Download textbook Clinical Trial Data Analysis With R And Sas Second Edition Chen ebook all chapter pdf

Clinical trial data analysis with R and
SAS Second Edition Chen

Visit to download the full and correct content document:
https://textbookfull.com/product/clinical-trial-data-analysis-with-r-and-sas-second-editi
on-chen/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...
Analysis of Correlated Data with SAS and R Mohamed M.

Shoukri
https://textbookfull.com/product/analysis-of-correlated-data-
with-sas-and-r-mohamed-m-shoukri/
Survival Analysis with Interval Censored Data A

Practical Approach with Examples in R SAS and BUGS 1st
Edition Kris Bogaerts
https://textbookfull.com/product/survival-analysis-with-interval-
censored-data-a-practical-approach-with-examples-in-r-sas-and-
bugs-1st-edition-kris-bogaerts/
Analysis of Clinical Trials Using SAS A Practical Guide

Second Edition Alex Dmitrienko
https://textbookfull.com/product/analysis-of-clinical-trials-
using-sas-a-practical-guide-second-edition-alex-dmitrienko/
Clinical Trial Optimization Using R 1st Edition Alex

Dmitrienko
https://textbookfull.com/product/clinical-trial-optimization-
using-r-1st-edition-alex-dmitrienko/
Graphical Data Analysis with R Antony Unwin
https://textbookfull.com/product/graphical-data-analysis-with-r-
antony-unwin/
Data Analysis with R 2nd Edition Tony Fischetti
https://textbookfull.com/product/data-analysis-with-r-2nd-
edition-tony-fischetti/
Applied time series analysis with R Second Edition

Elliott
https://textbookfull.com/product/applied-time-series-analysis-
with-r-second-edition-elliott/
Step by Step Programming with Base SAS 9 4 Second

Edition Sas Institute
https://textbookfull.com/product/step-by-step-programming-with-
base-sas-9-4-second-edition-sas-institute/
Applied Meta-Analysis with R and Stata (Chapman &

Hall/CRC Biostatistics Series) 2nd Edition Chen
https://textbookfull.com/product/applied-meta-analysis-with-r-
and-stata-chapman-hall-crc-biostatistics-series-2nd-edition-chen/
Clinical Trial
Data Analysis
Using R and SAS
Second Edition
Editor-in-Chief
Shein-Chung Chow, Ph.D., Professor, Department of Biostatistics and Bioinformatics,
Duke University School of Medicine, Durham, North Carolina
Series Editors
Byron Jones, Biometrical Fellow, Statistical Methodology, Integrated Information Sciences,
Novartis Pharma AG, Basel, Switzerland
Jen-pei Liu, Professor, Division of Biometry, Department of Agronomy,
National Taiwan University, Taipei, Taiwan
Karl E. Peace, Georgia Cancer Coalition, Distinguished Cancer Scholar, Senior Research Scientist
and Professor of Biostatistics, Jiann-Ping Hsu College of Public Health,
Georgia Southern University, Statesboro, Georgia
Bruce W. Turnbull, Professor, School of Operations Research and Industrial Engineering,
Cornell University, Ithaca, New York
Published Titles
Adaptive Design Methods in Clinical Basic Statistics and Pharmaceutical
Trials, Second Edition Statistical Applications, Second Edition
Shein-Chung Chow and Mark Chang James E. De Muth
Adaptive Designs for Sequential Bayesian Adaptive Methods for
Treatment Allocation Clinical Trials
Alessandro Baldi Antognini Scott M. Berry, Bradley P. Carlin,
and Alessandra Giovagnoli J. Jack Lee, and Peter Muller
Adaptive Design Theory and Bayesian Analysis Made Simple:
Implementation Using SAS and R, An Excel GUI for WinBUGS
Second Edition Phil Woodward
Mark Chang Bayesian Designs for Phase I–II
Advanced Bayesian Methods for Clinical Trials
Medical Test Accuracy Ying Yuan, Hoang Q. Nguyen,
Lyle D. Broemeling and Peter F. Thall
Analyzing Longitudinal Clinical Trial Data: Bayesian Methods for Measures
A Practical Guide of Agreement
Craig Mallinckrodt and Ilya Lipkovich Lyle D. Broemeling
Applied Biclustering Methods for Big Bayesian Methods for Repeated Measures
and High-Dimensional Data Using R Lyle D. Broemeling
Adetayo Kasim, Ziv Shkedy, Bayesian Methods in Epidemiology
Sebastian Kaiser, Sepp Hochreiter, Lyle D. Broemeling
and Willem Talloen
Bayesian Methods in Health Economics
Applied Meta-Analysis with R Gianluca Baio
Ding-Geng (Din) Chen and Karl E. Peace
Bayesian Missing Data Problems: EM,
Applied Surrogate Endpoint Evaluation Data Augmentation and Noniterative
Methods with SAS and R Computation
Ariel Alonso, Theophile Bigirumurame, Ming T. Tan, Guo-Liang Tian,
Tomasz Burzykowski, Marc Buyse, and Kai Wang Ng
Geert Molenberghs, Leacky Muchene,
Nolen Joy Perualila, Ziv Shkedy,
and Wim Van der Elst
Published Titles
Bayesian Modeling in Bioinformatics Computational Methods in Biomedical

Dipak K. Dey, Samiran Ghosh, Research
and Bani K. Mallick Ravindra Khattree and Dayanand N. Naik
Benefit-Risk Assessment in Computational Pharmacokinetics
Pharmaceutical Research and Anders Källén
Development Confidence Intervals for Proportions
Andreas Sashegyi, James Felli, and Related Measures of Effect Size
and Rebecca Noel Robert G. Newcombe
Benefit-Risk Assessment Methods in Controversial Statistical Issues in
Medical Product Development: Bridging Clinical Trials
Qualitative and Quantitative Assessments Shein-Chung Chow
Qi Jiang and Weili He
Data Analysis with Competing Risks
Bioequivalence and Statistics in Clinical and Intermediate States
Pharmacology, Second Edition Ronald B. Geskus
Scott Patterson and Byron Jones
Data and Safety Monitoring Committees
Biosimilar Clinical Development: in Clinical Trials, Second Edition
Scientific Considerations and New Jay Herson
Methodologies
Design and Analysis of Animal Studies
Kerry B. Barker, Sandeep M. Menon,
in Pharmaceutical Development
Ralph B. D’Agostino, Sr., Siyan Xu, and Bo Jin
Shein-Chung Chow and Jen-pei Liu
Biosimilars: Design and Analysis of
Design and Analysis of Bioavailability
Follow-on Biologics
and Bioequivalence Studies, Third Edition
Shein-Chung Chow
Shein-Chung Chow and Jen-pei Liu
Biostatistics: A Computing Approach
Design and Analysis of Bridging Studies
Stewart J. Anderson
Jen-pei Liu, Shein-Chung Chow,
Cancer Clinical Trials: Current and and Chin-Fu Hsiao
Controversial Issues in Design and
Design & Analysis of Clinical Trials for
Analysis
Economic Evaluation & Reimbursement:
Stephen L. George, Xiaofei Wang,
An Applied Approach Using SAS & STATA
and Herbert Pang
Iftekhar Khan
Causal Analysis in Biomedicine and
Design and Analysis of Clinical Trials
Epidemiology: Based on Minimal
for Predictive Medicine
Sufficient Causation
Shigeyuki Matsui, Marc Buyse,
Mikel Aickin
and Richard Simon
Clinical and Statistical Considerations in
Design and Analysis of Clinical Trials with
Personalized Medicine
Time-to-Event Endpoints
Claudio Carini, Sandeep Menon, and Mark Chang
Karl E. Peace
Clinical Trial Data Analysis Using R
Design and Analysis of Non-Inferiority Trials
Ding-Geng (Din) Chen and Karl E. Peace
Mark D. Rothmann, Brian L. Wiens,
Clinical Trial Data Analysis Using R and SAS, and Ivan S. F. Chan
Second Edition
Difference Equations with Public Health
Ding-Geng (Din) Chen, Karl E. Peace,
Applications
and Pinggao Zhang
Lemuel A. Moyé and Asha Seth Kapadia
Clinical Trial Methodology
Karl E. Peace and Ding-Geng (Din) Chen
Published Titles
DNA Methylation Microarrays: Interval-Censored Time-to-Event Data:
Experimental Design and Statistical Methods and Applications
Analysis Ding-Geng (Din) Chen, Jianguo Sun,
Sun-Chong Wang and Arturas Petronis and Karl E. Peace
DNA Microarrays and Related Genomics Introductory Adaptive Trial Designs:
Techniques: Design, Analysis, and A Practical Guide with R
Interpretation of Experiments Mark Chang
David B. Allison, Grier P. Page, Joint Models for Longitudinal and Time-
T. Mark Beasley, and Jode W. Edwards to-Event Data: With Applications in R
Dose Finding by the Continual Dimitris Rizopoulos
Reassessment Method Measures of Interobserver Agreement
Ying Kuen Cheung and Reliability, Second Edition
Dynamical Biostatistical Models Mohamed M. Shoukri
Daniel Commenges and Medical Biostatistics, Third Edition
Hélène Jacqmin-Gadda A. Indrayan
Elementary Bayesian Biostatistics Meta-Analysis in Medicine and
Lemuel A. Moyé Health Policy
Emerging Non-Clinical Biostatistics in Dalene Stangl and Donald A. Berry
Biopharmaceutical Development and Methods in Comparative Effectiveness
Manufacturing Research
Harry Yang Constantine Gatsonis and Sally C. Morton
Empirical Likelihood Method in Mixed Effects Models for the Population
Survival Analysis Approach: Models, Tasks, Methods
Mai Zhou and Tools
Essentials of a Successful Biostatistical Marc Lavielle
Collaboration Modeling to Inform Infectious Disease
Arul Earnest Control
Exposure–Response Modeling: Methods Niels G. Becker
and Practical Implementation Modern Adaptive Randomized Clinical
Jixian Wang Trials: Statistical and Practical Aspects
Frailty Models in Survival Analysis Oleksandr Sverdlov
Andreas Wienke Monte Carlo Simulation for the
Fundamental Concepts for New Clinical Pharmaceutical Industry: Concepts,
Trialists Algorithms, and Case Studies
Scott Evans and Naitee Ting Mark Chang
Generalized Linear Models: A Bayesian Multiregional Clinical Trials for
Perspective Simultaneous Global New Drug
Dipak K. Dey, Sujit K. Ghosh, and Development
Bani K. Mallick Joshua Chen and Hui Quan
Handbook of Regression and Modeling: Multiple Testing Problems in
Applications for the Clinical and Pharmaceutical Statistics
Pharmaceutical Industries Alex Dmitrienko, Ajit C. Tamhane,
Daryl S. Paulson and Frank Bretz
Inference Principles for Biostatisticians
Ian C. Marschner
Published Titles
Noninferiority Testing in Clinical Trials: Statistical Design and Analysis of Clinical

Issues and Challenges Trials: Principles and Methods
Tie-Hua Ng Weichung Joe Shih and Joseph Aisner
Optimal Design for Nonlinear Response Statistical Design and Analysis of
Models Stability Studies
Valerii V. Fedorov and Sergei L. Leonov Shein-Chung Chow
Patient-Reported Outcomes: Statistical Evaluation of Diagnostic
Measurement, Implementation and Performance: Topics in ROC Analysis
Interpretation Kelly H. Zou, Aiyi Liu, Andriy Bandos,
Joseph C. Cappelleri, Kelly H. Zou, Lucila Ohno-Machado, and Howard Rockette
Andrew G. Bushmakin, Jose Ma. J. Alvir, Statistical Methods for Clinical Trials
Demissie Alemayehu, and Tara Symonds Mark X. Norleans
Quantitative Evaluation of Safety in Drug Statistical Methods for Drug Safety
Development: Design, Analysis and Robert D. Gibbons and Anup K. Amatya
Reporting
Statistical Methods for Healthcare
Qi Jiang and H. Amy Xia
Performance Monitoring
Quantitative Methods for Traditional Alex Bottle and Paul Aylin
Chinese Medicine Development
Statistical Methods for Immunogenicity
Shein-Chung Chow
Assessment
Randomized Clinical Trials of Harry Yang, Jianchun Zhang, Binbing Yu,
Nonpharmacological Treatments and Wei Zhao
Isabelle Boutron, Philippe Ravaud,
Statistical Methods in Drug Combination
and David Moher
Studies
Randomized Phase II Cancer Wei Zhao and Harry Yang
Clinical Trials
Statistical Testing Strategies in the
Sin-Ho Jung
Health Sciences
Repeated Measures Design with Albert Vexler, Alan D. Hutson,
Generalized Linear Mixed Models for and Xiwei Chen
Randomized Controlled Trials
Statistics in Drug Research:
Toshiro Tango
Methodologies and Recent
Sample Size Calculations for Clustered Developments
and Longitudinal Outcomes in Clinical Shein-Chung Chow and Jun Shao
Research
Statistics in the Pharmaceutical Industry,
Chul Ahn, Moonseong Heo,
Third Edition
and Song Zhang
Ralph Buncher and Jia-Yeong Tsay
Sample Size Calculations in Clinical
Survival Analysis in Medicine and
Research, Second Edition
Genetics
Shein-Chung Chow, Jun Shao,
Jialiang Li and Shuangge Ma
and Hansheng Wang
Theory of Drug Development
Statistical Analysis of Human Growth
Eric B. Holmgren
and Development
Yin Bun Cheung Translational Medicine: Strategies and
Statistical Methods
Dennis Cosmatos and Shein-Chung Chow
Clinical Trial
Data Analysis
Using R and SAS
Second Edition
Ding-Geng (Din) Chen

Karl E. Peace
Pinggao Zhang
Harry Yang
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742
© 2017 by Taylor & Francis Group, LLC

CRC Press is an imprint of Taylor & Francis Group, an Informa business
No claim to original U.S. Government works
Printed on acid-free paper

Version Date: 20170407
International Standard Book Number-13: 978-1-4987-7952-4 (Hardback)
This book contains information obtained from authentic and highly regarded sources. Reasonable
efforts have been made to publish reliable data and information, but the author and publisher cannot
assume responsibility for the validity of all materials or the consequences of their use. The authors and
publishers have attempted to trace the copyright holders of all material reproduced in this publication
and apologize to copyright holders if permission to publish in this form has not been obtained. If any
copyright material has not been acknowledged please write and let us know so we may rectify in any
future reprint.
Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced,
transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or
hereafter invented, including photocopying, microfilming, and recording, or in any information
storage or retrieval system, without written permission from the publishers.
For permission to photocopy or use material electronically from this work, please access
www.copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc.
(CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization
that provides licenses and registration for a variety of users. For organizations that have been granted
a photocopy license by the CCC, a separate system of payment has been arranged.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and
are used only for identification and explanation without intent to infringe.
Library of Congress Cataloging-in-Publication Data
Names: Chen, Ding-Geng. | Peace, Karl E., 1941- | Zhang, Pinggao. | Chen,
Ding-Geng. Clinical trial data analysis using R.
Title: Clinical trial data analysis using R and SAS.
Other titles: Clinical trial data analysis using R
Description: Second edition / Ding-Geng Chen, Karl E. Peace, Pinggao Zhang. |
Boca Raton : CRC Press, 2017. | “Major updates to include SAS
programs”--Preface. | Previous edition: Clinical trial data analysis using
R / Ding-Geng Chen, Karl E. Peace (Boca Raton, Florida : CRC Press, 2011).
Identifiers: LCCN 2016057399 | ISBN 9781498779524 (hardback)
Subjects: LCSH: Clinical trials--Statistical methods. | R (Computer program
language) | SAS (Computer program language)
Classification: LCC R853.C55 C43 2017 | DDC 610.72/7--dc23
LC record available at https://lccn.loc.gov/2016057399
Visit the Taylor & Francis Web site at

http://www.taylorandfrancis.com
and the CRC Press Web site at
http://www.crcpress.com
Dedication
To my parents and parents-in-law who value high education and hard work,
and to my wife, Ke, my son, John D. Chen, and my daughter, Jenny K. Chen,
for their love and support.
Ding-Geng(Din) Chen
To the memory of my late mother, Elsie Mae Cloud Peace, my late wife, Jiann-
Ping Hsu (JP), and to my son, Christopher K. Peace, daughter-in-law, Ashley
Hopkins Peace, granddaughter, Camden, and grandson, Henry. Memories of
my mom and wife JP continue to sustain me. My son, Chris, and his family
provide much joy and esteem.
Karl E. Peace
To my wife, Jenny Z. Yang, and my sons, Andrew Y. Zhang and Anthony Y.

Zhang, for their love and support. I am particularly thankful to my wife for
her patience and understanding while I worked on the book.
Pinggao Zhang
ix
Contents
List of Figures xix
List of Tables xxi
Preface for the Second Edition xxiii
Preface for the First Edition xxvii
About the Authors xxxi
1 Introduction to R 1
1.1 What is R? . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Steps on Installing R and Updating R Packages . . . . . . . . 2
1.2.1 First Step: Install R Base System . . . . . . . . . . . . 3
1.2.2 Second Step: Installing and Updating R Packages . . . 3
1.2.3 Steps to Get Help and Documentation . . . . . . . . . 4
1.3 R for Clinical Trials . . . . . . . . . . . . . . . . . . . . . . . 5
1.4 A Simple Simulated Clinical Trial . . . . . . . . . . . . . . . 7
1.4.1 Data Simulation . . . . . . . . . . . . . . . . . . . . . 7
1.4.1.1 R Functions . . . . . . . . . . . . . . . . . . . 7
1.4.1.2 Data Generation and Manipulation . . . . . 8
1.4.1.3 Basic R Graphics . . . . . . . . . . . . . . . . 10
1.4.2 Data Analysis . . . . . . . . . . . . . . . . . . . . . . . 13
1.5 Summary and Recommendations for Further Reading . . . . 15
1.6 Appendix: SAS Programs . . . . . . . . . . . . . . . . . . . . 15
2 Overview of Clinical Trials 17

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.2 Phases of Clinical Trials and Objectives . . . . . . . . . . . . 17
2.2.1 Phase 0 Trials . . . . . . . . . . . . . . . . . . . . . . 17
2.2.2 Phase I Trials . . . . . . . . . . . . . . . . . . . . . . . 18
2.2.3 Phase II Trials . . . . . . . . . . . . . . . . . . . . . . 18
2.2.4 Phase III Trials . . . . . . . . . . . . . . . . . . . . . . 19
2.2.5 Phase IV Trials . . . . . . . . . . . . . . . . . . . . . . 19
2.3 The Clinical Development Plan . . . . . . . . . . . . . . . . . 20
2.4 Biostatistical Aspects of a Protocol . . . . . . . . . . . . . . 20
2.4.1 Background or Rationale . . . . . . . . . . . . . . . . 21
xi
xii Contents
2.4.2 Objective . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.4.3 Plan of Study . . . . . . . . . . . . . . . . . . . . . . . 22
2.4.3.1 Study Population . . . . . . . . . . . . . . . 22
2.4.3.2 Study Design . . . . . . . . . . . . . . . . . . 22
2.4.3.3 Problem Management . . . . . . . . . . . . . 24
2.4.4 Statistical Analysis Section . . . . . . . . . . . . . . . 24
2.4.4.1 Study Objectives as Statistical Hypotheses . 24
2.4.4.2 Endpoints . . . . . . . . . . . . . . . . . . . 25
2.4.4.3 Statistical Methods . . . . . . . . . . . . . . 25
2.4.4.4 Statistical Monitoring Procedures . . . . . . 26
2.4.5 Statistical Design Considerations . . . . . . . . . . . . 27
2.4.6 Subset Analyses . . . . . . . . . . . . . . . . . . . . . 28
2.5 Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . 29
3 Treatment Comparisons in Clinical Trials 31

3.1 Data from Clinical Trials . . . . . . . . . . . . . . . . . . . . 31
3.1.1 Diastolic Blood Pressure . . . . . . . . . . . . . . . . . 31
3.1.2 Clinical Trial on Duodenal Ulcer Healing . . . . . . . 33
3.2 Statistical Models for Treatment Comparisons . . . . . . . . 34
3.2.1 Models for Continuous Endpoints . . . . . . . . . . . . 34
3.2.1.1 Student’s t-Tests . . . . . . . . . . . . . . . . 34
3.2.1.2 One-Way Analysis of Variance (ANOVA) . . 36
3.2.1.3 Multi-Way ANOVA: Factorial Design . . . . 37
3.2.1.4 Multivariate Analysis of Variance (MANOVA)
. . . . . . . . . . . . . . . . . . . . . . . . . 38
3.2.2 Models for Categorical Endpoints: Pearson’s χ2 -test . 38
3.3 Data Analysis in R . . . . . . . . . . . . . . . . . . . . . . . 39
3.3.1 Analysis of the DBP Trial . . . . . . . . . . . . . . . . 39
3.3.1.1 Preliminary Data analysis . . . . . . . . . . . 39
3.3.1.2 t-test . . . . . . . . . . . . . . . . . . . . . . 40
3.3.1.3 Bootstrapping Method . . . . . . . . . . . . 43
3.3.1.4 One-Way ANOVA for Time Changes . . . . 44
3.3.1.5 Two-Way ANOVA for Interaction . . . . . . 47
3.3.1.6 MANOVA for Treatment Difference . . . . . 51
3.3.2 Analysis of Duodenal Ulcer Healing Trial . . . . . . . 52
3.3.2.1 Using Pearson’s χ2 -test . . . . . . . . . . . . 52
3.3.2.2 Using Contingency Table . . . . . . . . . . . 54
3.4 Summary and Conclusions . . . . . . . . . . . . . . . . . . . 56
4 Treatment Comparisons in Clinical Trials with Covariates 63

4.1.1 Diastolic Blood Pressure . . . . . . . . . . . . . . . . . 63
4.1.2 Clinical Trials for Betablockers . . . . . . . . . . . . . 64
4.1.3 Clinical Trial on Familial Adenomatous Polyposis . . . 64
Contents xiii
4.2 Statistical Models Incorporating Covariates . . . . . . . . . . 65

4.2.1 ANCOVA Models for Continuous Endpoints . . . . . . 65
4.2.2 Logistic Regression for Binary/Binomial Endpoints . . 68
4.2.3 Poisson Regression for Clinical Endpoint with Counts 70
4.2.4 Overdispersion . . . . . . . . . . . . . . . . . . . . . . 70
4.3.1 Analysis of DBP Trial . . . . . . . . . . . . . . . . . . 73
4.3.1.1 Analysis of Baseline Data . . . . . . . . . . . 73
4.3.1.2 ANCOVA of DBP Change from Baseline . . 76
4.3.1.3 MANCOVA for DBP Change from Baseline . 79
4.3.2 Analysis of Betablocker Trial . . . . . . . . . . . . . . 80
4.3.3 Analysis of Data from Familial Adenomatous Polyposis
Trial . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
5 Analysis of Clinical Trials with Time-to-Event Endpoints 95

5.1 Clinical Trials with Time-to-Event Data . . . . . . . . . . . . 96
5.1.1 Phase II Trial of Patients with Stage-2 Breast
Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . 96
5.1.2 Breast Cancer Trial with Interval-Censored Data . . . 97
5.2 Statistical Models . . . . . . . . . . . . . . . . . . . . . . . . 98
5.2.1 Primary Functions and Definitions . . . . . . . . . . . 98
5.2.1.1 The Hazard Function . . . . . . . . . . . . . 98
5.2.1.2 The Survival Function . . . . . . . . . . . . . 99
5.2.1.3 The Death Density Function . . . . . . . . . 99
5.2.1.4 Relationships between These Functions . . . 100
5.2.2 Parametric Models . . . . . . . . . . . . . . . . . . . . 100
5.2.2.1 The Exponential Model . . . . . . . . . . . . 100
5.2.2.2 The Weibull Model . . . . . . . . . . . . . . 100
5.2.2.3 The Rayleigh Model . . . . . . . . . . . . . . 101
5.2.2.4 The Gompertz Model . . . . . . . . . . . . . 101
5.2.2.5 The Lognormal Model . . . . . . . . . . . . . 101
5.3 Statistical Methods for Right-Censored Data . . . . . . . . . 101
5.3.1 Nonparametric Models: Kaplan-Meier Estimator . . . 101
5.3.2 Cox Proportion Hazards Regression . . . . . . . . . . 102
5.4 Statistical Methods for Interval-Censored Data . . . . . . . . 104
5.4.1 Turnbull’s Nonparametric Estimator . . . . . . . . . . 105
5.4.2 Parametric Likelihood Estimation with Covariates . . 106
5.4.3 Semiparametric Estimation: the IntCox . . . . . . . . 107
5.5 Step-by-Step Implementations in R . . . . . . . . . . . . . . 108
5.5.1 Stage-2 Breast Carcinoma . . . . . . . . . . . . . . . . 108
5.5.1.1 Fit Kaplan-Meier . . . . . . . . . . . . . . . 108
5.5.1.2 Fit Weibull Parametric Model . . . . . . . . 112
5.5.1.3 Fit Cox Regression Model . . . . . . . . . . . 114
xiv Contents
5.5.2 Breast Cancer with Interval-Censored Data . . . . . . 116

5.5.2.1 Fit Turnbull’s Nonparametric Estimator . . . 116
5.5.2.2 Fit Turnbull’s Nonparametric Estimator
Using R Package interval . . . . . . . . . . 122
5.5.2.3 Fitting Parametric Models . . . . . . . . . . 123
5.5.2.4 Testing Treatment Effect Using
Semiparametric Estimation: IntCox . . . . . 125
5.5.2.5 Testing Treatment Effect Using
Semiparametric Estimation: ictest . . . . . 128
5.6 Summary and Discussions . . . . . . . . . . . . . . . . . . . 129
6 Longitudinal Data Analysis for Clinical Trials 133

6.1 Clinical Trials . . . . . . . . . . . . . . . . . . . . . . . . . . 133
6.1.1 Diastolic Blood Pressure Data . . . . . . . . . . . . . 133
6.1.2 Clinical Trial on Duodenal Ulcer Healing . . . . . . . 134
6.2 Statistical Models . . . . . . . . . . . . . . . . . . . . . . . . 135
6.2.1 Linear Mixed Models . . . . . . . . . . . . . . . . . . . 135
6.2.2 Generalized Linear Mixed Models . . . . . . . . . . . . 137
6.2.3 Generalized Estimating Equation . . . . . . . . . . . . 138
6.3 Longitudinal Data Analysis for Clinical Trials . . . . . . . . 138
6.3.1 Analysis of Diastolic Blood Pressure Data . . . . . . . 138
6.3.1.1 Data Graphics and Response Feature Analysis 139
6.3.1.2 Longitudinal Modeling . . . . . . . . . . . . 146
6.3.2 Analysis of Cimetidine Duodenal Ulcer Trial . . . . . 152
6.3.2.1 Preliminary Analysis . . . . . . . . . . . . . 152
6.3.2.2 Fit Logistic Regression to Binomial Data . . 152
6.3.2.3 Fit Generalized Linear Mixed Model . . . . . 155
6.3.2.4 Fit GEE . . . . . . . . . . . . . . . . . . . . 157
6.4 Summary and Discussion . . . . . . . . . . . . . . . . . . . . 159
7 Sample Size Determination and Power Calculations in

Clinical Trials 165
7.1 Pre-requisites for Sample Size Determination . . . . . . . . . 165
7.2 Comparison of Two Treatment Groups with Continuous
Endpoints . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
7.2.1 Fundamentals . . . . . . . . . . . . . . . . . . . . . . . 167
7.2.2 Basic Formula for Sample Size Calculation . . . . . . 169
7.2.3 R Function power.t.test . . . . . . . . . . . . . . . . 170
7.2.4 Unequal Variance: samplesize Package . . . . . . . . 173
7.3 Two Binomial Proportions . . . . . . . . . . . . . . . . . . . 176
7.3.1 R Function power.prop.test . . . . . . . . . . . . . . 176
7.3.2 R Library: pwr . . . . . . . . . . . . . . . . . . . . . . 179
7.3.3 R Function nBinomial in gsDesign library . . . . . . 182
Contents xv
7.4 Time-to-Event Endpoint . . . . . . . . . . . . . . . . . . . . 186

7.5 Design of Group Sequential Trials . . . . . . . . . . . . . . . 190
7.5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . 190
7.5.2 gsDesign . . . . . . . . . . . . . . . . . . . . . . . . . 191
7.6 Longitudinal Trials . . . . . . . . . . . . . . . . . . . . . . . 197
7.6.1 Longitudinal Trial with Continuous Endpoint . . . . . 197
7.6.1.1 The Model Setting . . . . . . . . . . . . . . . 197
7.6.1.2 Sample Size Calculations . . . . . . . . . . . 198
7.6.1.3 Power Calculation . . . . . . . . . . . . . . . 199
7.6.1.4 Example and R Illustration . . . . . . . . . . 199
7.6.2 Longitudinal Binary Endpoint . . . . . . . . . . . . . 203
7.6.2.1 Approximate Sample Size Calculation . . . . 203
7.6.2.2 Example and R Implementation . . . . . . . 204
7.7 Relative Changes and Coefficient of Variation . . . . . . . . 205
7.7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . 205
7.7.2 Sample Size Calculation Formula . . . . . . . . . . . . 205
7.7.3 Example and R Implementation . . . . . . . . . . . . . 206
8 Meta-Analysis of Clinical Trials 215

8.1.1 Clinical Trials for Betablockers: Binary Data . . . . . 216
8.1.2 Data for Cochrane Collaboration Logo: Binary Data . 216
8.1.3 Clinical Trials on Amlodipine: Continuous Data . . . . 217
8.2 Statistical Models for Meta-Analysis . . . . . . . . . . . . . . 217
8.2.1 Clinical Hypotheses and Effect Size . . . . . . . . . . . 218
8.2.2 Fixed-Effects Meta-Analysis Model: The
Weighted Average . . . . . . . . . . . . . . . . . . . . 219
8.2.3 Random-Effects Meta-Analysis Model:
DerSimonian-Laird . . . . . . . . . . . . . . . . . . . . 221
8.2.4 Publication Bias . . . . . . . . . . . . . . . . . . . . . 223
8.3.1 Analysis of Betablocker Trials . . . . . . . . . . . . . . 223
8.3.1.1 Fitting the Fixed-Effects Model . . . . . . . 224
8.3.1.2 Fitting the Random-Effects Model . . . . . . 227
8.3.2 Meta-Analysis for Cochrane Collaboration Logo . . . 230
8.3.3 Analysis of Amlodipine Trial Data . . . . . . . . . . . 232
8.3.3.1 Load the Library and Data . . . . . . . . . . 232
8.3.3.2 Fit the Fixed-Effects Model . . . . . . . . . . 232
8.3.3.3 Fit the Random-Effects Model . . . . . . . . 235
xvi Contents
9 Bayesian Methods in Clinical Trials 241

9.1 Bayesian Models . . . . . . . . . . . . . . . . . . . . . . . . . 241
9.1.1 Bayes’ Theorem . . . . . . . . . . . . . . . . . . . . . 241
9.1.2 Posterior Distributions for Some Standard Distributions 243
9.1.2.1 Normal Distribution with Known Variance . 243
9.1.2.2 Normal Distribution with Unknown Variance 244
9.1.2.3 Normal Regression . . . . . . . . . . . . . . . 244
9.1.2.4 Binomial Distribution . . . . . . . . . . . . . 245
9.1.2.5 Multinomial Distribution . . . . . . . . . . . 245
9.1.3 Simulation from the Posterior Distribution . . . . . . 245
9.1.3.1 Direct Simulation . . . . . . . . . . . . . . . 246
9.1.3.2 Importance Sampling . . . . . . . . . . . . . 247
9.1.3.3 Gibbs Sampling . . . . . . . . . . . . . . . . 247
9.1.3.4 Metropolis-Hastings Algorithm . . . . . . . . 248
9.2 R Packages in Bayesian Modeling . . . . . . . . . . . . . . . 249
9.2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . 249
9.2.2 R Packages using WinBUGS . . . . . . . . . . . . . . . 250
9.2.2.1 R2WinBUGS . . . . . . . . . . . . . . . . . . . 250
9.2.2.2 BRugs . . . . . . . . . . . . . . . . . . . . . . 251
9.2.2.3 rbugs . . . . . . . . . . . . . . . . . . . . . . 251
9.2.2.4 Typical Usage . . . . . . . . . . . . . . . . . 251
9.2.3 MCMCpack . . . . . . . . . . . . . . . . . . . . . . . . . 252
9.3 MCMC Simulations . . . . . . . . . . . . . . . . . . . . . . . 253
9.3.1 Normal-Normal Model . . . . . . . . . . . . . . . . . . 253
9.3.2 Beta-Binomial Model . . . . . . . . . . . . . . . . . . 255
9.4 Bayesian Data Analysis . . . . . . . . . . . . . . . . . . . . . 259
9.4.1 Blood Pressure Data: Bayesian Linear Regression . . . 259
9.4.2 Binomial Data: Bayesian Logistic Regression . . . . . 261
9.4.3 Count Data: Bayesian Poisson Regression . . . . . . . 266
9.4.4 Comparing Two Treatments . . . . . . . . . . . . . . . 267
9.5 Summary and Discussion . . . . . . . . . . . . . . . . . . . . 271
10 Bioequivalence Clinical Trials 277

10.1 Data from Bioequivalence Clinical Trials . . . . . . . . . . . 277
10.1.1 Data from Chow and Liu (2009) . . . . . . . . . . . . 277
10.1.2 Bioequivalence Trial on Cimetidine Tablets . . . . . . 277
10.2 Bioequivalence Clinical Trial Endpoints . . . . . . . . . . . . 279
10.3 Statistical Methods to Analyze Bioequivalence . . . . . . . . 281
10.3.1 Decision CIs for Bioequivalence . . . . . . . . . . . . . 282
10.3.2 The Classical Asymmetric Confidence Interval . . . . 283
10.3.3 Westlake’s Symmetric Confidence Interval . . . . . . . 283
10.3.4 Two One-Sided Tests . . . . . . . . . . . . . . . . . . 284
10.3.5 Bayesian Approaches . . . . . . . . . . . . . . . . . . . 284
Contents xvii
10.3.6 Individual-Based Bienayme-Tchebycheff (BT)

Inequality CI . . . . . . . . . . . . . . . . . . . . . . . 285
10.3.7 Individual-Based Bootstrap CIs . . . . . . . . . . . . . 286
10.4 Step-by-Step Implementation in R . . . . . . . . . . . . . . . 286
10.4.1 Analyze the data from Chow and Liu (2009) . . . . . 286
10.4.1.1 Load the data into R . . . . . . . . . . . . . 286
10.4.1.2 Tests for Carryover Effect . . . . . . . . . . . 288
10.4.1.3 Test for Direct Formulation Effect . . . . . . 290
10.4.1.4 Analysis of Variance . . . . . . . . . . . . . . 292
10.4.1.5 Decision CIs . . . . . . . . . . . . . . . . . . 293
10.4.1.6 Classical Shortest 90% CI . . . . . . . . . . . 293
10.4.1.7 The Westlake CI . . . . . . . . . . . . . . . . 294
10.4.1.8 Two One-Sided Tests . . . . . . . . . . . . . 295
10.4.1.9 Bayesian Approach . . . . . . . . . . . . . . 295
10.4.1.10 Individual-Based BT CI . . . . . . . . . . . . 295
10.4.1.11 Bootstrap CIs . . . . . . . . . . . . . . . . . 296
10.4.2 Analyze the data from Cimetidine Trial . . . . . . . . 299
10.4.2.1 Clinical Trial Endpoints Calculations . . . . 299
10.4.2.2 ANOVA: Tests for Carryover and Other
Effects . . . . . . . . . . . . . . . . . . . . . 304
10.4.2.3 Decision CIs . . . . . . . . . . . . . . . . . . 308
10.4.2.4 Classical Shortest 90% CI . . . . . . . . . . . 308
10.4.2.5 The Westlake CI . . . . . . . . . . . . . . . . 309
10.4.2.6 Two One-Sided CIs . . . . . . . . . . . . . . 309
10.4.2.7 Bayesian Approach . . . . . . . . . . . . . . 310
10.4.2.8 Individual-Based BT CI . . . . . . . . . . . . 310
10.4.2.9 Bootstrap CIs . . . . . . . . . . . . . . . . . 311
10.6 Appendix: SAS Program . . . . . . . . . . . . . . . . . . . . 314
11 Adverse Events in Clinical Trials 315

11.1 Adverse Event Data from a Clinical Trial . . . . . . . . . . . 315
11.2 Statistical Methods . . . . . . . . . . . . . . . . . . . . . . . 317
11.2.1 Confidence Interval (CI) Methods . . . . . . . . . . . 318
11.2.1.1 Comparison Using Direct CI Method . . . . 318
11.2.1.2 Comparison Using Indirect CI Methods . . . 318
11.2.2 Significance Level Methods (SLMs) . . . . . . . . . . . 319
11.2.2.1 SLM using normal approximation . . . . . . 319
11.2.2.2 SLM using exact binomial distribution . . . 320
11.2.2.3 SLM using resampling from pooled samples . 320
11.2.2.4 SLM using resampling from pooled AE rates 321
11.3 Step-by-Step Implementation in R . . . . . . . . . . . . . . . 321
11.3.1 Clinical Trial Data Manipulation . . . . . . . . . . . . 321
11.3.2 R Implementations for CI Methods . . . . . . . . . . . 322
11.3.3 R Implementations for Indirect CI Methods . . . . . . 323
xviii Contents
11.3.4 R for Significant Level Methods . . . . . . . . . . . . . 327

11.3.4.1 R for SLM with normal approximation . . . 327
11.3.4.2 R for SLM with exact binomial . . . . . . . . 328
11.3.4.3 R for SLM using Sampling-Resampling . . . 330
11.4 Summary and Discussions . . . . . . . . . . . . . . . . . . . 333
12 Analysis of DNA Microarrays in Clinical Trials 337

12.1 DNA Microarray . . . . . . . . . . . . . . . . . . . . . . . . . 337
12.1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . 337
12.1.2 DNA, RNA, and Genes . . . . . . . . . . . . . . . . . 338
12.1.3 Central Dogma of Molecular Biology . . . . . . . . . . 338
12.1.4 Probes, Probesets, Mismatch, and Perfectmatch . . . 339
12.1.5 Microarray and Statistical Analysis . . . . . . . . . . . 340
12.1.6 Software: R/Bioconductor . . . . . . . . . . . . . . . 340
12.2 Breast Cancer Data . . . . . . . . . . . . . . . . . . . . . . . 340
12.2.1 Data Source . . . . . . . . . . . . . . . . . . . . . . . . 341
12.2.2 Low-Level Data Analysis . . . . . . . . . . . . . . . . 343
12.2.2.1 Introduction . . . . . . . . . . . . . . . . . . 343
12.2.2.2 Library affy . . . . . . . . . . . . . . . . . . 344
12.2.2.3 Quality Control . . . . . . . . . . . . . . . . 347
12.2.2.4 Background, Normalization, and
Summarization . . . . . . . . . . . . . . . . . 349
12.2.3 High-Level Analysis . . . . . . . . . . . . . . . . . . . 352
12.2.3.1 Statistical t-test . . . . . . . . . . . . . . . . 354
12.2.3.2 Model Fitting . . . . . . . . . . . . . . . . . 355
12.2.3.3 Number of Significantly Expressed Genes . . 360
12.2.4 Functional Analysis of Gene Lists . . . . . . . . . . . . 361
Bibliography 363
Index 373
List of Figures
1.1 Distributions for Data Generated for “Placebo” . . . . . . . . 10

1.2 Distributions for Data Generated for “Drug” . . . . . . . . . . 11
1.3 Data with Regression Line for Each Treatment . . . . . . . . 12
1.4 Diagnostics Plot . . . . . . . . . . . . . . . . . . . . . . . . . 14
3.1 Boxplot for Two Treatments . . . . . . . . . . . . . . . . . . . 40

3.2 Bootstrap Distribution of Mean Differences . . . . . . . . . . 44
3.3 Interaction Plots for DBP Trial . . . . . . . . . . . . . . . . . 48
3.4 Ulcer Healing Distribution . . . . . . . . . . . . . . . . . . . . 56
4.1 Boxplot for Baseline DBP . . . . . . . . . . . . . . . . . . . . 73

4.2 Baseline DBP as Function of “Age” and “Sex” . . . . . . . . . 76
4.3 DBP Change as Function of “Age” for each “TRT” . . . . . . 79
5.1 Kaplan-Meier Estimator for Survival Function . . . . . . . . . 109

5.2 Nonparametric Estimation for Cumulative Hazard Function . 111
5.3 Turnbull’s Nonparametric Estimator for Both Treatments . . 119
5.4 Turnbull’s Estimator Overlaid with Midpoint Kaplan-Meier Es-
timator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
5.5 Turnbull’s NPMLE from R Package interval . . . . . . . . . 123
5.6 Estimated Survival Functions from IntCox . . . . . . . . . . . 126
5.7 Bootstrapping Distribution for Treatment Difference . . . . . 128
6.1 DBP as a Function of Time for Each Patient . . . . . . . . . 140

6.2 DBP as a Function of Time Grouped for All Patients . . . . . 141
6.3 Boxplot by Treatments . . . . . . . . . . . . . . . . . . . . . . 142
6.4 Bivariate Plot for Intercept and Slope . . . . . . . . . . . . . 144
6.5 QQ-Plot for Model 3 . . . . . . . . . . . . . . . . . . . . . . . 150
6.6 Boxplot for Residuals for all Subjects by Treatment . . . . . 151
7.1 Graphical Features of Sample Size Determination . . . . . . . 168

7.2 Numerator in Sample Size Calculation . . . . . . . . . . . . . 171
7.3 Nonlinear Relationship between Sample Size to Power . . . . 174
7.4 Nonlinear Relationship between Power and Significance Level 180
7.5 Sample Size Calculations for 80% Power . . . . . . . . . . . . 185
7.6 Default Plot for gsDesign Boundaries . . . . . . . . . . . . . . 195
xix
xx List of Figures
7.7 Spending Function Plot for gsDesign . . . . . . . . . . . . . . 196

7.8 Perspective Surface for Power in Longitudinal Study . . . . . 202
7.9 Perspective Surface for Sample Size . . . . . . . . . . . . . . . 207
8.1 Forest Plot for the Betablockers Trial with 95% CIs from Fixed-
Effects Meta-Analysis . . . . . . . . . . . . . . . . . . . . . . 225
8.2 Funnel Plot for the Betablocker Trial . . . . . . . . . . . . . . 226
8.3 Forest Plot for the Betablocker Trial with 95% CIs from
Random-Effects Meta-Analysis . . . . . . . . . . . . . . . . . 228
8.4 Forestplot with More Options . . . . . . . . . . . . . . . . . 229
8.5 Forestplot for Cochrane Data . . . . . . . . . . . . . . . . . 231
8.6 A Detailed Forest Plot for the Angina Trial with 95% CIs . . 234
8.7 Funnel Plot for the Angina Trial . . . . . . . . . . . . . . . . 234
9.1 Distributions of Prior, Direct Simulation, and MCnormalnor-

mal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
9.2 Distributions of Prior, Direct Simulation, and MCbinomialbeta 259
9.3 MCMC Plots . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
9.4 MCMC Plots for Betablocker . . . . . . . . . . . . . . . . . . 265
9.5 MCMC Plots for Polyps Data . . . . . . . . . . . . . . . . . . 267
9.6 Distributions of Prior and the Four Treatments . . . . . . . . 270
10.1 Bootstrap Distribution for the Mean of the Individual Ratios 297
10.2 Bootstrap Distribution for the Ratio of the Means . . . . . . 298
10.3 QQ-Plot for the Ratio of the Means . . . . . . . . . . . . . . 299
10.4 The Concentration-by-Time for First Sequence and Second Pe-
riod. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
10.5 The Mean Concentration-by-Time for First Sequence and Sec-
ond Period . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
10.6 Bootstrap Distribution for the Mean of the Individual Ratios 312
10.7 Bootstrap Distribution for the Ratio of the Means . . . . . . 313
11.1 Plot of AE Rate for the Three Dose Treatments . . . . . . . . 317

11.2 The Estimated Significance Level α for Each Stage . . . . . . 329
11.3 The Estimated αs for Each Stage Under All Four Methods . 333
12.1 Boxplot for Original Expression Data . . . . . . . . . . . . . . 347

12.2 MA Plot for Two Arrays . . . . . . . . . . . . . . . . . . . . . 349
12.3 Boxplot after RMA . . . . . . . . . . . . . . . . . . . . . . . . 353
12.4 Volcano Plot for Gene Expression Data . . . . . . . . . . . . 359
12.5 Venn Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . 361
List of Tables
1.1 ANOVA Table for Simulated Clinical Trial Data . . . . . . . 14
3.1 Diastolic Blood Pressure Trial Data . . . . . . . . . . . . . . 32

3.2 Duodenal Ulcer Trial Data . . . . . . . . . . . . . . . . . . . 34
4.1 Data for Betablocker Clinical Trial . . . . . . . . . . . . . . . 64

4.2 Number of Polyps in FAP Clinical Trial . . . . . . . . . . . . 65
5.1 Breast Carcinoma Data . . . . . . . . . . . . . . . . . . . . . 96

5.2 Breast Cancer Data . . . . . . . . . . . . . . . . . . . . . . . 97
5.3 Kaplan-Meier Estimator for Survival Function . . . . . . . . . 110
6.1 Ulcer Trial Data . . . . . . . . . . . . . . . . . . . . . . . . . 135
8.1 Data for Cochrane Collaboration Logo . . . . . . . . . . . . . 216

8.2 Angina Trial Data . . . . . . . . . . . . . . . . . . . . . . . . 217
10.1 AUC Data from Chow and Liu (2009) . . . . . . . . . . . . . 278

10.2 AUC, CMAX, and TMAX Data from Cimetidine . . . . . . . 280
11.1 AE Data for Clinical Trial . . . . . . . . . . . . . . . . . . . . 315
12.1 Experimental Design and Data . . . . . . . . . . . . . . . . . 342
xxi
Preface for the Second Edition
Since the publication of the first edition of this book in 2011, we have received
extensive compliments on how well it was structured for use by clinical trial
statisticians and analysts in analyzing their own clinical trial data following the
detailed step-by-step illustrations using R. We have also received suggestions
and comments for further improvement among which is to add SAS to the new
edition. A feature of this second edition is to also illustrate data analyses using
the SAS system. Therefore, in this second edition, we have incorporated all
suggestions and comments from enthusiastic readers and corrected all errors
and typos in addition to including SAS programs for data analysis. The SAS
programs appear in the appendix of each chapter corresponding to the sections
where analyses using R were performed.
Another major update is to change the way data are loaded into R. In the
first edition, we used RODBC to read the dataset from an Excel book (named
as datR4CTDA.xlsx) where all data are stored. Many readers communicated
to us that they had difficulties in using RODBC. Therefore, in this edition, we
saved all the datasets into .csv (comma separated values) files and use the R
command read.csv to read the data into R. Readers can also use read.table
to read the data into R for analysis.
We have updated the chapters. In Chapter 3, we included the clinical
trial data analysis for correlated data using multivariate analysis of variance
(MANOVA) in Section 3.2.1.4 with R implementation of this MANOVA ap-
proach in Section 3.3.1.6. The associated SAS programs are included in an
appendix at the end of the chapter. In Chapter 4, we also included the clinical
trial data analysis for correlated data using multivariate analysis of covariance
(MANCOVA) in Section 4.3.1.3. The associated SAS programs are included
in an appendix at the end of the chapter.
In Chapter 5, the IntCox package is no longer supported, but can be ob-
tained from https://cran.r-project.org/web/packages/intcox/index.
html. So we kept Section 5.4.3 for the description of this method as well as the
R implementation in Section 5.5.2.4. However, we updated the analysis using
another R package of ictest to test treatment effect using semiparametric
estimation in Section 5.5.2.5. In addition, we updated the analysis using yet
another R package for interval-censored data (i.e., interval) to fit Turnbull’s
nonparametric estimator in Section 5.5.2.2. The SAS programs for all the
analyses are included in an appendix at the end of this chapter.
In Chapter 6, we updated the analysis using lmerTest. In analysis of
xxiii
xxiv Preface for the Second Edition
longitudinal data using mixed-effects modeling, typically two R packages of

nlme and lme4 are used with more updates from lme4 . However, this pack-
age does not list the p-values for their fixed-effects estimates as discussed by
the creator, Professor Bates in https://stat.ethz.ch/pipermail/r-help/
2006-May/094765.html. Dr. Alexandra Kuznetsova (alku@dtu.dk) expanded
lme4 to lmerTest with F -tests of types I-III hypotheses for the fixed-effects,
likelihood-ratio tests for the random-effects, least squares means (population
means), and differences of least squares means for the fixed effects factors
with corresponding plots. In this edition, we updated the analyses with this
package to illustrate longitudinal data analysis where all parameter estimates
are the same for both editions and p-values are given in this second edition.
We updated Chapter 7 to include power analysis using SAS. The SAS
procedure proc power is a very powerful and commonly used procedure for
statistical power analysis and sample size determination. In this chapter, we
updated all the power calculations from R in the first edition this book to
include SAS programs in this second edition. In Chapter 8 for meta-analysis,
we programmed the meta-analysis using SAS proc iml following the theory
of meta-analysis since there is no existing SAS procedure for this purpose.
We used the example in Section 8.3.3 for illustration purposes. Based on our
experience, we recommend interested readers use R for their meta-analysis due
to its extensive functionalities and ease of use of all the R packages designed
for meta-analysis. We also recommend our book (Chen and Peace (2013)) for
this purpose.
In Chapter 9 for Bayesian analysis, we make use of the SAS pro-
cedure proc MCMC which is commonly used in SAS for Bayesian model-
ing. We also illustrated the proc genmod with the option bayes to im-
plement Bayes modeling corresponding to the data analysis in R in this
chapter. Bioequivalence clinical trials have been commonly analyzed in SAS
and there are many SAS programs online to be used. Therefore, we do
not duplicate this effort in Chapter 10. Instead, we refer the reader to
the following online link: http://onbiostatistics.blogspot.com/2012/
04/cookbook-sas-codes-for-bioequivalence.html from Dr. Deng in his
“Cookbook SAS Codes for Bioequivalence Test in 2 × 2 × 2 Crossover Design”,
which is for the bioequivalence trials used in this chapter.
For analysis of adverse events in clinical trials in Chapter 11, there is
no SAS procedure specifically designed for this analysis. We thus made use
of SAS procedures of proc iml and proc model and programmed step-by-
step for the examples in the chapter for illustration. In Chapter 12 for
analysis of DNA microarray, we still highly recommend using R Biocon-
ductor from http://www.bioconductor.org described in this chapter to
analyze DNA microarray data. For readers who really like to use SAS,
there is an experimental procedure HPMIXED in SAS for this purpose as
seen in https://support.sas.com/documentation/cdl/en/statug/63033/
HTML/default/viewer.htm#hpmixed_toc.htm.
With these updates, the book is more suitable as a text for a course in
Preface for the Second Edition xxv
clinical trial data analysis at the graduate level (Master’s or Doctorate’s)

using R and SAS. In addition, the book should be a valuable reference for self-
study and a learning tool for clinical trial practitioners and biostatisticians
in public health, medical research universities, governmental agencies, and
the pharmaceutical industry, particularly those with little or no experience in
using R and SAS.
Readers may use the computer programs and datasets and modify the R
and SAS programs for their own applications. To facilitate the understanding
of implementation in R and SAS, we annotated all the R and SAS programs
with comments and explanations so that readers can easily understand the
meaning of the corresponding R and SAS programs.
We would like to express our gratitude to many individuals. First, thanks
to David Grubbs from Taylor & Francis for his encouragement in producing
this second edition. Thanks also go to our research assistant Lin Nie who
helped greatly with the SAS programs. Special thanks are due to professors
Robert Gentleman and Ross Ihaka who created the R language with visionary
open source, as well as to the developers and contributing authors in the R
community for their endless efforts and contributed packages.
As a special note to interested readers, all the datasets, R, and SAS pro-
grams in this book can be obtained from the first author (Professor Chen) by
email at: DrDG.Chen@gmail.com.
Ding-Geng(Din) Chen, Ph.D.

University of North Carolina, Chapel Hill, NC, USA
University of Pretoria, Pretoria, South Africa
Karl E. Peace, Ph.D.

Jiann-Ping Hsu College of Public Health
Georgia Southern University, Statesboro, GA, USA
Pinggao Zhang, Ph.D.

Shire Pharmaceuticals, Boston, MA, USA
Preface for the First Edition
With the exception of two chapters, our first book: “Clinical Trial Methodol-
ogy” (Peace and Chen (2010)) contained no statistical analysis software code
for the analysis results presented therein. In this book we provide a thor-
ough presentation of biostatistical analyses of clinical trial data with detailed
step-by-step illustrations on their implementation using R. In each chapter, ex-
amples of clinical trials based on the authors’ actual experience in many areas
of clinical drug development are presented. After understanding the applica-
tion, various biostatistical methods appropriate for analyzing data from the
clinical trials are identified. Then analysis code is developed using appropri-
ate R packages and functions to analyze the data. Analysis code development
and results are presented in a stepwise fashion. This stepwise approach should
enable readers to follow the logic and gain an understanding of the analysis
methods and the R implementation so that they may use R to analyze their
own clinical trial data.
Based on their experience in biostatistical research and working in clinical
development, the authors understand that there are gaps between developed
statistical methods and applications of statistical methods by students and
practitioners. This book is intended to fill this gap by illustrating the im-
plementation of statistical methods using R applied to real clinical trial data
following a step-by-step presentation style.
With this style, the book is suitable as a text for a course in clinical trial
data analysis at the graduate level (Master’s or Doctorate’s), particularly for
students seeking degrees in statistics or biostatistics. In addition, the book
should be a valuable reference for self-study and a learning tool for clinical
trial practitioners and biostatisticians in public health, medical research uni-
versities, governmental agencies and the pharmaceutical industry, particularly
those with little or no experience in using R.
R has become widely used in statistical modeling and computing since its
creation in the mid 1990s and it is now an integrated and essential software
for statistical analyses. Becoming familiar with R is then an imperative for the
next generation of statistical data analysts. In Chapter 1, we present a basic
introduction to the R system, where to get R, how to install R and how to
upgrade R packages. Readers who are already familiar with R may skip this
chapter and go directly to any of the remaining chapters.
In Chapter 2, we provide an overview of the phases and objectives of
clinical trials as well as biostatistical aspects of clinical trials.
xxvii
xxviii Preface for the First Edition
In Chapter 3, we consider basic treatment comparisons in clinical trials

using the R system. Datasets from two clinical efficacy trials are introduced;
“Diastolic Blood Pressure” data from a hypertension trial and “Duodenal Ulcer
Healing” data from a large duodenal ulcer trial. Statistical methods, such as
Student’s t-test, analysis of variance (ANOVA), bootstrapping, Pearson’s χ2
and other contingency table methods, appropriate for the type of data are
identified followed by step-by-step statistical analyses using R.
In Chapter 4, we present data analysis methods for treatment comparisons
in clinical trials adjusting for covariates using R. In addition to the “Diastolic
Blood Pressure” dataset in Chapter 3, datasets from two additional clinical
trials are introduced; one from a large trial of “Beta-Blockers” and one from
a trial of “Familial Andenomatous Polyposis”. We present statistical methods
such as analysis of covariance (ANCOVA), logistic regression for binomial data
and Poisson regression for count data appropriate for analyzing these types of
data using R. In the application of logistic regression, we emphasize diagnostics
for detecting overdispersion for correct modeling and present several remedies
whenever overdispersion is pinpointed.
Analysis methods using R for data from clinical trials with time-to-event
endpoints are presented in Chapter 5. In this chapter, we use data from a
Phase II trial of patients with Stage-2 breast carcinoma as an example of
right-censored time-to-event data. In addition, data from a publically avail-
able breast cancer trial is used as an example of interval-censored time-to-
event data. We then present the associated statistical models for analyzing
these data with appropriate R packages. Methods for right-censored data are
typically well-known; e.g. the nonparametric Kaplan-Meier estimator, semi-
parametric Cox regression and full parametric models such as the exponen-
tial, Weibull or other distributions. However, methods for analyzing interval-
censored data are less-known and sometimes not available in statistical soft-
ware packages. In this chapter, we include some up-to-date statistical methods
as well as the R packages for analyzing this type of data.
Analysis methods appropriate for data from longitudinal clinical trials are
presented in Chapter 6. Two datasets are used as examples in this chapter.
The first dataset is the “Diastolic Blood Pressure” used in Chapters 3 and
4 and reflects continuous data. The second dataset is from the clinical trial
of Duodenal Ulcer Healing of which a subset was analyzed in Chapter 3,
and reflects categorical data. These datasets are analyzed using longitudinal
statistical methods, such as linear mixed models, generalized linear mixed
models and generalized estimating equations and implemented in R packages.
Chapter 7 discusses sample size determination and power analysis in clini-
cal trials. We present an extensive list of methods as well as the R packages to
calculate sample size required under different data types and protocol design.
Meta-analysis of data from clinical trials is presented in Chapter 8. In
this chapter, we present both fixed-effects and random-effects models used in
meta-analysis with several R packages to implement these models.
Since Bayesian methods are being increasingly used in the design and
Another random document with
no related content on Scribd:
from one to the other, wondering at their tears, but
untouched by any sense of their sorrow. How vividly the
scene rose before him again! His mother's face, the well-
worn, shabby furniture, the very atmosphere of the old
home seemed about him for the moment.
Yet how long ago it was! If the little sister had lived, she
might now have been an anxious-looking mother herself,
with grown-up children. And Frank—baby Frank—what had
become of him? Dead, probably,—yes, surely he was dead,
and better dead. But Michael heaved a sigh as he thought of
his brother. He had not been so moved for years. Certainly
the visit of that little maiden had exercised a softening
influence upon him. How long it was since he had seen his
brother!
His life altogether seemed long, as he looked back on it.

Very, very old, the little girl had called him. She was wrong
there; he was not so very old. But he was getting old. He
could not deny that, and the thought caused him a throb of
pain. For the first time, he felt his life to be narrow and
contracted and unsatisfactory. He rose and walked up and
down the limited space in which it was possible to move
between the piled-up books. He drew aside the curtain
which hung over the little door, and looked up at the stars
shining brightly above the tall houses opposite. And he
sighed again. What was making him feel so unlike himself
to-night?
He half hoped that on the following day little Margery might

again make her appearance in his shop. Would she forget
her little debt to him? He hoped not. He did not care about
the fourpence so much; but he did want to see the pretty
little creature again. But throughout the day he looked for
her in vain. Nor did she appear on the next, nor the next.
He was conscious of disappointment. On the third day,
however, he had news which concerned her.
A well-known customer entered the shop. But though he

knew the man well as a customer, Michael knew very little
of him. He was not interested in the lives of his customers,
except where they touched his own. He knew this one to be
a minister of some kind, and from the alacrity with which he
dropped the theological books he desired, if their price were
at all high, the bookseller imagined that money was not
plentiful with him. The fact that this preacher often asked
for works by Mr. Spurgeon proved nothing, since Michael's
experience had taught him that the writings of that great
preacher have a fascination for every species of religious
teacher who tries to open his lips in public for the edification
of his brother man.
"Good day, sir," said Michael, as his customer entered the

shop. "What can I do for you to-day?"
"Good day, Mr. Betts. I really do not know that I want

anything in particular, only, as I was passing close by, I
could not resist the temptation to look in, just to see what
you have. Your wares are always tempting to me."
"I'm glad to hear it, sir. Look round, and welcome. Take
your time over it. There's a new lot of books here that I've
purchased lately. Maybe you'd fancy some of them."
"Thanks, thanks," said the other, turning eagerly towards

the books. He stood for some minutes examining them in
silence. Suddenly he said, "Ah! Here is a book, I see, by
Professor Lavers. It is very sad about him, is it not?"
"What about him, sir?" asked Michael, in surprise.

"What? You have not heard? How strange! And he, a
neighbour of yours!"
"He lives in Gower Street, certainly, if that makes him a

neighbour of mine," said Betts grimly; "but what is the
matter with him, sir?"
"He is very ill indeed; the doctors think he cannot recover.

It is a sad pity. A scientific man of such ability can ill be
spared."
Michael made no reply. He stood dismayed. A curious

sensation of pain smote him. He was not thinking of the
loss to science. He was thinking of a certain little fair-
haired, blue-eyed child, who would soon be fatherless, if
this news were true.
"He was taken ill very suddenly, I believe," said the

customer; "it is acute pneumonia. He has not been ill more
than three or four days. Still, I wonder you have not heard,
living so near."
"Such neighbourhood does not count for much in London,

sir," replied Michael, rousing himself from his abstraction.
"Persons may live next door to each other for years, and
never learn each other's names. I should know nothing of
Professor Lavers if he did not happen to be a customer of
mine."
"I suppose that is so," said the minister thoughtfully. "I

suppose nowhere can one attain such complete isolation as
in this London. Its life must tend to harden human hearts
into selfish indifference to the needs of others. Sad indeed
were the life of man, if he were left to the mercies of his
brother man. But it comforts me to think of the great love
of God enfolding the sinful, sorrowful city, and the heart of
God pitying the infinite struggles and woes of humanity. But
I must not linger now. Good day, Mr. Betts."
"Good day," said Michael. He had hardly followed what the

other was saying. It did not seem to him that the love of
God in any perceptible degree brightened the life of man.
But what could a man who was satisfied with himself, and
never done anything wrong in his life, know of the love of
God?
CHAPTER III
LITTLE MARGERY'S LOSS
"I WONDER if he is any better?" Michael Betts said to

himself as he rose the next morning.
It was something new for him to give a thought to any one,

save in the way of business. It was strange indeed that he
should actually feel anxious concerning the health of a
neighbour; but as he moved to and fro, coaxing his fire into
a blaze, and preparing his solitary meal, Michael was
exceedingly desirous of learning how the new day had found
Professor Lavers.
When the woman arrived who came every morning at nine

to clean up his place and do him such womanly services as
he required, he broke through the reserve he was wont to
maintain towards her, and asked her if she could tell him
how Professor Lavers was.
"How who is?" she asked, with an air of surprise.
"Professor Lavers."
"And who's 'e? I never 'eard of 'im," she said.
"Oh," Michael answered, with some embarrassment, "he

lives in Gower Street—No. 48. He's a very learned, noted
man. I thought you might know about him."
"I never 'eard of 'im," she said again. "Is 'e ill, then? What's
the matter with 'im?"
Michael answered her very curtly. Since she could not

satisfy his curiosity, he was not disposed to gratify hers. He
went back into the shop, and busied himself with his books.
About noon the bell over his door tinkled, and looking up he
saw with pleasure that little Margery was entering the shop,
accompanied by a servant maid, who carried several small
parcels.
"Good morning, Mr. Betts," she said, in her clear, high

tones. "I've come to pay you the fourpence I owe you."
"Thank you, missy," he said, looking with interest at the

sweet childish face and the blue eyes lifted so frankly to his.
"It's for the 'Pilgrim's Progress,' you know. I dare-say you

thought I had quite forgotten it, but I hadn't; only nurse
would not let me come before."
"It was no matter, miss. You need not have troubled about
it. And do you like the book as much as you thought you
would?"
"Oh yes; the pictures are lovely. But it is such a pity: we
can't have any nice plays now; we're in dreadful trouble at
home. My father is very ill, and Noel has been sent away to
Aunt Susie's because he would make a noise, and I'm all
alone, and I don't like it."
"Dear, dear! I'm very sorry to hear that," said Michael,

feeling more moved than he could have believed it possible
that he would have been by a matter which did not concern
himself in the least; "but I hope your father is a little better
this morning, my dear."
"I don't think so," said the little girl, with unshed tears in
her eyes as she lifted them to his, "for mother was crying
this morning, and she would not have cried if father had
been better. We're quite in the Slough of Despond at home,
aren't we, Jane?"
Jane smiled in response to the child's quaint words, but her

eyes had a troubled expression. She shook her head as she
met Michael's inquiring glance.
"He's no better," she said in a low tone, "and I'm sore afraid
he'll never be no better."
"It's horrid without Noel," said little Margery, as she sprang

lightly on to the top of a pile of big lexicons and then back
again to the floor. "I can't play alone, and Jane does not
know how to play properly. Besides, we must not make a
noise."
She stood for a moment with a troubled look on her pretty

pink and white face. Then, as she looked up at the old
bookseller, a new idea occurred to her.
"Had you ever a little brother or sister to play with you, Mr.
Betts?—when you were a little boy, I mean. Of course it's a
very long time ago."
"Well, yes, miss, I had a little brother once; but, as you say,
it's a long time ago."
"Then I suppose he is grown-up now. Where is he?"
"I don't know, miss."
"You don't know?" repeated the child in amazement. "You

don't know where your brother is?"
The face of old Betts flushed as he caught the surprise in

her tones.
"It's true, missy; I don't know where he is. Maybe he is

dead; but I can't say."
"Now, Miss Margery, it's time we were going," said Jane

quickly. "You know you promised me you would not stay a
minute if I let you come in."
"All right; I'm ready," responded Margery; but she turned

again to Michael ere she left the shop.
"Do you live here all by yourself, Mr. Betts? It's very lonely
for you, isn't it? But I suppose people don't mind that when
they get old."
He made no reply, except to bid her good day; and the next
minute the green cloak and long golden locks had floated on
the wind round the corner, and he was alone once more.
Was it very lonely for him? He had not thought so before;

but to-day, as he looked round on the dingy old shop, so
closely packed with books, and later, as he sat eating alone
with little appetite the ill-cooked, unsavoury meal which his
charwoman had prepared for him, he had a vague sense
that his life was empty, and dull, and unlovely, and that he
wanted something more for happiness than his trade could
give him, even though he was making a good thing of it.
Almost the first thing Michael Betts saw when he unfolded

his newspaper the next morning was the announcement of
the death of Professor Lavers. After he had read the brief
notice more than once, he read nothing more for some
time. He sat with his breakfast untasted before him, gazing
abstractedly at the row of bookshelves opposite. But he did
not see the titles printed on the dingy covers. He was
seeing a wee, winsome face, half hidden by drooping curls,
and hearing the music of a sweet, childish voice. When he
roused himself, it was to sigh heavily, and say half aloud,
"It's a sad pity. It's a sad pity for that sweet little maid."
CHAPTER IV
MICHAEL MAKES A GOOD BARGAIN
SOME weeks passed by, and Michael saw nothing more of

little Margery. He thought of her more than once, wondering
how it was with her and her little brother, now that their
father was no more. Sometimes when the sudden tinkling of
the bell over the shop door warned him of the approach of a
customer, he would look up, half hoping that he might see
the wee figure in the green cloak and close-fitting velvet
bonnet. But Margery did not come, and if she had, she
would not have worn the green cloak. That had been
exchanged for sombre black, which gave a new and pathetic
beauty to her sweet, round, pink and white face.
More than a month had passed since the professor's death,

when one evening the servant who had accompanied little
Margery when she came to pay the fourpence entered the
shop, and asked Michael if he could call at No. 48, Gower
Street, on the following morning, as her mistress wished to
see him. After a moment's reflection, Michael replied that
he would come, and added an inquiry as to the health of the
little lady who had accompanied her when she came before.
"Miss Margery is quite well, thank you," replied the maid,

"and as much of a chatterbox as ever. I never knew such a
child for asking you questions and saying queer things.
There's no knowing how to answer her."
"Did she grieve much for her father?" asked Michael.
"Well, yes, she cried a great deal. She was fond of her
father, was Miss Margery. And it upset her to see her
mother crying. But she got over it sooner than you would
think. Children quickly forget their troubles. If you could
have seen her and her little brother playing together on the
day their father was buried, you'd have been surprised. But,
there, it wasn't to be expected they could miss their father,
for they saw so little of him. He was always shut up in his
study with his books. A regular bookworm he was. You
couldn't call him nothing else."
Michael made no reply; but he wondered whether, if he had

been the father of a sweet little girl like Margery, he would
have been content to live shut up with his books all day.
"Then I may tell Mrs. Lavers that you will come to-morrow
morning at ten o'clock?" said the maid as she turned to go.
Michael assented. He was not surprised that Mrs. Lavers
should send for him. It frequently happened that his
attendance was requested at houses where there had been
a recent bereavement, necessitating considerable changes.
The professor's widow doubtless wished to dispose of some
or all of her husband's books.
Michael was not mistaken. On his arrival at the house he

was at once ushered into the late professor's library, and
whilst he waited there, he feasted his eyes on the contents
of the ample bookshelves, which were fully and richly
stocked. He had time to form a good estimate of the value
of the books ere Mrs. Lavers came into the room.
The professor's widow was a delicate, graceful woman of

about thirty-five years of age. Her face was pale, and it had
a very careworn, sorrowful expression; but it still bore the
traces of past beauty, and the fair, rippling hair which
showed beneath her widow's cap, and her soft blue eyes,
reminded Michael of little Margery. As she addressed him,
Michael knew instinctively that he had to deal with a true
lady in the highest sense of the term. And she, glancing at
him, was struck with his air of respectability, and believed
that she might trust in his integrity.
"I sent for you, Mr. Betts," she said, "because I am obliged
to part with my husband's books. I have to move into a
very small house, into which I cannot take them. And
indeed for other reasons I feel it my duty to sell them. I
have been advised to show them to you; I have been told
that I may trust you to give me a fair price for them."
"I can give you as good a price for them as any one in the
trade, madam," replied Michael promptly. "Though I say it
myself, it's true that I understand the second-hand book
market as well as any one can. Do I understand that you
wish to part with all the books in this room?"
"Yes, all of them," said the lady, with a sigh; "I have taken
away such as I want to retain."
"Then with your permission, madam, I will make a brief

inspection of the shelves, after which I shall be able to tell
you the sum I can offer."
The lady made a sign of assent, and sank wearily into a

large easy chair which stood by the hearth. She watched
him with a melancholy expression on her face as he slowly
directed his glance from shelf to shelf, occasionally pausing
to jot down certain titles in his notebook.
Michael had already made up his mind as to the probable

value of the books; but he was not going to commit himself
till he had made a thorough survey of the shelves. When at
last he had fully satisfied himself as to the number and
character of the books, he turned to the lady and named
the sum he was willing to give for them.
Mrs. Lavers' delicate cheek flushed as she heard it. She
looked at him with troubled eyes. "No more than that!" she
said. "It seems very little. Why, my husband spent pounds
and pounds every year upon books."
"' I've no doubt he did, madam; but buying and selling are
different things. I know what I may expect to make by
these books, and I assure you it would not pay me to give
more."
"No? Well, of course you know best." She looked

disappointed and anxious. "I must think more about it
before I decide."
"Certainly, madam. You may like to consult other dealers,

perhaps; but I do not think any one in the trade will offer
you more than I do."
With that he bade her good day, and went back to his shop.
Apparently his words proved true, and Mrs. Lavers found no

one willing to give a larger price for the professor's books,
for three days later she sent him an intimation that she was
willing to part with them for the sum he had named. She
requested that he would remove them as quickly as
possible, as she was about to quit the house.
Michael Betts lost no time in sealing the bargain, from

which he knew he should reap good profit. He promptly
waited on Mrs. Lavers and handed her the sum he had
agreed to give, and on the following evening he began to
remove the books, engaging for this purpose the help of a
man with a small hand-cart.
He did not see Mrs. Lavers ere he set about his task of
dismantling the bookshelves. Doubtless she felt it a cruel
necessity which forced her to dispose of her husband's
books. It was little wonder if she shrank from the pain of
seeing them carried out of the house, and therefore kept
out of the way.
Michael had removed the books from several shelves, when

he found lying on the top of a row of books, as if thrust
there by chance, the copy of the "Pilgrim's Progress" which
little Margery had purchased of him with such pleasure. He
took it up and looked at it in surprise. He could not be
mistaken in the book. He knew it by various tokens.
Moreover, a glance at the fly-leaf showed him that Margery
had written on it her own and her brother's name in a large,
round, childish hand.
"Margery and Noel Lavers—the book of them both," she had

written, that there might be no doubt as to the ownership of
the volume.
Surely it was by accident that the book lay there! It could

not be intended that he should carry it away with the
others. The children could hardly have tired so soon of the
book they had longed to possess. Persuaded that it was a
mistake, Michael laid the book carefully aside in a safe place
ere he went on with his task.
Presently he became aware of the sound of eager voices

and noisy little feet on the stairs, then suddenly the door of
the room in which he was working was flung back with a
jerk, and little Margery ran in, followed by a little boy with
dark, curly hair and dark eyes, who, however, started back
on seeing a stranger, and sped away as quickly as he had
come.
Margery looked astonished at finding the room occupied and

books lying about in piles on the floor. She stood still for
some moments, too surprised to utter a word. She wore a
black frock, and black shoes and stockings covered her
dainty feet and ankles; but a large white pinafore hid most
of the frock, and over it her golden curls fell in rich
profusion.
"Why, you're Mr. Betts, in whose shop I bought our

'Pilgrim's Progress,' she said at last. What are you doing
here in my father's library?"
"I'm looking over the books, my dear," replied Michael; after

a moment's hesitation. "Was that your little brother I saw
just now?"
"Yes, that's Noel; but he always runs away when he sees

anybody. It's riddiklus to be so shy. We were coming to look
for a sermon book, because we are going to play at church,
and I must have a book to read out of for the sermon—
something with plenty of long, hard words. It doesn't
matter if I don't say them properly, you know, for there's
only Noel there."
"I see. So you play at going to church. That's a strange

game."
"It's a very nice game," said Margery with eagerness. "I'm

the clergyman, and Noel's the congregation. Sometimes
Noel wants to preach; but he can't preach well. He means
to preach when he is a man, though. He says he shall drive
an omnibus all the week and preach on Sundays."
"Dear me!" said Michael, smiling, "that will be an unusual

kind of life. But look here, little missy, see what I've found!
Your 'Pilgrim's Progress!' You don't want to lose that, do
you?"
"Oh, where did you find it?" exclaimed the child. "Noel and I
have wanted it for days, and nurse said she did not know
where it was."
"It was here on this shelf," said Michael, as he gave it to

her.
"Oh, who could have put it there?" said little Margery,

taking the book and beginning to turn over its pages. "Do
you know, I like the pictures even better than I thought I
should. See, here is Giant Despair. Doesn't he look horrid!
But this is the picture I like best—'The Pilgrims Passing the
River.' Mother says that father has crossed the river and
gone to be with Jesus in the beautiful city. But I wish-oh, I
do wish that he could have stayed here with us!"
The child's voice had suddenly grown mournful, and tears

were dimming the blue eyes. A strange feeling came over
Michael Betts—a curious, choking sensation which he could
not understand. He longed to say something to comfort the
child; but what could he say?
"Mother says we must all cross the river some day,"

continued little Margery, after a moment's silence. "We
don't know when it will be. I should think you would cross
soon, Mr. Betts, for you are so very old. You are older than
my father was, aren't you?"
"I don't know, I am sure, miss," said Michael, turning

hastily to the shelves and beginning to lift down the books
without much heeding what he was doing.
"What are you going to do with those books, Mr. Betts?"

demanded the child.
"I am going to take them away, missy. You see, your poor
father won't want them any more, and they'd only be a
trouble to your mother, especially as she is leaving this
house, so I am going to take them to my shop."
Margery looked at him for a few moments in a troubled,

bewildered way. Then big tears gathered again in her eyes.
"Oh, I can't bear it!" she cried suddenly. "Father is gone,

and now his books are going, and everything will be
different. I cannot bear it."
And then, to Michael's consternation, she threw herself face
downwards on the rug and sobbed aloud with a child's
passionate vehemence. He was in utter dismay, not
knowing in the least what to say or do. It seemed to him
quite a long time that he stood there, helpless and
embarrassed; but in truth only a few minutes passed ere
the door opened and a woman wearing a white cap and
apron came quickly into the room.
"Come, come, Miss Margery, this won't do," she said, not
unkindly, though in a tone of remonstrance, as she bent
over the weeping child. "You mustn't give way like this.
Come, come now."
And taking the child in her arms, she carried her from the
room.
There were tears in Michael's eyes as he turned back to the

bookshelves. The hands which tried to lift the books shook
strangely. He hated his task now. He was thankful when he
had got through with it, and the last load was conveyed to
his shop.
He could not forget the child. He sat up late that night, still
busied with the books, for it was not easy to find room for
them all in the limited space which his premises afforded.
Margery's words kept ringing in his ears—"I should think
you would cross soon, Mr. Betts, for you are so very old.
You are older than my father was, aren't you?"
The child was right, though how she knew he could not
imagine. Michael had seen the professor's age recorded in
the newspaper—he was fifty-one, whereas Michael was fifty-
nine. But what of that?
No one but a child would think him old. Many men lived to
be eighty, and some even to ninety. And he was so well and
strong. No, he need not think yet of that dark, chill river of
death, the very thought of which made him shiver. But he
reflected, and the thought caused him to breathe more than
one heavy sigh, that when his time came to pass that river
there would be no one to go with him to the brink, no loving
voice to bid him farewell, no child to mourn for him as little
Margery mourned for her father. There had been no time in
his hard-working, self-centred, business absorbed life to
cultivate love; but Michael Betts was beginning to feel that
its absence made a sore and woeful lack in his life.
CHAPTER V
UNRIGHTEOUS GAIN
ON the following day, Michael was still busy with the late
professor's books. As he examined them more fully, he was
disposed to congratulate himself on the bargain he had
made. There were several valuable old books in the lot, and
others which, if less aged, were much in request. Michael
foresaw that he would make money by them. It was true his
returns would come in slowly; but nevertheless, he must in
time gain a handsome profit on the sum he had expended.
As he reflected on this, Michael's spirits rose. He forgot the

gloomy thoughts which had troubled him on the previous
evening. He ceased to think with pity of little Margery and
her mother. After all, theirs was the common lot. Men must
die, and women must weep. It was an ill wind which blew
nobody any good, and their wind of trouble had brought
him a good investment. Nothing pleased Michael more than
a prospect of making money. He loved to think that he was
accumulating capital.
He cherished the hope that he should die a rich man,

though he had no one to whom he could leave his savings
when death called him hence. He had never made a will. It
seemed so unnecessary to trouble about that yet. Some day
he would make one, of course. He had no intention of dying
intestate and letting the Crown seize his hardly earned
money. No, he thought he would leave his property to
charities. He had a vague idea that in this way he might
make amends for an uncharitable life. But it was rarely that
he gave the matter a serious thought. Why should he, when
death seemed so remote?
Michael began to realize money from the professor's books

sooner than he could have anticipated. Only a few days
later a gentleman came to the shop and asked for a copy of
an old but still valuable encyclopædia. Michael remembered
that there was one amongst his newly-acquired stock of
books. He looked for the work, and soon brought forward
the two strongly-bound, bulky volumes which formed it. He
was half afraid that his customer would be frightened at the
somewhat high price he felt obliged to ask for them. But the
gentleman made no demur. He seemed so pleased to obtain
them, indeed, that Michael half wished he had asked more.
"I'll take them with me, if you will just put a piece of paper
round them," the gentleman said. "But, stay what is this?"
He had been turning over the pages of one volume when he

came upon an envelope which seemed as if it had been
slipped between the leaves to mark a place.
"That, sir? Oh, I don't suppose it's anything of
consequence," said Michael, as he took it. As he turned it
over in his hand, he perceived, to his astonishment, what
appeared to be bank-notes within the envelope. With his
instinctive caution, however, he said nothing, but thrust the
envelope quickly out of sight. The gentleman concluded that
he had found its contents to be trivial, and said nothing
more about it; and as he waited whilst Michael did up the
parcel, he gave not another thought to the matter.
But as soon as his customer had quitted the shop, Michael

turned eagerly to the envelope. What was his amazement
when he drew from it five Bank of England notes for ten
pounds each! He could hardly believe his eyes. He looked at
them carefully, holding each note up to the light. There
could be no doubt that they were genuine. Fifty pounds!
What a treasure to light upon!
But how did the notes come to be within the old book? Who
had put them there? Had they belonged to Professor
Lavers?
In the first surprise of his discovery, Michael had not a

doubt as to the ownership of the money. His first impulse
was to return them to Mrs. Lavers forthwith. She certainly
had the first right to anything found within the books that
he had purchased of her. The thing to be done was clear
enough to Michael in the earliest moments which followed
his great surprise.
But later, as he looked at the notes and shook them in his

hand, and thought of all that might be done with them,
various doubts and possibilities presented themselves to his
mind. Who could say that the professor had put the notes
where he found them? The notes were not fresh and crisp;
they were soiled, and one was a little torn. They might have

Download textbook Clinical Trial Data Analysis With R And Sas Second Edition Chen ebook all chapter pdf

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Download textbook Clinical Trial Data Analysis With R And Sas Second Edition Chen ebook all chapter pdf

Uploaded by

Copyright:

Available Formats

Clinical trial data analysis with R and

SAS Second Edition Chen

Analysis of Correlated Data with SAS and R Mohamed M.

Survival Analysis with Interval Censored Data A

Analysis of Clinical Trials Using SAS A Practical Guide

Clinical Trial Optimization Using R 1st Edition Alex

Data Analysis with R 2nd Edition Tony Fischetti

Applied time series analysis with R Second Edition

Step by Step Programming with Base SAS 9 4 Second

Applied Meta-Analysis with R and Stata (Chapman &

Bayesian Modeling in Bioinformatics Computational Methods in Biomedical

Noninferiority Testing in Clinical Trials: Statistical Design and Analysis of Clinical

Ding-Geng (Din) Chen

© 2017 by Taylor & Francis Group, LLC

No claim to original U.S. Government works

Printed on acid-free paper

International Standard Book Number-13: 978-1-4987-7952-4 (Hardback)

Library of Congress Cataloging-in-Publication Data

Visit the Taylor & Francis Web site at

To my wife, Jenny Z. Yang, and my sons, Andrew Y. Zhang and Anthony Y.

List of Figures xix

List of Tables xxi

Preface for the Second Edition xxiii

Preface for the First Edition xxvii

About the Authors xxxi

2 Overview of Clinical Trials 17

3 Treatment Comparisons in Clinical Trials 31

4 Treatment Comparisons in Clinical Trials with Covariates 63

4.2 Statistical Models Incorporating Covariates . . . . . . . . . . 65

5 Analysis of Clinical Trials with Time-to-Event Endpoints 95

5.5.2 Breast Cancer with Interval-Censored Data . . . . . . 116

6 Longitudinal Data Analysis for Clinical Trials 133

7 Sample Size Determination and Power Calculations in

7.4 Time-to-Event Endpoint . . . . . . . . . . . . . . . . . . . . 186

8 Meta-Analysis of Clinical Trials 215

9 Bayesian Methods in Clinical Trials 241

10 Bioequivalence Clinical Trials 277

10.3.6 Individual-Based Bienayme-Tchebycheff (BT)

11 Adverse Events in Clinical Trials 315

11.3.4 R for Significant Level Methods . . . . . . . . . . . . . 327

12 Analysis of DNA Microarrays in Clinical Trials 337

1.1 Distributions for Data Generated for “Placebo” . . . . . . . . 10

3.1 Boxplot for Two Treatments . . . . . . . . . . . . . . . . . . . 40

4.1 Boxplot for Baseline DBP . . . . . . . . . . . . . . . . . . . . 73

5.1 Kaplan-Meier Estimator for Survival Function . . . . . . . . . 109

6.1 DBP as a Function of Time for Each Patient . . . . . . . . . 140

7.1 Graphical Features of Sample Size Determination . . . . . . . 168

7.7 Spending Function Plot for gsDesign . . . . . . . . . . . . . . 196

9.1 Distributions of Prior, Direct Simulation, and MCnormalnor-

11.1 Plot of AE Rate for the Three Dose Treatments . . . . . . . . 317

12.1 Boxplot for Original Expression Data . . . . . . . . . . . . . . 347

1.1 ANOVA Table for Simulated Clinical Trial Data . . . . . . . 14

3.1 Diastolic Blood Pressure Trial Data . . . . . . . . . . . . . . 32

4.1 Data for Betablocker Clinical Trial . . . . . . . . . . . . . . . 64

5.1 Breast Carcinoma Data . . . . . . . . . . . . . . . . . . . . . 96

6.1 Ulcer Trial Data . . . . . . . . . . . . . . . . . . . . . . . . . 135

8.1 Data for Cochrane Collaboration Logo . . . . . . . . . . . . . 216

10.1 AUC Data from Chow and Liu (2009) . . . . . . . . . . . . . 278

11.1 AE Data for Clinical Trial . . . . . . . . . . . . . . . . . . . . 315

12.1 Experimental Design and Data . . . . . . . . . . . . . . . . . 342

longitudinal data using mixed-effects modeling, typically two R packages of

clinical trial data analysis at the graduate level (Master’s or Doctorate’s)

Ding-Geng(Din) Chen, Ph.D.

Karl E. Peace, Ph.D.