Binary Data Analysis of Randomized Clinical Trials with Noncompliance E-Book

Kung-Jong Lui

• Herausgeber: John Wiley & Sons
• Kategorie: Wissenschaft und neue Technologien
• Serie: Statistics in Practice
• Sprache: Englisch

Seitenzahl: 417

Veröffentlichungsjahr: 2011

Beschreibung

It is quite common in a randomized clinical trial (RCT) to encounter patients who do not comply with their assigned treatment. Since noncompliance often occurs non-randomly, the commonly-used approaches, including both the as-treated (AT) and as-protocol (AP) analysis, and the intent-to-treat (ITT) (or as-randomized) analysis, are all well known to possibly produce a biased inference of the treatment efficacy. This book provides a systematic and organized approach to analyzing data for RCTs with noncompliance under the most frequently-encountered situations. These include parallel sampling, stratified sampling, cluster sampling, parallel sampling with subsequent missing outcomes, and a series of dependent Bernoulli sampling for repeated measurements. The author provides a comprehensive approach by using contingency tables to illustrate the latent probability structure of observed data. Using real-life examples, computer-simulated data and exercises in each chapter, the book illustrates the underlying theory in an accessible, and easy to understand way. Key features: * Consort-flow diagrams and numerical examples are used to illustrate the bias of commonly used approaches, such as, AT analysis, AP analysis and ITT analysis for a RCT with noncompliance. * Real-life examples are used throughout the book to explain the practical usefulness of test procedures and estimators. * Each chapter is self-contained, allowing the book to be used as a reference source. * Includes SAS programs which can be easily modified in calculating the required sample size. Biostatisticians, clinicians, researchers and data analysts working in pharmaceutical industries will benefit from this book. This text can also be used as supplemental material for a course focusing on clinical statistics or experimental trials in epidemiology, psychology and sociology.

Leseprobe

Contents

Cover

Statistics in Practice

Title Page

Copyright

Dedication

Preface

About the Author

1: Randomized clinical trials with noncompliance: issues, definitions and problems of commonly used analyses

1.1 Randomized encouragement design (RED)

1.2 Randomized consent designs

1.3 Treatment efficacy versus programmatic effectiveness

1.4 Definitions of commonly used terms and assumptions

1.5 Most commonly used analyses for a RCT with noncompliance

2: Randomized clinical trials with noncompliance under parallel groups design

2.1 Testing superiority

2.2 Testing noninferiority

2.3 Testing equivalence

2.4 Interval estimation

2.5 Sample size determination

2.6 Risk model-based approach

Appendix

3: Randomized clinical trials with noncompliance in stratified sampling

3.1 Testing superiority

3.2 Testing noninferiority

3.3 Testing equivalence

3.4 Interval estimation

3.5 Test homogeneity of index in large strata

Appendix

4: Randomized clinical trials with noncompliance under cluster sampling

4.1 Testing superiority

4.2 Testing noninferiority

4.3 Testing equivalence

4.4 Interval estimation

4.5 Sample size determination

4.6 An alternative randomization-based approach

Appendix

5: Randomized clinical trials with both noncompliance and subsequent missing outcomes

5.1 Testing superiority

5.2 Testing noninferiority

5.3 Testing equivalence

5.4 Interval estimation

5.5 Sample size determination

5.6 An alternative missing at random (MAR) model

Appendix

6: Randomized clinical trials with noncompliance in repeated binary measurements

6.1 Testing superiority

6.2 Testing noninferiority

6.3 Testing equivalence

6.4 Interval estimation

6.5 Sample size determination

References

Index

Statistics in Practice

Statistics in Practice

Series Advisors

Human and Biological Sciences Stephen SennUniversity of Glasgow, UK

Earth and Environmental Sciences Marian ScottUniversity of Glasgow, UK

Industry, Commerce and Finance Wolfgang JankUniversity of Maryland, USA

Statistics in Practice is an important international series of texts which provide detailed coverage of statistical concepts, methods and worked case studies in specific fields of investigation and study.

With sound motivation and many worked practical examples, the books show in down-to-earth terms how to select and use an appropriate range of statistical techniques in a particular practical field within each title's special topic area.

The books provide statistical support for professionals and research workers across a range of employment fields and research environments. Subject areas covered include medicine and pharmaceutics; industry, finance and commerce; public services; the earth and environmental sciences, and so on.

The books also provide support to students studying statistical courses applied to the above areas. The demand for graduates to be equipped for the work environment has led to such courses becoming increasingly prevalent at universities and colleges. It is our aim to present judiciously chosen and well-written workbooks to meet everyday practical needs. Feedback of views from readers will be most valuable to monitor the success of this aim.

A complete list of titles in this series appears at the end of the volume.

This edition first published 2011 © 2011 John Wiley & Sons, Ltd

Registered officeJohn Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, United Kingdom

For details of our global editorial offices, for customer services and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com.

The right of the author to be identified as the author of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher.

Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books.

Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought.

Library of Congress Cataloging-in-Publication Data

Lui, Kung-Jong, author. Binary data analysis of randomized clinical trials with noncompliance/Kung-Jong Lui, Department of Mathematics and Statistics, San Diego State University, USA. p. ; cm. Includes bibliographical references and index. ISBN 978-0-470-66095-9 (cloth) – ISBN 978-1-119-99160-1 (epdf) – ISBN 978-1-119-99161-8 (obook) – ISBN 978-1-119-99390-2 (epub) – ISBN 978-1-119-99391-9 (mobi) 1. Clinical trials–Statistical methods. 2. Drugs–Testing–Statistical methods. I. Title. [DNLM: 1. Randomized Controlled Trials as Topic–methods. 2. Medication Adherence. 3. Statistics as Topic–methods. QV 771] RM301.27.L85 2011 615.5072$′$4–dc22 2010051072

A catalogue record for this book is available from the British Library.

Print ISBN: 978-0-470-66095-9 ePDF ISBN: 978-1-119-99160-1 oBook ISBN: 978-1-119-99161-8 ePub ISBN: 978-1-119-99390-2 Mobi ISBN: 978-1-119-99391-9

Dedicated to

Professors William G. Cumberland and Abdelmonem A. Afifi at UCLA, as well as Professor Daniel McGee at Florida State University

Preface

In a randomized clinical trial (RCT), it is quite common to encounter patients who do not comply with their assigned treatment due to ethical reasons, patient's decision or the feature of a study design (such as pre-randomized consent designs). Since noncompliance often occurs non-randomly, the commonly-used subgroup analyses, including as-treated (AT) analysis and as-protocol (AP) analysis, are well known to produce a possibly biased inference of treatment efficacy due to the incomparability of the underlying prognostic conditions for patients between two comparison groups. To alleviate this concern, the intent-to-treat (ITT) (or as-randomized (AR)) analysis has been often suggested for a RCT with noncompliance. However, the ITT analysis estimates the programmatic effectiveness rather than the treatment efficacy. Although ITT analysis may provide us with unbiased test for assessing the superiority of an experimental treatment to a standard treatment, the ITT analysis tends to underestimate the relative treatment effect in the presence of noncompliance under certain commonly-assumed conditions. Thus, how to assess the treatment efficacy in a RCT with noncompliance becomes practically useful and important.

The analysis of data for a RCT with noncompliance is generally quite complicated even for the simplest case of a simple noncompliance RCT, in which only patients assigned to the experimental treatment can have access to the experimental treatment. Furthermore, the frequent involvement of sophisticated numerical iterative procedures based on likelihoods to obtain parameter estimates makes this topic even more challenging and difficult for many clinicians and data analysts to appreciate. This book is to focus attention on the level which clinicians with one year of solid training in biostatistics can comprehend, and provides readers with a simple, systematic, and organized approach to study treatment effect for a RCT with noncompliance when the patient response is dichotomous and the noncompliance status is all-or-none in a variety of situations. This book adopts an instructive and easily-understood approach by using contingency tables to explicitly lay down the latent probability structure of observed data so that readers can easily visualize the logics and the ideas behind the development of the proposed test procedures and estimators in a one unified model frame. By contrast, when using the proportion difference (PD) to measure the relative treatment effect, we assume the structural risk additive model based on the model-based approach. While using the proportion ratio (PR) to measure the relative treatment effect, we assume the structural risk multiplicative model. Furthermore, this book presents all test procedures, estimators and sample size calculation procedures in closed forms. Readers may simply use a hand calculator to calculate all the test statistics, interval estimators or sample size calculation formulae without the need of employing any iterative numerical procedures in the situations considered here. For the easy access of the particular topic of reader's interest, this book is written in such a constructive structure that the underlying assumptions, notation, test procedures and formulae in each chapter are self-contained. Readers may directly refer to the particular chapter without the need of reading the details in all the preceding chapters, although I must admit that some assumptions, definitions in notation, and important notes are repeated to avoid confusions in narrative or ambiguities in formulae and findings. Through some real-life examples and computer-simulated data, readers can appreciate the practical usefulness of the test procedures and estimators discussed in this book. The exercises given at the end of each chapter can further help readers better understand the underlying assumptions, the theory and limitations of the proposed test procedures and estimators, as well as other relevant issues and extensions. To facilitate the use of sample size determination presented here, we include in Appendix SAS programs that can be easily modified by readers to accommodate the situations in which they are interested. Despite the book generally adopting the principal stratification approach to account for the effect due to noncompliance, this book also briefly addresses use of the model-based approach (which is related to a quite general class of the structural mean models (SMMs) proposed elsewhere) and notes the relations of parameters and estimators between these two approaches. Because the discussion on the SMM is truly beyond the modest scope of this book, the SMM is not discussed in the book. Readers who are interested in this area may begin with reading a few key references regarding the SMMs cited here.

This book is intended for postgraduates, clinicians, biostatisticians and data analysts. This book can be used as supplemental material for an introductory-level course focusing on clinical statistics or experimental trials in Epidemiology, Psychology and Sociology. This book may also be used as a desk reference for clinicians or biostatisticians when they come across binary data in the presence of noncompliance. To clarify the main issues raised by noncompliance and strengthen the narrative, we explicitly define and discuss some common assumptions and terms encountered in a RCT with noncompliance, as well as include numerical examples to illustrate the bias of most commonly-used subgroup analyses in Chapter 1. Because testing superiority, non-inferiority and equivalence, interval estimation and sample size calculation are all the most fundamental statistical topics for analyzing clinical data, this book concentrates discussions on these when we use the PD, the PR and the odds ratio (OR) to measure treatment efficacy under various frequently-encountered situations. These include parallel groups design (Chapter 2), stratified sampling (Chapter 3), cluster sampling (Chapter 4), parallel sampling with subsequent missing outcomes (Chapter 5) and data in repeated binary measurements (Chapter 6). Clinicians and biostatisticians should find that this book is useful and handy.

I wish to express my indebtedness to my colleagues Drs. Richard Levine, Barbara Bailey and Kristin Duncan at San Diego State University and the five anonymous reviewers who generously provided valuable comments and suggestions on an early draft and outlines of contents of the manuscript. I also wish to thank my wife Jen-Mei, whose continued patience and understanding have endured throughout so many years and made the work much more pleasant than it otherwise would have been. I want to thank my brothers Dan-Yang, Kung-Yi and Kung-Jen for their encouragements in many years. Finally I want to express my deepest appreciation to my parents, Shung-Wu and Li-Ching for their endless love, spiritual support and guidance, which continue to last in my memory.

Kung-Jong Lui San Diego

About the Author

Kung-Jong Lui is a professor in the Department of Mathematics and Statistics at San Diego State University. He obtained his Ph.D. in biostatistics in 1982, M.S. in biostatistics in 1979, M.A. in Mathematics in 1977, all from UCLA, and B.S. in Mathematics in 1975 at Fu-Jen Catholic University at Taipei, Taiwan. He has had 150 publications in peer-reviewed journals, including Biometrics, Statistics in Medicine, Biometrical Journal, Computational Statistics and Data Analysis, Psychometrika, Journal of Biopharmaceutical Statistics, Drug Information Journal, Contemporary Clinical Trials, Journal of Applied Statistics, Statistical Methodology, Communications in Statistics, Theory and Methods, Science, Nature, Proceedings of National Academy of Sciences, Journal of Official Statistics, IEEE Transactions on Reliability, Environmetrics, Test, American Journal of Epidemiology, American Journal of Public Health, New England Journal of Medicine, Journal of the American Medical Association, etc. He is the author of the book Statistical Estimation of Epidemiological Risk published by Wiley in 2004. He is an Associate Editor for Biometrical Journal. He is a Fellow of the American Statistical Association, a Fellow of the American College of Epidemiology, and a life member of International Chinese Statistical Association.