This is the fifth edition of a very successful textbook on clinical trials methodology, written by recognized leaders who have long and extensive experience in all areas of clinical trials. The three authors of the first four editions have been joined by two others who add great expertise. A chapter on regulatory issues has been included and the chapter on data monitoring has been split into two and expanded. Many contemporary clinical trial examples have been added. There is much new material on adverse events, adherence, issues in analysis, electronic data, data sharing and international trials. This book is intended for the clinical researcher who is interested in designing a clinical trial and developing a protocol. It is also of value to researchers and practitioners who must critically evaluate the literature of published clinical trials and assess the merits of each trial and the implications for the care and treatment of patients. The authors use numerous examples of published clinical trials to illustrate the fundamentals. The text is organized sequentially from defining the question to trial closeout. One chapter is devoted to each of the critical areas to aid the clinical trial researcher. These areas include pre-specifying the scientific questions to be tested and appropriate outcome measures, determining the organizational structure, estimating an adequate sample size, specifying the randomization procedure, implementing the intervention and visit schedules for participant evaluation, establishing an interim data and safety monitoring plan, detailing the final analysis plan and reporting the trial results according to the pre-specified objectives. Although a basic introductory statistics course is helpful in maximizing the benefit of this book, a researcher or practitioner with limited statistical background would still find most if not all the chapters understandable and helpful. While the technical material has been kept to a minimum, the statistician may still find the principles and fundamentals presented in this text useful.

Clinical trials have become essential research tools for evaluating the benefits and risks of new interventions for the treatment and prevention of diseases, from cardiovascular disease to cancer to AIDS. Based on the authors’ collective experiences in this field, Introduction to Statistical Methods for Clinical Trials presents various statistical topics relevant to the design, monitoring, and analysis of a clinical trial.

After reviewing the history, ethics, protocol, and regulatory issues of clinical trials, the book provides guidelines for formulating primary and secondary questions and translating clinical questions into statistical ones. It examines designs used in clinical trials, presents methods for determining sample size, and introduces constrained randomization procedures. The authors also discuss how various types of data must be collected to answer key questions in a trial. In addition, they explore common analysis methods, describe statistical methods that determine what an emerging trend represents, and present issues that arise in the analysis of data. The book concludes with suggestions for reporting trial results that are consistent with universal guidelines recommended by medical journals.

Developed from a course taught at the University of Wisconsin for the past 25 years, this textbook provides a solid understanding of the statistical approaches used in the design, conduct, and analysis of clinical trials.

Table of Contents

PREFACE

Introduction to Clinical Trials

History and Background

Ethics of Clinical Research

Types of Research Design and Types of Trials

The Need for Clinical Trials

The Randomization Principle

Timing of a Clinical Trial

Trial Organization

Protocol and Manual of Operations

Regulatory Issues

Overview of the Book

Defining the Question

Statistical Framework

Elements of Study Question

Outcome or Response Measures

The Surrogate Outcome

Composite Outcomes

Summary

Problems

Study Design

Early Phase Trials

Phase III/IV Trials

Non-Inferiority Designs

Screening, Prevention, and Therapeutic Designs

Adaptive Designs

Conclusions

Problems

Sample Size

Sample Size versus Information

A General Setup for Frequentist Designs

Loss to Follow-up and Non-Adherence

Survival Data

Clustered Data

Tests for Interaction

Equivalence/Non-Inferiority Trials

Other Considerations

Problems

Randomization

The Role of Randomization

Fixed Randomization Procedures

Treatment- and Response-Adaptive Randomization Procedures

Covariate-Adaptive Randomization Procedures

Summary and Recommendations

Problems

Data Collection and Quality Control

Planning for Collection of Clinical Trial Data

Categories of Clinical Data

Data Quality Control

Conclusions

Survival Analysis

Background

Estimation of Survival Distributions

Comparison of Survival Distributions

Regression Models

Composite Outcomes

Summary

Problems

Longitudinal Data

A Clinical Longitudinal Data Example

The Subject-Specific Model

Two-Stage Estimation

The Random-Effects, Subject-Specific Model

The Population-Average (Marginal) Model

Restricted Maximum Likelihood Estimation (REML)

Standard Errors

Testing

Additional Levels of Clustering

Generalized Estimating Equations for Non-Normal Data

Missing Data

Summary

Quality of Life

Defining QoL

Types of QoL Assessments

Selecting a QoL Instrument

Developing a QoL Instrument

Quality of Life Data

Analysis of QoL Data

Summary

Data Monitoring and Interim Analysis

Data and Safety Monitoring

Examples

The Repeated Testing Problem

Group Sequential Tests

Triangular Test

Curtailment Procedures

Inference Following Sequential Tests

Discussion

Problems

Selected Issues in the Analysis

Bias in the Analysis of Clinical Trial Data

Choice of Analysis Population

Missing Data

Subgroup Analyses

Multiple Testing Procedures

Summary

Problems

Closeout and Reporting

Closing out a Trial

Reporting Trial Results

Problems

Appendix: Delta Method, Maximum Likelihood Theory, and Information

Delta Method

Asymptotic Theory for Likelihood-Based Inference

Hypothesis Testing

Computing the MLE

Information

Brownian Motion

REFERENCES

INDEX

Randomized clinical trials are the gold standard for establishing many clinical practice guidelines and are central to evidence based medicine. Obtaining the best evidence through clinical trials must be done within the boundaries of rigorous science and ethical principles. One fundamental principle is that trials should not continue longer than necessary to reach their objectives. Therefore, trials must be monitored for recruitment progress, quality of data, adherence to patient care or prevention standards, and early evidence of benefit or harm. Frequently, a group of external experts, independent from the investigators and trial sponsor, is charged with this monitoring responsibility, especially for safety and early benefit. This group is referred to by various names, such as a data monitoring committee or a data and safety monitoring board. This book, through a series of case studies presented by many distinguished clinical trial experts, illustrates the complexity of this monitoring process. The editors provide an overview of the process and a summary of a multitude of the lessons learned from the cases presented.

This book should be useful to anyone serving on a data and safety monitoring board, or planning to do so, for colleagues in academia, industry and governmental agencies, and for teaching students in biostatistics, epidemiology, clinical trials and medical ethics. No other text has as extensive a collection of cases which provide insight into the many issues, often conflicting, that must be examined before recommendations to continue or discontinue a trial can be made. While depth in statistical methods is not required, some familiarity with statistical design andanalysis issues in clinical trials is helpful. The cases cover trials which were terminated early for convincing evidence of benefit, or for harmful effects. Cases with complex issues are also included. This series of cases should provide broad background information for potential monitoring committee members and better prepare them for the challenges that may exist in the trials for which they are responsible.

The three editors have contributed two overview chapters as well as several case studies to go along with cases contributed by a distinguished group of colleagues experienced in the design, monitoring and analysis of clinical trials. Dr. David DeMets is currently Professor and Chair, Department of Biostatistics and Medical Informatics at the University of Wisconsin-Madison. He is past president of the Eastern North American Region (ENAR) of the International Biometric Society, a past member of the Board of Directors of the American Statistical Association and an elected Fellow. He recently received the Robert Gordon Lectureship Award, given by the National Institutes of Health, for significant contributions to the field of clinical trials. Dr. Curt Furberg, is currently Professor (and former Chair) of the Department of Public Health Sciences at Wake Forest University. Previously, he was Head of the Clinical Trials Branch and Associate Director of the Clinical Applications and Prevention Program at the National Heart, Lung, and Blood Institute. Dr. Lawrence Friedman is a former Director of the Division of Epidemiology and Clinical Applications and a former Assistant Director for Ethics and Clinical Research at the National Heart, Lung, and Blood Institute. All three arepast presidents of the Society of Clinical Trials. The editors have collaborated previously as coauthors of a text: the Fundamentals of Clinical Trials.