A Scoping Review on Artificial Intelligence-Supported Interventions for Nonpharmacologic Management of Chronic...

Area of Science:

• Rheumatology
• Artificial Intelligence
• Digital Health

Background:

• Chronic rheumatic diseases impact millions globally, necessitating innovative non-pharmacological management strategies.
• Traditional interventions often face challenges in adherence and long-term efficacy.
• Artificial intelligence (AI) presents a novel approach to support patient self-management and therapeutic adherence.

Purpose of the Study:

• To systematically review and summarize AI-supported non-pharmacological interventions for adults with chronic rheumatic diseases.
• To detail the components, AI methodologies, and evidence base for these interventions.
• To assess the impact of AI interventions on clinical, psychological, and behavioral outcomes.

Main Methods:

• Comprehensive literature search across Embase, PubMed, Cochrane, and Scopus databases.
• Inclusion of all publication types and intervention settings for adults with chronic rheumatic diseases.
• Data extraction on study characteristics, intervention components, AI methodologies, and outcomes by two independent reviewers.

Main Results:

• Fifteen AI-supported interventions were identified, predominantly for osteoarthritis (73%), focusing on education and exercise (67%).
• Rule-based expert systems (40%) were the most common AI tool, followed by NLP (33%) and machine learning (27%).
• Interventions showed high usability, engagement, and adherence, with reported improvements in exercise compliance, physical activity, and symptoms, though long-term sustainability was uncertain.

Conclusions:

• AI-supported interventions demonstrate potential for improving education, exercise, and behavioral guidance in chronic rheumatic diseases.
• Evidence suggests high usability and patient engagement with AI tools.
• The long-term clinical impact of AI interventions on symptom management remains uncertain and requires further research.