https://orcid.org/0000-0001-9106-2204
The introduction of randomized controlled trials in the 1940s and 1960s significantly contributed to advances in clinical medicine, including the screening, diagnosis, prevention, and treatment of health and wellbeing conditions. Edited by KyungMann Kim, Frank Bretz, Ying Kuen K. Cheung, and Lisa V. Hampson, this book offers an in-depth treatment of current statistical methods that are relevant to the planning, monitoring, and analysis of clinical trials, with a specific focus on randomized controlled trials. It talks about the history of the topic, different ways to look at results, experimental designs, figuring out sample sizes, interim analyses, and statistical issues such as multiple testing, subgroup analysis, competing risks, and joint models. It also talks about topics like multiple assignment randomization trials, safety outcome analysis, non-inferiority trials, using historical data, and surrogate outcome validation.
The handbook aims to serve as a reference text for individuals involved in design, monitoring, and analysis and can also function as a textbook for a graduate course in statistical methods for randomized controlled trials. The handbook acknowledges the importance of randomized controlled trials in answering urgent clinical questions, particularly during the COVID-19 pandemic, and highlights the evolution of statistical methods for design and analysis. The Handbook of Statistical Methods for Randomized Controlled Trials provides a comprehensive guide to statistical concepts essential for designing, monitoring, and analyzing clinical trials.
Part I RCTs, a type of clinical trial, have been around since the early 20th century, with the first modern trial being the British Medical Research Council’s Streptomycin Treatment of Pulmonary Tuberculosis trial. Modern trials in the United States began in the early 1960s, supported by the National Institutes of Health and the 1962 Kefauver-Harris amendments to the Food, Drug, and Cosmetic Act of 1938. Statistical concepts are fundamental to clinical trials, and their planning, monitoring, and analysis involve a logical and sequential approach. To make statistical analysis more useful, we need to think about things like competing risks, multiple testing, subgroup analyses, using historical data, sequential multiple assignment randomization trials, and joint models for longitudinal markers and clinical outcomes.
Parts II–VI of the guidebooks cover the remaining portion. Part II (Chapters 2–7) outlines the common results employed to investigate clinical questions and hypotheses in randomized controlled trials. It also discusses the statistical distributions that are appropriate for the outcome measures of interest. These outcomes lead to the formulation of statistical hypotheses, the deployment of statistical analysis techniques to test these hypotheses, and the utilization of statistical models in the study of statistical data. Chapter 2 will discuss statistical techniques appropriate for dichotomous (qualitative) and ordinal data. Chapter 3 will cover statistical techniques suitable for continuous (quantitative) outcomes. Chapter 4 will explain statistical techniques suitable for time-to-event outcomes subject to right censoring, commonly used in randomized controlled trials in chronic diseases. Chapter 5 will outline statistical techniques suitable for count data. Chapter 6 will elaborate on statistical techniques suitable for longitudinal data. Lastly, Chapter 7 will detail statistical techniques suitable for repeated event data. The following are the outcomes often used in randomized controlled trials across many diseases and health situations: Part II provides a preliminary explanation of how to determine the required sample size and conduct a power analysis, which will be further detailed in Part III.
Part III (Chapters 8–14) is all about getting randomized controlled trials ready. It covers trial design, figuring out the sample size, and power analysis based on the chosen primary endpoints. In Part II, we previously explained these endpoints. Chapter 8 will discuss cross-over design, which involves switching participants between different treatments or conditions. Chapter 9 will cover factorial design, which examines the effects of multiple independent variables on the dependent variable. Chapter 10 will explore cluster randomized design, an approach commonly used in health services research where groups or clusters randomly receive different interventions. Chapter 11 will explore various methods of treatment allocation, including randomization, stratification, and outcome-adaptive allocation. Chapter 12 will provide information on sample size estimation and power analysis for different types of data, including dichotomous, ordinal, continuous, and count data. Chapter 13 will discuss the process of determining the appropriate sample size for time-to-event data that is subject to right censoring. Chapter 14 will focus on the estimation of sample size and power analysis for longitudinal data.
Part IV (Chapters 15–17) provides a detailed account of the monitoring process in randomized controlled trials. It covers topics such as data and safety monitoring, interim analysis, strategies for early halting, sample size re-estimation, and adaptive designs. Chapter 15 will discuss various techniques for conducting interim analyses, including group sequential methods, triangular methods, and stochastic curtailment tests. Chapter 16 will focus on sample re-estimation during interim analyses. Lastly, Chapter 17 will provide an overview of adaptable designs.
Part V (Chapter 18–21) covers practical and significant matters in data analysis that extend beyond the scope of standard statistical procedures discussed in Part II. Chapter 18 will address the issue of multiple tests resulting from the presence of multiple outcomes. Chapter 19 will explain the correct methods for conducting subgroup analyses. Chapter 20 will tackle the complex problem of competing risks in the analysis of time-to-event data. Chapter 21 will explore joint models for longitudinal markers and clinical outcomes, typically observing them as time-to-event data with right censoring.
Part VI (Chapters 22–26) discusses several subjects related to the design and analysis of randomized controlled trials. Chapter 22 will cover sequential multiple assignment randomization trials for dynamic treatment allocation in certain types of randomized controlled trials, particularly those involving consecutive interventions, such as in mental health conditions. Chapter 23 will discuss the statistical analysis of safety data, specifically adverse events. Chapter 24 will cover the design and analysis of non-inferiority trials, which aim to establish or demonstrate that the new treatment is not inferior to the standard treatment. Chapter 25 will explain how to incorporate historical data into the design and analysis of randomized controlled trials. Lastly, Chapter 26 will address the validation of outcomes as surrogate measures for clinical outcomes.
This book has the power to be a reference for teaching or research. This book provides a comprehensive guide to statistical methods relevant to randomized, controlled clinical trials. The authors explain each chapter in detail, providing an in-depth understanding of the concepts and techniques involved. The authors in this book are leading experts in their fields. The authors have extensive experience and knowledge in statistics and randomized controlled clinical trials, thereby lending authority and reliability to the content presented. A good balance between the statistical theory underlying the method and practical application in the context of randomized controlled clinical trials. Readers can understand the theoretical basics and observe the practical application of the methods to real data. This book includes examples and case studies that use real data. Readers gain a practical understanding of the methods taught in the book and an overview of their application in randomized controlled clinical trials. Each chapter includes extensive references, providing additional resources for readers who wish to explore a particular topic further. Readers can use this reference as a starting point to explore more in-depth literature on statistics and randomized controlled clinical trials.
However, this book has several weaknesses. Due to its comprehensive nature, this book may not be able to provide very in-depth detail on every topic covered. Some readers may need additional sources or more detailed references to deepen their understanding of a particular topic. Readers with a basic understanding of statistics and randomized controlled clinical trials are the intended audience for this book. Readers who do not have a strong background in statistics may face challenges in understanding more advanced concepts or complex techniques. If you are interested in learning about randomized controlled clinical trials without a strong background in statistics, there are also other resources that may be a more suitable choice, such as more beginner-friendly books or learning resources in clinical statistics or randomized controlled clinical trials.
Overall, this book is a valuable and useful resource for those interested in the statistics of randomized controlled clinical trials. However, it is important to pay attention to your level of knowledge and personal needs before deciding whether this book is suitable or not. It is best to choose learning resources that match your level of knowledge and comfort in statistics to maximize understanding and benefits gained.
Department of Elementary Education, Universitas Negeri Yogyakarta: Yogyakarta
Special Region of Yogyaka, Indonesia
[email protected]
Yoppy Wahyu Purnomo
Department of Elementary Education, Universitas Negeri Yogyakarta: Yogyakarta
Special Region of Yogyaka, Indonesia
Fery Muhamad Firdaus
Department of Elementary Education, Universitas Negeri Yogyakarta: Yogyakarta
Special Region of Yogyaka, Indonesia