Bayesian Adaptive Methods for Clinical Trials

by Berry; Scott M.

• ISBN13: 9781439825488
• ISBN10: 1439825483
• Edition: 1st
• Format: Hardcover
• Copyright: 2010-07-19
• Publisher: CRC Press

Summary

Driven by modern computing capabilities, there has been a rapid increase of interest in Bayesian methods for the analysis of clinical trial data. This book provides an introduction to Bayesian adaptive methods for the design and analysis of clinical trials. It presents background material on Bayesian inference and covers clinical trials in Phases I, II, and III. The methods are illustrated with many detailed examples and case studies using real data obtained by the authors in their extensive work. The examples are implemented using R and WinBUGS software, which is made available on a supporting website.

Table of Contents

Foreword	p. xi
Preface	p. xiii
Statistical approaches for clinical trials	p. 1
Introduction	p. 1
Comparisons between Bayesian and frequentist approaches	p. 4
Adaptivity in clinical trials	p. 6
Features and use of the Bayesian adaptive approach	p. 8
The fully Bayesian approach	p. 8
Bayes as a frequentist tool	p. 10
Examples of the Bayesian approach to drug and medical device development	p. 12
Basics of Bayesian inference	p. 19
Introduction to Bayes' Theorem	p. 19
Bayesian inference	p. 26
Point estimation	p. 26
Interval estimation	p. 27
Hypothesis testing and model choice	p. 29
Prediction	p. 34
Effect of the prior: sensitivity analysis	p. 37
Role of randomization	p. 38
Handling multiplicities	p. 40
Bayesian computation	p. 42
The Gibbs sampler	p. 44
The Metropolis-Hastings algorithm	p. 45
Convergence diagnosis	p. 48
Variance estimation	p. 49
Hierarchical modeling and metaanalysis	p. 51
Principles of Bayesian clinical trial design	p. 63
Bayesian predictive probability methods	p. 64
Bayesian indifference zone methods	p. 66
Prior determination	p. 68
Operating characteristics	p. 70
Incorporating costs	p. 78
Delayed response	p. 81
Noncompliance and causal modeling	p. 82
Appendix: R Macros	p. 86
Phase I studies	p. 87
Rule-based designs for determining the MTD	p. 88
Traditional 3+3 design	p. 88
Pharmacologically guided dose escalation	p. 91
Accelerated titration designs	p. 92
Other rule-based designs	p. 92
Summary of rule-based designs	p. 92
Model-based designs for determining the MTD	p. 93
Continual reassessment method (CRM)	p. 94
Escalation with overdose control (EWOC)	p. 102
Time-to-event (TITE) monitoring	p. 105
Toxicity intervals	p. 109
Ordinal toxicity intervals	p. 113
Efficacy versus toxicity	p. 116
Trial parameters	p. 117
Joint probability model for efficacy and toxicity	p. 117
Defining the acceptable dose levels	p. 118
Efficacy-toxicity trade-off contours	p. 118
Combination therapy	p. 121
Basic Gumbel model	p. 122
Bivariate CRM	p. 126
Combination therapy with bivariate response	p. 127
Dose escalation with two agents	p. 129
Appendix: R Macros	p. 134
Phase II studies	p. 137
Standard designs	p. 137
Phase IIA designs	p. 138
Phase IIB designs	p. 140
Limitations of traditional frequentist designs	p. 142
Predictive probability	p. 142
Definition and basic calculations for binary data	p. 143
Derivation of the predictive process design	p. 146
Sequential stopping	p. 150
Binary stopping for futility and efficacy	p. 150
Binary stopping for futility, efficacy, and toxicity	p. 151
Monitoring event times	p. 154
Adaptive randomization and dose allocation	p. 155
Principles of adaptive randomization	p. 155
Dose ranging and optimal biologic dosing	p. 163
Adaptive randomization in dose finding	p. 167
Outcome adaptive randomization with delayed survival response	p. 168
Hierarchical models for phase II designs	p. 173
Decision theoretic designs	p. 176
Utility functions and their specification	p. 176
Screening designs for drug development	p. 179
Case studies in phase II adaptive design	p. 183
The Battle trial	p. 183
The I-SPY 2 trial	p. 189
Appendix: R Macros	p. 191
Phase III studies	p. 193
Introduction to confirmatory studies	p. 193
Bayesian adaptive confirmatory trials	p. 195
Adaptive sample size using posterior probabilities	p. 196
Futility analyses using predictive probabilities	p. 200
Handling delayed outcomes	p. 204
Arm dropping	p. 208
Modeling and prediction	p. 211
Prior distributions and the paradigm clash	p. 218
Phase III cancer trials	p. 221
Phase II/III seamless trials	p. 228
Example phase II/III trial	p. 230
Adaptive design	p. 231
Statistical modeling	p. 232
Calculation	p. 233
Simulations	p. 235
Case study: Ablation device to treat atrial fibrillation	p. 241
Appendix: R Macros	p. 247
Special topics	p. 249
Incorporating historical data	p. 249
Standard hierarchical models	p. 250
Hierarchical power prior models	p. 252
Equivalence studies	p. 260
Statistical issues in bioequivalence	p. 261
Binomial response design	p. 263
2 x 2 crossover design	p. 265
Multiplicity	p. 268
Assessing drug safety	p. 269
Multiplicities and false discovery rate (FDR)	p. 275
Subgroup analysis	p. 276
Bayesian approach	p. 276
Bayesian decision theoretic approach	p. 277
Appendix: R Macros	p. 280
References	p. 281
Author index	p. 297
Index	p. 303
Table of Contents provided by Ingram. All Rights Reserved.