (5 years).
A discount factor of 3.5% was applied to account for the individual’s time preference for costs to be incurred later rather than sooner. This follows guidance for best practice.

Tables 5 and 6 illustrate this method, incorporating the initial purchasing cost of £1,000,000 for the MIT-Manus robotic gym and £15,000 annual fees.

Table 5

Equivalent annual cost or equivalent annuity from Rodgers 2019.

Table 6

Illustrative cost of the MIT-Manus robotic gym per session from Rodgers 2019.

Resource use varied across studies included in the clinical review due to the following factors:

Variation in the frequency and duration of training time with the robot-assisted device, with sessions ranging from 20 minutes to 60 minutes, not including time spent receiving conventional rehabilitation therapy. In some instances, robot-assisted arm training added more intervention time, and, in these cases, there would be additional staff time costs. Sessions mostly occurred 3-5 days per week. In the included clinical studies, the interventions were delivered for between 2 weeks and 9 weeks and had follow-up periods from 3 weeks up to 8 months.
A small number of studies included other interventions being given with robot-assisted training (such as neuromuscular, transcranial and functional electrical stimulation) which would also be an additional cost.
Training was primarily supervised by a member of the rehabilitation team, such as occupational therapists and physiotherapists. However, one study from the clinical review (Budhota 2021^10xxx) reported that the training was minimally supervised by occupational therapists as well as bioengineers. Rodgers 2019⁸⁸ reported that therapists and therapy assistants delivered interventions.
The level of supervision differed across studies as well. Most studies were reported to have participants supervised by therapists, however, Hesse 2005³⁹ reported that while patients were left unsupervised during the training, a therapist remained ‘within shouting distance’ in case of problems and Housman 2009⁴¹ reported mixed supervision, where the first three sessions were supervised before offering intermittent supervision for the remaining sessions. Remy-Neris 2021⁴⁹ assessed a similar approach, where a therapist was present during the first 4 sessions but for the remaining sessions, the therapist set the patient up in the device, adjusted the device parameters, and programmed the exercises, but the participant then trained independently.
Additional resource use required as part of the intervention, such as staff-training costs.

Table 7

Unit costs of health care professionals who may be involved in delivering robot-assisted arm training interventions.

1.1.11. Evidence statements

Effectiveness/Qualitative

Economic

One cost-utility analysis found that robot-assisted arm training plus usual care was dominated (higher costs and lower QALYs) by usual care alone for people following a stroke. This analysis was assessed as directly applicable with potentially serious limitations.
One cost-utility analysis suggested that for people following stroke, usual rehabilitation plus an additional hour of games-based self-rehabilitation using an exoskeleton incurred lower costs and higher QALYs compared to usual rehabilitation alone, however total costs and QALY gains were not statistically significant between groups. This analysis was assessed as partially applicable with potentially serious limitations.

1.1.12. The committee’s discussion and interpretation of the evidence

1.1.12.1. The outcomes that matter most

The committee included the following outcomes: person/participant generic health-related quality of life, carer generic health-related quality of life, activities of daily living, arm function, arm muscle strength, spasticity, stroke-specific Patient-Reported Outcome Measures, withdrawal for any reason and adverse events (including cardiovascular events, injuries and pain and other reported adverse events). All outcomes were considered equally important for decision making and therefore have all been rated as critical.

This review updated a published Cochrane review⁷⁸. Therefore, the outcomes used in this review are the same as those reported in the Cochrane review with the inclusion of four additional outcomes which were agreed by the guideline committee: person/participant and carer generic health-related quality of life, stroke-specific Patient-Reported Outcome Measures and spasticity. Person/participant and carer generic health-related quality of life outcomes were added to this review as they are important outcomes for understanding the holistic impact of the treatment and to further understanding of the economic considerations. Similarly, stroke-specific Patient-Reported Outcome Measures were added as these provide insight into how the interventions affect the persons functional abilities or quality of life. The spasticity outcome was added as the committee deemed it important given the nature of the intervention and as previous research has highlighted increases in spasticity as a potential adverse effect of robot assisted therapy.

The committee chose to investigate these outcomes at post intervention and at more than and equal to 6 months follow up period as they considered that there could be a difference in the short-term and long-term effects of the intervention.

There was a large amount of evidence available for the majority of the outcomes at both follow up time points with the number of studies reporting each outcome ranged from 2 to 66. Evidence was more limited for person/participant health-related quality of life and cardiovascular adverse events, but there was sufficient evidence available for the committee to make a recommendation.

1.1.12.2. The quality of the evidence

Seventy-five randomised controlled trial studies were included in the review including six crossover RCTs (in which only the first phase was analysed as a parallel trial). Evidence was available for robot assisted arm training compared to any other intervention (including usual care, placebo and no treatment) at post-intervention and after 6 months follow up periods. Results from studies that compared robot assisted arm training to any of the above interventions were pooled together in the analysis as this was the method employed by the Cochrane review.

The evidence varied from high to very low quality, with the majority being of very low quality. Outcomes were commonly downgraded for risk of bias, inconsistency, indirectness and imprecision. Risk of bias was rated as a concern in the majority of the studies. This was generally due to bias in the randomisation procedure, bias arising due to deviations from the intended interventions, bias in the measurement of the reported result and bias arising from missing outcome data.

Inconsistency was present in many of the outcomes due to the large number of studies and the heterogeneity in the included evidence which reported different time periods post-stroke, doses of the intervention and sample sizes. Heterogeneity was investigated with sensitivity analyses and the pre-specified subgroup analyses. None of the analyses resolved the heterogeneity so these outcomes were downgraded for inconsistency. In several cases the heterogeneity was deemed to be due to differences in the study sample sizes (specifically Rodgers 2019 had a much larger population than any others in the review). Therefore, to avoid over emphasising the effects of the smaller unblinded studies in the analysis a fixed effects analysis was employed for these outcomes rather than using a random effects model.

Seven outcomes were downgraded due to outcome indirectness arising from a short follow up duration. As detailed in the protocol, any outcome reported after the post intervention follow up (and at the longest available follow up time point) was included in the more than and equal to 6 months follow up category. However, if these outcomes were reported at less than 6 months they were downgraded for indirectness. Imprecision was seen in several outcomes due to small sample sizes and uncertainty around the effect estimate.

The committee concluded that the evidence was of a sufficient quality to make recommendations. The committee noted that studies took place in a wide range of countries worldwide which in some of cases may limit applicability to the NHS. One lay member also voiced their concern that a number of studies have been funded by the manufacturers which may introduce further bias in these studies. However, a large multi-site NIHR funded study (⁸⁸) recently took place in the UK which included a health economic analysis. This study reported many of the outcomes included in this review and was of low risk of bias. Therefore, the committee gave this study greater consideration in their decision making.

1.1.12.3. Benefits and harms

The results showed that when robot assisted arm training was compared to any other intervention an inconsistent effect was seen. There was a clinically important benefit in some outcomes and no clinically important difference in other outcomes in arm function at more than and equal to 6 months and arm muscle strength at end of intervention and more than and equal to 6 months. An unclear effect where some outcomes showed a clinically important benefit, some no clinically important difference and one a clinically important harm was also seen in stroke-specific Patient-Reported Outcome Measures at end of intervention.

No clinically important difference was seen in person/participant generic health-related quality of life at end of intervention, arm function at end of intervention, spasticity at end of intervention and more than and equal to 6 months, stroke-specific Patient-Reported Outcome Measures at more than and equal to 6 months, withdrawal for any reason at end of intervention and more than and equal to 6 months and adverse events (including cardiovascular events, injuries and pain and other reported adverse events) at end of intervention and more than and equal to 6 months.

An inconsistent effect where some outcomes showed no clinically important difference and some showed a clinically important harm was seen in person/participant generic health-related quality of life at more than and equal to 6 months. The committee acknowledged that where there was evidence that robot arm therapy was worse than any other intervention at improving quality of life, this was based on evidence from the Rodgers 2019⁸⁸ study, which was a large RCT in which the 2 comparison groups (an enhanced upper limb therapy intervention and usual care) were combined for the analysis. The committee considered the fact that the enhanced therapy group received regular one on one, face-to-face physiotherapy treatment which seemed to be more intensive than the usual care provided in other studies. Hence, the committee suggested this may explain the benefit for the other interventions arm for this outcome. Furthermore, when the robot assisted arm training arm was compared to the usual care arm alone the results showed a small benefit for robot therapy in the post-intervention follow-up and no difference at more than and equal to 6 months.

The committee acknowledged the benefits reported for several of the arm muscle strength outcomes and concluded that robot assisted arm training may be appropriate for improving muscle strength alone. However, this does not appear to translate to functional gains, improvements in activities of daily living and ultimately in the person’s quality of life. These outcomes may be more important to the holistic wellbeing of the person and was considered in their deliberation. However, the committee agreed that improving upper limb strength may reduce pain and improve joint stability. Therefore, they suggested that these devices may be appropriate for strength training in a specific subset of patients who present with a motor deficit and in whom upper limb strengthening is the main goal of treatment. These findings were echoed in the experiences of one lay member who had used a robotic device during his rehabilitation and suggested that although it may have helped improve his strength in the short term it did not seem to have any lasting positive effects on his function.

The committee also discussed the results of the Rodgers 2019⁸⁸, study which found greater improvements in the enhanced upper limb therapy group when compared to the robot-assisted arm training group for many outcomes. This enhanced therapy arm included face-to-face functional task training delivered by a physiotherapist which was matched for time with the robot therapy arm. Based on these findings the committee argued that more intensive physiotherapy for the upper limb seems to be more effective than additional therapy delivered by the robot device. This view was supported by the lay members who preferred therapy sessions delivered by physiotherapists rather than ‘being left alone with a machine’. One lay member suggested that the personal relationships formed with the physiotherapist are crucial for building trust and increasing motivation to engage in therapy sessions. They also noted that technical issues with the devices along with time spent explaining and setting up the devices wasted valuable therapy time.

On reviewing the evidence, the committee considered the balance of benefits and harms and the large amount of evidence reporting no clinically important difference. Ultimately, they agreed that the evidence did not support a recommendation for the use of robot-assisted arm training. The committee were satisfied by the amount of evidence available and noted that the evidence encompassed a wide range of robotic devices performing different types of movement at different doses and in subacute/chronic time periods post stroke. Therefore, they did not feel that a research recommendation was necessary.

Despite the lack of evidence in support of robot assisted arm training there was also no evidence reporting a harm of the device. Therefore, the committee agreed that if services already have access to a robot device there is no clinical reason why they should avoid using it in specific circumstances (for example: people after stroke who present with upper limb motor problems in whom the main treatment goal is to improve upper limb strength). However, the overall clinical benefit of a sole improvement in muscle strength was unclear as there was no evidence to suggest that overarching outcomes which may be more important to people after stroke, such as quality of life, would be improved and there would be resource use implications. This time could also be used by a therapist for other therapy that may be able to achieve greater benefits in other areas that may impact quality of life more. Taking into account these factors, and the cost effectiveness evidence, the committee concluded that robot-assisted arm training should not be offered as part of an upper limb rehabilitation program.

1.1.12.4. Cost effectiveness and resource use

Two health economic analyses were identified for this review. The first study included in the review was a within-trial cost-utility analysis of an RCT included in the clinical review, which compared usual rehabilitation (1 hours, 5 days per week for 4 weeks) plus an additional daily hour of self-rehabilitation , consisting of basic stretching and active exercises for the control group versus usual rehabilitation plus an additional daily hour of self-rehabilitation consisting of gravity-supported, games-based training using an exoskeleton (Armeo®Spring). The results suggested that the Exo group intervention dominates usual care (lower costs and higher QALYs), however total costs and QALY gains were not statistically significant between groups. The study conclusions were shown to be robust following a probabilistic sensitivity analysis. The analysis was assessed as partially applicable as the study was set in the French healthcare system which may not reflect the current UK NHS context. In addition, the French population valuation tariff was used to estimate QALYs, but NICE reference case specifies that the UK tariff is preferred. Potentially serious limitations were identified as the study was a within-trial analysis of a single RCT which meant the results only reflect this study and not the wider evidence based identified in the clinical review. References for unit costs were not reported either which further limits the interpretation of the results for a UK context.

The second study was also a within-trial cost-utility analysis of a UK RCT included in the clinical review, where participants were randomised to one of three programmes over a 12-week period: usual care (45 minutes with a physiotherapist or occupational therapist, 5 days a week); robot-assisted training (45 minutes per day, 3 times per week) plus usual care or the EULT programme (45 minutes with a physiotherapist, 3 times per week) plus usual care. The results found that robot-assisted arm training was not cost-effective, as it incurred higher overall costs than both usual care and EULT, primarily due to having higher intervention costs. In addition, robot-assisted training was not associated with higher QALYs than usual care and resulted in lower QALYs than EULT.

There was low uncertainty in this conclusion. Note the conclusions about the EULT intervention are discussed in the intensity of rehabilitation evidence report. The analysis was assessed as directly applicable with minor limitations. Although it is a within-RCT analysis and so only reflects the results of this study, the RCT was a large, recent, NIHR funded, UK-based study that was considered highly applicable by the committee. In addition, while it had a limited follow-up period, sensitivity analyses that extrapolated the trial data to a 12-month time horizon did not change the study conclusions regarding robot arm training.

The committee were also presented with intervention costs from the NIHR study, which incorporated capital and maintenance costs for the robot as well as physiotherapy time to supervise the training. The estimated cost per session of the robot was £41 (assuming the robot is used for an average of 35 session per week for 52 weeks per year with capital costs spread over 5 years). This incorporated an initial purchase cost of £1,000,000 and £15,000 annual fees. Physiotherapy time with robot-assisted training was the same as for EULT and higher than usual care. The committee noted that they were unsure if people would receive supervision from a physiotherapist for the entire duration of robot-training if this was provided in clinical practice. Less staff supervision would suggest lower intervention costs than what is reported in the analysis but given that there wasn’t an increase in QALYs with robot-arm training it would still not be cost-effective in this case. It is also unknown if less supervision would affect clinical outcomes. The committee also highlighted that storage and space to use the devices in an NHS setting would likely be an issue. Costs related to this were not captured in the NIHR study as it was possible to install the robot in existing therapy rooms, however the committee did not think this would always be possible.

The committee stated that robot arm training is not commonly used in current practice, however it was noted that a few UK hospitals currently own a robot-training machine. They discussed that even where machines were already available there would be ongoing maintenance costs and use would require staff time for supervision of the intervention (and machines would ultimately need replacing if use continued). In addition, it was noted that if machines were only used in a small subset of patients and so could not be used to full capacity this may increase the cost per use and so overall intervention costs. The committee also highlighted that time was also required for setting-up the machine for each use and to teach the person how to use it. Staff training costs to use the machine may also be incurred. For these reasons the committee agreed that there would be a significant resource impact associated with robot arm training and alongside the limited clinical evidence the committee concluded the robot arm training was not cost-effective for the NHS and made a ‘do not offer’ recommendation.

1.1.13. Recommendations supported by this evidence review

This evidence review supports recommendation 1.13.18.

1.1.14. References

1.: Abdollahi F, Corrigan M, Lazzaro EDC, Kenyon RV, Patton JL. Error-augmented bimanual therapy for stroke survivors. NeuroRehabilitation. 2018; 43(1):51–61 [PMC free article: PMC8682915] [PubMed: 30040762]

2.: Abdullah HA, Tarry C, Lambert C, Barreca S, Allen BO. Results of clinicians using a therapeutic robotic system in an inpatient stroke rehabilitation unit. Journal of Neuroengineering and Rehabilitation. 2011; 8(1):1–12 [PMC free article: PMC3182973] [PubMed: 21871095]

3.: Amirabdollahian F, Loureiro R, Gradwell E, Collin C, Harwin W, Johnson G. Multivariate analysis of the Fugl-Meyer outcome measures assessing the effectiveness of GENTLE/S robot-mediated stroke therapy. Journal of Neuroengineering and Rehabilitation. 2007; 4(1):1–16 [PMC free article: PMC1819386] [PubMed: 17309791]

4.: Ang KK, Guan C, Phua KS, Wang C, Zhou L, Tang KY et al. Brain-computer interface-based robotic end effector system for wrist and hand rehabilitation: results of a three-armed randomized controlled trial for chronic stroke. Frontiers in Neuroengineering. 2014; 7:30 [PMC free article: PMC4114185] [PubMed: 25120465]

5.: Aprile I, Germanotta M, Cruciani A, Loreti S, Pecchioli C, Cecchi F et al. Upper Limb Robotic Rehabilitation After Stroke: A Multicenter, Randomized Clinical Trial. Journal of Neurologic Physical Therapy. 2020; 44(1):3–14 [PubMed: 31834217]

6.: Aprile I, Germanotta M, Cruciani A, Pecchioli C, Loreti S, Papadopoulou D et al. Poststroke shoulder pain in subacute patients and its correlation with upper limb recovery after robotic or conventional treatment: A secondary analysis of a multicenter randomized controlled trial. International Journal of Stroke. 2021; 16(4):396–405 [PubMed: 32640881]

7.: Bank of England. Inflation calculator. 2022. Available from: https://www.bankofengland.co.uk/monetary-policy/inflation/inflation-calculator Last accessed: 01/05/2023.

8.: Bishop L, Stein J, Schoenherr G, Chen C, Nilsen D, Beer R. Robot assisted hand exercise compared with conventional exercise therapy after ischemic stroke: a pilot study. Neurorehabilitation and Neural Repair. 2014; 28(9):919

9.: Brokaw EB, Nichols D, Holley RJ, Lum PS. Robotic therapy provides a stimulus for upper limb motor recovery after stroke that is complementary to and distinct from conventional therapy. Neurorehabilitation and Neural Repair. 2014; 28(4):367–376 [PubMed: 24297763]

10.: Budhota A, Chua KSG, Hussain A, Kager S, Cherpin A, Contu S et al. Robotic Assisted Upper Limb Training Post Stroke: A Randomized Control Trial Using Combinatory Approach Toward Reducing Workforce Demands. Frontiers in neurology [electronic resource]. 2021; 12:622014 [PMC free article: PMC8206540] [PubMed: 34149587]

11.: Burgar CG, Lum PS, Scremin AM, Garber SL, Van der Loos HF, Kenney D et al. Robot-assisted upper-limb therapy in acute rehabilitation setting following stroke: Department of Veterans Affairs multisite clinical trial. Journal of Rehabilitation Research and Development. 2011; 48(4):445–458 [PubMed: 21674393]

12.: Burgar CG, Lum PS, Shor M, Van der Loos HFM. Rehabilitation of upper limb dysfunction in chronic hemiplegia: robot-assisted movements vs. conventional therapy. Archives of Physical Medicine and Rehabilitation. 1999; 80(9):1121 [PubMed: 12098155]

13.: Burgar CG, Lum PS, Shor PC, Van der Loos HFM. Development of robots for rehabilitation therapy: The Palo Alto VA/Stanford experience. Journal of Rehabilitation Research and Development. 2000; 37(6):663–674 [PubMed: 11321002]

14.: Calabro RS, Accorinti M, Porcari B, Carioti L, Ciatto L, Billeri L et al. Does hand robotic rehabilitation improve motor function by rebalancing interhemispheric connectivity after chronic stroke? Encouraging data from a randomised-clinical-trial. Clinical Neurophysiology. 2019; 130(5):767–780 [PubMed: 30904771]

15.: Carpinella I, Lencioni T, Bowman T, Bertoni R, Turolla A, Ferrarin M et al. Effects of robot therapy on upper body kinematics and arm function in persons post stroke: a pilot randomized controlled trial. Journal of Neuroengineering and Rehabilitation. 2020; 17(1):10 [PMC free article: PMC6990497] [PubMed: 32000790]

16.: Chen YW, Chiang WC, Chang CL, Lo SM, Wu CY. Comparative effects of EMG-driven robot-assisted therapy versus task-oriented training on motor and daily function in patients with stroke: a randomized cross-over trial. Journal of Neuroengineering and Rehabilitation. 2022; 19(1):6 [PMC free article: PMC8762925] [PubMed: 35034664]

17.: Chen ZJ, Gu MH, He C, Xiong CH, Xu J, Huang XL. Robot-Assisted Arm Training in Stroke Individuals With Unilateral Spatial Neglect: A Pilot Study. Frontiers in neurology [electronic resource]. 2021; 12:691444 [PMC free article: PMC8297561] [PubMed: 34305798]

18.: Chen ZJ, He C, Guo F, Xiong CH, Huang XL. Exoskeleton-Assisted Anthropomorphic Movement Training (EAMT) for Poststroke Upper Limb Rehabilitation: A Pilot Randomized Controlled Trial. Archives of Physical Medicine and Rehabilitation. 2021; 102(11):2074–2082 [PubMed: 34174225]

19.: Chinembiri B, Ming Z, Kai S, Xiu Fang Z, Wei C. The fourier M2 robotic machine combined with occupational therapy on post-stroke upper limb function and independence-related quality of life: A randomized clinical trial. Topics in Stroke Rehabilitation. 2021; 28(1):1–18 [PubMed: 32434454]

20.: Conroy SS, Whitall J, Dipietro L, Jones-Lush LM, Zhan M, Finley MA et al. Effect of gravity on robot-assisted motor training after chronic stroke: a randomized trial. Archives of Physical Medicine and Rehabilitation. 2011; 92(11):1754–1761 [PMC free article: PMC4556241] [PubMed: 21849168]

21.: Coote S, Stokes EK. The effect of robot mediated therapy on upper extremity function following stroke?initial results. Irish Journal of Medical Science. 2003; 172(2):26–27

22.: Coote S, Stokes EK, Murphy BT, Harwin W. The effect of GENTLE/s robot mediated therapy on upper extremity function post stroke. [PubMed: 18441036]

23.: Coskunsu DK, Akcay S, Ogul OE, Akyol DK, Ozturk N, Zileli F et al. Effects of robotic rehabilitation on recovery of hand functions in acute stroke: a preliminary randomized controlled study. Acta Neurologica Scandinavica. 2022:499–511 [PubMed: 35855628]

24.: Daly JJ, Hogan N, Perepezko EM, Krebs HI, Rogers JM, Goyal KS et al. Response to upper-limb robotics and functional neuromuscular stimulation following stroke. Journal of Rehabilitation Research and Development. 2005; 42(6) [PubMed: 16680610]

25.: Daly JJ, Rogers J, McCabe J, Monkiewicz M, Burdsall R, Pundik S. Recovery of actual functional tasks in response to motor learning, robotics, and functional electrical stimulation.

26.: Daunoraviciene K, Adomaviciene A, Grigonyte A, Griskevicius J, Juocevicius A. Effects of robot-assisted training on upper limb functional recovery during the rehabilitation of poststroke patients. Technology and Health Care. 2018; 26(s2):533–542 [PubMed: 29843276]

27.: Dehem S, Gilliaux M, Stoquart G, Detrembleur C, Jacquemin G, Palumbo S et al. Effectiveness of upper-limb robotic-assisted therapy in the early rehabilitation phase after stroke: A single-blind, randomised, controlled trial. Annals of Physical and Rehabilitation Medicine. 2019; 62(5):313–320 [PubMed: 31028900]

28.: Drummond MF, Sculpher MJ, Torrance GW, O’Brien BJ, Stoddart GL. Methods for the economic evaluation of health care programme. Third ed. Oxford University Press. 2005.

29.: Fasoli SE, Krebs HI, Ferraro M, Hogan N, Volpe BT. Does shorter rehabilitation limit potential recovery poststroke? Neurorehabilitation and Neural Repair. 2004; 18(2):88–94 [PubMed: 15228804]

30.: Fazekas G, Horvath M, Troznai T, Toth A. Robot-mediated upper limb physiotherapy for patients with spastic hemiparesis: a preliminary study. Journal of Rehabilitation Medicine. 2007; 39(7):580–582 [PubMed: 17724559]

31.: Fernandez-Garcia C, Ternent L, Homer TM, Rodgers H, Bosomworth H, Shaw L et al. Economic evaluation of robot-assisted training versus an enhanced upper limb therapy programme or usual care for patients with moderate or severe upper limb functional limitation due to stroke: results from the RATULS randomised controlled trial. BMJ Open. 2021; 11(5):e042081 [PMC free article: PMC8154983] [PubMed: 34035087]

32.: Franceschini M, Mazzoleni S, Goffredo M, Pournajaf S, Galafate D, Criscuolo S et al. Upper limb robot-assisted rehabilitation versus physical therapy on subacute stroke patients: A follow-up study. Journal of Bodywork and Movement Therapies. 2020; 24(1):194–198 [PubMed: 31987544]

33.: Frisoli A, Barsotti M, Sotgiu E, Lamola G, Procopio C, Chisari C. A randomized clinical control study on the efficacy of three-dimensional upper limb robotic exoskeleton training in chronic stroke. Journal of Neuroengineering and Rehabilitation. 2022; 19(1):14 [PMC free article: PMC8817500] [PubMed: 35120546]

34.: Gandolfi M, Vale N, Dimitrova EK, Mazzoleni S, Battini E, Filippetti M et al. Effectiveness of Robot-Assisted Upper Limb Training on Spasticity, Function and Muscle Activity in Chronic Stroke Patients Treated With Botulinum Toxin: A Randomized Single-Blinded Controlled Trial. Frontiers in neurology [electronic resource]. 2019; 10:41 [PMC free article: PMC6365972] [PubMed: 30766508]

35.: Grigoras AV, Irimia DC, Poboroniuc MS, Popescu CD. Testing of a hybrid FES-robot assisted hand motor training program in sub-acute stroke survivors. Advances in Electrical and Computer Engineering. 2016; 16(4):89–95

36.: Gueye T, Dedkova M, Rogalewicz V, Grunerova-Lippertova M, Angerova Y. Early post-stroke rehabilitation for upper limb motor function using virtual reality and exoskeleton: equally efficient in older patients. Neurologia i Neurochirurgia Polska. 2021; 55(1):91–96 [PubMed: 33314016]

37.: Helbok R, Schoenherr G, Spiegel M, Sojer M, Brenneis C. Robot-assisted hand training (Amadeo) compared with conventional physiotherapy techniques in chronic ischemic stroke patients: a pilot study. DGNR Bremen, Nov. 2010;

38.: Hesse S, He A, Werner C C, Kabbert N, Buschfort Rd. Effect on arm function and cost of robot-assisted group therapy in subacute patients with stroke and a moderately to severely affected arm: a randomized controlled trial. Clinical Rehabilitation. 2014; 28(7):637–647 [PubMed: 24452706]

39.: Hesse S, Werner C, Pohl M, Rueckriem S, Mehrholz J, Lingnau ML. Computerized arm training improves the motor control of the severely affected arm after stroke: a single-blinded randomized trial in two centers. Stroke. 2005; 36(9):1960–1966 [PubMed: 16109908]

40.: Hollenstein C, Cabri C. Additional therapy with computer?aided training system compared to occupational therapy arm group therapy. Neuroreha. 2011; 3(1):40–42

41.: Housman SJ, Scott KM, Reinkensmeyer DJ. A randomized controlled trial of gravity-supported, computer-enhanced arm exercise for individuals with severe hemiparesis. Neurorehabilitation and Neural Repair. 2009; 23(5):505–514 [PubMed: 19237734]

42.: Hsieh Y-w, Liing R-j, Lin K-c, Wu C-y, Liou T-h, Lin J-c et al. Sequencing bilateral robot-assisted arm therapy and constraint-induced therapy improves reach to press and trunk kinematics in patients with stroke. Journal of Neuroengineering and Rehabilitation. 2016; 13(1):1–9 [PMC free article: PMC4802889] [PubMed: 27000446]

43.: Hsieh Y-w, Lin K-c, Horng Y-s, Wu C-y, Wu T-c, Ku F-l. Sequential combination of robot-assisted therapy and constraint-induced therapy in stroke rehabilitation: a randomized controlled trial. Journal of Neurology. 2014; 261(5):1037–1045 [PubMed: 24748465]

44.: Hsieh Y-w, Wu C-y, Liao W-w, Lin K-c, Wu K-y, Lee C-y. Effects of treatment intensity in upper limb robot-assisted therapy for chronic stroke: a pilot randomized controlled trial. Neurorehabilitation and Neural Repair. 2011; 25(6):503–511 [PubMed: 21436390]

45.: Hsu HY, Chiu HY, Kuan TS, Tsai CL, Su FC, Kuo LC. Robotic-assisted therapy with bilateral practice improves task and motor performance in the upper extremities of chronic stroke patients: A randomised controlled trial. Australian Occupational Therapy Journal. 2019; 66(5):637–647 [PubMed: 31317553]

46.: Hsu HY, Yang KC, Yeh CH, Lin YC, Lin KR, Su FC et al. A Tenodesis-Induced-Grip exoskeleton robot (TIGER) for assisting upper extremity functions in stroke patients: a randomized control study. Disability and Rehabilitation. 2021:1–9 [PubMed: 34586927]

47.: Hung JW, Yen CL, Chang KC, Chiang WC, Chuang IC, Pong YP et al. A Pilot Randomized Controlled Trial of Botulinum Toxin Treatment Combined with Robot-Assisted Therapy, Mirror Therapy, or Active Control Treatment in Patients with Spasticity Following Stroke. Toxins. 2022; 14(6) [PMC free article: PMC9228472] [PubMed: 35737076]

48.: Hwang CH, Seong JW, Son D-S. Individual finger synchronized robot-assisted hand rehabilitation in subacute to chronic stroke: a prospective randomized clinical trial of efficacy. Clinical Rehabilitation. 2012; 26(8):696–704 [PubMed: 22261813]

49.: Iwamoto Y, Imura T, Suzukawa T, Fukuyama H, Ishii T, Taki S et al. Combination of Exoskeletal Upper Limb Robot and Occupational Therapy Improve Activities of Daily Living Function in Acute Stroke Patients. Journal of Stroke and Cerebrovascular Diseases. 2019; 28(7):2018–2025 [PubMed: 31047819]

50.: Jiang S, You H, Zhao W, Zhang M. Effects of short-term upper limb robot-assisted therapy on the rehabilitation of sub-acute stroke patients. Technology and Health Care. 2021; 29(2):295–303 [PubMed: 33285652]

51.: Jones K, Burns A. Unit costs of health and social care 2021. Canterbury. Personal Social Services Research Unit University of Kent, 2021. Available from: https://www.pssru.ac.uk/project-pages/unit-costs/unit-costs-of-health-and-social-care-2021/

52.: Kahn LE, Averbuch M, Rymer WZ, Reinkensmeyer DJ. Comparison of robot-assisted reaching to free reaching in promoting recovery from chronic stroke.

53.: Kahn LE, Zygman ML, Rymer WZ, Reinkensmeyer DJ. Robot-assisted reaching exercise promotes arm movement recovery in chronic hemiparetic stroke: a randomized controlled pilot study. Journal of Neuroengineering and Rehabilitation. 2006; 3(1):1–13 [PMC free article: PMC1550245] [PubMed: 16790067]

54.: Kim JH, Ko MH, Park JW, Lee HJ, Nam KY, Nam YG et al. Efficacy of Electromechanically-Assisted Rehabilitation of Upper Limb Function in Post-Stroke Patients: A Randomized Controlled Study. Journal Of Rehabilitation Medicine Clinical Communications. 2021; 4:1000074 [PMC free article: PMC8628063] [PubMed: 34858561]

55.: Kim MS, Kim SH, Noh SE, Bang HJ, Lee KM. Robotic-Assisted Shoulder Rehabilitation Therapy Effectively Improved Poststroke Hemiplegic Shoulder Pain: A Randomized Controlled Trial. Archives of Physical Medicine and Rehabilitation. 2019; 100(6):1015–1022 [PubMed: 30878496]

56.: Kutner NG, Zhang R, Butler AJ, Wolf SL, Alberts JL. Quality-of-life change associated with robotic-assisted therapy to improve hand motor function in patients with subacute stroke: a randomized clinical trial. Physical Therapy. 2010; 90(4):493–504 [PMC free article: PMC2848350] [PubMed: 20185616]

57.: Lee KW, Kim SB, Lee JH, Lee SJ, Yoo SW. Effect of upper extremity robot-assisted exercise on spasticity in stroke patients. Annals of rehabilitation medicine. 2016; 40(6):961 [PMC free article: PMC5256323] [PubMed: 28119825]

58.: Lee MJ, Lee JH, Lee SM. Effects of robot-assisted therapy on upper extremity function and activities of daily living in hemiplegic patients: A single-blinded, randomized, controlled trial. Technology and Health Care. 2018; 26(4):659–666 [PubMed: 30124459]

59.: Lemmens RJM, Timmermans AAA, Janssen-Potten YJM, Pulles SA, Geers RPJ, Bakx WGM et al. Accelerometry measuring the outcome of robot-supported upper limb training in chronic stroke: a randomized controlled trial. PloS One. 2014; 9(5):e96414 [PMC free article: PMC4019639] [PubMed: 24823925]

60.: Lencioni T, Fornia L, Bowman T, Marzegan A, Caronni A, Turolla A et al. A randomized controlled trial on the effects induced by robot-assisted and usual-care rehabilitation on upper limb muscle synergies in post-stroke subjects. Scientific Reports. 2021; 11(1):5323 [PMC free article: PMC7935882] [PubMed: 33674675]

61.: Liao W-w, Wu C-y, Hsieh Y-w, Lin K-c, Chang W-y. Effects of robot-assisted upper limb rehabilitation on daily function and real-world arm activity in patients with chronic stroke: a randomized controlled trial. Clinical Rehabilitation. 2012; 26(2):111–120 [PubMed: 21840917]

62.: Lin Y, Li QY, Qu Q, Ding L, Chen Z, Huang F et al. Comparative Effectiveness of Robot-Assisted Training Versus Enhanced Upper Extremity Therapy on Upper and Lower Extremity for Stroke Survivors: a Multicentre Randomized Controlled Trial. Journal of Rehabilitation Medicine. 2022:jrm00314 [PMC free article: PMC9463635] [PubMed: 35892319]

63.: Linder SM, Rosenfeldt AB, Reiss A, Buchanan S, Sahu K, Bay CR et al. The home stroke rehabilitation and monitoring system trial: a randomized controlled trial. International Journal of Stroke. 2013; 8(1):46–53 [PMC free article: PMC4362540] [PubMed: 23280269]

64.: Lo AC, Guarino PD, Richards LG, Haselkorn JK, Wittenberg GF, Federman DG et al. Robot-assisted therapy for long-term upper-limb impairment after stroke. New England Journal of Medicine. 2010; 362(19):1772–1783 [PMC free article: PMC5592692] [PubMed: 20400552]

65.: Lum PS, Burgar CG, Shor PC, Majmundar M, Van der Loos M. Robot-assisted movement training compared with conventional therapy techniques for the rehabilitation of upper-limb motor function after stroke. Archives of Physical Medicine and Rehabilitation. 2002; 83(7):952–959 [PubMed: 12098155]

66.: Lum PS, Burgar CG, Van der Loos M, Shor PC, Majmundar M, Yap R. MIME robotic device for upper-limb neurorehabilitation in subacute stroke subjects: A follow-up study. Journal of Rehabilitation Research and Development. 2006; 43(5) [PubMed: 17123204]

67.: Ma D, Li X, Xu Q, Yang F, Feng Y, Wang W et al. Robot-Assisted Bimanual Training Improves Hand Function in Patients With Subacute Stroke: a Randomized Controlled Pilot Study. Frontiers in Neurology. 2022; 13 [PMC free article: PMC9298653] [PubMed: 35873779]

68.: Marganska VK, Blanco J, Campen K. Three-dimensional, task-specifi c robot therapy of the arm after stroke: a multicentre, parallel-group randomised tria. Lancet Neurology. 2014; 13(2):159–166 [PubMed: 24382580]

69.: Mashizume Y, Zenba Y, Takahashi K. Novel Mechanism of Action: Efficacy of Upper Extremity Robotic Therapy For Chronic Stroke Patients in Occupational Therapy. Archives of Physical Medicine and Rehabilitation. 2020; 101(11):e98

70.: Masiero S, Armani M, Ferlini G, Chiasera A, Rosati G, Rossi A. A novel robot?assisted upper?limb rehabilitation program in acute management of post?stroke patients: a randomized controlled trial. Neurorehabilitation and Neural Repair. 2012; 26(4):401

71.: Masiero S, Armani M, Ferlini G, Rosati G, Rossi A. Randomized trial of a robotic assistive device for the upper extremity during early inpatient stroke rehabilitation. Neurorehabilitation and Neural Repair. 2014; 28(4):377–386 [PubMed: 24316679]

72.: Masiero S, Armani M, Rosati G. Upper-limb robot-assisted therapy in rehabilitation of acute stroke patients: focused review and results of new randomized controlled trial. Journal of Rehabilitation Research and Development. 2011; 48(4):355–366 [PubMed: 21674388]

73.: Masiero S, Celia A, Rosati G, Armani M. Robotic-assisted rehabilitation of the upper limb after acute stroke. Archives of Physical Medicine and Rehabilitation. 2007; 88(2):142–149 [PubMed: 17270510]

74.: Mayr A, Kofler M, Saltuari L. ARMOR: an electromechanical robot for upper limb training following stroke. A prospective randomised controlled pilot study. Handchirurgie, Mikrochirurgie, Plastische Chirurgie: Organ der Deutschsprachigen Arbeitsgemeinschaft fur Handchirurgie: Organ der Deutschsprachigen Arbeitsgemeinschaft fur Mikrochirurgie der Peripheren Nerven und Gefasse: Organ der V. 2008; 40(1):66–73 [PubMed: 18322901]

75.: Mazzoleni S, Buono L, Dario P, Posteraro F. Upper limb robot-assisted therapy in subacute and chronic stroke patients: preliminary results on initial exposure based on kinematic measures.

76.: Mazzoleni S, Carrozza MC, Sale P, Franceschini M, Posteraro F, Tiboni M. Effects of upper limb robot-assisted therapy on motor recovery of subacute stroke patients: a kinematic approach. [PubMed: 24187318]

77.: McCabe J, Monkiewicz M, Holcomb J, Pundik S, Daly JJ. Comparison of robotics, functional electrical stimulation, and motor learning methods for treatment of persistent upper extremity dysfunction after stroke: a randomized controlled trial. Archives of Physical Medicine and Rehabilitation. 2015; 96(6):981–990 [PubMed: 25461822]

78.: Mehrholz J, Pohl M, Platz T, Kugler J, Elsner B. Electromechanical and robot-assisted arm training for improving activities of daily living, arm function, and arm muscle strength after stroke. Cochrane Database Syst Rev 2018, Issue DOI: 10.1002/14651858.cd006876.pub5. [PMC free article: PMC6513114] [PubMed: 30175845] [CrossRef]

79.: Merians AS, Fluet GG, Qiu Q, Yarossi M, Patel J, Mont AJ et al. Hand Focused Upper Extremity Rehabilitation in the Subacute Phase Post-stroke Using Interactive Virtual Environments. Frontiers in Neurology. 2020; 11 [PMC free article: PMC7726202] [PubMed: 33324323]

80.: National Institute for Health and Care Excellence. Developing NICE guidelines: the manual [updated January 2022]. London. National Institute for Health and Care Excellence, 2014. Available from: https://www.nice.org.uk/process/pmg20

81.: Organisation for Economic Co-operation and Development (OECD). Purchasing power parities (PPP). 2012. Available from: http://www.oecd.org/std/ppp Last accessed: 01/05/2023.

82.: Orihuela-Espina F, Rold?n GF, S?nchez-Villavicencio I, Palafox L, Leder R, Sucar LE et al. Robot training for hand motor recovery in subacute stroke patients: a randomized controlled trial. Journal of Hand Therapy. 2016; 29(1):51–57 [PubMed: 26847320]

83.: Padua L, Imbimbo I, Aprile I, Loreti C, Germanotta M, Coraci D et al. Cognitive reserve as a useful variable to address robotic or conventional upper limb rehabilitation treatment after stroke: a multicentre study of the Fondazione Don Carlo Gnocchi. European Journal of Neurology. 2020; 27(2):392–398 [PubMed: 31536677]

84.: Park JH. The effects of robot-assisted left-hand training on hemispatial neglect in older patients with chronic stroke: A pilot and randomized controlled trial. Medicine. 2021; 100(9):e24781 [PMC free article: PMC7939197] [PubMed: 33655943]

85.: Rabadi MH, Galgano M, Lynch D, Akerman M, Lesser M, Volpe BT. A pilot study of activity-based therapy in the arm motor recovery post stroke: a randomized controlled trial. Clinical Rehabilitation. 2008; 22(12):1071–1082 [PubMed: 19052246]

86.: Ranzani R, Lambercy O, Metzger JC, Califfi A, Regazzi S, Dinacci D et al. Neurocognitive robot-assisted rehabilitation of hand function: a randomized control trial on motor recovery in subacute stroke. Journal of Neuroengineering and Rehabilitation. 2020; 17(1):115 [PMC free article: PMC7444058] [PubMed: 32831097]

87.: Remy-Neris O, Le Jeannic A, Dion A, Medee B, Nowak E, Poiroux E et al. Additional, Mechanized Upper Limb Self-Rehabilitation in Patients With Subacute Stroke: The REM-AVC Randomized Trial. Stroke. 2021; 52(6):1938–1947 [PubMed: 33910364]

88.: Rodgers H, Bosomworth H, Krebs HI, van Wijck F, Howel D, Wilson N et al. Robot assisted training for the upper limb after stroke (RATULS): a multicentre randomised controlled trial. Lancet. 2019; 394(10192):51–62 [PMC free article: PMC6620612] [PubMed: 31128926]

89.: Rodgers H, Bosomworth H, Krebs HI, van Wijck F, Howel D, Wilson N et al. Robot-assisted training compared with an enhanced upper limb therapy programme and with usual care for upper limb functional limitation after stroke: the RATULS three-group RCT. Health Technology Assessment (Winchester, England). 2020; 24(54):1–232 [PMC free article: PMC7682262] [PubMed: 33140719]

90.: Sale P, Franceschini M, Mazzoleni S, Palma E, Agosti M, Posteraro F. Effects of upper limb robot-assisted therapy on motor recovery in subacute stroke patients. Journal of Neuroengineering and Rehabilitation. 2014; 11(1):1–8 [PMC free article: PMC4074149] [PubMed: 24946799]

91.: Sale P, Mazzoleni S, Lombardi V, Galafate D, Massimiani MP, Posteraro F et al. Recovery of hand function with robot-assisted therapy in acute stroke patients: a randomized-controlled trial. International Journal of Rehabilitation Research. 2014; 37(3):236–242 [PubMed: 24769557]

92.: Singh N, Saini M, Kumar N, Srivastava MVP, Mehndiratta A. Evidence of neuroplasticity with robotic hand exoskeleton for post-stroke rehabilitation: a randomized controlled trial. Journal of Neuroengineering and Rehabilitation. 2021; 18(1):76 [PMC free article: PMC8101163] [PubMed: 33957937]

93.: Straudi S, Baroni A, Mele S, Craighero L, Manfredini F, Lamberti N et al. Effects of a Robot-Assisted Arm Training Plus Hand Functional Electrical Stimulation on Recovery After Stroke: A Randomized Clinical Trial. Archives of Physical Medicine and Rehabilitation. 2020; 101(2):309–316 [PubMed: 31678222]

94.: Susanto EA, Tong RKY, Ockenfeld C, Ho NSK. Efficacy of robot-assisted fingers training in chronic stroke survivors: a pilot randomized-controlled trial. Journal of Neuroengineering and Rehabilitation. 2015; 12(1):1–9 [PMC free article: PMC4422529] [PubMed: 25906983]

95.: Takahashi K, Domen K, Sakamoto T, Toshima M, Otaka Y, Seto M et al. Efficacy of upper extremity robotic therapy in subacute poststroke hemiplegia: an exploratory randomized trial. Stroke. 2016; 47(5):1385–1388 [PubMed: 27006452]

96.: Takebayashi T, Takahashi K, Amano S, Gosho M, Sakai M, Hashimoto K et al. Robot-Assisted Training as Self-Training for Upper-Limb Hemiplegia in Chronic Stroke: a Randomized Controlled Trial. Stroke. 2022; 53(7):2182–2191 [PubMed: 35345897]

97.: Taravati S, Capaci K, Uzumcugil H, Tanigor G. Evaluation of an upper limb robotic rehabilitation program on motor functions, quality of life, cognition, and emotional status in patients with stroke: a randomized controlled study. Neurological Sciences. 2021; 11:11 [PubMed: 34247295]

98.: Taveggia G, Borboni A, Salvi L, Mul C, Fogliaresi S, Villafa?e JH et al. Efficacy of robot-assisted rehabilitation for the functional recovery of the upper limb in post-stroke patients: a randomized controlled study. European Journal of Physical and Rehabilitation Medicine. 2016; 52(6):767–773 [PubMed: 27406879]

99.: Timmermans AAA, Lemmens RJM, Monfrance M, Geers RPJ, Bakx W, Smeets RJEM et al. Effects of task-oriented robot training on arm function, activity, and quality of life in chronic stroke patients: a randomized controlled trial. Journal of Neuroengineering and Rehabilitation. 2014; 11(1):1–12 [PMC free article: PMC3973831] [PubMed: 24684808]

100.: Tomi TJD, Savi AM, Vidakovi AS, Rodi SZ, Isakovi MS, Rodr?guez-de-Pablo C et al. ArmAssist robotic system versus matched conventional therapy for poststroke upper limb rehabilitation: a randomized clinical trial. BioMed Research International. 2017; 2017 [PMC free article: PMC5306984] [PubMed: 28251157]

101.: Valles KB, Montes S, de Jesus Madrigal M, Burciaga A, Mart?nez MaE, Johnson MJ. Technology-assisted stroke rehabilitation in Mexico: a pilot randomized trial comparing traditional therapy to circuit training in a Robot/technology-assisted therapy gym. Journal of Neuroengineering and Rehabilitation. 2016; 13(1):1–15 [PMC free article: PMC5025604] [PubMed: 27634471]

102.: Vanoglio F, Bernocchi P, Mul C, Garofali F, Mora C, Taveggia G et al. Feasibility and efficacy of a robotic device for hand rehabilitation in hemiplegic stroke patients: a randomized pilot controlled study. Clinical Rehabilitation. 2017; 31(3):351–360 [PubMed: 27056250]

103.: Villafa?e JH, Taveggia G, Galeri S, Bissolotti L, Mull C, Imperio G et al. Efficacy of short-term robot-assisted rehabilitation in patients with hand paralysis after stroke: a randomized clinical trial. Hand. 2018; 13(1):95–102 [PMC free article: PMC5755871] [PubMed: 28719996]

104.: Volpe BT, Krebs HI, Hogan N, Edelstein L, Diels C, Aisen M. A novel approach to stroke rehabilitation: robot-aided sensorimotor stimulation. Neurology. 2000; 54(10):1938–1944 [PubMed: 10822433]

105.: Volpe BT, Lynch D, Rykman-Berland A, Ferraro M, Galgano M, Hogan N et al. Intensive sensorimotor arm training mediated by therapist or robot improves hemiparesis in patients with chronic stroke. Neurorehabilitation and Neural Repair. 2008; 22(3):305–310 [PMC free article: PMC4943019] [PubMed: 18184932]

106.: Wolf SL, Sahu K, Bay RC, Buchanan S, Reiss A, Linder S et al. The HAAPI (Home Arm Assistance Progression Initiative) trial: a novel robotics delivery approach in stroke rehabilitation. Neurorehabilitation and Neural Repair. 2015; 29(10):958–968 [PMC free article: PMC4573760] [PubMed: 25782693]

107.: Wu C-y, Chen M-D, Chen Y-t, Wu L-L, Lin K-c. Unilateral and Bilateral Robot-Assisted Arm Training Had Differential Effects on Upper Limb Function in Chronic Stroke Survivors.

108.: Wu C-y, Chuang L-l, Chen M-D, Chen Y-t, Lin K-c. Abstract P289: Therapist-Based and Robot-Assisted Physical Training Have Differential Effects on Motor Control of Upper Limb and Quality of Life after Chronic Stroke. 2012;

109.: Wu C-y, Yang C-l, Chuang L-l, Lin K-c, Chen H-c, Chen M-d et al. Effect of therapist-based versus robot-assisted bilateral arm training on motor control, functional performance, and quality of life after chronic stroke: a clinical trial. 2012; 92:1006–1016 [PubMed: 22517782]

110.: Xu Q, Li C, Pan Y, Li W, Jia T, Li Z et al. Impact of smart force feedback rehabilitation robot training on upper limb motor function in the subacute stage of stroke. NeuroRehabilitation. 2020; 47(2):209–215 [PubMed: 32741790]

111.: Yoo DH, Cha YJ, kyoung Kim S, Lee JS. Effect of three-dimensional robot-assisted therapy on upper limb function of patients with stroke. Journal of Physical Therapy Science. 2013; 25(4):407–409

112.: Zengin-Metli D, Ozbudak-Demir S, Eraktas I, Binay-Safer V, Ekiz T. Effects of robot assistive upper extremity rehabilitation on motor and cognitive recovery, the quality of life, and activities of daily living in stroke patients. Journal of Back and Musculoskeletal Rehabilitation. 2018; 31(6):1059–1064 [PubMed: 29966188]

Appendices

Appendix I. Health economic model

Health economic modelling was not undertaken for this review.

Appendix J. Excluded studies

Clinical studies

Table 11

Studies excluded from the clinical review.

Health Economic studies

Download PDF (157K)

Final

Evidence reviews underpinning recommendation 1.13.18 in the NICE guideline

These evidence reviews were developed by NICE

Disclaimer: The recommendations in this guideline represent the view of NICE, arrived at after careful consideration of the evidence available. When exercising their judgement, professionals are expected to take this guideline fully into account, alongside the individual needs, preferences and values of their patients or service users. The recommendations in this guideline are not mandatory and the guideline does not override the responsibility of healthcare professionals to make decisions appropriate to the circumstances of the individual patient, in consultation with the patient and/or their carer or guardian.

Local commissioners and/or providers have a responsibility to enable the guideline to be applied when individual health professionals and their patients or service users wish to use it. They should do so in the context of local and national priorities for funding and developing services, and in light of their duties to have due regard to the need to eliminate unlawful discrimination, to advance equality of opportunity and to reduce health inequalities. Nothing in this guideline should be interpreted in a way that would be inconsistent with compliance with those duties.

NICE guidelines cover health and care in England. Decisions on how they apply in other UK countries are made by ministers in the Welsh Government, Scottish Government, and Northern Ireland Executive. All NICE guidance is subject to regular review and may be updated or withdrawn.

Bookshelf ID: NBK601177PMID: 38442218

Table 1PICO characteristics of review question

Population	Inclusion: Adults (age ≥16 years) who have had a first or recurrent stroke (including people after subarachnoid haemorrhage). Exclusion: Children (age <16 years) People who had a transient ischaemic attack
Intervention	Robot-assisted arm training (all types pooled together)
Comparison	Any other intervention (including usual care and no treatment – all comparators pooled together) Confounding factors (for non-randomised studies only): Presence of comorbidities Stroke severity Time period since stroke
Outcomes	All outcomes are considered equally important for decision making and therefore have all been rated as critical: At the following time periods: Post-intervention (outcomes reported immediately after the intervention has finished). ≥6 months (the longest time period will be used for this outcome. If the outcome is less than 6 months, then it will be included but downgraded for indirectness). Person/participant generic health-related quality of life (continuous outcomes will be prioritised) Carer generic health-related quality of life (continuous outcomes will be prioritised) Activities of daily living (continuous outcomes will be prioritised) Arm function (continuous outcomes will be prioritised) Arm muscle strength (continuous outcomes will be prioritised) Spasticity (continuous outcomes prioritised) Stroke-specific Patient-Reported Outcome Measures (continuous outcomes will be prioritised) Withdrawal for any reason (dichotomous outcome) Adverse events (dichotomous outcomes) Cardiovascular events Injuries and pain Other reported adverse events
Study design	Systematic reviews of RCTs Parallel RCTs Cross over trials (only the first study period will be included) Non-randomised studies (if insufficient RCT evidence is available) Prospective cohort studies Retrospective cohort studies Case-control studies Published NMAs and IPDs will be considered for inclusion.

Table 2Summary of studies included in the evidence review

Study

Intervention and comparison

Population

Outcomes

Comments

Abdollahi 2018¹

Robot-assisted arm training (n=12)

Three 45-minute sessions per week (six sessions total). Duration 2 weeks.

Region of upper limb trained: Not stated/unclear

Level of supervision: Not stated/unclear

Type of movement delivered by robotic device: Not stated/unclear

Any other intervention (n=10)

Bilateral training without error augmentation

Three 45- minute sessions per week (six sessions total). Duration 2 weeks.

Concomitant therapy: Not reported.

People after a first or recurrent stroke

Mean age: 53.86 years

N = 26

Time after stroke: Chronic (≥6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Withdrawal for any reason at post-intervention

Setting: Outpatient rehabilitation hospital in the United States of America.

Sources of funding: Not reported.

Abdullah 2011²

Robot-assisted arm training (n=9)

45 minutes, 3 times a week for 8-11 weeks.

Region of upper limb trained: Proximal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Starting with passive and moving up to active assisted

Any other intervention (n=11)

Conventional arm therapy 45 minutes, 3 times a week for 8-11 weeks

Concomitant therapy: not available*

People after a first or recurrent stroke

Mean age: Not available*

N = 20

Time after stroke: Subacute (7 days - 6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Amirabdollahian 2007³

Subsidiary papers:

Coote²²

Coote 2003²¹

Robot-assisted arm training (n=16)

ABC - 3 weeks at baseline (phase A), then 3 weeks robot-mediated therapy (phase B) then 3 weeks sling suspension (phase C). Follow up at 6 weeks.

Region of upper limb trained: Proximal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Mixed

Any other intervention (conventional arm therapy) (n=11)

Sling suspension.

ACB - 3 weeks at baseline (phase A), then 3 weeks sling suspension (phase C), then 3 weeks robot-mediated therapy. Follow up at 6 weeks.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: Not available*

N = 31

Time after stroke: Chronic (≥6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Withdrawal for any reason at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Crossover study: First time period included only as per the original Cochrane review.

Ang 2014⁴

Robot-assisted arm training (n=15)

With the haptic knob robot with and without a brain computer interface. Total of 18 sessions over 6 weeks, 3 times per week, 90 min per session.

Region of upper limb trained: Mixed

Level of supervision: Supervised

Type of movement delivered by robotic device: Passive

Any other intervention (n=7)

Standard arm therapy. Total of 18 sessions over 6 weeks, 3 times per week, 90 min per session.

Concomitant therapy: therapist-assisted arm mobilisation for 30 minutes.

People after a first or recurrent stroke

Mean age: Not available*

N= 22

Mean time after stroke: Mixed >4 months

Subacute and Chronic.

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post intervention and ≥6 months

Withdrawal for any reason at post intervention and ≥6 months

Adverse events at post intervention and ≥6 months

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Aprile 2020⁵

Subsidiary papers:

Aprile 2021⁶

Padua 2020⁸³

Robot-assisted arm training (n=123)

Performed daily for 45 minutes, 5 days a week, for a total of 30 sessions.

Region of upper limb trained: Mixed.

Level of supervision: Supervised

Type of movement delivered by robotic device: Mixed

Any other intervention (n=124)

Conventional treatment performed daily for 45 minutes, 5 days a week, for a total of 30 sessions.

Concomitant therapy: Conventional rehabilitation sessions (6 times/week), lasting 45 minutes

Occupational and speech therapy were provided, if needed.

People after a first or recurrent stroke

Mean age (SD): 69.0 (11.2) years

N = 247

Time after stroke: Subacute (7 days - 6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Person/participant generic health related quality of life at post-intervention

Activities of daily living at post-intervention

Arm function at post-intervention

Arm muscle strength at post-intervention

Withdrawal for any reason at post-intervention

Setting: 8 rehabilitation centres of the Fondazione Don Carlo Gnocchi, in Italy.

Sources of funding: Not reported.

Bishop 2014⁸

Subsidiary study:

Helbok 2010³⁷

Robot-assisted arm training (n=16)

Robot therapy with the Amadeo Hand robot three times per week for eight weeks, for 60 minutes.

Region of upper limb trained: Distal limb

Level of supervision: Not stated/unclear

Type of movement delivered by robotic device: Not stated/unclear

Any other intervention (n=15)

Standard arm therapy for three times per week for eight weeks, for 60 minutes.

Concomitant therapy: No additional information.

People after a first or recurrent stroke

Mean age: Not stated/unclear

N = 31

Time after stroke: Chronic (at least 6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention

Arm function at post-intervention

Arm muscle strength at post-intervention

Withdrawal for any reason at post-intervention

Setting: United States of America and Austria.

Sources of funding: No additional information.

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Brokaw 2014⁹

Robot-assisted arm training (n=7)

Group AB: 12 hours of robotic training within a month (A) and 12 hours of conventional therapy within a month (B), separated by a month of wash-out period.

Region of upper limb trained: Mixed

Level of supervision: Supervised

Type of movement delivered by robotic device: Active-assisted

Any other intervention (n=5)

Group BA: 12 hours of conventional therapy within a month (B), and 12 hours of robotic training within a month (A) separated by a month of wash-out period.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: Not available*

N = 12

Time after stroke: Chronic (≥6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention

Withdrawal for any reason at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Crossover study: First time period included only.

Budhota 2021¹⁰

Robot-assisted arm training (n=22)

Robotic therapy with H-Man: 18 training sessions of 60 min each, followed by a 30 min 1:1 conventional therapy session three times a week and over a span of 6 weeks.

Region of upper limb trained: Not stated/unclear.

Level of supervision: Supervised

Type of movement delivered by robotic device: Not stated/unclear

Any other intervention (n=22)

Conventional therapy: 18 training sessions of 90 min each, three times a week and over a span of 6 weeks

Concomitant therapy: None reported.

People after a first or recurrent stroke

Mean age (SD): 55.5 (10.7) years

N = 44

Median time after stroke (IQR):

Intervention: 458 (451.3) days,

Control: 390 (327.5) days

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention and ≥6 months

Arm muscle strength at post-intervention and ≥6 months

Withdrawal for any reason at post-intervention and ≥6 months

Adverse events at post-intervention

Setting: conducted at the outpatient clinic of the Tan Tock Seng Hospital, Centre for Advanced Rehabilitation Therapeutics (TTSH-CART), Singapore.

Sources of funding: Supported by the National Medical Research Council (NMRC, NMRCB2b0006c) Singapore and the H-Man project (NMRC/BnB/0006b/2013), Ministry of Health, Singapore; Ageing Research Institute for Society and Education (ARISE), Singapore: M4082063 and Interdisciplinary Graduate School, Nanyang Technological University, Singapore.

Burgar 2011¹¹

Robot-assisted arm training (n=36)

15 x1 hour therapy sessions over a 3-week period (1 robot group received 30 1- hour therapy sessions over 3- week period).

Region of upper limb trained: Proximal

Level of supervision: Supervised

Type of movement delivered by robotic device: Mixed

Any other intervention (n=18)

15 x1 hour therapy sessions over a 3-week period

Concomitant therapy: None reported.

People after a first or recurrent stroke

Mean age (SD): Not available*

N = 54

Mean time after stroke: 11 days

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention and ≥6 months.

Arm function at post-intervention and ≥6 months.

Arm muscle strength at post-intervention and ≥6 months.

Spasticity at post-intervention and ≥6 months.

Withdrawal for any reason at post- intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Calabro 2019¹⁴

Robot-assisted arm training (n=25)

40 hand training sessions of 45 minutes each, 5 times a week, for 8 consecutive weeks.

Region of upper limb trained: Not stated/ unclear

Level of supervision: Supervised

Type of movement delivered by robotic device: Mixed

Any other intervention (n=25)

40 hand training sessions of 45 minutes each, 5 times a week, for 8 consecutive weeks.

Concomitant therapy: The patients were asked not to undertake other physiotherapy treatments during the 8-week training period.

People after a first or recurrent stroke

Mean age (SD): 64.5 (3.0) years

N = 50

Mean time after stroke (SD): 10 (2) months

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention

Withdrawal for any reason at post-intervention

Adverse events at post-intervention

Setting: In-patient, at the Neuro-robotic Rehabilitation Unit of the IRCCS Centro Neurolesi Bonino Pulejo, Italy.

Sources of funding: No funding.

Carpinella 2020¹⁵

Subsidiary paper:

Lencioni 2021⁶⁰

Robot-assisted arm training (n=20)

Using a planar robotic manipulandum 20 sessions of 45 minutes each, 5 times a week.

Region of upper limb trained: Proximal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Active assisted movement

Any other intervention (n=20)

Usual care arm-specific physiotherapy 20 sessions of 45 minutes each, 5 times a week.

Concomitant therapy: Participants in both groups received a rehabilitation treatment for the affected upper limb consisting of 20 sessions of 45 min each, 5 times a week by trained physiotherapists.

People after a first or recurrent stroke

Median age (IQR):

Intervention: 67 (58 to 70) years

Control: 59 (46 to 69) years

N = 40

Median time after stroke (IQR): intervention: 7 (1.7 to 11.9) months, control: 5.3 (1.9 to 89.6) months

Ethnicity: Not stated/unclear

Severity: Mild (or NIHSS 1-5)

Activities of daily living at post-intervention

Arm function at post-intervention

Withdrawal for any reason at post-intervention

Setting: 2 stroke rehabilitation hospitals in Italy.

Sources of funding: supported by the Italian Ministry of Health (Ricerca Corrente and Ricerca Finalizzata: grant no. GR-2011-02348942).

Chen 2022¹⁶

Robot-assisted arm training (n=16)

Hand of Hope robotic hand system. 12 sessions of robot-assisted intervention first, followed by a 1-month washout period, then 12 sessions (3 sessions per week for 4 consecutive weeks) of task-oriented interventions.

Region of upper limb trained: Distal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Mixed

Any other intervention (n=15)

Usual care. Task-oriented interventions. 12 sessions (3 sessions per week for 4 consecutive weeks). After which they had a 1-month washout period and then participated in 12 sessions of robot assisted arm training.

Concomitant therapy: No additional information.

People after a first or recurrent stroke

Mean age (SD): 59.6 (11.0) years

N = 31

Mean time after stroke (SD): 48.1 (40.6) months

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention

Withdrawal for any reason at post-intervention

Setting: Outpatient follow-up in Taiwan.

Sources of funding: This study was supported by Chang Gung Memorial Hospital (BMRP553, CMRPD1I0033), the Ministry of Science and Technology (MOST 109-2314-B-192-027-MY3) and Healthy Aging Research Center, Chang Gung University from the Featured Areas Research Center Program within the Framework of the Higher Education Sprout Project by the Ministry of Education in Taiwan (EMRPD1L0411).

Crossover study: First time period only included.

Chen 2021¹⁷

Robot-assisted arm training (n=10)

(Armule®, Intelbot intelligent machine Co., Ltd, Wuhan, China) 45 minutes daily, 5 days/week for 4 weeks.

Region of upper limb trained: Not stated/unclear

Level of supervision: Supervised

Type of movement delivered by robotic device: Mixed

Any other intervention (n=10)

Conventional therapy 45 min daily, 5 days/week for 4 weeks.

Concomitant therapy: Conventional rehabilitation programs continued as usual for all participants.

People after a first or recurrent stroke

Mean age (SD): 47.4 (8.47) years

N = 20

Mean time after stroke (SD): 91.8 (74.2) days

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention

Arm function at post-intervention

Withdrawal for any reason at post-intervention

Adverse events at post-intervention

Setting: Department of Rehabilitation Medicine, China.

Sources of funding: received financial support for the research and publication of this article from National Natural Science Foundation of China (U 1913601 and No. 91648203).

This study is reported in forest plots as Chen 2021A.

Chen 2021¹⁸

Robot-assisted arm training (n=10)

Exoskeleton-assisted anthropomorphic movement training

45 min daily, 5 days/week for 4 weeks.

Region of upper limb trained: Not stated/unclear

Level of supervision: Supervised

Type of movement delivered by robotic device: Mixed

Any other intervention (n=10)

Conventional therapy 45 min daily, 5 days/week for 4 weeks.

Concomitant therapy: All participants received routine multidisciplinary treatment, including medication and usual poststroke care.

People after a first or recurrent stroke

Mean age (SD): 51 (13.5) years

N = 20

Mean time after stroke (SD): 62.5 (48.7) days

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention

Arm function at post-intervention

Withdrawal for any reason at post-intervention

Adverse events at post-intervention

This study is reported in forest plots as Chen 2021B.

Setting: Department of Rehabilitation Medicine in China.

Sources of funding: Supported by the National Natural Science Foundation of China (grant nos. U 1913601, 91648203).

Chinembiri 2021¹⁹

Robot-assisted arm training (n=25)

Robot and occupational therapy. 50-70 minutes per day, 5 days a week for 6 weeks.

Region of upper limb trained: Mixed

Level of supervision: Supervised

Type of movement delivered by robotic device: Mixed

Any other intervention (n=25)

Usual care. Training involving self-range of motion and passive stretch exercises for the shoulder, elbow, wrist and thumb joints, and muscles (five sets of repetitions) for the first 10 minutes, then a larger selection of upper limb exercises for the next 40 minutes.

Concomitant therapy: Both groups received 30 training sessions lasting 50 minutes per day (for the control group and lower end of the intervention group), 5 days a week for a total of 6 weeks.

People after a first or recurrent stroke

Mean age (SD): 57.5 (8.4) years

N = 50

Mean time after stroke: Not stated/unclear

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention

Arm function at post-intervention

Withdrawal for any reason at post-intervention

Adverse events at post-intervention

Setting: Outpatients in China.

Sources of funding: Supported by the Jiangsu Provincial Medical Youth Talent under Grant (number QNRC2016376).

Conroy 2011²⁰

Robot-assisted arm training (n=41)

Over 6 weeks, 3 sessions per week for 1 hour.

Region of upper limb trained: Not stated/unclear

Level of supervision: Not stated/unclear

Type of movement delivered by robotic device: Active assisted movement

Any other intervention (n=21)

Intensive conventional arm exercise.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: not available*

N = 62

Mean (SD) time after stroke: Chronic (≥6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention and ≥6 months

Arm function at post-intervention and ≥6 months

Withdrawal for any reason at post-intervention and ≥6 months

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Coskunsu 2022²³

Robot-assisted arm training (n=12)

1 hour, 5 days per week for 3 weeks.

Region of upper limb trained: Distal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Mixed

Any other intervention (n=12)

Usual care.

Concomitant therapy: Everyone received rehabilitation exercises (including physiotherapy) for 1 hour (30 minutes for the upper extremity, 30 minutes for the lower extremity).

People after a first or recurrent stroke

Mean age (SD): 65.0 (15.0) years

N = 24

Time after stroke: Subacute (7 days – 6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention

Withdrawal for any reason at post-intervention

Setting: Inpatients in Turkey

Funding: Supported by the Rehab Robotic Company.

Daly 2005²⁴

Robot-assisted arm training (n = 7)

5 hours per day, 5 days per week for 12 weeks. 1.5 hours per day for robotic shoulder and elbow training.

Region of upper limb trained: Proximal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Active assisted movement

Any other intervention (n=6)

5 hours per day, 5 days per week for 12 weeks. 1.5 hours per day for functional electrical stimulation.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: Not available*

N = 13

Time after stroke: Chronic (≥6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention

Withdrawal for any reason at post-intervention

Adverse events at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Daunoraviciene 2018²⁶

Robot-assisted arm training (n=17)

Armeo Spring training for 30 minutes a day for 10 sessions (5 days a week).

Region of upper limb trained: Not stated/unclear

Level of supervision: Supervised

Type of movement delivered by robotic device: Not stated/unclear

Any other intervention (n=17)

30 minutes on 5 days a week of conventional functional rehabilitation.

Concomitant therapy: Conventional functional rehabilitation for 35-60 minutes/day in approximately 10 occupational therapy sessions.

People after a first or recurrent stroke

Mean age (SD): 65.7 (4.5) years

N = 34

Mean time after stroke (SD): 9.1 (5.1) weeks

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention

Arm function at post-intervention

Spasticity at post-intervention

Setting: Outpatients in Lithuania.

Sources of funding: No additional information.

Dehem 2019²⁷

Robot-assisted arm training (n=23)

REAplan(R) robot arm therapy. 45 minutes sessions supervised by a therapist. 4 sessions of conventional therapy per week was replaced and was completed for 9 weeks in total.

Region of upper limb trained: Distal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Active assisted movement

Any other intervention (n=22)

Conventional therapy focused on motor rehabilitation, matched with their personal needs and centre means.

Concomitant therapy: Both groups underwent their rehabilitation sessions during their hospitalisation with their regular physical therapists and occupational therapists.

People after a first or recurrent stroke

Mean age (SD): 67.9 (15.5) years

N = 45

Mean time after stroke (SD): Not reported

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention and at ≥6 months

Stroke-specific Patient-Reported Outcome Measures at post-intervention and at ≥6 months

Withdrawal for any reason at post-intervention and at ≥6 months

Adverse events at post-intervention and at ≥6 months

Setting: Three inpatient rehabilitation centres: Cliniques universitaries Saint-Luc (Brussels), Centre Hospitalier Valida (Brussels) and Centre Hospitalier Neurologique William Lennox (Ottignies) in Belgium.

Sources of funding: This work was supported by the Region Wallone, the Fondation Motrice and the Fondation Saint-Luc. The authors thank Axinesis (Wavre, Belgium) for development of the robot REAplan and Fishing Cactus (Mons, Belgium) for development of the game.

Fazekas 2007³⁰

Robot-assisted arm training (n=15)

30 minutes of Bobath therapy sessions on 20 consecutive days, plus an additional 30 minutes of robot therapy.

Region of upper limb trained: Proximal limb

Level of supervision: Not stated/unclear

Type of movement delivered by robotic device: Passive

Any other intervention (n=15)

30 minutes of Bobath therapy sessions on 20 consecutive days.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: Not available*

N = 30

Time after stroke: Chronic (>6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention

Arm function at post-intervention

Spasticity at post-intervention

Withdrawal for any reason at post-intervention

Adverse events at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Frisoli 2022³³

Robot-assisted arm training (n=13)

Robot therapy for 45 minutes, 3 times a week for 6 weeks.

Region of upper limb trained: Mixed

Level of supervision: Supervised

Type of movement delivered by robotic device: Active assisted movement

Any other intervention (n=13)

Manual rehabilitation including passive movement, goal directed movement and voluntary action for a matched amount of time.

Concomitant therapy: No additional information

People after a first or recurrent stroke

Mean age (SD): 66 (12) years

N = 26

Mean time after stroke (SD): 34 (22) months

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention

Spasticity at post-intervention

Withdrawal for any reason at post-intervention

Setting: Outpatient follow up in Italy.

Funding: Partially funded by SKILLS EU FP7 project. One author is funded by a postgraduate fellowship.

Gandolfi 2019³⁴

Robot-assisted arm training (n=16)

Passive mobilisation and stretching exercises for the affected upper limb (10 minutes) followed by robot-assisted exercises (35 minutes). 2 sessions per week for 5 consecutive weeks.

Region of upper limb trained: Mixed

Level of supervision: Supervised

Type of movement delivered by robotic device: Mixed

Any other intervention (n=16)

Conventional training consisting of upper limb passive mobilisation and stretching (10 minutes) followed by upper limb exercises (35 minutes), 2 sessions per week for 5 consecutive weeks.

Concomitant therapy: All people received botulinum toxin A treatment. The dose, volume and number of injection sites were set according to the severity of spasticity.

People after a first or recurrent stroke

Mean age (SD): 59.2 (14.7) years

N = 32

Mean time after stroke (SD): 5.6 (2.7) years

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention

Arm muscle strength at post-intervention

Spasticity at post-intervention

Withdrawal for any reason at post-intervention

Setting: People referred to the Neurorehabilitation Unit (AOUI Verona) and the Physical Medicine and Rehabilitation Section, “OORR” Hospital (University of Foggia) in Italy.

Sources of funding: No additional information.

Grigoras 2016³⁵

Robot-assisted arm training (n=13)

With hybrid FES exoskeleton system for hand rehabilitation. 12 sessions of 30 minutes for 2 weeks.

Region of upper limb trained: Distal

Level of supervision: Supervised

Type of movement delivered by robotic device: Active assisted movement

Any other intervention (n=12)

10 sessions of standard arm therapy (30 minutes for 2 weeks.)

Concomitant therapy: No information*

People after a first or recurrent stroke

Mean age: No information*

N = 25

Time after stroke: Subacute (7 days - 6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention

Stroke-specific Patient-Reported Outcome Measures at post-intervention

Withdrawal for any reason at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Gueye 2021³⁶

Robot-assisted arm training (n=25)

Virtual reality robot-assisted arm therapy using an Armeo Spring device: 45 minute sessions for 12 sessions over a three week period (4 sessions per week).

Region of upper limb trained: Not stated/unclear

Level of supervision: Not stated/unclear

Type of movement delivered by robotic device: Not stated/unclear

Any other intervention (n=25)

Usual care. An additional 45 minutes of physiotherapy for 12 sessions over a three week period (4 sessions per week).

Concomitant therapy: The programme consists of at least 3-4 hours of activity which includes one hour of physiotherapy twice a day, occupational therapy, therapies using passive or motor splints and individual or group therapy for speech and cognitive impairment.

People after a first or recurrent stroke

Mean age (SD): 67.3 (12.1) years

N = 50

Mean time after stroke (SD): 15.6 (6.9) days

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention

Arm function at post-intervention

Withdrawal for any reason at post-intervention

Setting: outpatients at the Stroke Rehabilitation Unit of the General University Hospital in Prague in the Czech Republic.

Sources of funding: No additional information.

Hesse 2014³⁸

Robot-assisted arm training (n=25)

Robot-assisted group therapy for 30 minutes plus individual arm therapy for 30 minutes, each workday for 4 weeks.

Region of upper limb trained: Distal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Mixed

Any other intervention (n=25)

Individual arm therapy for 2 × 30 minutes each workday for 4 weeks.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: Not available*

N = 50

Time after stroke: Subacute (7 days - 6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention and ≥6 months

Arm function at post-intervention and ≥6 months

Arm muscle strength at post-intervention and ≥6 months

Spasticity at post-intervention and ≥6 months

Withdrawal for any reason at post-intervention and ≥6 months

Adverse events at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Hesse 2005³⁹

Robot-assisted arm training (n=22)

Therapy with Bi-Manu Track robotic arm trainer 5 times a week for 6 weeks.

Region of upper limb trained: Distal limb

Level of supervision: Unsupervised

Type of movement delivered by robotic device: Mixed

Any other intervention (n=22)

Functional electrical stimulation (if possible EMG-initiated) for wrist extension 5 times a week for 6 weeks.

Concomitant therapy: Standard rehabilitation programme.

People after a first or recurrent stroke

Mean age: Not available*

N = 44

Time after stroke: Subacute (7 days - 6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention and ≥6 months

Arm muscle strength at post-intervention and ≥6 months

Spasticity at post-intervention and ≥6 months

Withdrawal for any reason at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Hollenstein 2011⁴⁰

Robot-assisted arm training (n=7)

With the Armeo device 5 times a week for 30 minutes over 2 weeks (10 times).

Region of upper limb trained: Proximal limb

Level of supervision: Not stated/unclear

Type of movement delivered by robotic device: Not stated/unclear

Any other intervention (n=6)

Arm group programme. Without device delivered by an occupational therapist for the same time and frequency as the robot therapy group.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: not available*

N = 13

Mean time after stroke: Not stated/unclear

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Crossover study: first time period only included.

Housman 2009⁴¹

Robot-assisted arm training (n=17)

With T-WREX device 3 times a week for 1 hour over 8-9 weeks.

Region of upper limb trained: Proximal limb

Level of supervision: Mixed

Type of movement delivered by robotic device: Passive movement

Any other intervention (n=17)

As above, but without the device.

Concomitant therapy:

Not available*

People after a first or recurrent stroke

Mean age: Not available*

N = 34

Time after stroke: Chronic (≥6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention and ≥6 months

Arm function at post-intervention and ≥6 months

Arm muscle strength at post-intervention and ≥6 months

Withdrawal for any reason at post-intervention and ≥6 months

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Hsieh 2014⁴³

Subsidiary paper:

Hsieh 2016⁴²

Robot-assisted arm training (n=32)

robot-assisted arm therapy (Bi-Manu-Track) with and without constraint-induced therapy-20 training sessions of 90 to 105 min/day, 5 days/ week for 4 weeks.

Region of upper limb trained: Distal limb

Level of supervision: Not stated/unclear

Type of movement delivered by robotic device: Mixed

Any other intervention (n=16)

Conventional therapy- 20 training sessions of 90 to 105 min/day, 5 days/ week for 4 weeks.

Concomitant therapy:

Not available*

People after a first or recurrent stroke

Mean age: Not available*

N = 48

Time after stroke: Chronic (≥6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention

Withdrawal for any reason at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Hsieh 2011⁴⁴

Robot-assisted arm training (n=12)

High/low intensity robot training with Bi-Manu Track.

20 training sessions for 90 to 105 minutes, 5days per week for 4 weeks. Participants also received 15-20 minutes of functional activities training.

Region of upper limb trained: Distal limb

Level of supervision: Not stated/unclear

Type of movement delivered by robotic device: Mixed

Any other intervention (n=6)

Participants received a structured protocol using conventional occupational therapy techniques.

Concomitant therapy:

Not available*

People after a first or recurrent stroke

Mean age: not available*

N = 18

Time after stroke: Chronic (≥6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention

Arm function at post-intervention

Arm muscle strength at post-intervention

Withdrawal for any reason at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Hsu 2019⁴⁵

Robot-assisted arm training (n=22)

With Bi-Manu Track three times per week for four weeks.

Region of upper limb trained: Distal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Mixed

Any other intervention (n=21)

40 minutes of therapist-facilitated task-specific training for the affected limb.

Concomitant therapy: All people received a 10-minute per-protocol sensorimotor stimulation session prior to the interventions as part of usual care.

People after a first or recurrent stroke

Mean age (SD): 52.9 (13.2) years

N = 43

Mean time after stroke (SD): 14.2 (11.1) months

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention and ≥6 months

Withdrawal for any reason at post-intervention and ≥6 months

Adverse events at post-intervention and ≥6 months

Setting: Outpatients in Taiwan.

Sources of funding: This work was supported by Chi Mei Medical Center and National Cheng Kung University under grant #CMNCKU10304. This work was also financially supported by the Medical Device Innovation Center, National Cheng Kung University from the Featured Areas Research Center Program within the framework of the Higher Education Sprout Project by the Ministry of Education in Taiwan.

Hsu 2021⁴⁶

Robot-assisted arm training (n=17)

An additional 20-minutes of robot-assisted arm training using TIGER (Tenodesis-induced-grip exoskeleton robot) Two sessions of training a week for 9 weeks.

Region of upper limb trained: Distal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Mixed

Any other intervention (n=17)

An additional 20-minutes of task-specific motor training through regular occupational therapy.

Concomitant therapy: All people received 20-minutes of regular task-specific motor training.

People after a first or recurrent stroke

Mean age (SD): 55.9 (15.0) years

N = 34

Mean time after stroke (SD): 30.0 (24.5) months

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention and ≥6 months

Spasticity at post-intervention and ≥6 months

Withdrawal for any reason at post-intervention and ≥6 months

Adverse events at post-intervention and ≥6 months

Setting: Outpatients in Taiwan.

Sources of funding: Financially supported by the Medical Device Innovation Center, National Cheng Kung University, from the Featured Areas Research Center Program within the framework of the Higher Education Sprout Project by the Ministry of Education in Taiwan. This project was supported in part by the Ministry of Science and Technology, Taiwan, under Grant MOST 108-2745-8-006-009 and in part by the National Cheng Kung University Hospital, Tainan, Taiwan under Grant NCKUH 10708003.

Hung 2022⁴⁷

Robot-assisted arm training (n=13)

75 minutes of training, 3 times a week for 8 weeks, with 45 minutes of the training being with a robot, and 30 minutes being functional activities practice.

Region of upper limb trained: Mixed

Level of supervision: Supervised

Type of movement delivered by robotic device: Mixed

Any other intervention (n=24)

Two groups combined. One received mirror therapy for 45 minutes and 30 minutes of functional activities practice. The other received conventional task-oriented practice for 45 minutes, and 30 minutes of functional activities practice.

Concomitant therapy: All people received an injection of botulinum toxin. All other routine rehabilitation that did not involve upper extremity training proceded as usual.

People after a first or recurrent stroke

Mean age (SD): 47.3 (11.5) years

N = 37

Mean time after stroke (SD): 34.8 (22.0) months

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention and ≥6 months

Spasticity at post-intervention and ≥6 months

Withdrawal for any reason at post-intervention and ≥6 months

Setting: Outpatients in Taiwan.

Funding: Government or Academic funding.

Hwang 2012⁴⁸

Robot-assisted arm training (n=9)

4 weeks (20 sessions) of active robot-assisted intervention with Amadeo (full-term intervention).

Region of upper limb trained: Distal limb

Level of supervision: Not stated/unclear

Type of movement delivered by robotic device: Not stated/unclear

Any other intervention (n=5)

2 weeks (10 sessions) of early passive therapy, followed by 2 weeks (10 sessions) of active robot-assisted intervention (the half term intervention) group. Data from the first 2 weeks of intervention were used.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: Not available*

N = 15

Mean time after stroke (SD): 6.5 (5.3) months

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention

Arm muscle strength at post-intervention

Spasticity at post-intervention

Stroke-specific Patient Reported Outcome Measure at post-intervention

Withdrawal for any reason at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Iwamoto 2019⁴⁹

Robot-assisted arm training (n =6)

Hybrid Assistive Limb 40 minutes per day for 6 weeks.

Region of upper limb trained: Distal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Active-assisted.

Any other intervention (n=6)

Occupational therapy for 6 weeks.

Concomitant therapy: The total time of combination therapy during A and occupational therapy during B was equivalent. In the current Japanese medical system, the medical doctor prescribes a rehabilitation programme, and rehabilitation therapists (occupational therapist, physiotherapist, and speech therapist) design individually tailored exercise programmes for acute stroke patients for up to 3 hours per day.

People after a first or recurrent stroke

Mean age (SD): 61.0 (18.9) years

N = 12

Mean time after stroke: Not stated/unclear

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention

Arm muscle strength living at post-intervention

Setting: Inpatient rehabilitation department of neurosurgical hospital.

Sources of funding: Not reported.

Crossover study: First time period included only.

Jiang 2021⁵⁰

Robot-assisted arm training (n=23)

Robot therapy (Armeo Spring) for 30 minutes twice a day, for 2 weeks.

Region of upper limb trained: Mixed

Level of supervision: Supervised

Type of movement delivered by robotic device: Not stated/unclear

Any other intervention (n=22)

Conventional rehabilitation for 30 minutes twice a day, for 2 weeks.

Concomitant therapy: All received conventional rehabilitation therapy for 30 minutes twice a day, for 2 weeks.

People after a first or recurrent stroke

Mean age (SD): 64.2 (11.5) years

N = 45

Mean time after stroke (SD): 19.8 (6.3) days

Ethnicity: Not stated/unclear

Mean severity (SD) – NIHSS: 6.1 (1.8)

Activities of daily living at post-intervention and ≥6 months

Arm function at post-intervention and ≥6 months

Arm muscle strength at post-intervention and ≥6 months

Spasticity at post-intervention and ≥6 months

Setting: Inpatient rehabilitation ward in China.

Sources of funding: Supported by a fund from the Lanzhou Science and Technology Bureau (document number: 2016-2-59).

Kahn 2006⁵³

Subsidiary paper:

Kahn 2001⁵²

Robot-assisted arm training (n=10)

8-week therapy programme involving 24 sessions, each lasting 45 minutes.

Region of upper limb trained: Proximal limb

Level of supervision: Not stated/unclear

Type of movement delivered by robotic device: Active-assisted

Any other intervention (n=9)

‘Free reaching training’ in an 8 week therapy programme involving 24 sessions, each lasting 45 minutes.

Concomitant therapy: Not available*

People after a first or recurrent stroke

N= 19

Mean age (SD): 64.1 (11.5) years

Mean time after stroke (SD): 19.8 (6.3) days

Ethnicity: Not stated/unclear

Mean severity (SD) – NIHSS: 6.1 (1.8)

Withdrawal for any reason at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Kim 2021¹⁷

Robot-assisted arm training (n=23)

Electromechanically assisted upper limb training using Camillo for 30 minutes a day, 5 days a week for 4 weeks.

Region of upper limb trained: Not stated/unclear

Level of supervision: Supervised

Type of movement delivered by robotic device: Not stated/unclear

Any other intervention (n=24)

Occupational therapist-assisted upper limb training for 30 minutes a day, 5 days a week for 4 weeks.

Concomitant therapy: All people underwent additional therapy for activities of daily living for 30 minutes daily during the study period.

People after a first or recurrent stroke

Mean age (SD): 59.7 (14.0) years

N = 47

Mean time after stroke (SD): 582.9 (1010.1) days

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Person/participant generic health-related quality of life at post-intervention

Arm function at post-intervention

Arm muscle strength at post-intervention

Spasticity at post-intervention

Withdrawal for any reason at post-intervention

Adverse events at post-intervention

Setting: Outpatients in the Republic of Korea.

Sources of funding: Supported by a grant of the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (grant number: HI15C1529). Device support from Man&Tel Co. Ltd, Gumi, Republic of Korea.

Kim 2019⁶⁹

Robot-assisted arm training (n=19)

Robot-assisted shoulder rehabilitation therapy for 30 minutes per day, 5 times per week for a total of 20 sessions for 4 weeks.

Region of upper limb trained: Proximal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Not stated/unclear

Any other intervention (n=19)

Conventional rehabilitation only

Concomitant therapy: all participants received usual care.

People after a first or recurrent stroke

Mean age (SD): 65.3 (8.9) years

N = 38

Mean time after stroke (SD): 3.3 (0.9) months

Ethnicity: Not stated/unclear

Mean severity (SD) – NIHSS: 9.2 (2.5)

Arm function at post-intervention and ≥6 months

Withdrawal for any reason at post-intervention and ≥6 months

Adverse events at post-intervention and ≥6 months

Setting: Outpatients in the Republic of Korea.

Sources of funding: Support by Wonkwang Institute of Clinical Medicine (2016-0669), Republic of Korea.

Kutner 2010⁵⁶

Robot-assisted arm training (n=10)

30 hours of repetitive task training plus 30 hours of robotic assisted training over 3 weeks.

Region of upper limb trained: Distal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Active-assisted

Any other intervention (n=11)

60 hours of repetitive task training over 3 weeks.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age (SD): Not available*

N = 21

Time after stroke: Mixed (3-9 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Stroke-specific Patient Reported Outcome at post-intervention and ≥6 months Measure

Withdrawal for any reason at post-intervention

Adverse events at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Lee 2016⁵⁷

Robot-assisted arm training (n=29)

With the robot Neuro-X over 20 sessions (30 minutes per session, 2 sessions per day, 5 days a week for 2 weeks).

Region of upper limb trained: Proximal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Mixed

Any other intervention (n=29)

Conventional upper extremity rehabilitation exercise twice daily.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age (SD): Not available*

N = 58

Time after stroke: Subacute (7 days - 6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention

Arm function at post-intervention

Arm muscle strength at post-intervention

Spasticity at post-intervention

Withdrawal for any reason at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Lee 2018⁵⁸

Robot-assisted arm training (n=15)

5, 30 min sessions REJOYCE robot treatment per week for 8 weeks.

Region of upper limb trained: Mixed

Level of supervision: Supervised

Type of movement delivered by robotic device: Active-assisted

Any other intervention (n=15)

5, 30 min sessions of general occupational therapy per week for 8 weeks.

Concomitant therapy: Both groups received general occupational therapy consisting of 5, 30 min sessions per week for 8 weeks.

People after a first or recurrent stroke

Mean age (SD): 51.2 (12.7) years

N = 30

Time after stroke: Chronic (≥6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention

Arm function at post-intervention

Withdrawal for any reason at post-intervention

Setting: Rehabilitation hospital in Korea.

Sources of funding: Not reported.

Liao 2012⁶¹

Subsidiary paper:

Hsieh 2011⁴⁴

Robot-assisted arm training (n=10)

With Bi-Manu - Track over 4 weeks, 5 days a week for 90 to 105 minutes per session. After robot training, participants received 15 minutes of training in functional activities.

Region of upper limb trained: Distal limb

Level of supervision: Not stated/unclear

Type of movement delivered by robotic device: Mixed

Any other intervention (n=10)

Protocol-based occupational therapy techniques. The control group received the same amount of therapy hours as the treatment group; after the active control therapy session the participants also received 15 minutes of training in functional activities.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: not available*

N = 20

Time after stroke: Chronic (≥6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention

Arm function at post-intervention

Withdrawal for any reason at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Lin 2022⁶²

Robot-assisted arm training (n=86)

Robot training for 30 minutes, 5 days a week for 3 weeks.

Region of upper limb trained: Mixed

Level of supervision: Supervised

Type of movement delivered by robotic device: Mixed

Any other intervention (n=86)

Enhanced occupational therapy that was time matched.

Concomitant therapy: All people received conventional rehabilitation, 5 days a week for 3 weeks divided into 30 minutes of physiotherapy and occupational therapy respectively.

People after a first or recurrent stroke

Mean age (SD): 59.0 (12.0) years

N = 172

Mean time after stroke (SD): 158.3 (170.9) days

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention

Activities of daily living at post-intervention

Withdrawal for any reason at post-intervention

Adverse events at post-intervention

Setting: Outpatients in China.

Funding: Government or academic funding.

Lo 2010⁶⁴

Robot-assisted arm training (n=49)

Maximum of 36 sessions over 12 weeks.

Region of upper limb trained: Mixed

Level of supervision: Supervised

Type of movement delivered by robotic device: Not stated/unclear

Any other intervention (n=78)

Intensive comparison therapy which matched the robot therapy in schedule and in form of intensity of movements. (n=50) Customary care (i.e. medical management, clinic visits needed and in some cases, rehabilitation services). (n=28) These groups were collapsed into one control group in analysis

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: Not available*

N = 127

Time after stroke: Chronic (≥6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention

Spasticity at post-intervention

Stroke-specific Patient Reported Outcome Measure at post-intervention

Withdrawal for any reason at post-intervention

Adverse events at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Lum 2002⁶⁵

Subsidiary papers:

Burgar 1999¹²

Burgar 2000¹³

Robot-assisted arm training (n=15)

Received bimanual and passive robot therapy by the MIME robot as per the control group.

Region of upper limb trained: Not stated/unclear

Level of supervision: Supervised

Type of movement delivered by robotic device: Mixed

Any other intervention (n=15)

Received 55 minutes of physiotherapy for the arm and 5 minutes of robot training for each of the 24 sessions over a 2 month period.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: not available*

N = 30

Time after stroke: Chronic (≥6 months)

Ethnicity: not stated/ unclear

Severity: not stated/ unclear

Activities of daily living at post-intervention and ≥6 months

Arm function at post-intervention and ≥6 months

Withdrawal for any reason at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Lum 2006⁶⁶

Robot-assisted arm training (n=24)

Unilateral/ bilateral or combined robot therapy, one hour per day for 6 weeks.

Region of upper limb trained: proximal

Level of supervision: Not stated/unclear

Type of movement delivered by robotic device: mixed

Any other intervention (n=6)

Received an equivalent intensity and duration of conventional therapy.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: not available*

N = 30

Time after stroke: Subacute (7 days - 6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention and ≥6 months

Arm function at post-intervention and ≥6 months

Arm muscle strength at post-intervention and ≥6 months

Spasticity at post-intervention and ≥6 months

Withdrawal for any reason at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Ma 2022⁶⁷

Robot-assisted arm training (n=13)

Robot therapy for 60 minutes, 5 days a week for 4 weeks.

Region of upper limb trained: Distal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Mixed

Any other intervention (n=13)

60 minutes of one-on-one conventional therapy for unilateral hand functional training.

Concomitant therapy: All people received 30 minutes of regular conventional therapy, 5 days a week for 4 weeks.

People after a first or recurrent stroke

Mean age (SD): 57.7 (9.8) years

N = 26

Mean time after stroke (SD): 10.2 (6.1) weeks

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention

Withdrawal for any reason at post-intervention

Setting: Inpatients in Taiwan.

Funding: Government or academic funding.

Marganska 2014⁶⁸

Robot-assisted arm training (n=39)

Robotic therapy with ARMin. Therapy was given 3 times a week for a period of 8 weeks (sum of 24 sessions). Minimum session time was 45 minutes.

Region of upper limb trained: Proximal limb

Level of supervision: Not stated/unclear

Type of movement delivered by robotic device: Not stated/unclear

Any other intervention (n=38)

Conventional therapy 3 times a week for a period of 8 weeks (sum of 24 sessions). Minimum session time was 45 minutes.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: not available*

N = 77

Time after stroke: Chronic (≥6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention and ≥6 months

Arm muscle strength at post-intervention and ≥6 months

Withdrawal for any reason at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Masiero 2011⁷²

Subsidiary papers:

Masiero 2012⁷⁰

Masiero 2014⁷¹

Robot-assisted arm training (n=11)

Received robotic training with the NeReBot, twice a day for 20 minutes, and 40 minutes conventional training, 5 days a week for at least 5 weeks.

Region of upper limb trained: Proximal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Not stated/unclear

Any other intervention (n=10)

80 minutes per day (including proprioceptive exercises, functional re-education, gait training, occupational therapy, and passive and active-assisted mobilisation of the hand and wrist) but without specifically exercising the proximal paretic arm.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: Not available*

N = 21

Time after stroke: Mixed (within 20 days of stroke)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention and ≥6 months

Arm function at post-intervention and ≥6 months

Arm muscle strength at post-intervention

Spasticity at post-intervention

Withdrawal for any reason at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Masiero 2007⁷³

Robot-assisted arm training (n =17)

Received additional early sensorimotor robotic training with the NeReBot, robot training twice a day, 5 days a week for at least 5 weeks.

Region of upper limb trained: Proximal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Passive

Any other intervention (n=18)

Received similar exposure to the robot (30 minutes twice per week) except that the exercises were performed with the unimpaired arm.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: not available*

N = 35

Time after stroke: Acute (72 hours - 7 days)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention and ≥6 months

Arm function at post-intervention

Arm muscle strength at post-intervention

Spasticity at post-intervention and ≥6 months

Withdrawal for any reason at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Mayr 2008⁷⁴

Robot-assisted arm training (n=4)

group AB: the participants received over 2 weeks, 5 times per week robot-assisted therapy with the ARMOR device, then 2 weeks with no intervention, and then over 2 weeks, 5 times per week EMG-initiated functional electrical stimulation.

Region of upper limb trained: Proximal limb

Level of supervision: Not stated/unclear

Type of movement delivered by robotic device: Not stated/unclear

Any other intervention (n=4)

Functional electrical stimulation.

group BA: the participants received 5 times per week over 2 weeks EMG-initiated functional electrical stimulation, then 2 weeks no intervention, and then 5 times per week over 2 weeks robot-assisted therapy.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: not available*

N = 8

Time after stroke: Subacute (7 days - 6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention

Arm muscle strength at post-intervention

Withdrawal for any reason at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Crossover study: First time period included only.

McCabe 2015⁷⁷

Subsidiary paper: Daly 2010²⁵

Robot-assisted arm training (n=12)

Motor Learning Programme in a 1:3 group paradigm for 3.5 hours per day + robotic-assisted arm training with the InMotion2 Shoulder-Elbow Robot 1.5 hours per day for 12 weeks.

Region of upper limb trained: Proximal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Not stated/unclear

Any other intervention (n=27)

Motor Learning Programme in a 1:3 group paradigm for 3.5 hours per day + functional electrical stimulation for1.5 hours per day for 12 weeks. Motor Learning Programme in a 1:3 group paradigm for 5 hours per day for 12 weeks. The 2 groups were combined for analysis.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: not available*

N = 39

Time after stroke: Chronic (≥6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention

Withdrawal for any reason at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Orihuela-Espina 2016⁸²

Robot-assisted arm training (n=9)

Robot therapy with the Amadeo (Inc. Typromotion) for 40 sessions 5 times a week for about 60 minutes.

Region of upper limb trained: Distal limb

Level of supervision: Not stated/unclear

Type of movement delivered by robotic device: Mixed

Any other intervention (n=9)

Classic occupational therapy 40 sessions 5 times a week for about 60 minutes.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: not available*

N = 17

Time after stroke: Subacute (7 days - 6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention

Arm muscle strength at post-intervention

Withdrawal for any reason at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Park 2021⁸⁴

Robot-assisted arm training (n=12)

20 sessions (five days a week for four weeks) of robot-assisted hand training using the Amadeo Robotic device (Trymotion GmbH, Graz, Austria)

Region of upper limb trained: Distal limb

Level of supervision: Not stated/unclear

Type of movement delivered by robotic device: Passive

Any other intervention (n=12)

20 sessions of conventional treatments that lasted 30 minutes each session

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age (SD): 70.3 (4.2) years

N = 24

Mean time after stroke (SD): 9.3 (2.4) months

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Withdrawal for any reason at post-intervention

Setting: Rehabilitation hospital in South Korea

Sources of funding: This work was supported by the Soonchunhyang University Research Fund. This work was supported by the Korea Institute for Advancement of Technology(KIAT) grant funded by the Korea Government(MOTIE) (P0012724, The Competency Development Program for Industry Specialist)

Rabadi 2008⁸⁵

Robot-assisted arm training (n=10)

Standard occupational and physical therapy for 3 hours per day plus 12 additional sessions of 40 minutes of robotic-assisted arm training with the MIT-Manus 5 days per week.

Region of upper limb trained: Proximal limb

Level of supervision: Not stated/unclear

Type of movement delivered by robotic device: Mixed

Any other intervention (n=20)

Group 1: standard occupational and physical therapy for 3 hours per day plus 12 additional sessions of 40 minutes of occupational therapy 5 days per week. Group 2: standard occupational and physical therapy for 3 hours per day plus 12 additional sessions of 40 minutes of arm ergometry 5 days per week. The 2 groups were combined for analysis.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: Not available*

N = 30

Time after stroke Subacute (7 days - 6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention

Arm function at post-intervention

Arm muscle strength at post-intervention

Spasticity at post-intervention

Withdrawal for any reason at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Ranzani 2020⁸⁶

Robot-assisted arm training (n=17)

45 minute sessions, for 6 weeks duration.

Region of upper limb trained: Distal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Mixed

Any other intervention (n=16)

Usual care.

Concomitant therapy: Both groups received conventional neurocognitive therapy sessions that included two or three exercises depending on the session duration (i.e., 30 or 45 min).

People after a first or recurrent stroke

Mean age (SD): 68.8 (12.2) years

N = 33

Mean time after stroke (SD): 3.1 (1.4) weeks

Ethnicity: Not stated/unclear

Mean severity (SD) - NIHSS: 1.52 (0.91)

Arm function at post-intervention and ≥6 months

Spasticity at post-intervention and ≥6 months

Withdrawal for any reason at post-intervention and ≥6 months

Setting: Rehabilitation centre in Switzerland.

Sources of funding: This work was supported by the National Center of Competence in Research on Neural Plasticity and Repair of the Swiss National Science Foundation (NCCR Neuro), the ETH CHIRP1 Research Grant on Cortically-Driven Assistance Adaptation during Sensorimotor Training, the Olga Mayenfisch Stiftung, the ETH Zurich Foundation in collaboration with Hocoma AG, and the Clinica Hildebrand Centro di Riabilitazione Brissago, Switzerland.

Remy-Neris 2021⁴⁹

Robot-assisted arm training (n=107)

Robotic training with ArmeoSpring exoskeleton device.

Duration 6 weeks.

Region of upper limb trained: Proximal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Active-assisted

Any other intervention (n=108)

Self-rehabilitation.

Duration 6 weeks.

Concomitant therapy:

Usual rehabilitation for all participants, followed by an additional daily hour of self-rehabilitation (two 30-minute sessions)

People after a first or recurrent stroke

Mean age (SD): 58.3 (13.7) years.

N = 215

Mean time after stroke (SD): 54.8 (22.2) days

Ethnicity: Not stated/unclear

Mean severity (SD) - NIHSS: 5.2 (2.4)

Person/participant health related quality of life at post-intervention and ≥6 months

Activities of daily living at post-intervention and ≥6 months

Arm function at post-intervention and ≥6 months

Stroke-specific Patient Reported Outcome Measure at post-intervention and ≥6 months

Withdrawal for any reason at post-intervention and ≥6 months

Adverse events at post-intervention

Setting: 21 inpatient rehabilitation centres in France.

Sources of funding: This study was supported by the French Ministry of Health: EMREM_AVC CHU BREST 20 220.

Rodgers 2019⁸⁸

Subsidiary papers:

Rodgers 2020⁸⁹

Fernandez-Garcia 2021³¹

Robot-assisted arm training (n=257)

Training with MIT-Manus robotic gym.

45 min of face-to-face therapy, three times per week for 12 weeks.

Region of upper limb trained: Proximal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Active-assisted

Any other intervention (n=513)

Repetitive functional task practice or usual care. 45 min of face-to-face therapy, three times per week for 12 weeks.

Concomitant therapy: Usual post-stroke care.

People after a first or recurrent stroke

Mean age (SD): 60.6 (13.5) years

N = 770

Median time after stroke (IQR): 233 (102 to 549) days

Ethnicity: Not stated/unclear

Mean severity (SD): 5.7 (3.2)

Person/participant health related quality of life at post-intervention and ≥6 months

Activities of daily living at post-intervention and ≥6 months

Arm function at post-intervention and ≥6 months

Stroke-specific Patient Reported Outcome Measure at post-intervention and ≥6 months

Withdrawal for any reason at post-intervention and ≥6 months

Adverse events at post-intervention

Setting: Four National Health Service (NHS) centres in the UK. Each centre comprised a stroke service in an NHS hospital with an MIT-Manus robotic gym system (InMotion commercial version, Interactive Motion Technologies, Watertown, MA, USA), plus stroke services in adjacent NHS Trusts and community services.

Sources of funding: National Institute for Health Research Health Technology Assessment Programme.

Sale 2014⁹⁰

Subsidiary papers:

Franceschi 2020³²

Mazzoleni⁷⁵

Mazzoleni⁷⁶

Sale 2014⁹¹

Robot-assisted arm training (n=26)

30 sessions of robot-assisted therapy (5 days a week for 6 weeks).

Region of upper limb trained: Proximal limb

Level of supervision: Not stated/unclear

Type of movement delivered by robotic device: Active-assisted

Any other intervention (n=27)

Conventional rehabilitative treatment 30 sessions (5 days a week for 6 weeks)

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: Not available*

N = 53

Time after stroke: Subacute (7 days - 6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention

Arm muscle strength at post-intervention

Spasticity at post-intervention

Withdrawal for any reason at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Singh 2021⁹²

Robot-assisted arm training (n=13)

Robot therapy sessions were conducted for 45 min per day for 5 days a week for 4 weeks.

Region of upper limb trained: Distal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Active-assisted

Any other intervention (n=14)

The conventional therapy session was conducted for 45 min per day for 5 days a week for 4 weeks.

Concomitant therapy: No additional information.

People after a first or recurrent stroke

Mean age (SD): 41.9 (11.2) years

N = 23

Mean time after stroke (SD): 12.0 (7.5) months

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention

Arm function at post-intervention

Spasticity at post-intervention

Withdrawal for any reason at post-intervention

Setting: Outpatient clinic in India.

Sources of funding: financially supported by SERB, DST India (YSS/2015/000697) and IIT Delhi, MFIRP (Project no. AI-19).

Straudi 2020⁹³

Robot-assisted arm training (n=20)

1 hour and 40 minutes of hand functional electrical stimulation + robot-assisted arm therapy for each session (5 times/week over 6 weeks).

Region of upper limb trained: Mixed

Level of supervision: Supervised

Type of movement delivered by robotic device: Active-assisted

Any other intervention (n=20)

1 hour and 40 minutes of conventional therapy (5 times/week over 6 weeks).

Concomitant therapy: In addition to arm rehabilitation, all patients received multidisciplinary rehabilitation based on an individualized approach.

People after a first or recurrent stroke

Median age (IQR):

Intervention: 68 (56 to 71) years

Control: 68 (58.5 to 73) years

N = 40

Median time after stroke (IQR):

Intervention: 39 (21 to 62) days

Control: 32.5 (20 to 51) days

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Withdrawal for any reason at post-intervention

Setting: Inpatient Rehabilitation at a University Hospital in Italy.

Sources of funding: Not reported.

Susanto 2015⁹⁴

Robot-assisted arm training (n=9)

Hand exoskeleton robot-assisted training for 10 1-hour sessions. Duration 5 weeks

Region of upper limb trained: Distal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Passive

Any other intervention (n=10)

Non-assisted training, 20 1- hour sessions for 5 weeks.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: Not available*

N = 19

Time after stroke: Chronic (≥6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention and ≥6 months

Withdrawal for any reason at post-intervention and ≥6 months

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Takebayashi 2022⁹⁶

Robot-assisted arm training (n=87)

Two groups combined. One group (n=44) received robot therapy and therapist led occupational therapy for 20 minutes. One group (n=43) received robot therapy and therapist led constrain induced movement therapy. Robot therapy was for 40 minutes, 3 days a week for 10 weeks.

Region of upper limb trained: Proximal limb

Level of supervision: Unsupervised

Type of movement delivered by robotic device: Mixed

Any other intervention (n=42)

40 minutes of self training followed by 20 minutes of therapist-led occupational therapy, 3 days a week for 10 weeks.

Concomitant therapy: No additional information.

People after a first or recurrent stroke

Mean age (SD): 59 (11) years

N = 129

Mean time after stroke (SD): 36.6 (51.3) months

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention

Arm muscle strength at post-intervention

Spasticity at post-intervention

Stroke-specific Patient-Reported Outcome Measures at post-intervention

Withdrawal for any reason at post-intervention

Adverse events at post-intervention

Setting: Outpatients in Japan.

Funding: Funded by Teijin Pharma Limited.

Takahashi 2016⁹⁵

Robot-assisted arm training (n=30)

40 minutes of standard therapy plus robot therapy with ReoGo for 40 additional minutes, 7 times a week for 6 weeks.

Region of upper limb trained: Proximal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Passive

Any other intervention (n=30)

40 minutes of standard therapy plus therapist-directed self-training for 40 additional minutes, 7 times a week for 6 weeks.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: Not available*

N = 60

Time after stroke: Subacute (7 days - 6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention

Arm function at post-intervention

Arm muscle strength at post-intervention

Spasticity at post-intervention

Withdrawal for any reason at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Taravati 2021⁹⁷

Robot-assisted arm training (n=22)

Hand-arm robotic assisted therapy for 30-45 min, 5 days a week for 4 weeks.

Region of upper limb trained: Mixed

Level of supervision: Supervised

Type of movement delivered by robotic device: Active-assisted

Any other intervention (n=23)

Conventional physiotherapy was provided for 5 days a week and for 4 weeks.

Concomitant therapy: The control group received only conventional therapy for 5 days a week and 4 weeks, while the study groups received the same amount of conventional therapy in addition to rehabilitation with the robotic rehabilitation.

People after a first or recurrent stroke

Mean age (SD): 53.4 (14.8) years

N = 45

Mean time after stroke (SD): 11.8 (8.3) months

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention

Arm function at post-intervention

Arm muscle strength at post-intervention

Spasticity at post-intervention

Stroke specific quality of life at post-intervention

Withdrawal for any reason at post-intervention

Setting: Rehabilitation hospital in Turkey.

Sources of funding: Not reported.

Taveggia 2016⁹⁸

Robot-assisted arm training (n=27)

Robot therapy with the Armeo Spring for 30 minutes per session, 5 times per week for 6 weeks.

Region of upper limb trained: Not stated/unclear

Level of supervision: Not stated/unclear

Type of movement delivered by robotic device: Passive

Any other intervention (n=27)

Physical rehabilitation therapy according to the Bobath concept for 30 minutes per session, 5 times a week for 6 weeks.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: Not available*

N = 54

Time after stroke: Mixed (0.5-12 months post-stroke)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention and ≥6 months

Arm muscle strength at post-intervention and ≥6 months

Spasticity at post-intervention and ≥6 months

Withdrawal for any reason at post-intervention and ≥6 months

Adverse events at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Timmermans 2014⁹⁹

Subsidiary paper:

Lemmens 2014⁵⁹

Robot-assisted arm training (n=11)

With end-effector robot HapticMaster 4 times/ week, twice a day for 30 minutes (separated by 0.5 hour to 1 hour of rest).

Region of upper limb trained: Proximal limb

Level of supervision: Not stated/unclear

Type of movement delivered by robotic device: Mixed

Any other intervention (n=11)

Arm-hand training programme 4 times/ week, twice a day for 30 minutes (separated by 0.5 hour to 1 hour of rest).

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: Not available*

N = 22

Time after stroke: Chronic (≥6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Person/participant generic health-related quality of life at post-intervention and ≥6 months

Arm function at post-intervention and ≥6 months

Withdrawal for any reason at post-intervention and ≤6 months

Adverse events at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Tomic 2017¹⁰⁰

Robot-assisted arm training (n=13)

Additional robot therapy with the ArmAssist (AA) for 30 minutes administered over 15 sessions each lasting 30 minutes, scheduled 5 days per week for 3 weeks

Region of upper limb trained: Proximal limb

Level of supervision: Not stated/unclear

Type of movement delivered by robotic device: Active-assisted

Any other intervention (n=13)

Additional occupational therapy for 30 minutes that was matched in its structure and amount to the AA training as close as possible and administered over 15 sessions each lasting 30 minutes, scheduled 5 days per week for 3 weeks

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: not available*

N = 26

Time after stroke: Subacute (7 days - 6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention

Arm function at post-intervention

Withdrawal for any reason at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Valles 2016¹⁰¹

Robot-assisted arm training (n = 13)

24, 2 hour therapy sessions over a 6-8 week period.

Region of upper limb trained: Mixed

Level of supervision: Supervised

Type of movement delivered by robotic device: Mixed

Any other intervention (n=14)

Standard rehabilitation therapy- 24 2-hour therapy sessions over a 6-8 week period.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: Not available*

N = 27

Mean time after stroke: Not stated/unclear (inclusion criteria says a minimum of 6 months post stroke)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention

Withdrawal for any reason at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Vanoglio 2017¹⁰²

Robot-assisted arm training (n=15)

Robot therapy with the Gloreha Professional consisted of a total of 30 sessions, lasting 40 minutes per day, for 5 days per week.

Region of upper limb trained: Distal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Passive

Any other intervention (n=15)

Passive arm therapy for 30 sessions, lasting 40 minutes per day, for 5 days per week

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: not available*

N = 30

Time after stroke: Subacute (7 days - 6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention

Arm muscle strength at post-intervention

Withdrawal for any reason at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Villafane 2018¹⁰³

Robot-assisted arm training (n=16)

Robot therapy with the hand Gloreha for 30 minutes for 3 days per week

Region of upper limb trained: Distal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Passive

Any other intervention (n=16)

Physical and occupational arm therapy for 30 minutes 3 days per week

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: Not available*

N = 32

Time after stroke: Subacute (7 days - 6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention

Arm function at post-intervention

Arm muscle strength at post-intervention

Withdrawal for any reason at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Volpe 2000¹⁰⁴

Subsidiary paper:

Fasoli 2004²⁹

Robot-assisted arm training (n=30)

The treatment group used the MIT-Manus device for arm training for 1 hour per day, 5 days a week (for at least 25 sessions)

Region of upper limb trained: Mixed

Level of supervision: Supervised

Type of movement delivered by robotic device: Active-assisted

Any other intervention (n=26)

Placebo. The control group had similar initial exposure to the robot with the exception that half the tasks were performed with the unimpaired arm, and when the participant could not perform the task with the affected limb, the unimpaired limb was used to complete the task or the technician assisted the movement. The robot never actively moved the limbs of participants in the control group. Participants were exposed to the robot 1 hour per week.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: Not available*

N = 56

Time after stroke: Subacute (7 days - 6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention

Arm function at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Volpe 2008¹⁰⁵

Robot-assisted arm training (n=11)

Robotic training with the InMotion2 robot (the commercial version of MIT-Manus). All participants had an identical number of treatment sessions, and the sessions were of the same duration (1 hour per session, 3 times a week for 6 weeks).

Region of upper limb trained: Mixed

Level of supervision: Supervised

Type of movement delivered by robotic device: Passive

Any other intervention (n=10)

Conventional therapy. Intensive movement protocol with a trained physiotherapist. All participants had an identical number of treatment sessions, and the sessions were of the same duration (1 hour per session, 3 times a week for 6 weeks).

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: not available*

N = 21

Time after stroke: Chronic (≥6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention

Spasticity at post-intervention

Stroke-specific Patient Reported Outcome Measures at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Wolf 2015¹⁰⁶

Subsidiary paper:

Linder 2013⁶³

Robot-assisted arm training (n=51)

Robot therapy with the Hand Mentor Pro (Kinetic Muscles Incs) for 60 minutes over a 8 (to 12) weeks period.

Region of upper limb trained: Mixed

Level of supervision: Supervised

Type of movement delivered by robotic device: Active-assisted

Any other intervention (n=48)

Conventional therapy. Home exercises for the arm therapy for 60 minutes over an 8 (to 12) week period

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: not available*

N = 99

Time after stroke: Subacute (7 days - 6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention

Withdrawal at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Wu 2012¹⁰⁹

Subsidiary papers:

Wu¹⁰⁷

Wu 2012¹⁰⁸

Robot-assisted arm training (n=14)

Robot-assisted (Bi-Manu-Track) arm trainer (RAT Group). Each group received treatment for 90 to 105 minutes per session, 5 sessions on weekdays, for 4 weeks.

Region of upper limb trained: Mixed

Level of supervision: Supervised

Type of movement delivered by robotic device: Mixed

Any other intervention (n=28)

A combination of two arms. 1) therapist-mediated bilateral arm training and 2)conventional therapy. Each group received treatment for 90 to 105 minutes per session, 5 sessions on weekdays, for 4 weeks.

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: Not available*

N = 42

Time after stroke: Chronic (≥6 months) Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Arm function at post-intervention

Stroke-specific Patient Reported Outcome Measures at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Xu 2020¹¹⁰

Robot-assisted arm training (n=22)

Robot training was provided in addition 20 min/time, once/day and five days/week.

Region of upper limb trained: Proximal limb

Level of supervision: Supervised

Type of movement delivered by robotic device: Mixed

Any other intervention (n=23)

Traditional occupational therapy.

Control group was trained with traditional exercises, 40 min, once/day, and five days/week.

Concomitant therapy: The patients in both groups received regular neurological medical and physical therapy with equal treatment volume. A 6 weeks rehabilitation programme was designed for all the patients.

People after a first or recurrent stroke

Mean age (SD): 61.4 (10.4) years

N = 55

Time after stroke (SD): 49.1 (21.8) days

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention

Arm function at post-intervention

Withdrawal for any reason at post-intervention

Setting: Rehabilitation hospital in China.

Sources of funding: The study was supported by the Beijing Municipal Administration of hospitals youth programme (No. QML2019002).

Yoo 2013¹¹¹

Robot-assisted arm training (n=11)

3-dimensional robot-assisted therapy (RAT) and conventional rehabilitation therapy (CT) for a total of 90 minutes (RAT: 30 minutes, CT: 60 minutes) a day with 10 minutes rest halfway through the session, received training 3 days a week for 6 weeks

Region of upper limb trained: Not stated/unclear

Level of supervision: Supervised

Type of movement delivered by robotic device: Mixed

Any other intervention (n=11)

The control group received only CT for 60 minutes a day on the same days as the first group

Concomitant therapy: Not available*

People after a first or recurrent stroke

Mean age: Not available*

N = 22

Time after stroke: Chronic (≥6 months)

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Activities of daily living at post-intervention

Arm function at post-intervention

Arm muscle strength at post-intervention

*This study was included in the original Cochrane review that was updated in this review. For further details see Mehrholz 2018⁷⁸.

Zengin-Metli 2018¹¹²

Robot-assisted arm training (n=20)

Robot therapy with Armeo Spring HocomAG Inc.for 3 weeks

Region of upper limb trained: Mixed

Level of supervision: Supervised

Type of movement delivered by robotic device: Active-assisted

Any other intervention (n=15)

Conventional program consisted of neurophysiological exercises with Brunnstrom approach, range of motion exercises and postural education for 3 weeks.

Concomitant therapy: Conventional program consisted of neurophysiological exercises with Brunnstrom approach, range of motion exercises and postural education.

People after a first or recurrent stroke

Mean age (SD): 61.0 (6.9) years

N = 35

Mean time after stroke (SD): 11.0 (5.1) weeks

Ethnicity: Not stated/unclear

Severity: Not stated/unclear

Person/participant generic health related quality of life at post-intervention

Activities of daily living

Arm function at post-intervention

Arm muscle strength at post-intervention

Setting: Stroke rehabilitation centre in Turkey.

Sources of funding: Not reported.

Table 3Clinical evidence summary: Robot-assisted arm training compared to any other intervention

Outcomes	№ of participants (studies) Follow-up	Certainty of the evidence (GRADE)	Relative effect (95% CI)	Anticipated absolute effects		Comments
Outcomes	№ of participants (studies) Follow-up	Certainty of the evidence (GRADE)	Relative effect (95% CI)	Risk with any other intervention	Risk difference with robot-assisted arm training	Comments
Person/participant health related quality of life (SF-36 PCS, 0-100, higher values are better, change score) at end of intervention	215 (2 RCTs) follow-up: mean 5 weeks	⨁◯◯◯ Very low^a^,^b	-	The mean person/participant health related quality of life at end of intervention was 1.37	MD 0.73 higher (0.81 lower to 2.27 higher)	MID = 2 (SF-36 established MID)
Person/participant health related quality of life (SF-36 MCS, 0-100, higher values are better, change score) at end of intervention	215 (2 RCTs) follow-up: mean 5 weeks	⨁◯◯◯ Very low^a^,^b	-	The mean person/participant health related quality of life at end of intervention was 3.84	MD 1.14 lower (3.5 lower to 1.22 higher)	MID = 3 (SF-36 established MID)
Person/participant health related quality of life (EQ5D, −0.11-1, higher values are better, final values and change scores) at end of intervention	716 (2 RCTs) follow-up: mean 4 weeks	⨁◯◯◯ Very low^b^,^c	-	The mean person/participant health related quality of life at end of intervention was 0.23	MD 0.01 higher (0.02 lower to 0.03 higher)	MID = 0.03 (EQ-5D established MID)
Person/participant health related quality of life (EQ5D, 0-100, higher values are better, change score) at ≥6 months	194 (1 RCT) follow-up: 12 months	⨁◯◯◯ Very low^b^,^d	-	The mean person/participant health related quality of life at ≥6 months was 19.08	MD 4.67 lower (10.58 lower to 1.24 higher)	MID = 9.9 (0.5 x median baseline SD)
Person/participant health related quality of life (EQ5D, −0.11-1, higher values are better, final values) at ≥6 months	625 (1 RCT) follow-up: 6 months	⨁⨁◯◯ Low^b^,^e	-	The mean person/participant health related quality of life at ≥6 months was 0.5	MD 0.04 lower (0.09 lower to 0.01 higher)	MID = 0.03 (EQ-5D established MID)
Activities of daily living (Barthel index, functional independence measure, stroke impact scale, MAL, Frenchay arm test, ABILHAND [different scale ranges], higher values are better, change scores) at end of intervention	1318 (25 RCTs) follow-up: mean 5 weeks	⨁◯◯◯ Very low^a^,^b^,^f	-	-	SMD 0.41 SD higher (0.16 higher to 0.67 higher)	MID = 0.5 SD (SMD)
Activities of daily living (Barthel index, functional independence measure, Motor activity log [different scale ranges], higher values are better, final values) at end of intervention	988 (11 RCTs) follow-up: mean 5 weeks	⨁⨁⨁⨁ High	-	-	SMD 0.14 SD higher (0.01 higher to 0.27 higher)	MID = 0.5 SD (SMD)
Activities of daily living (Barthel index, functional independence measure, motor activity log [different scale ranges], higher values are better, change scores) at ≥6 months	469 (9 RCTs) follow-up: mean 6 months	⨁⨁◯◯ Low^b^,^f	-	-	SMD 0.28 SD higher (0.09 higher to 0.46 higher)	MID = 0.5 SD (SMD)
Activities of daily living (Barthel index, Functional Independence Measure [different scale ranges], higher values are better, final values) at ≥6 months	670 (2 RCTs) follow-up: mean 4 months	⨁⨁◯◯ Low^f	-	-	SMD 0.02 SD higher (0.14 lower to 0.17 higher)	MID = 0.5 SD (SMD)
Arm function (FMA UE, Quick DASH, manual function test [different scale ranges], higher values are better, change scores) at end of intervention	2167 (48 RCTs) follow-up: mean 5 weeks	⨁⨁◯◯ Low^f^,^g	-	-	SMD 0.34 SD higher (0.26 higher to 0.43 higher)	MID = 0.5 SD (SMD)
Arm function (FMA UE, Chedoke Arm and Hand Activity [different scale ranges], higher values are better, final values) at end of intervention	1496 (24 RCTs) follow-up: mean 6 weeks	⨁⨁◯◯ Low^b^,^f	-	-	SMD 0.2 SD higher (0.09 higher to 0.31 higher)	MID = 0.5 SD (SMD)
Arm function (FMA UE, 0-66, higher values are better, change scores) at ≥6 months	517 (11 RCTs) follow-up: mean 6 months	⨁⨁⨁◯ Moderate^h	-	The mean arm function at ≥6 months was 9.01	MD 1.08 higher (0.09 higher to 2.07 higher)	MID = 6.6 (Fugl-Meyer upper extremity = Difference by 10% of the total scale)
Arm function (FMA UE, Korean DASH [different scale ranges], higher values are better, final values) at ≥6 months	930 (9 RCTs) follow-up: mean 4 months	⨁◯◯◯ Very low^b^,^f^,^g	-	-	SMD 0.61 SD higher (0.18 higher to 1.03 higher)	MID = 0.5 SD (SMD)
Arm muscle strength (Motricity index, MRC, manual muscle test, MRC total motor power [different scale ranges], higher values are better, change scores) at end of intervention	1019 (21 RCTs) follow-up: mean 5 weeks	⨁◯◯◯ Very low^a^,^b^,^f	-	-	SMD 0.45 SD higher (0.17 higher to 0.72 higher)	MID = 0.5 SD (SMD)
Arm muscle strength (Motricity index, MRC [different scale ranges], higher values are better, final values) at end of intervention	107 (3 RCTs) follow-up: mean 4 weeks	⨁◯◯◯ Very low^a^,^b^,^f	-	-	SMD 0.89 SD higher (0.19 higher to 1.6 higher)	MID = 0.5 SD (SMD)
Arm muscle strength (grip strength [kg], higher values are better, change scores and final values) at end of intervention	123 (5 RCTs) follow-up: mean 5 weeks	⨁⨁⨁◯ Moderate^b	-	The mean arm muscle strength at end of intervention was 3.48	MD 0.92 higher (0.39 lower to 2.22 higher)	MID = 1.83 (0.5 x median baseline SD)
Arm muscle strength (grip strength [Newton meter], higher values are better, change score and final value) at end of intervention	114 (2 RCTs) follow-up: mean 6 weeks	⨁⨁⨁⨁ High	-	The mean arm muscle strength at end of intervention was 6.8	MD 0.64 lower (4.18 lower to 2.91 higher)	MID = 4.3 (0.5 x median baseline SD)
Arm muscle strength (MRC total, MRC total motor power [different scale ranges], higher values are better, change scores) at ≥6 months	164 (4 RCTs) follow-up: mean 5 months	⨁◯◯◯ Very low^f^,ⁱ^,^j	-	-	SMD 0.48 SD higher (0.57 lower to 1.53 higher)	MID = 0.5 SD (SMD)
Arm muscle strength (MRC total, MI [different scale ranges], higher values are better, final value) at ≥6 months	84 (2 RCTs) follow-up: mean 2 months	⨁◯◯◯ Very lowⁱ^,^k	-	-	SMD 1.05 SD higher (0.59 higher to 1.51 higher)	MID = 0.5 SD (SMD)
Arm muscle strength (grip strength [kg], higher values are better, change score and final value) at ≥6 months	71 (2 RCTs) follow-up: mean 6 months	⨁⨁⨁◯ Moderate^b	-	The mean arm muscle strength at ≥6 months was 5.17	MD 1.06 higher (1.02 lower to 3.14 higher)	MID = 1.83 (0.5 x median baseline SD)
Spasticity (MAS, MAS total [different scale ranges], lower values are better, change scores) at end of intervention	761 (16 RCTs) follow-up: mean 5 weeks	⨁⨁◯◯ Low^f^,^l	-	-	SMD 0.23 SD lower (0.46 lower to 0.01 lower)	MID = 0.5 SD (SMD)
Spasticity (MAS, MAS total [different scale ranges], lower values are better, final values) at end of intervention	356 (10 RCTs) follow-up: mean 5 weeks	⨁⨁◯◯ Low^k	-	-	SMD 0.21 SD lower (0.42 lower to 0)	MID = 0.5 SD (SMD)
Spasticity (MAS, MAS total [different scale ranges], lower values are better, change scores) at ≥6 months	247 (7 RCTs) follow-up: mean 5 months	⨁⨁◯◯ Low^j^,^l	-	-	SMD 0.09 SD lower (0.34 lower to 0.17 higher)	MID = 0.5 SD (SMD)
Spasticity (MAS, MAS total [different scale ranges], lower values are better, final values) at ≥6 months	153 (4 RCTs) follow-up: mean 3 months	⨁◯◯◯ Very low^b^,ⁱ^,^k	-	-	SMD 0.2 SD lower (0.52 lower to 0.12 higher)	MID = 0.5 SD (SMD)
Stroke-specific Patient-Reported Outcome Measures (Stroke Impact Scale total, 0-100, higher values are better, final values and change scores) at end of intervention	284 (5 RCTs) follow-up: mean 7 weeks	⨁⨁◯◯ Low^b^,^l	-	The mean stroke-specific Patient-Reported Outcome Measures at end of intervention was 37.40	MD 5.31 higher (2.6 higher to 8.02 higher)	MID = 6.12 (0.5 x median control group SD)
Stroke-specific Patient-Reported Outcome Measures (Stroke Impact Scale - hand function domain [different scale ranges], higher values are better, change scores) at end of intervention	382 (5 RCTs) follow-up: mean 3 weeks	⨁◯◯◯ Very low^b^,^f^,^g	-	-	SMD 0.8 SD higher (0.31 lower to 1.91 higher)	MID = 0.5 SD (SMD)
Stroke-specific Patient-Reported Outcome Measures (SS-QOL, 49-245, higher values are better, final value) at end of intervention	37 (1 RCT) follow-up: 4 weeks	⨁◯◯◯ Very low^b^,^m	-	The mean stroke-specific Patient-Reported Outcome Measures at end of intervention was 140.8	MD 2.21 lower (23.36 lower to 18.94 higher)	MID = 14.1 (0.5 x median baseline SD)
Stroke-specific Patient-Reported Outcome Measures (Stroke Impact Scale - strength domain, 0-100, higher values are better, change score) at end of intervention	117 (1 RCT) follow-up: 10 weeks	⨁⨁⨁⨁ High	-	The mean stroke-specific Patient-Reported Outcome Measures at end of intervention was 4.43	MD 3.45 higher (2.58 higher to 4.32 higher)	MID = 1.03 (0.5 x median control group SD)
Stroke-specific Patient-Reported Outcome Measures (Stroke Impact Scale - memory domain, 0-100, higher values are better, change score) at end of intervention	117 (1 RCT) follow-up: 10 weeks	⨁⨁⨁◯ Moderate^b	-	The mean stroke-specific Patient-Reported Outcome Measures at end of intervention was 1.4	MD 0.19 higher (0.52 lower to 0.9 higher)	MID = 0.84 (0.5 x median control group SD)
Stroke-specific Patient-Reported Outcome Measures (Stroke Impact Scale - emotion domain, 0-100, higher values are better, change score) at end of intervention	117 (1 RCT) follow-up: 10 weeks	⨁⨁⨁◯ Moderate^b	-	The mean stroke-specific Patient-Reported Outcome Measures at end of intervention was 0.78	MD 1.24 lower (1.7 lower to 0.78 lower)	MID = 0.91 (0.5 x median control group SD)
Stroke-specific Patient-Reported Outcome Measures (Stroke Impact Scale - communication domain, 0-100, higher values are better, change score) at end of intervention	117 (1 RCT) follow-up: 10 weeks	⨁⨁⨁◯ Moderate^b	-	The mean stroke-specific Patient-Reported Outcome Measures at end of intervention was 0.9	MD 0.32 lower (0.94 lower to 0.3 higher)	MID = 0.8 (0.5 x median control group SD)
Stroke-specific Patient-Reported Outcome Measures (Stroke Impact Scale - ADL domain, 0-100, higher values are better, change scores and final value) at end of intervention	742 (3 RCTs) follow-up: mean 8 weeks	⨁◯◯◯ Very low^f^,ⁿ	-	The mean stroke-specific Patient-Reported Outcome Measures at end of intervention was 20.8	MD 0.12 higher (4.56 lower to 4.8 higher)	MID = 7.44 (0.5 x median control group SD)
Stroke-specific Patient-Reported Outcome Measures (Stroke Impact Scale - mobility domain, 0-100, higher values are better, change score and final value) at end of intervention	725 (2 RCTs) follow-up: 12 weeks	⨁⨁⨁⨁ High	-	The mean stroke-specific Patient-Reported Outcome Measures at end of intervention was 32.2	MD 0.44 higher (3.91 lower to 4.79 higher)	MID = 6.5 (0.5 x median control group SD)
Stroke-specific Patient-Reported Outcome Measures (Stroke Impact Scale - social participation domain, 0-100, higher values are better, change score and final value) at end of intervention	721 (2 RCT) follow-up: 12 weeks	⨁⨁⨁◯ Moderate^b	-	The mean stroke-specific Patient-Reported Outcome Measures at end of intervention was 25.5	MD 2.81 higher (5.98 lower to 11.6 higher)	MID = 6.7 (0.5 x median control group SD)
Stroke-specific Patient-Reported Outcome Measures (Stroke Impact Scale - stroke recovery domain, 0-100, higher values are better, change score) at end of intervention	117 (1 RCT) follow-up: 10 weeks	⨁⨁⨁◯ Moderate^b	-	The mean stroke-specific Patient-Reported Outcome Measures at end of intervention was 7.43	MD 1.11 higher (0.21 higher to 2.01 higher)	MID = 1.18 (0.5 x median control group SD)
Stroke-specific Patient-Reported Outcome Measures (Stroke Impact Scale - physical domain, 0-100, higher values are better, change score) at end of intervention	117 (1 RCT) follow-up: 10 weeks	⨁⨁⨁⨁ High	-	The mean stroke-specific Patient-Reported Outcome Measures at end of intervention was 2.28	MD 3.52 higher (2.99 higher to 4.05 higher)	MID = 0.68 (0.5 x median control group SD)
Stroke-specific Patient-Reported Outcome Measures (Stroke Impact Scale - hand function domain, 0-100, higher values are better, final value) at end of intervention	608 (1 RCT) follow-up: mean 12 weeks	⨁⨁⨁◯ Moderate^o	-	The mean stroke-specific Patient-Reported Outcome Measures at end of intervention was 18.1	MD 2.6 lower (6.75 lower to 1.55 higher)	MID = 13.0 (0.5 x median control group SD)
Stroke-specific Patient-Reported Outcome Measures (Stroke Impact Scale total, 0-100, higher values are better, change score and final value) at ≥6 months	90 (2 RCTs) follow-up: mean 5 months	⨁⨁◯◯ Low^k^,^p	-	The mean stroke-specific Patient-Reported Outcome Measures at ≥6 months was 25.1	MD 4.36 higher (1.64 lower to 10.36 higher)	MID = 12.0 (0.5 x median baseline SD)
Stroke-specific Patient-Reported Outcome Measures (Stroke Impact Scale - hand function domain, 0-100, higher values are better, final values and change scores) at ≥6 months	819 (3 RCTs) follow-up: mean 7 months	⨁⨁⨁◯ Moderateⁿ	-	The mean stroke-specific Patient-Reported Outcome Measures at ≥6 months was 23.22	MD 0.27 lower (3.98 lower to 3.45 higher)	MID = 8.3 (0.5 x median baseline SD)
Stroke-specific Patient-Reported Outcome Measures (Stroke Impact Scale - ADL domain, higher values are better, change score and final value) at ≥6 months	625 (2 RCTs) follow-up: mean 4 months	⨁⨁⨁◯ Moderateⁿ	-	The mean stroke-specific Patient-Reported Outcome Measures at ≥6 months was 29.87	MD 2.21 lower (5.71 lower to 1.28 higher)	MID = 8.2 (0.5 x median control group SD)
Stroke-specific Patient-Reported Outcome Measures (Stroke Impact Scale - mobility domain, higher values are better, final value) at ≥6 months	608 (1 RCT) follow-up: 6 months	⨁⨁⨁◯ Moderate^o	-	The mean stroke-specific Patient-Reported Outcome Measures at ≥6 months was 63.4	MD 1.7 lower (5.77 lower to 2.37 higher)	MID = 11.9 (0.5 x median control group SD)
Stroke-specific Patient-Reported Outcome Measures (Stroke Impact Scale - social participation domain, higher values are better, final value) at ≥6 months	604 (1 RCT) follow-up: 6 months	⨁⨁⨁◯ Moderate^o	-	The mean stroke-specific Patient-Reported Outcome Measures at ≥6 months was 50.0	MD 3 lower (7.23 lower to 1.23 higher)	MID = 12.1 (0.5 x median control group SD)
Withdrawal for any reason at end of intervention	3954 (72 RCTs) follow-up: mean 6 weeks	⨁◯◯◯ Very low^q^,^r	RD 0.00 (−0.02 to 0.02)	86 per 1,000	0 fewer per 1,000 (20 fewer to 20 more)^s	Precision calculated through Optimal Information Size (OIS) due to zero events in some studies. OIS determined power for the sample size = 0.04 (0.8-0.9 = serious, <0.8 = very serious).
Withdrawal for any reason at ≥6 months	1672 (21 RCTs) follow-up: mean 6 months	⨁◯◯◯ Very low^q^,^r	RD −0.02 (−0.04 to 0.01)	84 per 1,000	20 more per 1,000 (40 fewer to 10 more)^s	Precision calculated through Optimal Information Size (OIS) due to zero events in some studies. OIS determined power for the sample size = 0.13 (0.8-0.9 = serious, <0.8 = very serious).
Adverse events (cardiovascular events) at end of intervention	770 (1 RCT) follow-up: 3 months	⨁⨁◯◯ Low^b	RR 4.99 (0.97 to 25.55)	4 per 1,000	16 more per 1,000 (0 fewer to 96 more)	MID (precision) = RR 0.80 – 1.25.
Adverse events (cardiovascular events) at ≥6 months	770 (1 RCT) follow-up: 6 months	⨁⨁◯◯ Low^b	RR 2.00 (0.28 to 14.09)	4 per 1,000	4 more per 1,000 (3 fewer to 51 more)	MID (precision) = RR 0.80 – 1.25.
Adverse events (injuries and pain) at end of intervention	555 (5 RCTs) follow-up: mean 7 weeks	⨁◯◯◯ Very low^q^,^r	RD 0.03 (−0.07 to 0.13)	311 per 1,000	30 more per 1,000 (70 fewer to 130 more)^s	Precision calculated through Optimal Information Size (OIS) due to zero events in some studies. OIS determined power for the sample size = 0.12 (0.8-0.9 = serious, <0.8 = very serious).
Adverse events (injuries and pain) at ≥6 months	299 (3 RCTs) follow-up: mean 6 months	⨁⨁⨁⨁ High	RD 0.00 (−0.02 to 0.02)	0 per 1,000	0 fewer per 1,000 (20 fewer to 20 more)^s	Sample size used to determine precision: 75-150 = serious imprecision, <75 = very serious imprecision. MID (clinical importance) = 50 per 1,000.
Adverse events (other reported adverse events) at end of intervention	1736 (19 RCTs) follow-up: mean 6 weeks	⨁◯◯◯ Very low^q^,^r	RD 0.01 (−0.01 to 0.04)	87 per 1,000	10 more per 1,000 (10 fewer to 40 more)^s	Precision calculated through Optimal Information Size (OIS) due to zero events in some studies. OIS determined power for the sample size = 0.25 (0.8-0.9 = serious, <0.8 = very serious).
Adverse events (other reported adverse events) at ≥6 months	1274 (10 RCTs) follow-up: mean 5 months	⨁◯◯◯ Very low^q^,^r	RD 0.00 (−0.03 to 0.04)	113 per 1,000	0 fewer per 1,000 (30 fewer to 40 more)^s	Precision calculated through Optimal Information Size (OIS) due to zero events in some studies. OIS determined power for the sample size = 0.52 (0.8-0.9 = serious, <0.8 = very serious).

a: Downgraded by 2 increments as the majority of the evidence was of very high risk of bias (due to bias arising from the randomisation process, bias due to deviation from the intended intervention, bias due to missing outcome data and bias in measurement of the outcome)

b: Downgraded by 1 increment if the confidence interval crossed one MID or by 2 increments if the confidence interval crossed both MIDs

c: Downgraded by 2 increments as the majority of the evidence was of very high risk of bias (due to bias in the randomisation process, bias due to missing outcome data and bias in measurement of the reported result)

d: Downgraded by 2 increments as the majority of the evidence was of very high risk of bias (due to bias in measurement of the outcome and bias in selection of the reported result)

e: Downgraded by 1 increment as the majority of the evidence was of high risk of bias (due to bias in measurement of the outcome)

f: Downgraded by 1 or 2 increments because heterogeneity, unexplained by subgroup analysis

g: Downgraded by 1 increment as the majority of the evidence was of high risk of bias (due to a mixture of bias arising from the randomisation process, bias due to deviations from the intended intervention, bias due to missing outcome data, bias in measurement of the outcome and bias in selection of the reported result)

h: Downgraded by 1 increment as the majority of the evidence was of high risk of bias (due to a mixture of bias arising from the randomisation process, bias due to deviations from the intended intervention and bias due to missing outcome data)

i: Downgraded by 1 increment as the majority of the evidence was of high risk of bias (due to a mixture of bias arising from the randomisation process, bias due to deviations from the intended intervention, bias due to missing outcome data and bias in measurement of the outcome)

j: Downgraded by 1 increments due to outcome indirectness (as the majority of evidence was reported at a follow up of less than 6 months)

k: Downgraded by 2 increments as the majority of the evidence was of very high risk of bias (due to bias arising from the randomisation process, bias due to deviation from the intended intervention, bias due to missing outcome data, bias in measurement of the outcome and bias in selection of the reported result)

l: Downgraded by 1 increment as the majority of the evidence was of high risk of bias (due to a mixture of bias arising from the randomisation process, bias due to deviations from the intended intervention, bias due to missing outcome data and bias in measurement of the outcome)

m: Downgraded by 2 increments as the majority of the evidence was of very high risk of bias (due to bias due to deviation from the intended intervention and bias due to missing outcome data)

n: Downgraded by 1 increment as the majority of the evidence was of high risk of bias (due to a mixture of bias arising from the randomisation process and bias in measurement of the outcome)

o: Downgraded by 1 increment as the majority of the evidence was of high risk of bias (due to bias in measurement of the outcome)

p: Downgraded by 1 increment as the majority of the evidence was of high risk of bias (due to a mixture of bias arising from the randomisation process, bias due to deviations from the intended intervention and bias due to missing outcome data)

q: Downgraded for heterogeneity due to conflicting number of events in different studies (zero events in one or more studies)

r: Downgraded by 1 to 2 increments for imprecision due to zero events and small sample size

s: Absolute effect calculated by risk difference due to zero events in at least one arm of one study

Table 4Health economic evidence profile: Robot-assisted arm training versus usual care

Study

Applicability

Limitations

Other comments

Incremental cost

Incremental effects

Cost effectiveness

Uncertainty

Fernandez-Garcia 2021³¹

(UK)

Directly applicable

Minor limitations^(a)

Within-trial analysis of RATULS RCT⁸⁸ (n=768)
Cost-utility analysis (QALYs)
Population: adults with moderate or severe upper limb functional limitation as a result of first-ever stroke that had occurred between 1 week and 5 years before randomisation
Comparators:
1. Usual care (45 minutes with a physiotherapist or occupational therapist, 5 days a week)
2. More intensive – robot-assisted training (45 minutes per day, 3 times per week) plus usual care
3. More intensive – enhanced upper limb therapy (EULT) (45 minutes with a physiotherapist, 3 times per week) plus usual care.
Time horizon: 6 months

2-1: £1601^(b)

3-1: £741^(b)

3-2: Saves −£936^(b)

2-1: 0.00 QALYs

3-1: 0.01 QALYs

3-2: 0.02 QALYs

2-1: More intensive rehabilitation (robot arm training) was dominated by usual care.

3-1 (More intensive (EULT) vs usual care): £74,100 per QALY gained

3-2: EULT dominated Robot (lower costs and higher QALYs)

Probability cost effective (£20K threshold):

usual care 81%
more intensive (robot) 0%
more intensive (EULT) 19%

Sensitivity analyses around missing data and robot costs did not change conclusions.

Extrapolation of data to 12-month time horizon made more intensive rehabilitation (EULT) cost effective compared to usual care (£6,095; probability cost effective 55%). More intensive (robot) remained dominated by usual care.

Remy-Neris 2021⁴⁹

Partially applicable^(c)

Potentially serious limitations^(d)

Within-trial analysis of an RCT (n=215) included in the clinical review (same paper) with no modelled extrapolation.
Cost-utility analysis (QALYs)
Population: Adults, 3 weeks to 3 months post-stroke, with an FMA score of 10 to 40 points.
Comparators:
1)
Control group (n=108) was provided with usual rehabilitation for 1 hour, 5 days per week plus an additional daily hour of self-rehabilitation consisting of basic stretching and active exercises for 4 weeks.
2)
Exo group (n=107) was provided with usual rehabilitation for 1 hour, 5 days per week plus an additional daily hour of self-rehabilitation consisting of gravity-supported, games-based training using an exoskeleton (Armeo®Spring) for 4 weeks
Follow-up: 12 months

2−1: Saves £99^(e)

2−1^(f): 0.01 QALYs

Results suggested that the Exo group intervention dominates usual care (lower costs and higher QALYs), however total costs and QALY gains were not statistically significant between groups.

Probability of cost effective (£20K/£30K threshold): NR

Results were robust to probabilistic sensitivity analysis, where uncertainty on the ICER was described using 1000 bootstrap replications on the cost-effectiveness plane.

: Abbreviations: 95% CI= 95% confidence interval; EQ-5D-3L= EuroQol 5 dimensions 3 levels (scale: 0.0 [death] to 1.0 [full health], negative values mean worse than death); FIM= functional independence measure (scale 0-18, higher values are better); FMA UE= Fugl-Meyer Assessment Upper Extremity (scale 0-66, higher scores are better); ICER= incremental cost-effectiveness ratio; NA= not applicable; NR= not reported; QALYs= quality-adjusted life years; RCT= randomised controlled trial; SIS hand function= stroke Impact Scale - hand function domain (scale 0-100, higher values are better).

(a): Within-trial analysis based on RATULS RCT and so only reflects this study and not the wider evidence base identified in the clinical review.

(b): 2018 UK pounds. Cost components incorporated: intervention costs, follow-up costs, primary care, therapy and community-based, services, secondary care, residential and nursing home care, social services, medication costs. Unit costs were taken from 2017/18 NHS reference costs and 2017 PSSRU unit costs (which were inflated to 2018 prices using the Bank of England inflator⁷)

(c): French healthcare system may not reflect current UK NHS context. EQ-5D-3L French tariff was used to estimate QALYs but NICE reference case specifies that the UK tariff is preferred.

(d): Within-trial analysis based on a single-blinded RCT, therefore results only reflect this study and not the wider evidence base identified in the clinical review. References for unit costs were not reported which limits interpretation of results for UK context. Probability that intervention was cost-effective at £20K threshold was not reported.

(e): 2018 euros (€) converted to UK pounds purchasing power parities.⁸¹ References for unit costs were not reported but 2018 was assumed based on the study completion date. No significant between-group differences were reported for total costs (p=0.99). Cost components incorporated: Armeo®Spring exoskeleton (device cost, 5-year linear depreciation, maintenance, and physical therapist for patient training). Resource use estimates included inpatient rehabilitation days, outpatient physiotherapy, GP and specialist consultations and transportation costs.

(f): Mean difference taken from Figure 4 of guideline clinical review. There were no significant between-group differences in changes for any of the reported outcomes at any time point (p>0.05).

Table 5Equivalent annual cost or equivalent annuity from Rodgers 2019⁸⁸

Year	Discount factor at 3.5%^(a)	Equivalent annual cost of £263,084^(b) (£)
1	0.9662	272,292
2	0.9335	138,487
3	0.9019	93,904
4	0.8714	71,625
5	0.8420	58,268

(a): Discount factor (Dⁿ) = 1(1 + r)ⁿ, where r = discounting rate (e.g. 3.5%)

(b): Equivalent annual cost (A_n) = r/(1 – D_n).

Table 6Illustrative cost of the MIT-Manus robotic gym per session from Rodgers 2019⁸⁸

Cost of robot per session	Cost (£)
Capital
Opportunity cost of the capital (£58,268 × 5)^*	291,340
Annual cost of robotic gym	58,268
Annual cost of robotic gym per week (assume 52 weeks)	1121
Cost of robot per session – assuming an average of seven sessions per day	32
Maintenance
Annual maintenance costs	16,234
Maintenance costs per week (52 weeks)	312
Maintenance costs per session (35 sessions in 1 week)	8.92
Total
Robotic gym cost per session (capital + maintenance)	41

*: Authors did not provide rationale for applying the discounted 5-year annual cost to opportunity cost estimations.

Table 7Unit costs of health care professionals who may be involved in delivering robot-assisted arm training interventions

Resource	Cost per working hour (hospital / community)^(a)	Source
Band 6 PT/OT	£53 / £55	PSRRU 2021⁵¹
Band 7 PT/OT	£64 / £66	PSRRU 2021⁵¹
Rehabilitation assistant	£33/£32	PSRRU 2021⁵¹, estimated based on agenda for change band 3 salary^(b)

: Abbreviations: OT= occupational therapist; PT= physiotherapist; PSSRU= personal social services research unit.

(a): Note: Costs per working hour include salary, salary oncosts, overheads (management and other non-care staff costs including administration and estates staff), capital overheads and qualification costs.

(b): Band 3 PT/OT not in PSSRU 2021 so salary was assumed to equal Band 3 Mean annual basic pay per FTE for administration and estates staff, NHS England (PSSRU2021 p.149).

Table 11Studies excluded from the clinical review

Study	Code [Reason]
(2020) Correction…McCabe J, Monkiewicz M, Holcomb J, et al. Comparison of Robotics, Functional Electrical Stimulation, and Motor Learning Methods for Treatment of Persistent Upper Extremity Dysfunction After Stroke: A Randomized Controlled Trial. Archives of Physical Medicine and Rehabilitation. 2015;96(6):981-990. Archives of Physical Medicine & Rehabilitation 101(4): 730–730 [PubMed: 25461822]	- Correction only
(2019) Does emg-driven robotic treatment have effect for the recovery of the hand 9 years after stroke?. Gait & Posture 73: 362–363	- Full text paper not available
Adamovich, S. V. (2018) Robot assisted virtual rehabilitation for the hand post stroke (RAVR).	- Full text paper not available
Adomaviciene, A., Daunoraviciene, K., Kubilius, R. et al. (2019) Influence of New Technologies on Post-Stroke Rehabilitation: A Comparison of Armeo Spring to the Kinect System. Medicina 55(4): 09 [PMC free article: PMC6524064] [PubMed: 30970655]	- Comparator in study does not match that specified in this review protocol
Aguiar, L. T., Nadeau, S., Martins, J. C. et al. (2020) Efficacy of interventions aimed at improving physical activity in individuals with stroke: a systematic review. Disability & Rehabilitation 42(7): 902–917 [PubMed: 30451539]	- Systematic review used as source of primary studies
Akcay, S, Karagozoglu, Coskunsu D, Erkan, Ogul O et al. (2020) The effect of robotic rehabilitation for recovery of hand functions in patients with acute stroke: Pilot study. Gait and Posture 81(s1): 6–7.	- Conference abstract
Ambrosini, E., Gasperini, G., Zajc, J. et al. (2021) A Robotic System with EMG-Triggered Functional Eletrical Stimulation for Restoring Arm Functions in Stroke Survivors. Neurorehabilitation & Neural Repair 35(4): 334–345 [PubMed: 33655789]	- Study does not contain an intervention relevant to this review protocol
Anonymous (2020) Correction: Comparison of Robotics, Functional Electrical Stimulation, and Motor Learning Methods for Treatment of Persistent Upper Extremity Dysfunction After Stroke: A Randomized Controlled Trial (Archives of Physical Medicine and Rehabilitation (2015) 96(6) (981-990), (S0003999314012283), (10.1016/j.apmr.2014.10.022)). Archives of Physical Medicine and Rehabilitation 101(4): 730	- Correction only
Arya, K. N.; Pandian, S.; Puri, V. (2018) Rehabilitation methods for reducing shoulder subluxation in post-stroke hemiparesis: a systematic review. Topics in Stroke Rehabilitation 25(1): 68–81 [PubMed: 29017429]	- Systematic review used as source of primary studies
Bajaj, P. and Contractor, A. (2022) EFFICACY OF UPPER EXTREMITY ROBOTIC REHABILITATION IN ADDITION TO CONVENTIONAL REHABILITATION IN FUNCTIONAL IMPROVEMENT IN CHRONIC STROKE IN A TERTIARY CARE HOSPITAL IN INDIA. International Journal of Stroke 17(3supplement): 127	- Conference abstract
Baniqued, P. D. E., Stanyer, E. C., Awais, M. et al. (2021) Brain-computer interface robotics for hand rehabilitation after stroke: a systematic review. Journal of Neuroengineering & Rehabilitation 18(1): 15 [PMC free article: PMC7825186] [PubMed: 33485365]	- Systematic review used as source of primary studies
Bayindir, O; Akyuz, G; Sekban, N (2022) The effect of adding robot-assisted hand rehabilitation to conventional rehabilitation program following stroke: a randomized-controlled study. Turkish journal of physical medicine and rehabilitation 68(2): 254–261 [PMC free article: PMC9366479] [PubMed: 35989963]	- Data not reported in an extractable format or a format that can be analysed Medians and interquartile ranges
Bernhardt, J and Mehrholz, J (2019) Robotic-assisted training after stroke: RATULS advances science. Lancet 394(10192): 6–8. [PubMed: 31128923]	- Conference abstract
Bressi, F., Bravi, M., Campagnola, B. et al. (2020) Robotic treatment of the upper limb in chronic stroke and cerebral neuroplasticity: a systematic review. Journal of Biological Regulators & Homeostatic Agents 34(5suppl3): 11–44. Technology in Medicine [PubMed: 33386032]	- Systematic review used as source of primary studies
Bumin, G.; Colak, F.D.; Yasar, E. (2022) THE EFFECT OF UPPER EXTREMITY ROBOTIC REHABILITATION ON ACTIVITY PERFORMANCE IN INDIVIDUALS WITH STROKE. International Journal of Stroke 17(3supplement): 252–253	- Conference abstract
Burridge, Jane and Hughes, Ann-Marie (2020) Robot-assisted training offers little improvement in severe arm weakness and functions after stroke. Frontline (20454910) 26(1): 42–43	- Conference abstract
Cantillo-Negrete, J., Carino-Escobar, R. I., Carrillo-Mora, P. et al. (2021) Brain-Computer Interface Coupled to a Robotic Hand Orthosis for Stroke Patients’ Neurorehabilitation: A Crossover Feasibility Study. Frontiers in Human Neuroscience 15 (no pagination) [PMC free article: PMC8215105] [PubMed: 34163342]	- Study design not relevant to this review protocol
Carpinella, I, Lencioni, T, Bowman, T et al. (2019) Planar robotic training versus arm-specific physiotherapy: effects on arm function and motor strategies in post-stroke subjects. Gait and Posture 74(s): 7	- Full text paper not available
Cecchi, F., Germanotta, M., Macchi, C. et al. (2021) Age is negatively associated with upper limb recovery after conventional but not robotic rehabilitation in patients with stroke: a secondary analysis of a randomized-controlled trial. Journal of Neurology 268(2): 474–483 [PubMed: 32844309]	- Secondary publication of an included study that does not provide any additional relevant information
Chen, Z., Wang, C., Fan, W. et al. (2020) Robot-Assisted Arm Training versus Therapist-Mediated Training after Stroke: A Systematic Review and Meta-Analysis. Journal of Healthcare Engineering 2020: 8810867 [PMC free article: PMC7641296] [PubMed: 33194159]	- Systematic review used as source of primary studies
Chen, Z., Xia, N., He, C. et al. (2021) Action observation treatment-based exoskeleton (AOT-EXO) for upper extremity after stroke: study protocol for a randomized controlled trial. Trials [Electronic Resource] 22(1): 222 [PMC free article: PMC7981809] [PubMed: 33743788]	- Study design not relevant to this review protocol
Chien, W. T., Chong, Y. Y., Tse, M. K. et al. (2020) Robot-assisted therapy for upper-limb rehabilitation in subacute stroke patients: A systematic review and meta-analysis. Brain and Behavior 10(8): e01742 [PMC free article: PMC7428503] [PubMed: 32592282]	- Systematic review used as source of primary studies
Chinembiri, B. (2019) Comparing the effects of Fourier M2 robotic rehabilitation machine combined with conventional occupational therapy on hand function and quality of life in patients whose arms have been affected by a recent first stroke.	- Full text paper not available
Chua, K., Kuah, C., Ng, C. et al. (2018) Clinical and kinematic evaluation of the H-Man arm robot for post-stroke upper limb rehabilitation: preliminary findings of a randomised controlled trial. Annals of physical and rehabilitation medicine	- Full text paper not available
Comino-Suarez, N., Moreno, J. C., Gomez-Soriano, J. et al. (2021) Transcranial direct current stimulation combined with robotic therapy for upper and lower limb function after stroke: a systematic review and meta-analysis of randomized control trials. Journal of Neuroengineering & Rehabilitation 18(1): 148 [PMC free article: PMC8474736] [PubMed: 34565399]	- Systematic review used as source of primary studies
D’Anci, K. E., Uhl, S., Oristaglio, J. et al. (2019) Treatments for Poststroke Motor Deficits and Mood Disorders: A Systematic Review for the 2019 U.S. Department of Veterans Affairs and U.S. Department of Defense Guidelines for Stroke Rehabilitation. Annals of Internal Medicine 171(12): 906–915 [PubMed: 31739315]	- Systematic review used as source of primary studies
Da-Silva, R. H.; Moore, S. A.; Price, C. I. (2018) Self-directed therapy programmes for arm rehabilitation after stroke: a systematic review. Clinical Rehabilitation 32(8): 1022–1036 [PubMed: 29756513]	- Systematic review used as source of primary studies
de Sousa, D. G., Harvey, L. A., Dorsch, S. et al. (2018) Interventions involving repetitive practice improve strength after stroke: a systematic review. Journal of Physiotherapy 64(4): 210–221 [PubMed: 30245180]	- Systematic review used as source of primary studies
de-la-Torre, R., Ona, E. D., Balaguer, C. et al. (2020) Robot-Aided Systems for Improving the Assessment of Upper Limb Spasticity: A Systematic Review. Sensors 20(18): 14 [PMC free article: PMC7570987] [PubMed: 32937973]	- Systematic review used as source of primary studies
Dehem, S., Gilliaux, M., Stoquart, G. et al. (2018) Effectiveness of upper limb robotic-assisted therapy in the early phase of stroke rehabilitation: a single-blind, randomised, controlled trial. Annals of physical and rehabilitation medicine [PubMed: 31028900]	- Duplicate reference
Dixit, S. and Tedla, J. S. (2019) Effectiveness of robotics in improving upper extremity functions among people with neurological dysfunction: a systematic review. International Journal of Neuroscience 129(4): 369–383 [PubMed: 30311823]	- Systematic review used as source of primary studies
Duret, C. (2018) Robotic rehabilitation of the upper limb after a stroke (ROBOASSIST).	- Full text paper not available
Ellis, M. D., Carmona, C., Drogos, J. et al. (2018) Progressive Abduction Loading Therapy with Horizontal-Plane Viscous Resistance Targeting Weakness and Flexion Synergy to Treat Upper Limb Function in Chronic Hemiparetic Stroke: A Randomized Clinical Trial. Frontiers in neurology [electronic resource]. 9: 71 [PMC free article: PMC5825888] [PubMed: 29515514]	- Study does not contain an intervention relevant to this review protocol
Esquenazi, A., Lee, S., Watanabe, T. et al. (2018) Abstract edited–Supplemental therapeutic conventional vs. robotic upper limb exercise in acute stroke rehabilitation: a randomized, blinded assessor study. Annals of physical and rehabilitation medicine 61(suppl1): e95	- Full text paper not available
Ferreira, Fmrm, Chaves, M. E. A., Oliveira, V. C. et al. (2018) Effectiveness of robot therapy on body function and structure in people with limited upper limb function: A systematic review and meta-analysis. PLoS ONE [Electronic Resource] 13(7): e0200330 [PMC free article: PMC6042733] [PubMed: 30001417]	- Systematic review used as source of primary studies
Fonte, C., Varalta, V., Rocco, A. et al. (2021) Combined transcranial Direct Current Stimulation and robot-assisted arm training in patients with stroke: a systematic review. Restorative Neurology & Neuroscience 39(6): 435–446 [PubMed: 34974446]	- Systematic review used as source of primary studies
García-Rudolph, A.; Bernabeu-Guitart, M.; Opisso, E. (2020) [Intensities in the application of robotic technologies in upper extremity rehabilitation after a stroke: a systematic review of randomised controlled clinical trials]. Revista de neurologia 70(12): 434–443 [PubMed: 32500522]	- Full text paper not available
Gasperini, G., Rossini, M., Proserpio, D. et al. (2018) Hybrid robotic system combining passive exoskeleton and functional electrical stimulation for upper limb stroke rehabilitation: preliminary results of the retrainer multi-center randomized controlled trial. Annals of physical and rehabilitation medicine	- Conference abstract
Germanotta, M, Pecchioli, C, Cruciani, A et al. (2019) Efficacy of upper limb robot-assisted therapy compared with conventional therapy in stroke patients: preliminary results on a daily task assessed by means of motion analysis. Gait and Posture 74(s): 18	- Full text paper not available
Hameed, Husamuldeen K., Hassan, Wan Zuha Wan, Shafie, Suhaidi et al. (2020) A Review on Surface Electromyography-Controlled Hand Robotic Devices Used for Rehabilitation and Assistance in Activities of Daily Living. Journal of Prosthetics & Orthotics (JPO) 32(1): 3–13	- Review article but not a systematic review
Hayward, K. S., Kramer, S. F., Thijs, V. et al. (2019) A systematic review protocol of timing, efficacy and cost effectiveness of upper limb therapy for motor recovery post-stroke. Systematic Reviews 8(1): 187 [PMC free article: PMC6657039] [PubMed: 31345263]	- Study design not relevant to this review protocol
Hsieh, Y. W., Lin, K. C., Wu, C. Y. et al. (2018) Comparison of proximal versus distal upper-limb robotic rehabilitation on motor performance after stroke: a cluster controlled trial. Scientific Reports 8(1): 2091 [PMC free article: PMC5794971] [PubMed: 29391492]	- Study design not relevant to this review protocol
Hu, X. (2020) Upper limb rehabilitation after stroke assisted with a hybrid electrical stimulation (ES)-robot system.	- Full text paper not available
Hung, C. S., Hsieh, Y. W., Wu, C. Y. et al. (2019) Comparative Assessment of Two Robot-Assisted Therapies for the Upper Extremity in People With Chronic Stroke. American Journal of Occupational Therapy 73(1): 7301205010p1–7301205010p9 [PubMed: 30839256]	- Data not reported in an extractable format or a format that can be analysed Reports median and interquartile ranges only
Hung, J. W., Chen, Y. W., Chen, Y. J. et al. (2021) The Effects of Distributed vs. Condensed Schedule for Robot-Assisted Training with Botulinum Toxin A Injection for Spastic Upper Limbs in Chronic Post-Stroke Subjects. Toxins 13(8): 01 [PMC free article: PMC8402581] [PubMed: 34437410]	- Comparator in study does not match that specified in this review protocol
Hung, J. W., Wu, C. Y., Chang, K. C. et al. (2018) Comparative hybrid effects of combining botulinum toxin A injection with bilateral robot-assisted, mirror or task-oriented therapy for upper extremity spasticity in patients with chronic stroke. Annals of physical and rehabilitation medicine	- Conference abstract
Iamsirikij, C. (2018) Effects of upper limb rehabilitation robot EnMotion® in subacute stroke patients: a single blind randomized controlled trial (robotic rehab, stroke, subacute).	- Full text paper not available
Kagiyama, T. and Mukae, N. (2020) Clinical research for efficacy and safety of hand-finger rehabilitation robot SMOVE in the goods operation training for the patients with upper limb paresis after recovery stage stroke patients: a pilot study under single center, open-label, randomized, standard therapy controlled trial - (SMOVE pilot study_02).	- Full text paper not available
Kang, T. W., Oh, D. W., Lee, J. H. et al. (2018) Effects of integrating rhythmic arm swing into robot-assisted walking in patients with subacute stroke: a randomized controlled pilot study. International Journal of Rehabilitation Research 41(1): 57–62 [PubMed: 29140826]	- Study does not contain an intervention relevant to this review protocol
Keeling, A. B., Piitz, M., Semrau, J. A. et al. (2021) Robot enhanced stroke therapy optimizes rehabilitation (RESTORE): a pilot study. Journal of NeuroEngineering and Rehabilitation 18(1) [PMC free article: PMC7819212] [PubMed: 33478563]	- Study design not relevant to this review protocol Non-randomised study with sufficient randomised evidence included in the review
Khalid, S., Alnajjar, F., Gochoo, M. et al. (2021) Robotic assistive and rehabilitation devices leading to motor recovery in upper limb: a systematic review. Disability & Rehabilitation Assistive Technology: 1–15 [PubMed: 33861684]	- Systematic review used as source of primary studies
Kim, S. B., Lee, K. W., Lee, J. H. et al. (2018) Effect of Combined Therapy of Robot and Low-Frequency Repetitive Transcranial Magnetic Stimulation on Hemispatial Neglect in Stroke Patients. Annals of Rehabilitation Medicine 42(6): 788–797 [PMC free article: PMC6325312] [PubMed: 30613071]	- Comparator in study does not match that specified in this review protocol
Kim, W. S. (2018) Anodal tDCS over the contralesional hemisphere with robotic arm training in subacute stroke patients with severe upper limb hemiparesis.	- Full text paper not available
Kuo, L-C, Yang, K-C, Lin, Y-C et al. (2022) Internet of Things (IoT) Enables Robot-Assisted Therapy as a Home Program for Training Upper Limb Functions in Chronic Stroke: a Randomized Control Crossover Study. Archives of physical medicine and rehabilitation [PubMed: 36122608]	- Study design not relevant to this review protocol Crossover trial that only reports results for all of the participants together (does not report results for the first trial period only) - therefore it is not possible to extract results by the methods described in the Cochrane review and so the study is excluded
Lee, H. C., Kuo, F. L., Lin, Y. N. et al. (2021) Effects of Robot-Assisted Rehabilitation on Hand Function of People With Stroke: A Randomized, Crossover-Controlled, Assessor-Blinded Study. American Journal of Occupational Therapy 75(1): 7501205020p1–7501205020p11 [PubMed: 33399050]	- Study design not relevant to this review protocol
Lee, S. H., Kim, W. S., Park, J. et al. (2020) Effects of anodal transcranial direct current stimulation over the contralesional hemisphere on motor recovery in subacute stroke patients with severe upper extremity hemiparesis: Study protocol for a randomized controlled trial. Medicine 99(14): e19495 [PMC free article: PMC7220659] [PubMed: 32243365]	- Study design not relevant to this review protocol
Leem, Min Jeong, Kim, Gyu Seong, Kim, Kee Hoon et al. (2019) Predictors of functional and motor outcomes following upper limb robot-assisted therapy after stroke. International Journal of Rehabilitation Research 42(3): 223–228 [PubMed: 30932930]	- Study design not relevant to this review protocol
Lin, I. H., Tsai, H. T., Wang, C. Y. et al. (2019) Effectiveness and Superiority of Rehabilitative Treatments in Enhancing Motor Recovery Within 6 Months Poststroke: A Systemic Review. Archives of Physical Medicine & Rehabilitation 100(2): 366–378 [PubMed: 30686327]	- Systematic review used as source of primary studies
Lin, J. C. (2018) Robot-assisted hand rehabilitation for patients with stroke.	- Full text paper not available
Lin, K. C. (2018) Synergistic bilateral upper-limb stroke rehabilitation based on robotic priming technique.	- Full text paper not available
Liu, L. Y.; Li, Y.; Lamontagne, A. (2018) The effects of error-augmentation versus error-reduction paradigms in robotic therapy to enhance upper extremity performance and recovery post-stroke: a systematic review. Journal of Neuroengineering & Rehabilitation 15(1): 65 [PMC free article: PMC6033222] [PubMed: 29973250]	- Systematic review used as source of primary studies
Lo, K.; Stephenson, M.; Lockwood, C. (2019) The economic cost of robotic rehabilitation for adult stroke patients: a systematic review. JBI Database Of Systematic Reviews And Implementation Reports 17(4): 520–547 [PubMed: 30973526]	- Systematic review used as source of primary studies
Marotta, N., Demeco, A., Moggio, L. et al. (2021) The adjunct of transcranial direct current stimulation to Robot-assisted therapy in upper limb post-stroke treatment. Journal of Medical Engineering & Technology 45(6): 494–501 [PubMed: 34038313]	- Systematic review used as source of primary studies
Mashizume, Y; Zenba, Y; Takahashi, K (2020) Novel Mechanism of Action: Efficacy of Upper Extremity Robotic Therapy For Chronic Stroke Patients in Occupational Therapy. Archives of Physical Medicine and Rehabilitation 101(11): e98	- Conference abstract
Mehrholz, J., Pollock, A., Pohl, M. et al. (2020) Systematic review with network meta-analysis of randomized controlled trials of robotic-assisted arm training for improving activities of daily living and upper limb function after stroke. Journal of Neuroengineering & Rehabilitation 17(1): 83 [PMC free article: PMC7325016] [PubMed: 32605587]	- Systematic review used as source of primary studies
Merians, A. S., Fluet, G. G., Qiu, Q. et al. (2020) Hand Focused Upper Extremity Rehabilitation in the Subacute Phase Post-stroke Using Interactive Virtual Environments. Frontiers in Neurology 11 (no pagination) [PMC free article: PMC7726202] [PubMed: 33324323]	- Study design not relevant to this review protocol
Meyer, S., Verheyden, G., Kempeneers, K. et al. (2021) Arm-Hand Boost Therapy During Inpatient Stroke Rehabilitation: A Pilot Randomized Controlled Trial. Frontiers in Neurology 12 (no pagination) [PMC free article: PMC7952763] [PubMed: 33716948]	- Comparator in study does not match that specified in this review protocol
Moggio, L., de Sire, A., Marotta, N. et al. (2021) Exoskeleton versus end-effector robot-assisted therapy for finger-hand motor recovery in stroke survivors: systematic review and meta-analysis. Topics in Stroke Rehabilitation: 1–12 [PubMed: 34420498]	- Systematic review used as source of primary studies
Morone, G., Palomba, A., Martino Cinnera, A. et al. (2021) Systematic review of guidelines to identify recommendations for upper limb robotic rehabilitation after stroke. European journal of physical & rehabilitation medicine. 57(2): 238–245 [PubMed: 33491943]	- Review article but not a systematic review
Mubin, O., Alnajjar, F., Jishtu, N. et al. (2019) Exoskeletons With Virtual Reality, Augmented Reality, and Gamification for Stroke Patients’ Rehabilitation: Systematic Review. JMIR Rehabilitation And Assistive Technologies 6(2): e12010 [PMC free article: PMC6779025] [PubMed: 31586360]	- Systematic review used as source of primary studies
Park, S. W.; Kim, J. H.; Yang, Y. J. (2018) Mental practice for upper limb rehabilitation after stroke: a systematic review and meta-analysis. International Journal of Rehabilitation Research 41(3): 197–203 [PubMed: 29912022]	- Systematic review used as source of primary studies
Patel, J., Fluet, G., Qiu, Q. et al. (2019) Intensive virtual reality and robotic based upper limb training compared to usual care, and associated cortical reorganization, in the acute and early sub-acute periods post-stroke: a feasibility study. Journal of Neuroengineering & Rehabilitation 16(1): 92 [PMC free article: PMC6637633] [PubMed: 31315612]	- Study design not relevant to this review protocol
Perini, G., Bertoni, R., Thorsen, R. et al. (2021) Sequentially applied myoelectrically controlled FES in a task-oriented approach and robotic therapy for the recovery of upper limb in post-stroke patients: A randomized controlled pilot study. Technology & Health Care 29(3): 419–429 [PubMed: 33386831]	- Study does not contain an intervention relevant to this review protocol
Perini, G, Lencioni, T, Bertoni, R et al. (2019) Rehabilitation of upper limb in chronic stroke patients: pilot study of functional and neuromotor outcome of a task oriented approach including MeCFES and robotic treatment. Gait and Posture 74(s): 29–30.	- Full text paper not available
Quaglia, D., Gasperi, M., Coser, R. et al. (2018) Robotic rehabilitation effect on upper limb recovery in post-acute stroke. Gait & Posture 66: S31–S32	- Conference abstract
Reis, S. B., Bernardo, W. M., Oshiro, C. A. et al. (2021) Effects of Robotic Therapy Associated With Noninvasive Brain Stimulation on Upper-Limb Rehabilitation After Stroke: Systematic Review and Meta-analysis of Randomized Clinical Trials. Neurorehabilitation & Neural Repair 35(3): 256–266 [PubMed: 33522417]	- Systematic review used as source of primary studies
Remy-Neris, O., Medee, B., Bensmail, D. et al. (2018) Rehabilitation robotics of the upper limb after stroke. The REM_AVC trial. Annals of physical and rehabilitation medicine	- Conference abstract
Rintala, A, Paivarinne, V, Hakala, S et al. (2019) Effectiveness of Technology-Based Distance Physical Rehabilitation Interventions for Improving Physical Functioning in Stroke: A Systematic Review and Meta-analysis of Randomized Controlled Trials. Archives of Physical Medicine and Rehabilitation 100(7): 1339–1358. [PubMed: 30529323]	- Systematic review used as source of primary studies
Rosenthal, O. (2018) Performance-based selective training for robot-mediated upper limb rehabilitation after stroke.	- Full text paper not available
Rosenthal, O., Wing, A. M., Wyatt, J. L. et al. (2019) Boosting robot-assisted rehabilitation of stroke hemiparesis by individualized selection of upper limb movements - A pilot study. Journal of NeuroEngineering and Rehabilitation 16(1) [PMC free article: PMC6425657] [PubMed: 30894192]	- Comparator in study does not match that specified in this review protocol
Rosenthal, Orna, Wing, Alan M., Wyatt, Jeremy L. et al. (2019) Correction to: Boosting robot-assisted rehabilitation of stroke hemiparesis by individualized selection of upper limb movements - a pilot study. 16: N.PAG-N.PAG [PMC free article: PMC6466732] [PubMed: 30987648]	- Correction only
Rozevink, S. G., Hijmans, J. M., Horstink, K. A. et al. (2021) Effectiveness of task-specific training using assistive devices and task-specific usual care on upper limb performance after stroke: a systematic review and meta-analysis. Disability & Rehabilitation Assistive Technology: 1–14 [PubMed: 34788166]	- Systematic review used as source of primary studies
Serrezuela, R. R., Quezada, M. T., Zayas, M. H. et al. (2020) Robotic therapy for the hemiplegic shoulder pain: a pilot study. Journal of Neuroengineering & Rehabilitation 17(1): 54 [PMC free article: PMC7178610] [PubMed: 32321536]	- Data not reported in an extractable format or a format that can be analysed
Shin, J. H. (2019) Effects of upper extremity rehabilitation robot and transcranial direct current stimulation in chronic stroke.	- Full text paper not available
Suarez-Escobar, M. and Rendon-Velez, E. (2018) An overview of robotic/mechanical devices for post-stroke thumb rehabilitation. Disability & Rehabilitation Assistive Technology 13(7): 683–703 [PubMed: 29334274]	- Review article but not a systematic review
Takebayashi, T., Takahashi, K., Amano, S. et al. (2018) Assessment of the Efficacy of ReoGo-J Robotic Training Against Other Rehabilitation Therapies for Upper-Limb Hemiplegia After Stroke: Protocol for a Randomized Controlled Trial. Frontiers in neurology [electronic resource]. 9: 730 [PMC free article: PMC6121101] [PubMed: 30210446]	- Study design not relevant to this review protocol
Takebayashi, T., Takahashi, K., Domen, K. et al. (2020) Impact of initial flexor synergy pattern scores on improving upper extremity function in stroke patients treated with adjunct robotic rehabilitation: A randomized clinical trial. Topics in Stroke Rehabilitation 27(7): 516–524 [PubMed: 32151236]	- Secondary publication of an included study that does not provide any additional relevant information
Terranova, T. T., Simis, M., Santos, A. C. A. et al. (2021) Robot-Assisted Therapy and Constraint-Induced Movement Therapy for Motor Recovery in Stroke: Results From a Randomized Clinical Trial. Frontiers in Neurorobotics 15: 684019 [PMC free article: PMC8335542] [PubMed: 34366819]	- Comparator in study does not match that specified in this review protocol
Terranova, T., Simis, M., Santos, A. et al. (2018) Comparing effects of constraint-induced movement therapy and robotic therapy: randomized clinical trial. Annals of physical and rehabilitation medicine	- Conference abstract
Tramontano, M., Morone, G., Palomba, A. et al. (2020) Effectiveness of a sensor-based technology in upper limb motor recovery in post-acute stroke neurorehabilitation: a randomized controlled trial. Journal of Biological Regulators & Homeostatic Agents 34(5suppl3): 165–174. Technology in Medicine [PubMed: 33386046]	- Full text paper not available
Tsuchimoto, S., Shindo, K., Hotta, F. et al. (2019) Sensorimotor Connectivity after Motor Exercise with Neurofeedback in Post-Stroke Patients with Hemiplegia. Neuroscience 416: 109–125 [PubMed: 31356896]	- Data not reported in an extractable format or a format that can be analysed
Valdes, B. A. and Van der Loos, H. F. M. (2018) Biofeedback vs. game scores for reducing trunk compensation after stroke: a randomized crossover trial. Topics in Stroke Rehabilitation 25(2): 96–113 [PubMed: 29078743]	- Comparator in study does not match that specified in this review protocol
Valkenborghs, Sarah R., Callister, Robin, Visser, Milanka M. et al. (2019) Interventions combined with task-specific training to improve upper limb motor recovery following stroke: a systematic review with meta-analyses. Physical Therapy Reviews 24(34): 100–117	- Systematic review used as source of primary studies
Wright, Zachary A., Majeed, Yazan A., Patton, James L. et al. (2020) Key components of mechanical work predict outcomes in robotic stroke therapy. Journal of NeuroEngineering & Rehabilitation (JNER) 17(1): 1–12 [PMC free article: PMC7175566] [PubMed: 32316977]	- Data not reported in an extractable format or a format that can be analysed
Wu, C. Y. (2020) Robot-assisted therapy combined with mirror priming in upper limb training in stroke.	- Full text paper not available
Wu, J., Cheng, H., Zhang, J. et al. (2021) Robot-Assisted Therapy for Upper Extremity Motor Impairment After Stroke: A Systematic Review and Meta-Analysis. Physical Therapy 101(4): 04 [PubMed: 33454787]	- Systematic review used as source of primary studies
Xu, Q., Li, C., Pan, Y. et al. (2020) Impact of smart force feedback rehabilitation robot training on upper limb motor function in subacute stage of stroke. NeuroRehabilitation [PubMed: 32741790]	- Full text paper not available
Yuan, R. and Wang, H. (2022) TU-173. The effect of upper limb rehabilitation robot training on the motor function and neuroelectrophysiology of stroke patients. Clinical Neurophysiology 141(supplement): 29	- Conference abstract
Yáñez-Sánchez, A. and Cuesta-Gómez, A. (2020) [Effectiveness of the Armeo ® device in the rehabilitation of the upper limb of stroke’s patients. A review of the literature]. Revista de neurologia 70(3): 93–102 [PubMed: 31994166]	- Full text paper not available

Evidence reviews for robot-assisted arm training

1. Robot-assisted arm training

1.1. Review question

1.1.1. Introduction

1.1.2. Summary of the protocol

1.1.3. Methods and process

1.1.4. Effectiveness evidence

1.1.4.1. Included studies

Indirectness

Inconsistency

1.1.4.2. Excluded studies

1.1.5. Summary of studies included in the effectiveness evidence

1.1.6. Summary of the effectiveness evidence

1.1.7. Economic evidence

1.1.7.1. Included studies

1.1.7.2. Excluded studies

1.1.8. Summary of included economic evidence

1.1.9. Economic model

1.1.10. Unit costs

1.1.11. Evidence statements

Effectiveness/Qualitative

Economic

1.1.12. The committee’s discussion and interpretation of the evidence

1.1.12.1. The outcomes that matter most

1.1.12.2. The quality of the evidence

1.1.12.3. Benefits and harms

1.1.12.4. Cost effectiveness and resource use

1.1.13. Recommendations supported by this evidence review

1.1.14. References

Appendices

Appendix A. Review protocols

Appendix B. Literature search strategies

Appendix C. Effectiveness evidence study selection

Appendix D. Effectiveness evidence

Appendix E. Forest plots

Appendix F. GRADE tables

Appendix G. Economic evidence study selection

Appendix H. Economic evidence tables

Appendix I. Health economic model

Appendix J. Excluded studies

Clinical studies

Health Economic studies

Table 1PICO characteristics of review question

Table 2Summary of studies included in the evidence review

Table 3Clinical evidence summary: Robot-assisted arm training compared to any other intervention

Table 4Health economic evidence profile: Robot-assisted arm training versus usual care

Table 5Equivalent annual cost or equivalent annuity from Rodgers 201988

Table 6Illustrative cost of the MIT-Manus robotic gym per session from Rodgers 201988

Table 7Unit costs of health care professionals who may be involved in delivering robot-assisted arm training interventions

Table 11Studies excluded from the clinical review

Table 5Equivalent annual cost or equivalent annuity from Rodgers 2019⁸⁸

Table 6Illustrative cost of the MIT-Manus robotic gym per session from Rodgers 2019⁸⁸