Translating Concussion Gait Evaluations To A Mobile Approach: Reliability And Test Environment Differences

PURPOSE: To examine the reliability of an at-home gait assessment performed weekly via smartphone application, and to examine differences in gait performance compared to in-person assessment in a laboratory environment. METHODS: Participants underwent an in-person gait assessment at two timepoints 8 weeks apart by completing a 20-meter self-paced walking task under single/dual-task conditions. Between visits, participants self-administered at-home gait assessments weekly (IMPROVE, Control One LLC). Gait assessments were conducted with a smartphone affixed to the lumbar spine (in-person) or placed in a back pocket (at-home). The primary outcome for single/dual-task gait was gait velocity (m/s). Reliability was evaluated using one-way ANOVA and intraclass correlation coefficient (ICC). Testing environments were compared using a paired samples t-test. RESULTS: Thirty-seven uninjured young adults (age = 26.1 ± 3.2 years; 70% female) were assessed in-person at two timepoints (mean: 55.32.7 days apart). Participants completed 82% and 75% of possible weekly single/dual-task gait assessments at-home, respectively. At-home single-task gait had good reliability (Figure 1A) and at-home dual-task gait had excellent reliability (Figure 1B). Participants walked faster in-person under single-task (Figure 1C) and dual-task (Figure 1D) conditions compared to at-home testing. CONCLUSIONS: At-home gait assessments demonstrated high reliability across time during single/dual-task gait mobile testing in uninjured adults. Remote monitoring may be useful in an injured population during recovery when patients do not have easy/regular access to medical care. Significantly faster laboratory gait velocity may be partially explained by the Hawthorne effect, as participants may have performed differently under observation. Clinicians should consider these differences if comparing in-person and at-home task performance.