Statistics and Data Science Seminar: Prof Yinghui Wei (University of Exeter)

Routine data, also known as out-of-trial or real-world data, refer to information collected through administrative, clinical, and operational processes. As these data are generated continuously, at scale and across diverse populations, they provide a rich data resource for understanding population health, evaluating services, and supporting operational delivery. They are particularly valuable for studying real-world evidence, exploring patterns of care, assessing health inequalities, and analysing system performance. However, routine data are not primarily designed for research, and this introduces challenges, such as variable data quality, incomplete records, and computational capacity. In this work, we explore the opportunities, challenges and limitations of using routine data in health research, drawing on real examples, to illustrate both the potential and the complexities of working with these datasets.