PhD defence: Multi-Database Methodologies to Enhance Real-World Evidence Generation

This thesis discusses the challenges and strategies in conducting multi-database real-world evidence (RWE) studies. It highlights the advantages of RWE studies in the estimation of drug effects in real-world populations, particularly those underrepresented in traditional randomised controlled trials. Key issues addressed included the legal and practical difficulties of creating centralised repositories, data harmonisation strategies and managing database heterogeneity and missing data. The thesis presents case studies comparing different common data models (CDMs) and evaluates the effects of database heterogeneity on study outcomes.