SaeboGlove therapy for Upper limb disability and Severe Hand Impairment after stroke

Alexander, Jen (2024) SaeboGlove therapy for Upper limb disability and Severe Hand Impairment after stroke. PhD thesis, University of Glasgow.

Abstract

Background:
Upper limb movement impairment affects around 80% of people experiencing a stroke acutely. Severe movement impairment including impaired active digital extension are particularly common, seriously hinder access to meaningful, selfdirected, functional, guideline-based rehabilitation and predict poor recovery. Upper limb movement impairment involves severe chronic disability and reduced quality of life for the majority affected and an evidence-based intervention to improve function when active digital extension is lacking has yet to be identified. Non-robotic Dynamic Hand Orthoses (DHOs) are novel rehabilitation aids intended to assist digital extension and access to functional rehabilitation, yet little is known about their use or effectiveness in stroke rehabilitation. The work in this thesis is informed by two intervention development activities conducted by the author preceding it. The activities include the design, delivery, and successful piloting of a complex intervention involving a dynamic hand orthosis (SaeboGlove) (The EMPOWER Help for Hands Programme) and a subsequent related patient and public involvement activity.

Methods & results:
The thesis followed the Medical Research Council framework for the development and evaluation of complex interventions and aimed to appraise and progress the existing evidence base on the use of DHOs to improve upper limb recovery after stroke.

Reviews to appraise the existing evidence: The research commenced with two literature reviews on the use of non-robotic DHOs in stroke rehabilitation; a scoping review to map existing research (Objective 1a) and a systematic review with meta-analysis of the randomised controlled trials (RCTs) identified in the scoping review to synthesise evidence on the effects of DHOs on upper limb recovery (Objective 1b). Risk of bias in the systematic review was determined using The Cochrane risk of bias tool 1.0.
The scoping review identified 23 studies (n=229) in adult stroke survivors, all considered to have moderate to severe upper limb functional capacity. The evidence base on the use of non-robotic DHOs in stroke was lacking. There was (1) an absence of any systematic reviews, (2) a shortage of RCTs (only four small RCTs found) and (3) limited research in general with (i) adequate reporting of key information (participants, interventions, outcomes, and stakeholder involvement of any kind), (ii) high quality designs (e.g. multiple recruitment sites, adequate power to establish effectiveness, long term follow-up) and (iii) stroke survivors in the acute/subacute phase of recovery.
The systematic review and meta-analysis included the four RCTs (n=48) identified in the scoping review. The research gaps identified were consistent with those in the scoping review. No safety concerns were reported. Heterogeneity was low and a positive effect in favour of the use of non-robotic DHOs after stroke was observed in comparison to usual care, and usual care plus placebo regarding upper limb function and dexterity (primary outcomes) but not impairment (secondary outcome). However, a high risk of bias due, largely, to poor methodological reporting and a lack of high quality RCTs (e.g. with multiple recruitment sites, adequate power to establish effectiveness, long term follow-up) meant that there was uncertainty around the positive effects observed.

Evaluation to progress the existing evidence: The previously piloted EMPOWER intervention and study design underwent minor refinement and subsequent evaluation using learning gained from the development activities preceding this doctorate and the reviews conducted in it. The evaluation aimed to address identified gaps/limitations thereby progressing the existing evidence.
An outcome evaluation evaluated effects of the EMPOWER intervention on upper limb function using a pragmatic, multisite, parallel group RCT with blinded outcome assessment, adequate statistical power and long term follow up (Objective 2a). Adult stroke survivors from two Scottish Health Boards with upper limb disability and severe hand impairment including reduced or absent active digital extension, 7-60 days post stroke, were randomly assigned (1:1) via a central online service to receive six weeks of (1) usual care plus the EMPOWER intervention (individualised self-directed, repetitive, functional-based practice involving a SaeboGlove) (SaeboGlove group), or (2) usual care alone (Usual Care group).
The primary outcome was upper limb function measured by the Action Research Arm Test (ARAT) 6 weeks post randomisation. Secondary outcomes were the ARAT at 14 weeks and upper extremity impairment (Fugl Meyer Upper Extremity Score – Motor Total and Hand subscore, Sensory Total), self-perceived habitual functional upper limb use (Motor Activity Log), upper limb pain intensity (Visual Analogue Scale), degree of new disability (modified Rankin Scale (mRS), selfreported performance in activities of daily living (Barthel Index) and selfreported quality of life / health status (Stroke Impact Scale (Total plus Hand subscore) and EuroQol (EQ-5D-5L)) at 6 and 14 weeks. Linear regression analyses were performed on an intention to treat basis, adjusting for baseline value of the outcome, study site, time since stroke (≤1 month, >1 month) and severity of upper limb function (ARAT 0-10, 11-28, 29-46). Results are reported as adjusted mean between group differences with 95% confidence intervals and p-values.
78 participants were randomised (39 per group) between November 2019 and October 2022 from six study sites in Scotland. At randomisation 63.0 (80.8%) suffered ischaemic stroke, 28 (35.9%) were female, mean (SD) for age was 63.3 years (13.2), time from stroke was 28.6 days (15.1), and ARAT Total score was 13.3 (11.8).
Although the Usual Care group reported spending more upper limb therapy time with a therapist over the intervention period (weeks 1-6), they reported spending much less time on upper limb therapy overall (with and without a therapist) during weeks 1-6 (mean and median of 8.7 hours and 12.0 hours less respectively) and 1-14 (mean and median total of 6.8 hours and 18.9 hours less).
Outcome completion was good at the 6 (91.0%) and 14 week (82.1%) visits. The EMPOWER intervention led to a significant improvement in the primary outcome of upper limb function (ARAT Total) at 6 weeks (15.8 versus 23.8 points respectively, adjusted between group difference 7.1 points, 95% CI 0.22 to 14, p=0.047). ARAT Grasp, Pinch, and Grip subscores were also significantly improved at 6 weeks while the Gross subscore was not.
The change in the FMUE Motor Hand and Total scores were also both statistically significantly improved in the SaeboGlove compared to the Usual Care group at the 6 week visit. However, no other secondary outcome measures were statistically significant at any point in time. No safety concerns were identified for either intervention.
A qualitative process evaluation embedded within the RCT explored factors relating to support that moderate implementation of/engagement with the EMPOWER intervention using semi-structured interviews with a purposive sample of stroke survivor receiving it (n=7), their carers (n=5), and therapists (n=6). Thematic analysis and a generic inductive approach were used to identify frequent and noteworthy themes to inform future implementation/engagement (Objective 2b).
Four main themes influencing therapy implementation/engagement were identified: (1) Rehabilitation is hard work, (2) Stroke severity heightens the rehabilitation challenge and support required, (3) More meaningful upper limb therapy is needed/wanted, and (4) The EMPOWER intervention is acceptable and HELPs support engagement. Recommendations to optimise successful implementation/engagement therefore include: (1) adhering to the intended intervention implementation/engagement, (2) prior to therapy commencing ensuring that (i) staff have sufficient clinical expertise to establish realistic expectations in stroke survivors on what therapy engagement/recovery will involve and (ii) the work of rehabilitation is managed with tailored support based on individual needs/preferences, particularly when stroke symptoms are more severe, (3) utilising meaningful upper limb therapy tasks and objects, ideally before imminent discharge home, (4) continuing the EMPOWER programme without a dynamic hand orthosis when the six week intervention period ends to optimise continuity if functional upper limb rehabilitation goals are ongoing, and (5) on occasion, if clinically justified, extending the intervention by a couple of weeks.

Conclusions:
The existing research on non-robotic DHOs in stroke, with no systematic review and only four small RCTs, was limited and involved poor reporting and research designs. The SUSHI trial responded to these limitations and increases confidence in the systematic review findings. Individualised self-directed, repetitive, functional-based practice involving a SaeboGlove, started between 7 and 60 days after stroke in people with severe hand impairment including reduced/absent active digital extension led to significant improvements in ARAT score at 6 weeks after randomisation. The EMPOWER intervention was felt to improve the physical and emotional support to engage in upper limb therapy compared to usual care therapy. Stroke survivors receiving the EMPOWER intervention also recorded higher upper limb therapy engagement despite receiving less therapist input. Thus, the EMPOWER Help for Hands programme appears to be a more efficient intervention in terms of recovery time, therapist time (therapy input) and outcome. The improved rehabilitation support provided likely explains the increased upper limb therapy engagement and outcomes observed, novel findings in this challenging, hard to reach group.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Additional Information: Supported by funding from Chest Heart and Stroke Scotland and the Chief Scientist Office.
• Colleges/Schools: College of Medical Veterinary and Life Sciences > School of Cardiovascular & Metabolic Health
• Supervisor's Name: Dawson, Professor Jesse, Langhorne, Professor Peter and van Wijck, Professor Frederike
• Date of Award: 2024
• Depositing User: Theses Team
• Unique ID: glathesis:2024-84219
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 09 Apr 2024 15:42
• Last Modified: 22 Oct 2025 08:48
• Thesis DOI: 10.5525/gla.thesis.84219
• URI: https://theses.gla.ac.uk/id/eprint/84219

Related URLs

• Article DOI
• Article DOI