JMIR Publications

ABSTRACT

Background:

Demographic change and the increasing prevalence of mobility limitations and cognitive impairments in older adults place significant strain on long-term care systems. Wearable sensor technologies, such as sensor-equipped shoes, offer promising opportunities for fall risk assessment, mobility monitoring, and detection of agitation-related motor behavior. However, little is known about their feasibility, acceptance, and integration in institutional care settings.

Objective:

With this study, we aim to generate initial insights into the detectability and classification of movement patterns in older adults exhibiting agitated behavior using a sensor-based wearable insole system, alongside assessing the feasibility and acceptability of its application in everyday care practice.

Methods:

We conducted an exploratory four-week feasibility study in two long-term care facilities in Eastern Switzerland. Six residents with mild to moderate cognitive impairment and increased fall risk were recruited. Participants wore sensor-equipped shoes that continuously recorded gait and movement data. Weekly walking tests, structured simulations of agitation-related behaviors, and observational data from nursing staff were analyzed using AI-supported algorithms. The evaluation followed the RE-AIM framework to assess effectiveness, implementation, adoption, and potential for maintenance.

Results:

The system successfully captured gait parameters and differentiated agitation-related movement patterns. Random forest classifiers achieved high accuracy in detecting predefined agitation states, while Hidden Markov Models demonstrated cost-effective potential with larger datasets. Implementation revealed substantial organizational and technical challenges, including battery charging, staff communication, and shoe fit. Resident acceptance varied: while some participants reported increased perceived safety, others found the shoes unfamiliar or the vibration feature irritating. Staff reported perceived improvements in stability and reduced agitation, though effects could not be consistently attributed to the vibration function.

Conclusions:

Sensor-equipped shoes proved technically feasible for monitoring gait and agitation-related patterns in long-term care and were generally well received by nursing staff, who valued their potential to support daily practice. While some challenges arose, particularly regarding organizational routines, fluctuating resident health, and practical aspects such as regular charging, these were context-dependent rather than fundamental barriers. While residents expressed mixed views on comfort and acceptance, these insights underline the need for context-sensitive, participatory implementation strategies. Future research should develop scalable integration models to ensure that sensor-based monitoring technologies can be feasibly embedded into daily care routines, enhancing their acceptance and long-term sustainability not only from a technical but also from an ethical, organizational, and practical perspective.