JMIR Publications

ABSTRACT

Background:

Medication adherence is a critical challenge for people living with dementia (PLwD) and their caregivers. Standard care relies on appropriate medication management, yet there are few effective options for PLwD beyond manual pill counting approaches and caregiver administrated dosing. These methods are prone to errors and impose significant burden. Technologically enhanced adherence tools include smart caps, reminder apps, and electronic dispensers which have improved tracking and provided basic alerts, but continue to depend on manual interaction, lack integration with clinical systems, and are often unsuitable for individuals with cognitive decline. The HiDO Home Care System represents a fourth-generation artificial intelligence (AI) self-medication device (SMD), advancing the field by removing manual pill counting, automating chain-of-custody, verifying consumption, and logging medication adherence through neuroscience-based logic and real-time monitoring.

Objective:

This study evaluated the feasibility, usability, and performance of the HCS for at-home medication management in dyads of PLwD and their caregivers. We specifically examined set up, accuracy of dispensing medication, efficiency of task completion, and satisfaction with the device.

Methods:

A pooled analysis usability study was conducted with 35 caregiver-patient dyads at Rush University Medical Center. Participant dyads completed device set up, medication dispensing, and simulated medication use using the system’s automated logging and dual-camera verification. Dyads repeated dispensing and simulated medication use following automated reminders sent to their mobile device. Dyads were encouraged to repeat dispensing tasks multiple time. Quantitative measures included time to set-up the device, time from reminder to dispensation, number of successful attempts, and device reliability and system usability scores (SUS). Qualitative measures captured caregiver perceptions of usability, acceptability, and burden.

Results:

All 35 dyads successfully dispensed medications using the HCS. The average time for the first dispense attempt was 1:41 minutes (n=35). The second attempt averaged 1:35 minutes (n=21). Attempt 3 averaged 2:03 minutes (n=6). The system maintained accuracy across all users, with some variability in timing across age groups. The HCS received an overall mean System Usability Scale (SUS) score of 70.2 (n=34). Caregivers reported access to a system like HCS could reduce stress associated with medication administration and recommended improvements to specific design elements.

Conclusions:

The HCS demonstrated early feasibility, accuracy, and usability as an advanced SMD tailored to the unique needs of PLwD. By automating the medication safety chain from delivery through consumption, HCS reduces caregiver workload and enhances patient safety and medication management. These findings support HCS as a viable next-generation adherence solution, addressing limitations of prior devices and advancing dementia care. Larger-scale, longitudinal studies are planned to examine clinical outcomes, caregiver burden, and cost-effectiveness within real-world home and community settings.