Feasibility and efficacy of incorporating an exoskeleton in gait training during subacute stroke rehabilitation

Abstract: Introduction: Hemiparesis is the most common acute manifestation of stroke and often has a strong negative impact on walking ability leaving one third of patients dependent in walking activities outside one’s home. Improved methods for training of gait during stroke rehabilitation could tackle the challenge of achieving independent walking and promote better outcomes. Several studies have explored the value of introducing electromechanical gait machines in stroke rehabilitation to enhance gait training. One example is the exoskeleton Hybrid Assistive Limb (HAL). The HAL system has been found feasible to use during rehabilitation in the chronic stage after stroke, however knowledge of the feasibility in the subacute stage after stroke and its efficacy compared to evidence-based conventional gait training is still limited. Aim: The overall aim of this thesis was to evaluate the safety and feasibility of HAL for gait training in the subacute stage after stroke and the effect of HAL training on functioning, disability and health compared to conventional gait training, as part of an inpatient rehabilitation program in patients with severe limitations in walking in the subacute stage after stroke. Methods: This thesis contains two studies where one is a safety and feasibility study (Study I) and one is a prospective, randomized, open labeled, blinded evaluation study (Study II). In Study I, eight patients performed HAL training 5 days/week. The number of training sessions were adjusted individually and varied from 6 to 31 (median 16). Safety and feasibility aspects of the training were evaluated as well as clinical outcomes on functioning and disability (e.g. independence in walking, walking speed, balance, movement functions and activities of daily living), assessed before and after the intervention period. In Study II, 32 patients were randomized to either conventional training only or HAL training in addition to the conventional training, 4 days per week for 4 weeks. Within and between- group differences in independence in walking, walking speed/endurance, balance, movement functions and activities of daily living were investigated before and after the intervention period, as well as 6 months post stroke. In addition, gait pattern functions were evaluated after the intervention in a three-dimensional gait laboratory. At 6 months post stroke self- perceived aspects on functioning disability and health were assessed and subsequently correlated to the clinical assessments. Results: In Study I HAL was found to be safe and feasible for gait training after stroke in patients with hemiparesis, unable to walk independently, undergoing an inpatient rehabilitation program. All patients improved in walking independence and speed, movement function, and activities of daily living during the intervention period. In addition, it was found that patients walked long distances during the HAL sessions, suggesting that HAL training may be an effective method to enhance gait training during rehabilitation of patients in the subacute stage after stroke. In Study II substantial but equal improvements in the clinically evaluated outcomes in the two intervention groups were found. At six months post stroke, two thirds of patients were independent in walking, and a younger age but not intervention group served as the best predictor. Gait patterns were similarly impaired in both groups and in line with previous reports on gait patterns post stroke. Further, self-perceived ratings on functioning, disability and health were explained by the ability to perform self-care activities and not by intervention group. Conclusion: To incorporate gait training with HAL is safe and feasible during inpatient rehabilitation in the subacute stage after stroke and may be a way to increase the dose (i.e. number of steps) in gait training in the subacute stage after stroke. Among these included younger patients with hemiparesis and severe limitations in walking in the subacute stage after stroke, substantial improvements in body function and activity as well as equally impaired gait patterns were observed both after incorporated HAL training and after conventional gait training only, but without between-group differences. In future studies, potential beneficial effects on cardiovascular, respiratory, and metabolic functions should be addressed. Further, as the stroke population is heterogeneous, potential subgroups of patients who may benefit the most from electromechanically-assisted gait training should be identified.