OBJECTIVES: This systematic review evaluates the diverse applications of Artificial Intelligence (AI) in generating real-world evidence (RWE) within healthcare. The aim is to synthesize findings, assess methodologies, and identify challenges in AI’s application for robust RWE.

METHODS: A comprehensive literature search was conducted across PubMed/MEDLINE, Google Scholar, and Web of Science, focusing on studies published from 2014 to 2024 that employed AI techniques to generate RWE. Key search terms included "Artificial Intelligence," "Real-World Evidence," "Health Data," and "Machine Learning." The review adhered to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, emphasizing the methodologies and outcomes of the selected studies.

RESULTS: From a pool of 1123 initially identified studies, including database and grey literature sources, 76 underwent full-text review, with 26 meeting final inclusion criteria. The review showcased a global distribution of studies, with a notable concentration in the United States (n=12), followed by Europe (n=8), and Asia (n=6). AI applications in generating RWE were diverse, spanning cardiovascular and respiratory (n=6), neurological Conditions (n=5), endocrine conditions (n=5), general and Miscellaneous Conditions (n=10). Methodologies included machine learning algorithms (n=15), natural language processing (NLP) (n=7), deep learning techniques (n=3), and Random Forest, XGBoost (n=1). These approaches were applied across various data sources such as electronic health records (EHRs), claims data, and patient-reported outcomes, demonstrating AI's potential to enhance RWE accuracy in data extraction, disease prediction, and treatment outcomes. Challenges identified included data quality, integration issues across healthcare systems, and the need for more extensive long-term validation studies to ensure reliability.

CONCLUSIONS: AI offers promise in advancing RWE, supporting informed healthcare decisions and personalized care. Addressing data integration, ethical considerations, and validating AI-generated RWE are crucial for maximizing AI’s impact in healthcare. Continued advancements and rigorous evaluations are essential for effective AI integration into the healthcare landscape.