Does This Treatment Cause That Outcome? The Science of Estimating a Treatment Effect and Why It Matters

Part 1: Some History begins with Chapter 1: The Goal of Science, which includes subtopics 1.1 Cause-and-Effect, 1.2 The Light Switch, 1.3 Clinical Trials as Scientific Experiments, 1.4 The Treatment Effect(s), 1.5 The Crucial Equivalence, and 1.6 Summary So Far. Chapter 2: How Did We Get Here? covers 2.1 Some History of Clinical Research, 2.2 Statistics and Clinical Research, 2.3 The University Group Diabetes Program (UGDP), 2.4 More Statistical Rigor, 2.5 Intent to Treat as the Default Standard, 2.6 The National Research Council Report on Missing Data, 2.7 The Estimand, 2.8 Dapagliflozin and the FDA Advisory Committee Meeting, 2.9 Estimands – Broadening the Perspective, 2.10 The Stuff of Life, and 2.11 Summary So Far. Chapter 3: ICH E9(R1) Addendum on Estimands includes 3.1 What is the Question? An Example, 3.2 Another Example: Alzheimer’s Disease, and 3.3 Causal Inference. Part 2: ICH E9(R1) Explained begins with Chapter 4: The Four Attributes, which includes 4.1 Background, 4.2 Intercurrents Events, 4.3 Attribute 1: What Is the Treatment?, 4.4 Examples of Treatment Descriptions in the Medical Literature, 4.5 The Estimand-Defined Study Treatment, 4.6 A (Very Big) Missed Piece, 4.7 What Is the Treatment Effect Questions?, 4.8 Attribute 2: What is the Population?, 4.9 Attribute 3: What is the Variable?, 4.10 Attribute 4: What Is the Population-level Summary Measure, and 4.11 Summary. Chapter 5: The Five Strategies for Intercurrent Events includes 5.1 A Thought Experiment, 5.2 Some Early Commentary on the ITT Approach, 5.3 Summary So Far, 5.4 The Five Strategies, 5.5 Incomplete Data and Intercurrent Events, 5.6 Somethin’s Gotta Give, 5.7 Strategic Thinking, and 5.8 Summary. Chapter 6: The Tripartite Estimand Approach includes 6.1 Mixture Distribution, 6.2 An Illustrative Example, 6.3 The Tripartite Estimand Approach (TEA), 6.4 Visualization, and 6.5 Summary. Part 3: Weaving the Golden Thread begins with Chapter 7: Implementation of the Estimand Framework, which includes 7.1 The Story So Far, 7.2 The Succinct Estimand Framework, 7.3 An Illustrative Example Based in Reality, and 7.4 Summary. Chapter 8: Examples of the Golden Thread includes 8.1 Preamble, 8.2 Example 1: Covid-19 – A Vaccine Strategy, 8.3 Example 2: Graft versus Host Disease (GVHD), 8.4 Example 3: Major Depressive Disorder, 8.5 Example 4: Hepatocellular Carcinoma, 8.6 Example 5: Age-Related Macular Degeneration, and 8.7 Example 6: Chimeric Antigen Receptor T-cell Therapy. Part 4: Epistemology - What is the Truth? begins with Chapter 9: What Do We Mean by the Mean?, which includes 9.1 Suppose We Knew the Truth, 9.2 In Search of the Truth!, 9.3 What Do We Mean by the Null Hypothesis?, and 9.4 Summary. Chapter 10: Potential Explanations of the Estimand includes 10.1 Background, 10.2 Potential Outcomes, 10.3 Principal Stratification, 10.4 The Connection to ICH E9(R1), 10.5 The Tripartite Estimand Approach, and 10.6 Summary. The Epilogue includes Chapter 11: Epilogue, with subtopics 11.1 Principles, 11.2 Process, 11.3 Practicalities, and 11.4 Progress, followed by References.

Biography

Dr. Stephen Ruberg received a BA in mathematics from Thomas More College, an MS in Statistics from Miami of Ohio, and a PhD in Biostatistics from the University of Cincinnati.

Dr. Ruberg was in the pharma industry for 38 years and worked across drug development and commercialization – from R&D to Business Analytics. Throughout his career, Steve had senior leadership roles, including VP of Statistics and Data Management at several companies. While at Lilly, he formed the Advanced Analytics Hub and was its Scientific Leader. He was ultimately named a Distinguished Research Fellow in Lilly R&D. Dr. Ruberg served in many leadership roles in the pharmaceutical industry and statistical profession. He co-authored ICH-E9 Statistical Principles for Clinical Trials, and most notably, Steve served on a select Advisory Committee to the Secretary of Health and Human Services during the Bush administration for advancing the use of electronic medical records.

After retiring from Lilly in 2017, Dr. Ruberg has founded his own consulting firm, Analytix Thinking, LLC, which focuses on consulting and teaching pharma companies big and small, as well as lecturing and publishing on important statistical topics. Dr. Ruberg’s current research interests include estimands, subgroup identification, Bayesian methods for clinical drug development, and digital medicine. He has been a Fellow of the American Statistical Association since 1994, was given the Career Achievement Award by Quantitative Scientists in the Pharmaceutical Industry and was elected a Fellow of International Statistics Institute.

“Without a roadmap, you get lost before you know in the maze of clinical trials, treatment effects, estimands, intercurrent events, causal inference, missing data, and ICH E9(R1). Thankfully, Steve Ruberg’s opus is a masterfully crafted aid, built upon decades of original thinking, practice, and leadership in industry and the profession. It all clicks into place, thanks to Dr. Ruberg’s engaging, witty, and accessible writing style, his eye for detail and talent for clarity, without overly mathematizing, his attention for a logical taxonomical framework. Understanding is catalyzed by the author’s examples and experience, and well-chosen metaphors. This is a rewarding read for the practicing trialist, as well as for the master and PhD level student and instructor.”-
~Geert Molenberghs, Universiteit Hasselt & KU Leuven, Belgium

 “After more than five decades interpreting clinical trial outcomes, including as Center Director of the FDA CDER, now comes a book that adds clarity to my understanding of what treatment has caused a trial outcome.  Stephen Ruberg's book clarifies learnings from clinical trials that often confuse initiation of treatment with actually taking the treatment. Expanding on the International Council on Harmonization guideline ICH-E9(R1) description of an estimand, this treatise lucidly emphasizes the importance of precise definitions of “treatment” and “effect.” Clarifying the limitations of Intention to Treat (ITT) analysis, Ruberg describes alternative inference approaches for interpreting treatment-induced clinical trial outcomes, accounting for non-adherence to assigned treatments and the myriad real-world events that obfuscate interpretation of clinical trial outcomes. This book is required reading for all biostatisticians and clinical trialists who aim to precisely define which treatments cause observed outcomes in modern clinical trials.”
~Carl Peck, MD, University of California at San Francisco, USA; Founder and Former Chairman of NDA Partners (a ProPharma company)

“This book focuses on the essentials of clinical trial design and how to optimally describe these in a clinical trial protocol. It offers a thoughtful exploration of the ICH E9 (R1) addendum to the ICH E9 guideline on statistical principles for clinical trials. Written as a practical tool, it provides valuable insights for anyone involved in designing and interpreting clinical trials. Ruberg emphasizes the importance of starting the trial design process with a clear discussion of the clinical question the trial aims to answer. His easy-to-read style, supported by recurring examples, illustrates the need for the estimand framework and addresses its key elements. The book’s golden thread is guiding readers from identifying a relevant clinical question to concluding on a causal treatment effect. While some ideas may spark debate, Ruberg encourages open discussion and challenges to his propositions. Written with non-statisticians in mind, the book is accessible and thought-provoking, offering fresh perspectives on trial design. For clinical researchers, it’s a call to engage with the ICH E9 (R1) addendum, embrace teamwork, and recognize that estimand is not merely a statistical concept but a critical summary of the clinical question and how the trial will address it.”
~Nanco Hefting, Chief Specialist, Global Clinical Development, H. Lundbeck A/S, Denmark

“This book is an essential guide for clinical research professionals. With a clear and engaging style, it explores ICH E9(R1) and the estimand framework, offering practical strategies to simplify implementation in clinical trials. Dr. Ruberg emphasizes the importance of defining treatment conditions precisely and framing meaningful clinical questions within their context. Through relatable examples – such as the challenge of estimating hiking time up Mount Kilimanjaro – and rich historical insights on cause and effect, the book creates a “golden thread” from clinical question to statistical interpretation. Bridging the gap between clinical and statistical communities, it advocates for common sense, clarity, and flexibility in defining treatment effects. As a statistician who has been working for 20+ years in the pharmaceutical industry, I strongly recommend this book as an indispensable resource for researchers seeking to better understand treatment effects and design trials that accurately estimate them.”
~Elena Polverejan, PhD

"This is an insightful book for diligent scientists doing the serious work of clinical research.  Dr. Ruberg, a highly-experienced statistical scientist in the context of clinical trials, pushes hard against the status quo when dealing with the complexities of trying to make inference about the effect of a treatment that can be validly deduced from clinical trials. He addresses the complicated world of the design, implementation, and review of clinical trial data from first principles, tying in the philosophy of statistical inference and medical science and addressing regulatory precedent and practice. He establishes order to quiet confusion, and strict logic and process to subdue uncertainty and misunderstanding.  The end result is a calm clarity and precision to the scientific thinking surrounding the assessment of a treatment effect. Grounded in the principles of statistical inference, medical science, and regulatory practice, the book challenges readers to move beyond superficial solutions and embrace the complexity required to uncover rigorous, defendable answers. For those committed to excellence in clinical trials, this book offers the tools and inspiration to meet the highest standards of cause and effect that ultimately serves patients better.”
~Rafe Donahue, PhD, Retired Senior Statistical Director, GlaxoSmithKline

“Steve Ruberg has written a tour de force on estimands with clear explanations and real-world examples to guide the reader through this important but challenging statistical concept. He has accomplished one of the key goals of regulatory agencies around the globe in developing international guidance on estimands, ICH E9(R1), specifically, to encourage conversation and collaboration among clinicians and statisticians in determining what the research question is and what quantity should be estimated and tested in response to that question. Hats off to Steve for producing such an accessible text on what is often viewed as a less-than-accessible topic! I strongly recommend this book for any graduate level clinical trials course taught today or for anyone involved in clinical drug development.”
~Lisa M. LaVange, PhD, Professor Emerita and former Chair of Biostatistics, University of North Carolina at Chapel Hill, USA (and former Director of the Office of Biostatistics in the Center for Drug Evaluation and Research (CDER) at FDA)