Exploring Usability Standards in Robotic Rehabilitation: A Meta-Analysis of System Usability Scale Scores

Increasing integration of robotic devices in rehabilitation necessitates robust usability assessments to ensure effective human-robot interaction. This study presents a meta-analysis of System Usability Scale (SUS) scores reported in robotics-assisted rehabilitation, to evaluate whether the usability of these devices meets the standard SUS benchmark (mean = 68, SD = 12.5) and to identify any factors influencing usability across different users and device types. Following a systematic literature search, we analyzed 67 studies evaluating usability of robotic devices. Devices were categorized into 8 types, including upper and lower-limb exoskeletons, end effectors and sensor-based systems. From each study, we extracted SUS means, user information (healthy individuals, healthcare practitioners or patients), and scoring origin (developer vs. external investigator). Our results indicate that robotic devices in rehabilitation generally achieve good usability (74 on average), but current studies have a limited sample size (13 on average). Comparisons with the standard SUS benchmark showed a statistically significant deviation in most device categories, thus underscoring the importance of tailored usability standards in robotics-assisted rehabilitation. Future research should focus on larger, more diverse sample sizes and explore usability challenges specific to certain devices and user roles to enhance the development of accessible, user-friendly rehabilitation robots.