Bias and Randomization in Clinical Trials: 1980s–2020s–2060s

On the 80th anniversary of the Deming Conference, it is useful to look to the past, present and future of clinical trial methodology. In 1980 I published the paper “Patient Registration in a Cooperative Oncology Group” in the first volume of Controlled Clinical Trials. Here I reflect on how much has changed and how much remains the same. Although oncology trials are referenced here the issues can be interpreted more broadly. The history of randomization 1980s–2020s reflects the changes in what the research and regulatory communities considered as persuasive evidence of treatment efficacy and safety. This chapter will encourage the reader to think about what they consider persuasive evidence. Our journey will be guided by four “heroes” and will begin with the primitive oncology trials of the1970s and ensuing historical control arguments, before arriving at the highly regulated and complex trial design and analysis paradigm of today. The role of randomization is considered among the challenges of defining estimands, historical controls, real world data, Bayesian borrowing, and pragmatic trials. The chapter will conclude with imagining clinical trials in the 2060s where machine learning using aggregate data may change our definition of persuasive evidence. Also explored, are the health challenges Deming Conference presenters of the 2060s will likely be addressing. The question is raised of whether different communities (industry, academia, regulators, clinicians, patients) can co-exist each with their own definitions of persuasive evidence and how will randomization fit into these definitions.