Multicentric investigation on the safety, feasibility and usability of the ABLE lower-limb robotic exoskeleton for individuals with spinal cord injury: a framework towards the standardisation of clinical evaluations

Abstract

Background: Robotic lower-limb exoskeletons have the potential to provide additional clinical benefits for persons with spinal cord injury (SCI). However, high variability between protocols does not allow the comparison of study results on safety and feasibility between different exoskeletons. We therefore incorporated key aspects from previous studies into our study protocol and accordingly conducted a multicentre study investigating the safety, feasibility and usability of the ABLE Exoskeleton in clinical settings.

Methods: In this prospective pretest-posttest quasi-experimental study across two SCI centres in Germany and Spain, in- and outpatients with SCI were recruited into a 12-session training and assessment protocol, utilising the ABLE Exoskeleton. A follow-up visit after 4 weeks was included to assess after-training outcomes. Safety outcomes (device-related adverse events (AEs), number of drop-outs), feasibility and usability measures (level of assistance, donning/doffing-time) were recorded at every session together with changes in gait parameters and function. Patient-reported outcome measures including the rate of perceived exertion (RPE) and the psychosocial impact of the device were performed. Satisfaction with the device was evaluated in both participants and therapists.

Results: All 24 participants (45 ± 12 years), with mainly subacute SCI (< 1 year after injury) from C5 to L3, (ASIA Impairment Scale A to D) completed the follow-up. In 242 training sessions, 8 device-related AEs (pain and skin lesions) were reported. Total time for don and doff was 6:50 ± 2:50 min. Improvements in level of assistance and gait parameters (time, steps, distance and speed, p < 0.05) were observed in all participants. Walking function and RPE improved in participants able to complete walking tests with (n = 9) and without (n = 6) the device at study start (p < 0.05). A positive psychosocial impact of the exoskeleton was reported and the satisfaction with the device was good, with best ratings in safety (participants), weight (therapists), durability and dimensions (both).

Conclusions: Our study results prove the feasibility of safe gait training with the ABLE Exoskeleton in hospital settings for persons with SCI, with improved clinical outcomes after training. Our study protocol allowed for consistent comparison of the results with other exoskeleton trials and can serve as a future framework towards the standardisation of early clinical evaluations. Trial Registration https://trialsearch.who.int/ , DRKS00023503, retrospectively registered on November 18, 2020.

Keywords: Exoskeleton; Feasibility; Gait; Rehabilitation; Robotics; Safety; Spinal cord disorder; Spinal cord injury; Standardisation; Study protocol; Usability.

Conflict of interest statement

AC, AMV and JLP were employees of ABLE Human Motion S.L., the sponsor of the clinical investigation when the study was conducted and the publication prepared.

© 2023. The Author(s).

Figures

Fig. 1

Overview of the hard- and software of the ABLE Exoskeleton. A The ABLE Exoskeleton with the knee motors and passive hip joints. B The graphical user interface of the ABLE Care mobile app

Fig. 2

Overview of all study visits and measurements based on the International Classification of Functioning, Disability and Health (ICF) domains. Abbreviations are defined in the Abbreviation section

Fig. 3

Study flow chart of recruited participants in line with the STROBE statement (http://www.strobestatement.org)

Fig. 4

Progression of the Level of Assistance in donning and doffing. Level of Assistance (LoA) for A Donning and B Doffing is shown over the course of the training sessions

Fig. 5

Progression of the Level of Assistance in therapy tasks. The Level of Assistance (LoA) for A Sit-to-Stand, B Walk 10m, C Turn 180°, and D Stand-to-Sit tasks is shown over the course of the training sessions

Fig. 6

Average therapy time progression. The average total therapy time (time spent standing, walking or sitting in the exoskeleton) together with the standing and walking time for each session of the training programme of the 17 participants that completed sessions 2–11 is shown. Error bars represent standard deviation

Fig. 7

Progression of number of steps with the ABLE Exoskeleton. A boxplot of the number of steps for the 17 participants, who completed sessions 2–11, is shown

Fig. 8

QUEST 2.0 mean scores for device satisfaction items. The figure shows the satisfaction with the device for participants and therapists at follow-up. Blue and red dots represent the standard deviation both for study participants’ and therapists’ scores, respectively

Fig. 9

Importance of QUEST 2.0 categories for participants and therapists. The percentage of participants and therapists that selected each QUEST 2.0 category as the most important is shown