Effectiveness of the AI-Supporter in Reducing Urinary Tract Infections

Overview

The "AI Supporter," an intelligent excretion management robot, leverages artificial intelligence-based vision recognition to autonomously detect and cleanse affected areas, followed by drying and changing the diaper, thereby reducing caregiver strain and enhancing care quality. This study aims to assess the efficacy of the "AI Supporter" in decreasing the incidence of urinary tract infections and incontinence-associated dermatitis among incontinent patients, in addition to exploring its cost-effectiveness. Adopting an experimental (two groups) and longitudinal design, this research utilizes both convenience and random sampling strategies. The study anticipates recruiting 60 female subjects who have been confined to bed for more than three months with urinary and/or fecal incontinence. Participants will intermittently use the AI Supporter over a 14-day period. Measurement tools include routine urine analysis.

Background: As Taiwan progresses medically, the aging demographic has become a significant challenge, leading to an escalation in the disabled population. The lack of caregiving manpower represents a critical bottleneck in the provision of long-term care. Diaper changing, a daily and labor-intensive task for caregivers, involves bending motions that pose a risk of musculoskeletal injuries. Consequently, the imperative development of automated caregiving technologies has emerged. The "AI Supporter," an intelligent excretion management robot, leverages artificial intelligence-based vision recognition to autonomously detect and cleanse affected areas, followed by drying and changing the diaper, thereby reducing caregiver strain and enhancing care quality. Objective: This study aims to assess the efficacy of the "AI Supporter" in decreasing the incidence of urinary tract infections and incontinence-associated dermatitis among incontinent patients, in addition to exploring its cost-effectiveness. Methods: Adopting an experimental (two groups) and longitudinal design, this research utilizes both convenience and random sampling strategies. Scheduled from November 2024 to October 2025 at a residential long-term care facility in Central Taiwan, the study anticipates recruiting 60 female subjects who have been confined to bed for more than three months with urinary and/or fecal incontinence. Participants will intermittently use the AI Supporter over a 14-day period. Measurement tools include routine urine analysis, incontinence-associated dermatitis rating scales, pressure sore assessments, skin pH measurements, caregiver hours, and cost analyses pertaining to diapers and the AI Supporter. The principal analytical method employed will be Generalized Estimating Equations (GEE), with statistical significance defined at p \< 0.05. Expected Outcomes: The AI Supporter is expected to significantly reduce the occurrence of urinary tract infections and incontinance-associated dermatitis in patients, concurrently alleviating caregiver workload and diminishing associated costs.