Secondary use of real world data: challenges and opportunies in the era of artificial intelligence

The exponential growth of electronic health records (EHRs) and clinical data warehouses has created unprecedented opportunities for advancing medical research and improving patient care through secondary data use. As we enter an era where artificial intelligence (AI) transforms healthcare, the systematic exploitation of real-world data presents both remarkable potential and significant challenges that demand innovative solutions.

This presentation explores the evolving landscape of secondary use of clinical data, drawing from extensive experience in mining electronic health records across multiple domains, including cancer research and diagnostic decision support. We will examine the challenges of secondary use of data on quality, and explore how modern AI techniques, particularly advanced natural language processing, can modernize our approach to extracting actionable insights from complex, heterogeneous clinical datasets. Finally, we will present the CARPEM master observational trial, an AI-driven cancer data warehouse.