Artificial Intelligence Methods and Digital Intervention Strategies for Predicting and Managing Chronic Obstructive...

Area of Science:

• Pulmonary Medicine
• Digital Health
• Artificial Intelligence

Background:

• Chronic Obstructive Pulmonary Disease (COPD) poses a significant global health challenge, with acute exacerbations worsening progression and increasing hospitalizations.
• Emerging technologies like wearable biosensors, AI, and digital health tools offer potential for proactive COPD management.

Purpose of the Study:

• To synthesize evidence on AI-driven prediction of COPD exacerbations using wearable biosignals.
• To evaluate the effectiveness of digital health interventions on COPD management, quality of life, and medication adherence.

Main Methods:

• An umbrella review systematically searched PubMed, Scopus, and Web of Science (2015-2025).
• Methodological quality was assessed using AMSTAR-2, and study overlap quantified with CCA.
• A narrative synthesis addressed the two research questions, with the protocol registered in PROSPERO.

Main Results:

• Twenty-seven reviews indicated AI models have promising internal predictive accuracy but lack external validation and clinical integration.
• Digital health interventions showed small to moderate improvements in quality of life and physical function, with significant heterogeneity (I² = 0-94%).
• Evidence for improved medication adherence and reduced exacerbations was inconsistent, and most reviews had low methodological quality.

Conclusions:

• AI and digital tools, especially using wearable biosignals, show potential for proactive COPD management, surpassing traditional assessments.
• Clinical readiness is hampered by small-scale studies, interpretability issues, inconsistent outcomes, and insufficient external validation.
• Future research must focus on large-scale, rigorous, equitable studies with standardized methods and robust generalizability testing for clinical translation and regulatory approval.