Robotic-assisted Rehabilitation of the Extremities

1. patients diagnosed with cerebral vascular accident,
2. effects of robot-assisted therapy for the upper limb, and
3. outcomes reported in terms of motor and/or functional recovery of the upper paretic limb. 

1.  37 patients were treated with RAGT, and
2. 30 were treated with regular physiotherapy. 

1. robotic therapy,
2. electrical stimulation or
3. "other" therapy. 

1. feasibility of incorporating the device into an inpatient rehabilitation program (compliance with training schedule, reduction in therapist time required and subject questionnaires) and
2. efficacy of the robotic rehabilitation for improving functional outcomes (Graded and Redefined Assessment of Strength, Sensibility and Prehension (GRASSP), action research arm test, grip dynamometry and range of motion).

1. some trials investigated people who were independent in walking at the start of the study,
2. these researchers found variations between the trials with respect to devices used and duration and frequency of treatment, and
3. some trials included devices with functional electrical stimulation.

1. the experimental group, which received combined HEP and HMP for 3 hours/day × 5 days × 8 weeks, or
2. the control group, which received HEP only at an identical dosage. 

In a prospective, open, blinded end-point, randomized, multi-center exploratory clinical trial, Takahashi et al (2016) assessed the effectiveness of robotic therapy as an adjuvant to standard therapy during post-stroke rehabilitation. A total of 60 individuals with mild-to-moderate hemiplegia 4 to 8 weeks post-stroke were randomized to receive standard therapy plus 40 minutes of either robotic or self-guided therapy for 6 weeks (7 days/week); UE impairment before and after intervention was measured using the Fugl-Meyer assessment, Wolf Motor Function Test, and Motor Activity Log.  Robotic therapy significantly improved Fugl-Meyer assessment flexor synergy (2.1 ± 2.7 versus -0.1 ± 2.4; p < 0.01) and proximal UE (4.8 ± 5.0 versus 1.9 ± 5.5; p < 0.05) compared with self-guided therapy.  No significant changes in Wolf Motor Function Test or Motor Activity Log were observed.  Robotic therapy also significantly improved Fugl-Meyer assessment proximal UE among low-functioning patients (baseline Fugl-Meyer assessment score of less than 30) and among patients with Wolf Motor Function Test greater than or equal to 120 at baseline compared with self-guided therapy (p < 0.05 for both).  The authors concluded that robotic therapy as an adjuvant to standard rehabilitation may improve UE recovery in moderately impaired post-stroke patients.  However, they stated that results of this exploratory study should be interpreted with caution.

In a RCT, Taveggia et al (2016) evaluated the effectiveness of robotic-assisted motion and activity in additional to Physical and Rehabilitation Medicine (PRM), of the upper limb in post-stroke inpatients. A total of 54 patients (57 % female, mean ± SD age of 71 ± 12 years), with upper limb function deficit post-stroke were included in this study.  The experimental group received a passive mobilization of the upper limb through the robotic device ARMEO Spring and the control group received PRM for 6 consecutive weeks (5 days/week) in addition to traditional PRM.  These investigators evaluated the impact on functional recovery (Functional Independence Measure-FIM scale), strength (ARM Motricity Index [MI]), spasticity (Modified Ashworth Scale [MAS]) and pain (Numeric Rating Pain Scale [NRPS]).  All patients were evaluated by a blinded observer using the outcomes tests at enrollment (T0), after the treatment (T1) and at follow-up 6 weeks later (T2).  Both control and experimental groups evidenced an improvement of the outcomes after the treatment (MI, MAS and NRPS with p < 0.05).  The experimental group showed further improvements after the follow-up (all outcomes with p < 0.01).  The authors concluded that in the treatment of pain, disability and spasticity in upper limb post-stroke, robot-assisted mobilization associated to PRM is as effective as traditional rehabilitation.  The main drawbacks of this study were its relatively small sample size and short-term follow-up.  These findings need to be validated by further research.

Pirondini et al (2016) noted that exoskeletons for lower and upper extremities have been introduced in neuro-rehabilitation because they can guide the patient's limb following its anatomy, covering many degrees of freedom and most of its natural workspace, and allowing the control of the articular joints. These researchers evaluated the possible use of a novel exoskeleton, the Arm Light Exoskeleton (ALEx), for robot-aided neuro-rehabilitation and examined the effects of some rehabilitative strategies adopted in robot-assisted training.  They studied movement execution and muscle activities of 16 upper limb muscles in 6 healthy subjects, focusing on end-effector and joint kinematics, muscle synergies, and spinal maps.  The subjects performed three dimensional (3D) point-to-point reaching movements, without and with the exoskeleton in different assistive modalities and control strategies.  The results showed that ALEx supported the upper limb in all modalities and control strategies: it reduced the muscular activity of the shoulder's abductors and it increased the activity of the elbow flexors.  The different assistive modalities favored kinematics and muscle coordination similar to natural movements, but the muscle activity during the movements assisted by the exoskeleton was reduced with respect to the movements actively performed by the subjects.  Moreover, natural trajectories recorded from the movements actively performed by the subjects seemed to promote an activity of muscles and spinal circuitries more similar to the natural one.  The authors concluded that the preliminary analysis on healthy subjects supported the use of ALEx for post-stroke upper limb robotic-assisted rehabilitation, and it provided clues on the effects of different rehabilitative strategies on movement and muscle coordination.

Morone and associates (2016) noted that patients affected by mild stroke benefit more from physiological over-ground walking training than walking-like training performed in place using specific devices.  These researchers evaluated the effects of over-ground robotic walking training performed with the servo-assistive robotic rollator (i-Walker) on walking, balance, gait stability and falls in a community setting in patients with mild subacute stroke.  A total of 44 patients were randomly assigned to 2 different groups that received the same therapy in 2 daily 40-min sessions 5 days a week for 4 weeks; 20 sessions of standard therapy were performed by both groups.  In the other 20 sessions the subjects enrolled in the i-Walker-Group (iWG) performed with the i-Walker and the Control-Group patients (CG) performed the same amount of conventional walking oriented therapy.  Clinical and instrumented gait assessments were made pre- and post-treatment.  The follow-up observation consisted of recording the number of fallers in the community setting after 6 months.  Treatment effectiveness was higher in the iWG group in terms of balance improvement (Tinetti: 68.4 ± 27.6 % versus 48.1 ± 33.9 %, p = 0.033) and 10-meter and 6-min timed walking tests (significant interaction between group and time: F(1,40) = 14.252, p = 0.001; and F(1,40) = 7.883, p = 0.008, respectively).  When measured, latero-lateral upper body accelerations were reduced in iWG (F = 4.727, p = 0.036), suggesting increased gait stability, which was supported by a reduced number of falls at home.  The authors concluded that a robotic servo-assisted i-Walker improved walking performance and balance in patients affected by mild/moderate stroke, leading to increased gait stability and reduced falls in the community.

The authors stated that 2 main limitations were that the study was registered only after the end of data collection and that the follow-up assessment was limited to records concerning falls and no clinical or instrumental tool was used to assess balance and walking capabilities.  Further, the number of falls was self-reported by patients; therefore, it was conceivable that subjects under-reported the incidence of falls.  Another drawback was that it was unclear whether the improvements obtained using the i-Walker could also have been achieved by the control group if their training had been performed in more variable contexts.  In any case, this would have been very difficult to obtain because it would have involved greater effort on the part of the physiotherapist (or the intervention of more than 1 therapist) and could have led to safety problems related to patients’ falling (or fear of falling).  The authors stated that future research should evaluate the effect of the i-Walker in a larger sample and should include a follow-up group; it would be useful to explore the usefulness of the i-Walker as an assistive device for use in the home.

Lehman and colleagues (2017) stated that RAGT affords an opportunity to increase walking practice with mechanical assistance from robotic devices, rather than therapists, where the child may not be able to generate a sufficient or correct motion with enough repetitions to promote improvement.  However the devices are expensive and clinicians and families need to understand if the approach is worthwhile for their children, and how it may be best delivered.  These researchers appraised the existing evidence for the effectiveness of RAGT for pediatric gait disorders, including modes of delivery and potential benefit.  A total of 6 databases were searched from 1980 to October 2016, using relevant search terms.  Any clinical trial that evaluated a clinical aspect of RAGT for children/adolescents with altered gait was selected for inclusion.  Data were extracted following the PRISMA approach.  A total of 17 trials were identified, assessed for level of evidence and risk of bias, and appropriate data extracted for reporting; 3 RCTs were identified, with the remainder of lower level design.  Most individual trials reported some positive benefits for RAGT with children with CP, on activity parameters such as standing ability, walking speed and distance.  However a meta-analysis of the 2 eligible RCTs did not confirm this finding (p = 0.72).  Training schedules were highly variable in duration and frequency and adverse events (AEs) were either not reported or were minimal.  There was a paucity of evidence for diagnoses other than CP.  The authors concluded that there is weak and inconsistent evidence regarding the use of RAGT for children with gait disorders.  If clinicians (and their clients) choose to use RAGT, they should monitor individual progress closely with appropriate outcome measures including monitoring of AEs.  They stated that further research is needed using higher level trial design, increased numbers, in specific populations and with relevant outcome measures to both confirm effectiveness and clarify training schedules.

Dierick and co-workers (2017) compared gait and posture outcome measures between ambulatory hemorrhagic patients and ischemic patients, who received a similar 4 weeks' intervention blending a conventional bottom-up physiotherapy approach and an exoskeleton top-down RAGT approach with Lokomat.  A total of 40 adult hemiparetic stroke inpatient subjects were recruited: 20 hemorrhagic and 20 ischemic, matched by age, gender, side of hemisphere lesion, stroke severity, and locomotor impairments.  Functional Ambulation Category, Postural Assessment Scale for Stroke, Tinetti Performance Oriented Mobility Assessment, 6- Minute Walk Test (6MWT), Timed Up and Go and 10-Meter Walk Test were performed before and after a 4-week long intervention.  Functional gains were calculated for all tests.  Hemorrhagic and ischemic subjects showed significant improvements in Functional Ambulation Category (p < 0.001 and p = 0.008, respectively), Postural Assessment Scale for Stroke (p < 0.001 and p = 0.003), 6MWT (p = 0.003 and p = 0.015) and 10-Meter Walk Test (p = 0.001 and p = 0.024).  Ischemic patients also showed significant improvements in Timed Up and Go.  Significantly greater mean Functional Ambulation Category and Tinetti Performance Oriented Mobility Assessment gains were observed for hemorrhagic compared to ischemic, with large (dz = 0.81) and medium (dz = 0.66) effect sizes, respectively.  The authors concluded that overall, both groups exhibited quasi similar functional improvements and benefits from the same type, length and frequency of blended conventional physiotherapy and RAGT protocol.  They stated that the use of intensive treatment plans blending top-down physiotherapy and bottom-up robotic approaches is promising for post-stroke rehabilitation.

Borboni and colleagues (2017) examined if passive robotic-assisted hand motion, in addition to standard rehabilitation, would reduce hand pain, edema, or spasticity in all patients following acute stroke, in patients with and without hand paralysis.  A total of 35 participants, aged 45 to 80 years, with functional impairments of their upper extremities after a stroke were recruited for the study from September 2013 to October 2013.  One group consisted of 16 patients (mean age ± SD of 68 ± 9 years) with full paralysis and the other groups included 14 patients (mean age ± SD of 67 ± 8 years) with partial paralysis.  Patients in the both groups used the Gloreha device for passive mobilization of the hand twice-daily for 2 consecutive weeks.  The primary outcome measure was hand edema; secondary outcome measures included pain intensity and spasticity.  All outcome measures were collected at baseline and immediately after the intervention (2 weeks).  Analysis of variance revealed that the partial paralysis group experienced a significantly greater reduction of edema at the wrist (p = 0.005) and pain (p = 0.04) when compared with the full paralysis group.  Other outcomes were similar for the groups.  The authors concluded that the findings of this study suggested that the partial paralysis group experienced a significantly greater reduction of edema at the wrist and pain when compared with the full paralysis group; however, the reduction in pain did not meet the threshold of a minimal clinically important difference.

In a systematic review, Mehrholz and associates (2017) compared the effectiveness of BWSTT and RAGT with over-ground gait training and other forms of physiotherapy in individuals with traumatic SCI.  These researchers performed an extensive search for RCTs involving people with traumatic SCI that compared either BWSTT or RAGT with over-ground gait training and other forms of physiotherapy.  The 2 outcomes of interest were walking speed (m s-1) and walking distance (m).  BWSTT and RAGT were analyzed separately, and data were pooled across trials to derive mean between-group differences using a random-effects model.  A total of 13 RCTs involving 586 people were identified; 10 trials involving 462 participants compared BWSTT to over-ground gait training and other forms of physiotherapy, but only 9 trials provided useable data.  The pooled mean (95 % CI) between-group differences for walking speed and walking distance were -0.03 m s-1 (-0.10 to 0.04) and -7 m (-45 to 31), respectively, favoring over-ground gait training; 5 trials involving 344 participants compared RAGT to over-ground gait training and other forms of physiotherapy, but only 3 provided useable data.  The pooled mean (95 % CI) between-group differences for walking speed and walking distance were -0.04 m s-1 (95 % CI: -0.21 to 0.13) and -6 m (95 % CI: -86 to 74), respectively, favoring over-ground gait training.  The authors concluded that BWSTT and RAGT did not increase walking speed more than over-ground gait training and other forms of physiotherapy did, but their effects on walking distance were unclear.

Cheung and colleagues (2017) examined the effects of robotic-assisted training on the recovery of people with SCI; RCTs or quasi-RCTs involving people with SCI that compared robotic-assisted upper limbs or lower limbs training to a control of other treatment approach or no treatment were selected for analysis.  These investigators included studies involving people with complete or incomplete SCI.  They searched in Medline, Cumulative Index to Nursing and Allied Health Literature (CINAHL), Cochrane Central Register of Controlled Trials (Cochrane Library) and Excerpta Medica dataBASE (Embase) to August, 2016.  Bibliography of relevant articles on the effect of BWSTT on SCI subjects were screened to avoid missing relevant articles from the search of databases.  All kinds of objective assessments concerning physical ability, mobility and/or functional ability were included.  Assessments could be clinical tests (namely 6MWT and Functional Independence Measure) or laboratory test (i.e., gait analysis).  Subjective outcome measures were excluded from the present review.  A total of 11 RCT studies involving 443 subjects were included in the study.  Meta-analysis was performed on the included studies.  Walking independence (3.73 with 95 % CI: -4.92 to -2.53; p < 0.00001; I2 = 38 %) and endurance (53.32 m with 95 % CI: - 73.15 to -33.48; p < 0.00001; I2 = 0 %) were found to have better improvement in robotic-assisted training groups.  Lower limb robotic-assisted training was also found to be as effective as other types of body weight supported training.  There is a lack of upper limb robotic-assisted training studies, so a meta-analysis was not possible to be performed.  The authors concluded that robotic-assisted training is an adjunct therapy for physical and functional recovery for patients with SCI.  Moreover, they stated that future high-quality studies are needed to investigate the effects of robotic-assisted training on functional and cardiopulmonary recovery of SCI patients.

Intrepid Dynamic Exoskeletal Orthosis (IDEO)

The Intrepid Dynamic Exoskeletal Orthosis (IDEO) is a customized energy storing ankle-foot orthosis (AFO) developed by the U.S. Army for individuals who have suffered massive tissue, nerve and bone damage to supposedly return capabilities to the injured ankle.  It is designed to support and protect an extensive array of lower extremity limb injuries.  The device supposedly supplies energy storage and return capabilities that an injured ankle is no longer able to provide.  Purportedly, the individual can return to a high level of activity, such as running. The IDEO device is molded out of lightweight black carbon that includes a foot plate and a strut that runs up the back of the calf to a cuff that is situated just below the knee. Reportedly, when force is applied to the foot plate, the strut bends. As the individual steps down, it bends the foot plate, transferring energy forward. 

The IDEO is intended to return functionality to patients who have undergone ankle fusion procedures and to enable some patients with nerve and muscle loss to forgo ankle fusion or tendon transfer.  The IDEO is modular throughout the rehabilitation period to adapt to a patient’s changes in strength and motion.  Once the patient has progressed to an adequate level of recovery, the initial modular IDEO is replaced with a lighter, more dynamic definitive IDEO system.  Currently, there is insufficient evidence to support the clinical value of this device.

The PRIORITI-MTF Study-Testing Patient Response to the IDEO trial is currently recruiting participants.  The primary objective is to examine if a new type of custom designed brace (IDEO) along with a physical therapy program (Return to Run) improves physical function.  (Last updated November 10, 2014). 

The Influence of Heel Wedge Properties on Roll-over of the Intrepid Dynamic Exoskeletal Orthosis (IDEO) is not yet open for participant recruitment.  The primary objective is to determine how heel wedge properties may contribute to the smoothness of roll-over during gait.  Insight into the effects of heel wedge properties on roll-over will help optimize the design of the IDEO-heel wedge-shoe "system" and may produce guidelines for the customization of these features.  (Last verified July 2015). 

Powered Hip Orthoses in Persons with Spinal Cord Injury

Arazpour and colleagues (2014) stated that gait training has been shown to improve the walking performance of spinal cord-injured (SCI) patients. The use of powered hip orthoses (PHO) during gait training is one approach which could potentially improve rehabilitative outcomes for such subjects. These researchers evaluated the influence of a PHO on the kinematics and temporal-spatial parameters of walking by SCI patients. The PHO used for the gait training of the volunteer SCI patients was custom made and fitted according to patient’s lower limb length, and the ankle foot orthosis portions of the orthosis were made after casting the lower extremities. The structure of orthosis and its efficacy when worn by one SCI patient has been described in a previously publication (Arazpour, et al., 2012). A total of 4 SCI patients participated in this study.  Gait evaluation was performed at baseline and at 10 weeks following intervention with the use of a PHO and gait re-training.  Walking speed, step length, vertical and horizontal compensatory motions and hip joint kinematics were analyzed prior to and following the training regime.  Significant increases in walking speed and step length were demonstrated by the SCI patients when walking with the PHO following orthotic gait training.  Sagittal plane hip range of motion also increased, but not significantly.  However, vertical and horizontal compensatory motions decreased significantly.  The authors concluded that positive effects on the kinematics and temporal-spatial parameters of gait by SCI subjects were demonstrated following a period of gait training with a PHO.  Moreover, they stated that further studies are needed to confirm their long-term effects on the rehabilitation of SCI subjects.

Arazpour and associates (2015) noted that powered orthoses are a new generation of assistive devices for people with SCI, which are designed to induce motion to paralyzed lower limb joints using external power via electric motors or pneumatic or hydraulic actuators. Powered gait orthoses provide activated movement of lower limb joints to limit the forces applied through the upper limb joints and trunk muscles during ambulation due to the need to use an external walking aid, while simultaneously improving the kinetics and kinematics of walking in subjects with SCI. These investigators reviewed the walking efficacy of powered orthoses when used by people with paraplegia.  A literature search was performed in ISI Web of Knowledge, PubMed, Google Scholar, ScienceDirect, and Scopus databases.  Efficacy was demonstrated in producing activated motion of lower limb joints.  Powered gait orthoses have a beneficial effect on the kinetics, kinematics, and temporal-spatial parameters of gait, but their effect on muscle activity in individuals with SCI is still unclear.  The authors concluded that further research is needed regarding the design and construction of powered gait orthoses using significant power application to the ankle joints and their effect on lower limb muscle activity and gait patterns in subjects with SCI.

Ahmadi Bani and co-workers (2015) stated that the most simple and common approach in providing standing and walking by subjects with SCI is the use of mechanical orthoses. These include traditional orthoses, medial linkage orthoses (MLOs) and reciprocating gait orthoses (RGOs). Independence, energy expenditure, gait parameters, system reliability and cosmesis are important factors in orthotic design.  These researchers compared the evidence of existing mechanical orthoses to that of other types regarding these factors.  The preferred reporting items for systematic reviews and meta-analyses (PRISMA) method was used by an experience researcher based on selected keywords and their composition and an electronic search was performed in well-known databases.  A total of 20 articles were selected for final evaluation.  Many were case studies, and also had limited and heterogeneous sample sizes with different instruments used for evaluation.  The results of the analysis demonstrated that independence and cosmesis were improved when using MLOs, but gait parameters, energy expenditure and stability were all improved when using RGOs.  The authors concluded that those mechanical orthoses which have reciprocal motion and congruency between the anatomical and orthotic joints have been shown to provide positive effects on patient lifestyles.  However, they stated that further improvement is needed to more effectively meet the needs of SCI patients.

Arazpour et al (2016) noted that orthoses for various joints sections are considered to greatly influence the gait function and energy expenditure in SCI patients. These investigators determined the influence of orthoses characteristics and options on the improvement of walking in patients with SCI.  They performed a search using the Population Intervention Comparison Outcome (PICO) method, based on selected keywords; studies were identified electronically in the Science Direct, Google Scholar, Scopus, Web of Knowledge and PubMed databases.  The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) method was used to report the results.  Assessment of the quality of all articles was performed based on the Physiotherapy Evidence Database (PEDro scale).  A total of 12 studies evaluated the effects of different hip joint options on walking parameters and energy expenditure; 5 studies investigated the role of knee joint options on gait parameters and compensatory trunk motion.  Only 5 studies analyzed modified ankle joints on gait parameters in SCI patients; 9 studies analyzed gait parameters in SCI patients as powered orthoses and exoskeleton.  These studies had a low level of evidence according to the PEDro score (2/10).

Spinal Cord Injuries

Singh and associates (2018) provided an overview of the feasibility and outcomes of robotic-assisted upper extremity training for individuals with cervical SCI, and to identify gaps in current research and articulate future research directions.  A systematic search was conducted using Medline, Embase, PsycINFO, CCTR, CDSR, CINAHL and PubMed on June 7, 2017.  Search terms included 3 themes: robotics; SCI; and upper extremity. Studies using robots for upper extremity rehabilitation among individuals with cervical SCI were included. Identified articles were independently reviewed by two researchers and compared to pre-specified criteria.  Disagreements regarding article inclusion were resolved through discussion.  The modified Downs and Black checklist was used to assess article quality.  Participant characteristics, study and intervention details, training outcomes, robot features, study limitations and recommendations for future studies were abstracted from included articles.  A total of 12 articles (1 randomized clinical trial, 6 case series, 5 case studies) met the inclusion criteria; 5 robots were exoskeletons and 3 were end-effectors.  Sample sizes ranged from 1 to 17 subjects.  Articles had variable quality, with quality scores ranging from 8 to 20.  Studies had a low internal validity primarily from lack of blinding or a control group.  Individuals with mild-moderate impairments showed the greatest improvements on body structure/function and performance-level measures.  This review was limited by the small number of articles, low-sample sizes and the diversity of devices and their associated training protocols, and outcome measures.  The authors concluded that preliminary evidence suggested that robot-assisted interventions are safe, feasible and can reduce active assistance provided by therapists.  Moreover, they stated that future research in robotics rehabilitation with individuals with SCI is needed to determine the optimal device and training protocol as well as effectiveness.

Hayes and colleagues (2018) systematically reviewed the literature and identify if over-ground or treadmill based RAGT use in SCI individuals elicited differences in temporal-spatial characteristics and functional outcome measures.  A systematic search of the literature investigating over-ground and treadmill RAGT in SCIs was undertaken excluding case-studies and case-series.  Studies were included if the primary outcomes were temporal-spatial gait parameters.  Study inclusion and methodological quality were assessed and determined independently by 2 reviewers.  Methodological quality was assessed using a validated scoring system for randomized and non-randomized trials.  A total of 12 studies met all inclusion criteria.  Participant numbers ranged from 5 to 130 with injury levels from C2 to T12, American Spinal Injuries Association A-D; 3 studies used over-ground RAGT systems and the remaining 9 focused on treadmill based RAGT systems.  Primary outcome measures were walking speed and walking distance.  The use of treadmill or over-ground based RAGT did not result in an increase in walking speed beyond that of conventional gait training and no studies reviewed enabled a large enough improvement to facilitate community ambulation.  The authors concluded that the use of RAGT in SCI individuals has the potential to benefit upright locomotion of SCI individuals.  Moreover, they stated that its use should not replace other therapies but be incorporated into a multi-modality rehabilitation approach.

Yozbatiran and Francisco (2019) noted that tetraplegia resulting from cervical injury is the most frequent neurologic category following SCI and causes substantial disability.  The residual strength of partially paralyzed muscles is an important determinant of independence and function in tetraplegia.  Small improvements in UE function can make a clinically significant difference in daily activities.  Major advances in rehabilitation technologies over the past 20 years have allowed testing of robotic devices in rehabilitation of motor impairments.  The authors provided an overview of robotic-assisted training research for improving arm and hand functions following cervical SCI.  They concluded that robot-assisted rehabilitation of the UE following cervical SCI is safe and feasible; but lacks sufficient evidence of clinical effectiveness.

Stroke

Gandolfi and colleagues (2018) noted that bilateral arm training (BAT) has shown promise in expediting progress toward upper limb recovery in chronic stroke patients, but its neural correlates are poorly understood.  In a pilot study, these researchers evaluated changes in upper limb function and electroencephalographic (EEG) power after a robot-assisted BAT in chronic stroke patients.  In a within-subject design, 7 right-handed chronic stroke patients with upper limb paresis received 21 sessions (3 days/week) of the robot-assisted BAT.  The outcomes were changes in score on the upper limb section of the Fugl-Meyer assessment (FM), Motricity Index (MI), and Modified Ashworth Scale (MAS) evaluated at the baseline (T0), post-training (T1), and 1-month follow-up (T2).  Event-related desynchronization/synchronization were calculated in the upper alpha and the beta frequency ranges.  Significant improvement in all outcomes was measured over the course of the study.  Changes in FM were significant at T2, and in MAS at T1 and T2.  After training, desynchronization on the ipsilesional sensorimotor areas increased during passive and active movement, as compared with T0.  The authors concluded that a repetitive robotic-assisted BAT program may improve upper limb motor function and reduce spasticity in the chronically impaired paretic arm.  Effects on spasticity were associated with EEG changes over the ipsilesional sensorimotor network.  These researchers speculated that the reduction in spasticity may have facilitated EEG changes over the ipsilesional sensorimotor network.  They stated that the utility of a bilateral repetitive robot-assisted program as an adjuvant to physical therapy needs further consideration.

The authors stated that the main drawback of the present study was its small sample size (n = 7).  Moreover, the lack of homogeneity between brain lesion size and location would have precluded statistically significant results.  The event-related desynchronization (ERD; i.e., power band decrease)/event-related synchronization (ERS; i.e., power band increase) maps of the patients shown were different among themselves, and each map was also different with the control group in different ways.  The current results varied across patients and precluded to conclude.  According to these preliminary results, future studies would enroll larger sample size, and patients would be stratified according to lesion features.  It would allow discussing EEG power changes after specific robot-assisted upper limb training (i.e., active, passive, unilateral, and bilateral).  Other drawbacks were the lack of follow-up beyond 1 month and the lack of a control group receiving conventional therapy.  Moreover, control group should be age-matched, in order to compare it to a group of subjects affected by stroke.  However, since the peak frequency of the mu wave increased with age until maturation into adulthood, when it reached its final and stable frequency of 8 to 13 Hz, the age of the subjects should not significantly affect the EEG desynchronization process during movement.

Cho and associates (2018) evaluated the effects of RAGT on gait-related function in patients with acute/subacute stroke.  These researchers conducted a systematic review of RCTs published between May 2012 and April 2016.  This search included 334 articles (Cochrane, 51 articles; Embase, 175 articles; PubMed, 108 articles).  Based on the inclusion and exclusion criteria, 7 studies were selected for this review.  These investigators performed a quality evaluation using the PEDro scale.  In this review, 3 studies used an exoskeletal robot, and 4 studies used an end-effector robot as interventions.  As a result, RAGT was found to be effective in improving walking ability in subacute stroke patients.  Significant improvements in gait speed, functional ambulatory category, and Rivermead mobility index were found with RAGT compared with conventional physical therapy (p < 0.05).  The authors concluded that aggressive weight support and gait training at an early stage using a robotic device are helpful, and robotic intervention should be applied according to the patient's functional level and onset time of stroke.

The authors stated that the main drawback of this review was that it included a small number of studies (and subjects).  They stated that future review studies should include a qualitative analysis of the frequency and intensity of interventions, including more studies on subacute stroke patients.  In addition they stated that further studies are needed to demonstrate the effectiveness of RAGT according to the functional level of stroke patients in not only the subacute phase but also the chronic phase.

Dehem and associates (2018) stated that the impact of transcranial direct current stimulation (tDCS) is controversial in the neurorehabilitation literature.  It has been suggested that tDCS should be combined with other therapy to improve their efficacy.  In a randomized, controlled, double-blind, cross-over study, these researchers examined the effectiveness of a single-session of upper limb robotic-assisted therapy (RAT) combined with real or sham-tDCS in chronic stroke patients.  A total of 21 hemiparetic chronic stroke patients were included in this trial.  Participants underwent 2 sessions 7 days apart in a randomized order: 20-min of real dual-tDCS associated with RAT (REAL+RAT); and 20-min of sham dual-tDCS associated with RAT (SHAM+RAT).  Patient dexterity (Box and Block and Purdue Pegboard tests) and upper limb kinematics were evaluated before and just after each intervention.  The assistance provided by the robot during the intervention was also recorded.  Gross manual dexterity (1.8 ± 0.7 blocks, p = 0.008) and straightness of movement (0.01 ± 0.03, p < 0.05) improved slightly after REAL+RAT compared with before the intervention.  There was no improvement after SHAM+RAT.  The post-hoc analyses did not indicate any difference between interventions: REAL+RAT and SHAM+RAT (p > 0.05).  The assistance provided by the robot was similar during both interventions (p > 0.05).  The authors concluded that the results showed a slight improvement in hand dexterity and arm movement following the REAL+RAT tDCS intervention.  They noted that the observed effect after a single-session was small and not clinically relevant; repetitive sessions could increase the benefits of this combined approach.

Rodgers and colleagues (2019) noted that loss of arm function is a common problem following stroke.  Robot-assisted training might improve arm function and ADL. In a multi-center RCT, these researchers compared the effectiveness of robot-assisted training using the MIT-Manus robotic gym with an enhanced upper limb therapy (EULT) program based on repetitive functional task practice and with usual care.  This trial was carried out at 4 UK centers.  Stroke patients aged at least 18 years with moderate or severe upper limb functional limitation, between 1 week and 5 years after their 1st stroke, were randomly assigned (1:1:1) to receive robot-assisted training, EULT, or usual care.  Robot-assisted training and EULT were provided for 45 mins, thricely-weekly for 12 weeks.  Randomization was internet-based using permuted block sequences.  Treatment allocation was masked from outcome assessors but not from participants or therapists.  The primary outcome was upper limb function success (defined using the Action Research Arm Test [ARAT]) at 3 months.  Analyses were performed on an intention-to-treat (ITT) basis.  Between April 14, 2014, and April 30, 2018, a total of 770 participants were enrolled and randomly assigned to either robot-assisted training (n = 257), EULT (n = 259), or usual care (n = 254).  The primary outcome of ARAT success was achieved by 103 (44 %) of 232 patients in the robot-assisted training group, 118 (50 %) of 234 in the EULT group, and 85 (42 %) of 203 in the usual care group.  Compared with usual care, robot-assisted training (adjusted OR [aOR] 1.17 [98.3 % CI: 0.70 to 1.96]) and EULT (aOR 1.51 [0.90 to 2.51]) did not improve upper limb function; the effects of robot-assisted training did not differ from EULT (aOR 0.78 [0.48 to 1.27]).  More participants in the robot-assisted training group (39 [15 %] of 257) and EULT group (33 [13 %] of 259) had serious AEs than in the usual care group (20 [8 %] of 254), but none was attributable to the intervention.  The authors concluded that robot-assisted training and EULT did not improve upper limb function following stroke compared with usual care for patients with moderate or severe upper limb functional limitation.  These researchers stated that these findings did not support the use of robot-assisted training as provided in this trial in routine clinical practice.

Humeral Fracture

Nerz and colleagues (2017) noted that the incidence of proximal humeral fractures (PHFs) increases with age.  The functional recovery of the upper arm after such fractures is slow, and results are often disappointing.  Treatment of PHF is associated with long immobilization periods.  Evidence-based exercise guidelines are missing.  Loss of muscle mass as well as reduced ROM and motor performance are common consequences.  These losses could be partly counteracted by training interventions using robot-assisted arm support of the affected arm derived from neurorehabilitation.  Thus, shorter immobilization could be reached; however, this approach has been tested in only a few small studies.  These researchers examined if assistive robotic training augmenting conventional occupational and physical therapy can improve functional shoulder outcomes.  Patients aged between 35 and 66 years with PHF and surgical treatment will be recruited at 3 different clinics in Germany and randomized into an intervention group and a control group (n = 26 for each group).  Participants will be assessed before randomization and followed after completing an intervention period of 3 weeks and additionally after 3, 6 and 12 months.  The baseline assessment will include cognition (Short Orientation-Memory-Concentration Test); level of pain in the affected arm; ability to work; gait speed (10-m walk); disability of the arm, shoulder and hand (Disabilities of the Arm, Shoulder and Hand Outcome Measure [DASH]); ROM of the affected arm (goniometer measurement); visual acuity; and motor function of orthopedic patients (Wolf Motor Function Test–Orthopedic version [WMFT-O]).  Clinical follow-up directly after the intervention will include assessment of DASH as well as ROM and motor function (WMFT-O).  The primary outcome parameter will be the DASH, and the secondary outcome parameter will be the WMFT-O.  The long-term results will be assessed prospectively by postal follow-up.  All patients will receive conventional occupational and physical therapy.  The intervention group will receive additional robot-assisted training using the Armeo®Spring robot for 3 weeks.  The authors concluded that this study protocol describes a phase-II, randomized, controlled, single-blind, multi-center intervention study.  The results will guide and possibly improve methods of rehabilitation after proximal humeral fracture.

The authors stated that this trial/protocol have several limitations: First, the sample size is relatively small (n = 26 for each group) and powered only for a functional end-point (DASH).  Social or economic end-points such as return to work or cost-effectiveness of this intervention would require larger sample sizes and is not the purpose of this early-stage RCT.  Second, the set-up with 3 different associated centers will lead to some difficulties in the standardization of the assessments.  It is expected that different observers and therapists in the 3 institutions have subjective differences in the accomplishment of the tests.  Assessment of the WMFT will be videotaped to standardize this as much as possible.  Lastly, heterogeneous treatment modalities will also exist to some extent for rehabilitation after PHF in the conventional occupational and physical therapy in terms of duration, intensity and frequency. Theoretically, participation in the robotic group could lead to increased or decreased motivation to participate in other types of exercises. Adherence will therefore be documented.

Furthermore, UpToDate reviews on "Proximal humeral fractures in adults" (Bassett, 2018a), "Midshaft humeral fractures in adults" (Bassett , 2018b), "Proximal humeral fractures in children" (Ryan, 2018a), and "Midshaft humeral fractures in children" (Ryan, 2018b) do not mention robotic-assisted rehabilitation as a management option.

Myoelectric Elbow-Wrist-Hand Orthosis

In an observational, cohort study, Peters and colleagues (2017) determined the immediate effect of a portable, myoelectric elbow-wrist-hand orthosis (MEWHO) on paretic UE impairment in chronic, stable, moderately impaired stroke survivors (n = 18).  Participants were administered a battery of measures testing UE impairment and function.  They then donned a fabricated MEWHO and were again tested on the same battery of measures while wearing the device.  The primary outcome measure was the UE Section of the Fugl-Meyer Scale.  Subjects were also administered a battery of functional tasks and the Box and Block (BB) test.  Subjects exhibited significantly reduced UE impairment while wearing the MEWHO (FM: t17 = 8.56, p < 0.0001) and increased quality in performing all functional tasks while wearing the MEWHO, with 3 sub-tasks showing significant increases (feeding [grasp]: z = 2.251, p = 0.024; feeding [elbow]: z = 2.966, p = 0.003; drinking [grasp]: z = 3.187, p = 0.001).  Additionally, subjects showed significant decreases in time taken to grasp a cup (z = 1.286, p = 0.016) and increased gross manual dexterity while wearing a MEWHO (BB test: z = 3.42, p < 0.001).  The authors concluded that findings of this study suggested that UE impairment, as measured by the Fugl-Meyer Scale, was significantly reduced when donning a MEWHO, and these changes exceeded the Fugl-Meyer Scale's clinically important difference threshold.  Furthermore, utilization of a MEWHO significantly increased gross manual dexterity and performance of certain functional tasks.  These researchers stated that future work will integrate education sessions to increase subjects' ability to perform multi-joint functional movements and attain consistent functional changes.  The authors noted that this was the first study comparing subjects with and without a MEWHO.  Well-designed studies with large sample sizes and control groups are needed.

Dunaway and associates (2017) described the application of a commercially available, custom MEWHO, on a veteran diagnosed with chronic stroke with residual left hemiparesis.  The MEWHO provides powered active assistance for elbow flexion/extension and 3 jaw chuck grip.  It is a non-invasive orthosis that is driven by the user's EMG signal.  Experience with the MEWHO and associated outcomes were reported.  The participant completed 21 out-patient occupational therapy sessions that incorporated the use of a custom MEWHO without grasp capability into traditional occupational therapy interventions.  He then up-graded to an advanced version of that MEWHO that incorporated grasp capability and completed an additional 14 sessions.  Strength, ROM, spasticity (MAS), the BB test, the Fugl-Meyer assessment and observation of functional tasks were used to track progress.  The participant also completed a home log and a manufacturers' survey to track usage and user satisfaction over a 6-month period.  Active left UE ROM and strength increased significantly (both with and without the MEWHO) and tone decreased, demonstrating both a training and an assistive effect.  The participant also demonstrated an improved ability to incorporate his affected extremity (with the MEWHO) into a wide variety of bilateral, gross motor ADLs such as carrying a laundry basket, lifting heavy objects (e.g., a chair), using a tape measure, meal preparation, and opening doors.  The authors concluded that custom myoelectric orthoses offer an exciting opportunity for individuals diagnosed with a variety of neurological conditions to make advancements toward their recovery and independence, and warrant further research into their training effects as well as their use as assistive devices.