ZEPU-SGI1PLUS Hand Training Study (ZEPUSGI1RCT)

Upper limb motor dysfunction following stroke, incomplete spinal cord injury, traumatic brain injury, and orthopedic conditions is a major contributor to long-term disability. Conventional rehabilitation, while foundational, is often limited in the intensity, repetition, and task specificity needed to optimally drive neuroplasticity-dependent motor recovery. Electromechanical and robotic rehabilitation systems have emerged as adjuncts capable of delivering standardized, repetitive, programmable movement training with real-time feedback and objective performance monitoring, and evidence from high-income settings suggests such systems can improve motor function and functional outcomes when combined with conventional therapy. However, data on the feasibility, safety, and effectiveness of these technologies in low- and middle-income country (LMIC) settings, including Bangladesh, remain limited.

Bangladesh Medical University (BMU) has recently introduced the ZEPU-SG1PLUS Hand Functional Comprehensive Training System at its Robotic Rehabilitation Centre. Before this technology is integrated more broadly into clinical practice, this pilot study will generate foundational local evidence on its feasibility, safety, and preliminary efficacy, and will inform the design and sample-size estimation of a future adequately powered randomized controlled trial.

Device: The ZEPU-SG1PLUS is a non-powered, mechanically assisted training system built around a stable steel-frame table integrated with twelve dedicated training modules targeting distinct hand and upper-limb movement patterns, including thumb-to-palm motion, wrist flexion-extension, forearm rotation, finger flexion/extension, pinching, side pinching, spherical grip rotation, ulnar/radial deviation, horizontal stretching, and hand lifting. Resistance is provided by an adjustable counterweight system (four independent columns, 250 g increments, 500 g-3000 g range), allowing progressive, individualized resistance training. A tablet-based interactive software interface supports user identification, session setup, timing control, and recording of training activity, with session duration configurable from 1-99 minutes.

Design: This is a pilot randomized controlled trial conducted at the Robotic Rehabilitation Centre, Department of Physical Medicine and Rehabilitation, BMU. Eligible participants (adults with upper-limb motor dysfunction due to stroke, incomplete spinal cord injury, traumatic brain injury, or orthopaedic/post-surgical conditions) will be randomly allocated 1:1 to an Intervention Arm (standard-of-care rehabilitation plus ZEPU-SG1PLUS training) or a Control Arm (standard-of-care rehabilitation alone). Robotic sessions will be delivered 3 times weekly for 12 weeks (30-45 minutes per session), under supervision of physiatrists and physiotherapists, with training parameters progressively adjusted based on individual performance and tolerance. Assessments occur at baseline and at two-week intervals through the intervention and follow-up periods.

Assessment methods: Muscle strength is assessed by Manual Muscle Testing, grip strength by dynamometer, spasticity by the Modified Ashworth Scale, and general health status by the EQ-VAS. Device-recorded training parameters (e.g., repetitions, resistance level, session duration) are captured as objective indicators of rehabilitation progress. Feasibility and usability are evaluated through screening/enrollment flow, session adherence and dropout tracking, therapist training and device set-up time, level of assistance required for donning/doffing, and a Post-Study System Usability Questionnaire administered at study completion.

Safety monitoring: Pre-session screening of vital signs and skin integrity, use of a safety harness/tracking system during initial sessions, continuous therapist supervision, post-session skin and fatigue checks, and an emergency-stop mechanism on the device are built into the protocol. All adverse events and serious adverse events are recorded and reported to the safety monitor and the BMU Institutional Review Board.

Statistical analysis: Descriptive statistics will summarize safety outcomes (AEs/SAEs, adherence, dropouts). Continuous efficacy variables will be summarized as mean ± SD (or median/IQR for non-normal data) at each time point, with paired/unpaired t-tests or Wilcoxon signed-rank tests used to compare baseline versus follow-up findings as appropriate. The primary analysis will follow an intention-to-treat approach including all randomized participants, with a secondary per-protocol analysis restricted to those completing the intervention per protocol; missing follow-up data will be handled using group-wise mean substitution. A two-tailed p-value < 0.05 will be considered statistically significant. Analyses will be performed using SPSS.

Inclusion criteria

• Presence of upper-limb motor dysfunction, with or without concurrent lower-limb impairment, due to one of the following conditions: (a) stroke, with onset between 2 and 24 months prior to enrollment; (b) incomplete spinal cord injury (neurological level C5 and below), classified as ASIA Impairment Scale C or D; (c) traumatic brain injury with persistent upper-limb functional limitation; or (d) orthopedic or post-surgical conditions involving the shoulder, elbow, wrist, or hand with persistent functional impairment for > 3 months
• Age between 18 and 60 years, inclusive
• Ability to understand study procedures and follow instructions, and to provide written informed consent (or assent with guardian consent where applicable)
• Body weight and limb anthropometry compatible with the ZEPU-SG1PLUS Hand Functional Comprehensive Training System per manufacturer specifications
• Medically stable and cleared by a physician for robotic rehabilitation, with no active infection, uncontrolled cardiac or respiratory disease, severe osteoporosis, uncontrolled epilepsy, or untreated deep vein thrombosis

Exclusion criteria

• Complete spinal cord injury or profound motor paralysis preventing safe interaction with robotic devices
• Severe cognitive impairment, defined as a Mini-Mental State Examination (MMSE) score < 24, that would compromise safe participation
• Severe spasticity of the upper limb, defined as a Modified Ashworth Scale score > 3
• Unstable fractures, severe fixed joint contractures of the shoulder, elbow, wrist, or hand (e.g., > 30°), or severe pain limiting safe robotic training
• Uncontrolled cardiac arrhythmia, presence of a pacemaker, or other implanted electronic medical device incompatible with robotic sensors
• Pregnancy