Smart Nursing AI Agent for Health Self-Management

Overview

This study aims to develop and evaluate a smart nursing AI agent to support health self-management among middle-aged and older patients, and further to examine its effectiveness in home care and health self-management. The goal is to propose innovative strategies for middle-aged and older care that enhance the quality of home care and strengthen patients' health self-management abilities. A randomized controlled trial design will be adopted, with a planned enrollment of 102 participants. Participants will be randomly assigned to either the experimental group or the control group. Regardless of group assignment, their routine medical care will remain unchanged. Both groups will participate in the study for 12 weeks (8 weeks of intervention and 4 weeks of follow-up), with identical frequency and number of follow-ups, interviews, and evaluations. The experimental group will receive an 8 week intervention using the smart nursing AI agent care model, which provides daily interaction, reminder systems, personalized recommendations, activity monitoring, goal setting, motivational support, and health self-management education. The control group will continue with the existing standard health education approach. Data collection will include demographic information, medical history, physiological indicators, functional activity measures (e.g., steps, distance, exercise duration), sarcopenia screening, nutritional assessment, physical activity levels, health self-management ability, health literacy, quality of life, and digital usage behaviors. Statistical analyses will be conducted to evaluate the effectiveness, feasibility, and acceptability of the intervention.

Taiwan has entered a super-aged society, where many middle-aged and older adults experience activity limitations due to muscle loss, joint degeneration, or declining physical strength. These conditions often lead to gait instability, fatigue, and an increased risk of falls, which affect daily life and raise safety concerns. Traditionally, hospital discharge or outpatient education is delivered through one-time paper handouts, which patients often forget or find difficult to follow at home. Digital nursing technologies, such as mobile applications, telehealth, and wearable devices, can overcome geographical barriers, provide continuous support, and enhance health self-management abilities. Therefore, this study aims to develop and evaluate a smart nursing AI agent to support health self-management among middle-aged and older patients, addressing challenges related to patients' declining functional capacity and nurse workforce shortages. The goal is to propose innovative strategies to improve the quality of home care and strengthen patients' self-management of health. The inclusion criteria for this study are patients aged 50 years or older who present with mobility limitations such as gait instability, sarcopenia, joint degeneration, chronic pain, or functional decline, and who can ambulate and are alert and oriented. Exclusion criteria include patients with terminal cancer or major illness with a life expectancy of less than six months, those with moderate to severe dementia or language impairment, those with severe visual or hearing impairment preventing AI interaction, and those unable to complete walking tests independently. This study adopts a randomized controlled trial design with a planned enrollment of 102 participants. Participants will be randomly assigned to either the experimental group or the control group, while their routine medical care will remain unchanged. Both groups will participate in the study for 12 weeks, including 8 weeks of intervention and 4 weeks of follow-up, with an identical frequency of follow-ups, interviews, and evaluations. The experimental group will receive an 8-week intervention using a smart nursing AI agent model, which provides daily interaction, reminder systems, personalized recommendations, activity monitoring, goal setting, motivational support, and health self-management education. In contrast, the control group will continue with standard health education practices. Data collection will include demographic information, medical history, physiological indicators, functional activity measures (such as steps, distance, and exercise duration), sarcopenia screening, nutritional assessment, physical activity levels, health self-management ability, health literacy, quality of life, and digital usage behaviors. Participants will undergo routine assessments, including physiological indicators, sarcopenia screening, nutritional assessment, and physical activity levels, twice (baseline and three months post-intervention), and complete additional questionnaires, including health self-management ability, health literacy, and quality of life, three times (baseline, one month post-intervention, and two months post-intervention). Statistical analyses will be conducted to evaluate the effectiveness, feasibility, and acceptability of the intervention. This study expects that the introduction of a smart nursing AI agent model will provide middle-aged and older patients with more continuous and effective support, thereby promoting health self-management and improving quality of life.