Effects of Wearable Exoskeleton Robots on Gait and Functional Outcomes in Stroke Patients: A Systematic Review of Randomized Controlled Trials

This systematic review will evaluate the effects of wearable exoskeleton robots on gait and functional recovery outcomes in stroke patients. The review will include controlled trials (randomized and, where applicable, quasi-randomized or non-randomized controlled designs) published within the last 10 years. Primary outcomes include gait speed (10-Meter Walk Test), walking endurance (6-Minute Walk Test), and balance ability (Berg Balance Scale). Secondary outcomes include biomechanical parameters (EMG, kinematics), functional independence, and quality of life. Risk of bias will be assessed using the Cochrane Risk of Bias 2.0 tool (RoB 2) for randomized trials and ROBINS-I for non-randomized controlled studies. Findings will be synthesized narratively and, where data permit, analyzed quantitatively through meta-analysis. PICOS CRITERIA P (Population): Stroke Patients • All phases included: acute, subacute, and chronic stroke • Walking status: independent ambulators OR those capable of gait training with assistance • Exclusions: severe cognitive impairment, severe paralysis preventing gait training I (Intervention): Wearable Exoskeleton Robots • For overground walking and functional mobility training • Includes: lower limb, hip, knee, and ankle assistive robots • Exclusions: treadmill-based systems (e.g., Lokomat), end-effector devices C (Comparison): Control Groups • Conventional physical therapy or standard care • Traditional orthoses (AFO, canes, walkers) • No-robot conditions O (Outcomes): Gait and Functional Outcomes • Primary: Gait speed (10MWT), walking endurance (6MWT), balance (BBS) • Secondary: Biomechanical measures (EMG, kinematics), functional independence, quality of life S (Study Design): Controlled Trials • Randomized controlled trials (RCTs) • Controlled clinical trials (CCTs) • Exclusions: case reports, case series without controls ELIGIBILITY CRITERIA Inclusion: - Adult stroke patients (≥18 years) at any post-stroke phase - Wearable exoskeleton robots for overground gait training - Controlled study designs (RCT, CCT) - Reports at least one primary outcome (10MWT, 6MWT, or BBS) - Published in English or Korean - Published within last 10 years (no specific start date restriction, but focus on contemporary devices) Exclusion: - Non-stroke populations or mixed populations without separate stroke data - Treadmill-based robotic systems (Lokomat, LOPES, G-EO) - End-effector devices - Upper limb exoskeletons - Uncontrolled study designs (case reports, single-arm studies) - No relevant gait/functional outcomes reported SEARCH STRATEGY Databases: - PubMed/MEDLINE - Embase - Cochrane CENTRAL - CINAHL - Web of Science - IEEE Xplore - Korean databases (KoreaMed, KMBASE) Additional Sources: - Clinical trial registries (ClinicalTrials.gov, WHO ICTRP) - Conference proceedings (ICORR, ICRA, World Stroke Congress) - Reference lists and forward citation tracking - Manufacturer contacts for unpublished data KEY SEARCH TERMS Population: stroke, cerebrovascular accident, CVA, hemiplegia, hemiparesis Intervention: exoskeleton, wearable robot, robotic gait, powered orthosis, lower limb robot Outcomes: gait, walking, ambulation, balance, mobility, functional recovery Design: randomized, controlled trial, RCT, CCT OUTCOMES Primary Outcomes: 1. Gait speed: 10-Meter Walk Test (10MWT) in m/s 2. Walking endurance: 6-Minute Walk Test (6MWT) in meters 3. Balance: Berg Balance Scale (BBS, 0-56 points) Secondary Outcomes: 1. Biomechanical parameters: EMG (muscle activity), kinematics (joint angles), spatiotemporal gait parameters 2. Functional independence: Functional Independence Measure (FIM), Barthel Index 3. Quality of life: Stroke Impact Scale (SIS), SF-36, EQ-5D 4. Adverse events and safety METHODOLOGY Screening: - Dual independent screening (title/abstract, then full-text) - Covidence or Rayyan screening platform - Disagreements resolved by discussion; third reviewer if needed Data Extraction: - Dual independent extraction using standardized forms - Study characteristics, participant details, intervention/comparator details, outcomes, risk of bias items - Missing data: author contact attempted Risk of Bias Assessment: - Cochrane RoB 2 for RCTs - ROBINS-I for non-randomized controlled trials - GRADE for certainty of evidence Synthesis: - Narrative synthesis for all included studies - Random-effects meta-analysis if ≥3 homogeneous studies available - Subgroup analyses: stroke chronicity, exoskeleton type, training duration (if ≥10 studies) - Sensitivity analyses: excluding high risk of bias studies, fixed vs random effects - Heterogeneity assessed using I² and Chi² tests - Publication bias: funnel plots and Egger's test if ≥10 studies REGISTRATION AND TRANSPARENCY - Protocol registered on OSF (Open Science Framework) before search execution - PRISMA 2020 reporting guidelines followed - All data extraction forms, search strategies, and analysis scripts shared on OSF - Deviations from protocol documented and reported