ISSUE: Both randomized controlled trials and observational studies provide evidence regarding pharmaceutical efficacy or effectiveness and safety. Can high-quality observational studies replace randomized controlled trials to determine relative treatment effects?

OVERVIEW: Traditional randomized controlled trial (RCT) evidence has long been considered the gold standard to determine relative treatment effects and inform clinical decision-making, regulatory approval, and value assessment. However, there is growing interest in using real-world observational evidence to evaluate relative treatment effects, both in a regulatory and a health technology assessment (HTA) context. For example, we see conditional and accelerated regulatory approvals based on single arm trials compared with historical controls based on observational data. Yet, many would argue that the risk of bias, incomplete data, possibility for data dredging or p-hacking, and data silos make it difficult or impossible to trust the results of analyses involving observational data. This panel will discuss the tension between the need for credible evidence and the demand for faster access to promising new treatment alternatives, and the current and potential future role of observational evidence. Lucinda Orsini will provide an overview of efforts and initiatives where RCTs have been compared with observational evidence. Bill Crown will argue that high quality real-world evidence is a robust alternative to RCTs and can be used to understand causal inferences of treatment effects in a more efficient and timely manner. Leanne Larson will outline the spectrum of pragmatic approaches to study design and why each is important. Jeroen Jansen will illustrate how randomized and observational study findings can be formally integrated to make optimal use of the available evidence base.