D. Schliessmann, C. Schuld, H. J. Gerner, N. Weidner declare no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. Three co-authors, R. Rupp, E. P. Hofer and M. Knestel are holders of a European patent (No. EP000001959909A1) for the MoreGait device, does not alter the authors' adherence to PLOS ONE policies on sharing data and materials. The authors confirm that the fact that one of the co-authors, N. Weidner, serves as an editor for PLOS ONE does not alter the authors' adherence to PLOS ONE editorial policies and criteria.
Conceived and designed the experiments: RR HP CS HJG MK EPH. Performed the experiments: HP MK. Analyzed the data: DS RR CS NW HJG. Contributed reagents/materials/analysis tools: MK EPH RR DS CS. Wrote the paper: RR DS HP CS HJG NW EPH MK.
The compact
Prospective, pre-post intervention, proof-of-concept study to test the feasibility of an unsupervised home-based application of five MoreGait prototypes in subjects with incomplete spinal cord injury (iSCI).
Twenty-five (5 tetraplegia, 20 paraplegia) participants with chronic (mean time since injury: 5.8 ± 5.4 (standard deviation, SD) years) sensorimotor iSCI (7 ASIA Impairment Scale (AIS) C, 18 AIS D; Walking Index for Spinal Cord Injury (WISCI II): Interquartile range 9 to 16) completed the training (45 minutes / day, at least 4 days / week, 8 weeks). Baseline status was documented 4 and 2 weeks before and at training onset. Training effects were assessed after 4 and 8 weeks of therapy.
After therapy, 9 of 25 study participants improved with respect to the dependency on walking aids assessed by the WISCI II. For all individuals, the short-distance walking velocity measured by the 10-Meter Walk Test showed significant improvements compared to baseline (100%) for both self-selected (Mean 139.4% ± 35.5% (SD)) and maximum (Mean 143.1% ± 40.6% (SD)) speed conditions as well as the endurance estimated with the six-minute walk test (Mean 166.6% ± 72.1% (SD)). One device-related adverse event (pressure sore on the big toe) occurred in over 800 training sessions.
Home-based robotic locomotion training with MoreGait is feasible and safe. The magnitude of functional improvements achieved by MoreGait in individuals with iSCI is well within the range of complex locomotion robots used in hospitals. Thus, unsupervised MoreGait training potentially represents an option to prolong effective training aiming at recovery of locomotor function beyond in-patient rehabilitation.
German Clinical Trials Register (DKRS)
Loss of mobility has devastating effects on the quality of life of those affected and their ability to remain independent in the community. This applies to individuals with lesions of the central nervous system (CNS) sustained for example through stroke or spinal cord injury (SCI). In subjects with incomplete SCI (iSCI), intensive gait training leads to substantial improvements in walking function [
Several factors of motor learning—task specificity, repetition, active participation and appropriate intrinsic and extrinsic feedback—have been identified as contributing to the long-term retention of a newly acquired skill [
Patients can most easily incorporate practice in their daily lives in a home-based training regimen. This may offer the advantage of practice within their personal space, where problem-solving is highly motivated [
The
The aim of this prospective, pre-post intervention proof-of-concept study was to test the safety of autonomous locomotor training with the MoreGait prototypes at the homes of individuals with sensorimotor iSCI and to obtain preliminary data about its efficacy. The pilot study results indicate that home-based robotic locomotion training with MoreGait is feasible and safe. The magnitude of functional improvements achieved by MoreGait training in individuals with iSCI is well within the range of complex locomotion robots used in hospitals.
For this prospective, baseline-controlled, single center cohort proof-of-concept study, inclusion criteria were (1) age between 18 and 60, (2) chronic (at least 1 year after trauma), (3) traumatic or ischemic/haemorrhagic sensorimotor iSCI (ASIA Impairment Scale (AIS) C, D [
Between January 2009 and January 2011, 46 individuals were screened, from whom 35 were included in the study. The intended number of study participants finalizing the training was set to 30 prior to the start of study due to the limited number of available prototypes and due to funding and time constraints. Twenty-five individuals (11 female, 14 male; 5 tetraplegic, 20 paraplegic; mean age: 44.0 ± 12.4 (SD) years) with chronic (mean time since injury: 5.8 ± 5.4 (SD) years) sensorimotor iSCI (7 AIS C, 18 AIS D; WISCI II from 5 to 19) completed the training procedure (
The numbers of subjects involved in the different phases of the study are shown in the diagram.
The study was approved by the Ethics Committee of Heidelberg University Hospital (vote no. MV-174/2007) and was conducted according to the World Medical Association Declaration of Helsinki and the Guidelines for Good Clinical Practice. The protocol submitted to the ethical committee for this clinical trial and supporting TREND checklist are available as supporting information (see
It has been registered at the German Institute for Medical Documentation and Information (DIMDI) as a clinical trial (registration no. 9053) with a novel medical product according to the guidelines of the European Medicinal Devices Act and with the main ID DRKS00005587 in the the German Clinical Trials Register (DKRS). Participants gave written informed consent prior to study inclusion. The individual shown in
A subject during training in the MoreGait device (A), top (B) and front (C) view of the mediolateral bars of the stimulative shoe, user interface and feedback screen (D).
The MoreGait device used in this study consists of a special seat combined with an inclined backrest, a pneumatically driven gait orthosis for each side to assist movements of both legs in the sagittal plane (independently driven knee and ankle joint, hip joint mechanically linked to knee joint via a fixed kinematic chain that allows only horizontal movements of the ankle joint) and a dedicated mechanical foot stimulation unit. Its dimensions are 172 x 70 x 130 cm (l x w x h) and the total weight is approximately 115 kg (
Pneumatic fluidic muscles (Festo AG & Co. KG, Esslingen, Germany) were selected as actuators. They offer the advantage of inherent low stiffness, which results in soft, safe and comfortable movements. Additionally, control parameters were set to allow for deviations of up to 5° from the predefined movement trajectory [
For safety reasons, the user’s body is placed in a semi-supine position. In this configuration, sufficient loading of the foot sole during stance phase cannot be generated by the user’s own body weight. Therefore a novel device—a “stimulative shoe”—was developed to mimic the loading of the foot sole without requiring the patient to be completely verticalized. This mechanical stimulation unit consists of 10 mediolateral plastic bars, which are mounted on pairs of pneumatically driven short-stroke cylinders (
One of the key factors for any kind of locomotion therapy to succeed is active participation by the patients. In order to continuously provide the users with information about whether they are performing training correctly, a feedback functionality was implemented. A rating measure is calculated from the user’s estimated active torques, and both the progress of the training and the absolute performance level are visualized on a display (
The right orthosis as well as the backrest can be lowered manually, enabling the patient to transfer autonomously to the lowered backrest. After successful transfer, the backrest can be inclined and the orthosis can be lifted in the training position. Leaving the device is performed in reverse order.
After screening and study inclusion, 3 assessment visits(0, 2 and 4 weeks) within a 4-week baseline period were planned prior to training onset, followed by assessments in the middle and at the end of an 8-week training period and a follow-up assessment 3 months after the end of training (
Three baseline (BL) assessments within the first 4 weeks are followed by 2 assessments at the middle and end of the training period. The follow-up assessment was carried out 3 months after the end of training. Assessments included the Walking Index for Spinal Cord Injury II (WISCI II), Timed Up and Go Test (TUG), 10-MeterWalk Test (10MWT) at both self-selected and maximum speed, six-minute walk test (6-MIN-TEST), as well as International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI) assessments and measurement of spasticity according to the modified Ashworth scale (MAS).
The assessments of each visit consisted of a set of well-established functional and neurological tests. The assessment scheme and outcome parameter were defined prospectively before the start of the study and was not changed over the course of the study. The self-selected WISCI II was defined as the primary outcome measure of the study due to its importance for the participants [
Dermatomes and myotomes as defined by the International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI) [
During the 8-week therapy period, individuals trained with the MoreGait device for 30–45 minutes per day, 4 to 6 days per week. Users were instructed to set step frequency at a comfortable level to avoid fatigue during each session. The training took place at the participants’ homes without supervision of the study personnel (mean distance from the spinal cord injury center in Heidelberg 209.12 ± 162.16 km; Google Maps (
To assess the users’ satisfaction with technical design, safety and therapeutic functionality of the MoreGait device, a user survey was made. The survey consisted of a paper questionnaire which was sent to each study participant who finished the 8-weeks of therapy. The proprietary questionnaire consisting of 51 questions was designed for obtaining dedicated user feedback on details of the MoreGait device. It mainly used a five-point scale for rating of answers, which is known from other standardized surveys on assistive technology [
Twenty of the twenty-five study participants, who completed the 8-weeks MoreGait training, replied to the survey. The results were grouped into the three main categories “Therapy outcome”, “Transfers, fastening and release” and “Training experience” including perception of safety. Results were analyzed on a descriptive basis eg, with boxplots.
Statistical analysis was performed with R 2.15.1 [
Baseline stability was tested in the context of the Friedman’s test. The baseline is considered stable, if pairwise post hoc tests of all three baseline assessments are not significant.
If not stated otherwise, numbers are displayed as mean ± SD. To allow for a more generalized analysis of the assessments of different study participants at 4 and 8 weeks of therapy, percentage values are calculated which are normalized to the mean baseline representing 100%.
The follow-up analysis was reduced to descriptive statistics because of the low number of available datasets for this stage (N = 10). Data from individuals that did not complete the study were not included in the analysis.
One device-related adverse event (Grade 2–3 [
Baseline stability was shown for all outcome measures, since all baseline post hoc comparisons were not significant (see
A numerical overview of the absolute outcome measures is given in
Visit | Baseline 1 | Baseline 2 | Baseline 3 | Mean / Median Baseline* | 4 weeks therapy | 8 weeks therapy | 8 weeks therapy |
Follow-up(20 weeks) |
---|---|---|---|---|---|---|---|---|
Outcome measure | ||||||||
WISCI II [score] |
12 (9–16) | 12 (9–16) | 12 (9–16) | 12 (9–16) | 16 (11.75–16); {24} | 16 (15–16);{24} | 16 (12–18.25) | 14 (11.25–16) |
MAS [score] |
2 (0–9) | 1 (0–4) | 2 (0–4.5); {24} | 1 (0–6.33) | 0 (0–6) | 1 (0–4.5); {24} | 0 (0–8.25) | 1 (0–7.0) |
10MWT (sss) [m/s] | 0.37 ± 0.27 | 0.36 ± 0.24 | 0.37 ± 0.24 | 0.37 ± 0.25 | 0.42 ± 0.30;{24} | 0.47 ± 0.27;{24} | 0.52 ± 0.35 | 0.59 ± 0.42 |
10MWT (ms) [m/s] | 0.47 ± 0.43 | 0.45 ± 0.35 | 0.46 ± 0.37 | 0.46 ± 0.38 | 0.52 ± 0.42; {24} | 0.57 ± 0.35; {24} | 0.66 ± 0.46 | 0.71 ± 0.49 |
6-MIN-TEST [m] | 117.04 ± 103.87; {24} | 115.63 ± 103.08; {24} | 119.46 ± 109.78; {24} | 117.38 ± 104.59 | 142.74 ± 107.01; {23} | 164.74 ± 115.03; {23} | 185.1 ± 154.88 | 210.70 ± 162.96 |
TUG [s] | 67.71 ± 57.89; {24} | 61.60 ± 51.62 | 56.18 ± 38.34 | 61.34 ± 48.02 | 50.83 ± 36.10; {24} | 37.21 ± 22.71; {24} | 37 ± 25.36 | 50.80 ± 54.96 |
LEMS [points] | 30.32 ± 9.33 | 29.72 ± 8.99 | 29.92 ± 9.24 | 29.99 ± 8.97 | 33.36 ± 10.87 | 36.42 ± 9.77; {24} | 35.1 ± 11.28 | 33.70 ± 11.49 |
MS L2 [points] | 7.48 ± 1.05 | 7.32 ± 1.35 | 6.84 ± 1.46 | 7.21 ± 1.13 | 7.76 ± 0.93 | 7.88 ± 1.20 | 8.1 ± 1.37 | 8.20 ± 1.69 |
MS L3 [points] | 7.60 ± 1.32 | 7.36 ± 1.70 | 7.48 ± 1.42 | 7.48 ± 1.24 | 8.40 ± 1.58 | 8.52 ± 1.56 | 9.1 ± 0.99 | 8.60 ± 1.51 |
MS L4 [points] | 4.84 ± 3.21 | 4.68 ± 3.22 | 5.04 ± 3.36 | 4.85 ± 3.14 | 5.80 ± 3.56 | 6.48 ± 3.33 | 6.7 ± 3.30 | 6.00 ± 3.37 |
MS L5 [points] | 5.04 ± 3.61 | 4.96 ± 3.48 | 5.00 ± 3.42 | 5.00 ± 3.44 | 5.44 ± 3.63 | 6.08 ± 3.98 | 5.1 ± 4.58 | 5.00 ± 4.55 |
MS S1 [points] | 5.16 ± 3.22 | 5.32 ± 3.18 | 5.60 ± 3.11 | 5.36 ± 3.02 | 6.24 ± 3.46 | 6.80 ± 3.34 | 6.1 ± 3.87 | 5.90 ± 4.01 |
Mean ± standard deviation of Walking Index for Spinal Cord Injury II (WISCI II), modified Ashworth scale (MAS), 10-Meter Walk Test (10MWT)—self-selected speed (sss) and maximum speed (ms), six-minute walk test (6-MIN-TEST), Timed Up and Go Test (TUG) and lower extremity motor scores (LEMS) are listed chronologically. Segmental motor scores (MS) for myotomes L2—S1 are also provided. Sample sizes are displayed in “{}” where they deviate from N = 25.
aMedian and 25%-75% quartiles are given for ordinal scales eg, WISCI II and MAS.
bAs follow-up data have a sample size of N = 10 the corresponding subset of patients within the 8-week therapy assessment is presented for better comparison.
Percent of the mean baseline ± standard error (SE) along the course of the study in relation to the mean baseline for (A) Lower extremity motor score (LEMS), (B) Timed Up and Go Test (TUG), (C) 10-MeterWalk Test (10MWT) at maximum speed (-ms) and self-selected speed (-sss), (D) six-minute walk test (6-MIN-TEST). Horizontal bars mark significant differences (p < 0.05), which were determined on the basis of the absolute values in
The primary outcome measure, WISCI II, showed overall significant results (p < 0.0006) and increased from 12 (9 to 16) to 16 (15 to 16) (baseline to 8 weeks therapy; median (interquartile range (IQR))). Nine participants (2/7 AIS C, 7/18 AIS D) were less dependent on walking aids after therapy than before. A qualitative analysis of the AIS subgroups revealed that subjects classified as AIS C show a trend towards a higher improvement than those classified as AIS D (median of increase from baseline to end of therapy in AIS C = 7 and in AIS D = 4). Post hoc tests in WISCI II revealed a significantly (
The results of the 10MWT for both self-selected (139.4% ± 35.5%; 8-weeks assessment) and maximum (143.1% ± 40.6%; 8-weeks assessment) speed conditions showed a significant overall improvement (both p < 0.0001) in short-distance walking velocity. Post hoc analysis revealed a significant increase of self-selected (
There were significant overall results in LEMS (p < 0.0001). Post hoc analysis showed significant (
The MAS of all study participants did not show any significant differences over the course of the therapy (p = 0.2379). However, in the 7 participants in whom spasticity was present (defined by a mean MAS > 4) at baseline (median at baseline: 16 (IQR 9.5 to 16)), a trend towards decreased spasticity (median: -3 (IQR-5.5 to -1.5)) was observed.
Two temporary conversions in the AIS (B to C and back to B and D to C and back to D) were detected during the therapy (onset, 4 weeks, 8 weeks) period [
Baseline and follow-up assessments (
WISCI II scores displayed as (A) relative values ± standard error (SE) and (B) absolute values ± SE for participants (N = 10) who attended the follow-up assessment (green) and for participants (N = 15) who did not attend the follow-up assessment (red).
Analysis of the WISCI II follow-up assessments shows that after the end of therapy, 1 subject further improved (9 to 16), 7 remained at the same level and 2 became worse (19 to 5, 16 to 11). Five months after therapy onset, 7 of the 10 follow-up visitors were less dependent on walking aids compared to baseline.
Individuals were overall satisfied with their training experience (3.80 ± 0.85,
Boxplots showing survey results on a 5 point scale for the categories (A) Training experience, (B) Transfers, fastening and release, and (C) Therapy outcome. The survey was completed by twenty of the twenty-five participants who finished the MoreGait study. Sample sizes are displayed in “{}”, where they deviate from N = 20.
Transfers on the device and back to the wheelchair, as well as fastening and releasing leg straps were rated as between moderate and easy (
Study participants were overall satisfied with the outcome of the therapy (
We investigated the safety and efficacy of 5 prototypes of the novel MoreGait robotic locomotor training device, which has been explicitly developed and designed for autonomous use in the home environment. This is to our knowledge the world’s first locomotor training device dedicated to this purpose [
We found that home-based training with this compact device is feasible and effective, and could be handled well by the users. A study assessing the safety of the supervised application of the driven gait orthosis Lokomat in children and adolescents reported 5 adverse events requiring discontinuation of therapy in about 1.400 training sessions [
The participants in our study with chronic iSCI showed no signs of neurological or functional recovery during the 4-week baseline period. This clearly demonstrates that the documented improvements following MoreGait training were therapy related and not spontaneous. The mean gain of 2.08 ± 3.82 levels in WISCI II clearly exceeds the recently reported clinically meaningful threshold of 1 level in iSCI [
The users benefited from the home-based robotic locomotion therapy in a variety of ways. The dependency on walking aids—a highly relevant issue for individuals with iSCI—was remarkably reduced in 9 participants after the MoreGait therapy. The importance of this aspect for the study participants is underlined by the outcome of the descriptive subgroup analysis of follow-up attendees vs. non-attendees. Attendees of the follow-up assessment had a much higher increase in WISCI II levels than non-attendees. The therapy-induced gain in WISCI II levels and
During the therapy period, the participants’ walking ability increased considerably. This was reflected by improvements in short-distance gait speed (10MWT) by approximately 40%, endurance (6-MIN-TEST) by roughly 65% and standing up, turning, and sitting down (TUG) by around 30%. Those improvements were also seen in the 9 individuals, who needed less support by walking aids over the course of the therapy.
A direct comparison of the study results with other studies is very difficult due to differences in patient populations (type/severity of lesion, functional status, exclusion of spontaneous recovery,) and therapy regimens (frequency, duration) [
A number of studies utilizing body weight-supported training for improving walking in individuals with SCI have reported improvements in lower-limb strength in patients with chronic SCI that are in the range of our results [
There was a trend towards decreased spasticity at the end of therapy in study participants with a mean MAS at baseline greater than 4. However, this finding has to be interpreted very carefully due to the low reliability of the MAS to detect subtle changes in spasticity [
The following limitations of the study have to be considered: Besides screening of the internal medical database, study participants were recruited by advertising the study on the university hospital’s website and in a magazine for people with disabilities focusing on individuals with SCI [
We did not document the type and focus of concomitant therapies and medication. Although study participants were asked not to modify their physical therapy, unsupervised training program, or antispastic medication, it cannot be excluded that changes in the regimes of these therapies throughout the MoreGait training period contributed to the improvements. The MoreGait therapy was applied at home as an add-on therapy. Therefore, we cannot exclude that the gait improvements were simply caused by the higher training intensity. On the other hand, this regimen of use best reflects the intended application scenario, in which MoreGait is the key component for allowing a higher intensity of gait training at home. Randomized controlled studies are necessary in the future to show the superiority of the MoreGait training in comparison to other, more simple home-based therapies.
Due to the safety-driven design of the MoreGait the user is put in a semi-reclined position. The influence of this non-physiological posture during the locomotion therapy with MoreGait on balance needs to be determined in future studies.
While baseline and follow-up assessments were performed at the Spinal Cord Injury Center, the majority of the 4-weeks and 8-weeks assessments took part in the participant’s home environment. While no influence of a community environment is reported on the 10MWT, positive effects are described on the gait endurance assessed by the 6-MIN-TEST [
The findings of the present study demonstrate that a robotic device reduced to a technical minimum can be introduced into a feasible, safe and effective gait rehabilitation therapy at home and thus might influence future robotic gait-rehabilitation strategies. A randomized-controlled trial investigating the effects of MoreGait therapy in acute iSCI is currently underway. Other neurological disease conditions affecting locomotor function may also benefit from this kind of robotic therapy, and thus warrant future investigation. The MoreGait device represent a valuable platform for future investigations on systematic identification and ranking of the therapeutic impact of machine parameters like degree of foot loading, inclination of the backrest or the prolonged therapy time.
Robotic home-based locomotion therapy with MoreGait allows patients to continue high-frequency training of locomotor function based on principles of activation of spinal locomotor networks and of motor learning after discharge from rehabilitation centers. The functional improvements following 8 weeks of MoreGait therapy in individuals with chronic sensorimotor iSCI are well within the range of those achieved with complex locomotion robots used at hospitals [
(PDF)
(PDF)
Overall p-values, as well as p-values and confidence intervals (ci) of the post-hoc comparisons of all baseline (BL) measurements, each BL with 4 weeks, and each BL with 8 weeks outcomes of the Walking Index for Spinal Cord Injury II (WISCI II), 10-Meter Walk Test (10MWT)—self-selected speed (sss) and maximum speed (ms), six-minute walk test (6-MIN-TEST), Timed Up and Go Test (TUG) and lower extremity motor scores (LEMS) are listed chronologically. Also the mean and standard deviations of p-values of the comparison among all BL, BL with 4 weeks, and BL with 8 weeks are provided. All significant differences (p < 0.05) are marked in red.
(XLSX)
The top row contains the descriptor of each of the assessments. The rest of the rows of the table contain the assessment data of each study participant.
(XLSX)
The top row contains the items of the end user survey. Each of the rows of the rest of the table contains the answers of each end user to each of the items.
(XLSX)
The authors thank all study participants, as well as F. Degenhard, M. Niess, O. Betz, J. Nückles and W. Roth for their support in maintenance and repair of the MoreGait devices during the study.