Test–Retest Reliability of the Sway Mobile Application for Balance and Cognitive Testing
Abstract
Mobile applications like Sway Medical are increasingly used for sideline and clinical concussion assessments. The purpose of this study was to evaluate the test–retest reliability, systematic error, and reliable change indices of the Sway Medical mobile application. A total of 79 healthy adults (age = 23 ± 6 years; 42 males, 37 females) completed three testing sessions, each testing session consisting of the eight Sway Medical mobile application modules including symptoms, balance, and cognitive tasks (reaction time, impulse control, inspection time, memory, reverse number counting, and Stroop). Intraclass correlation coefficients and Friedman’s analyses of variance were used to assess reliability and learning effects. Most measures demonstrated moderate-to-excellent reliability (intraclass correlation coefficients: .47–.91). No significant systematic error was found for balance, impulse control, or memory. However, significant learning effects were observed for Stroop ( p < .001) and reaction time ( p = .011). Reliable change indices were established at 70%, 80%, and 90% confidence intervals to assist in identifying clinical decline. The Sway Medical mobile application is a stable and reliable tool for most balance and cognitive assessments in healthy young adults. Clinicians should account for potential learning effects in reaction time and Stroop tasks by utilizing practice trials.