User-Oriented Evaluation of a Robotic Rollator That Provides Navigation Assistance in Frail Older Adults with and without Cognitive Impairment

Background: Navigational skills decline with age, and this decline is even more pronounced in cognitively impaired (CI) older adults. Navigation assistance is an emerging functionality of robotic rollators (RRs). The evidence on the effectiveness of RR-integrated navigation systems in potential end-users is, however, scarce. Objective: To determine whether RR-provided navigation assistance improves navigation within a real-life environment in the intended user group of frail older adults with and without cognitive impairment currently using a rollator in daily life. Methods: A randomized, between-subject, 2 × 2 factorial design was conducted to test the effects of navigation assistance and cognitive status on participants' navigation performance. Twenty CI (Mini-Mental State Examination [MMSE] 17-26) and 22 not cognitively impaired (NCI; MMSE >26) older rollator users (age 82.5 ± 8.7 years) were included. Participants were matched for cognitive status (CI vs. NCI) and randomized to one of two conditions: RR (1) with or (2) without activated navigation system. All participants had to complete a two-section navigation path with the RR in an unfamiliar, real-life environment. Participants with RR-assisted navigation were supported in wayfinding by directional audio cues of the RR-integrated navigation system. Participants without RR-assisted navigation had to complete the sections by orienting themselves along conventional signposts. Outcomes were success rate, completion and stopping time, number of stops, walking distance, and gait speed. Results: The navigation assistance condition had no significant effect on the success rate in the CI, NCI, or total group. We found significant interactions between navigation assistance and cognitive status for both sections (p = 0.002-0.040), such that RR-assisted navigation reduced the completion time (both sections), stopping time (section 1), and number of stops (section 2) in the CI (p ≤ 0.001-0.014) but not in the NCI group. On the more complex section 2, RR-assisted navigation led to a reduced stopping time and walking distance in the total group (p = 0.014-0.016). Conclusion: The RR-integrated navigation system was effective for improving navigation within a real-life environment in potential end-users, especially in those with cognitive impairment. This is the first study to provide statistical evidence on the effectiveness of an RR-integrated navigation system in the intended user group.