Statistical Issues in Drug Development

Professor Stephen Senn (MSc, PhD, CStat) is a statistical consultant, researcher and blogger. He has extensive experience in both academia and industry, and is recognized worldwide for his studies in statistical methodology applied to drug development.Professor Senn has been the recipient of national and international awards, including the 1st George C Challis award for Biostatistics at the University of Florida, and the Bradford Hill Medal of the Royal Statistical Society. He is a Fellow of the Royal Society of Edinburgh and an honorary life member of Statisticians in the Pharmaceutical Industry (PSI) and the International Society for Clinical Biostatistics (ISCB) and has honorary professorships in statistics at The University of Sheffield and the University of Edinburgh.

• Preface to the Third EditionPreface to the Second Edition xiiiPreface to the First Edition xviiAcknowledgements xxi1 Introduction 11.1 Drug development 11.2 The role of statistics in drug development 21.3 The object of this book 31.4 The author’s knowledge of statistics in drug development 41.5 The reader and his or her knowledge of statistics 41.6 How to use the book 5References 6Part 1 Four Views of Statistics in Drug Development: Historical, Methodological, Technical and Professional 92 A Brief and Superficial History of Statistics for Drug Developers 112.1 Introduction 112.2 Early Probabilists 122.3 James Bernoulli (1654–1705) 132.4 John Arbuthnott (1667–1753) 142.5 The mathematics of probability in the late 17th, the 18th and early 19th centuries 142.6 Thomas Bayes (1701–1761) 152.7 Adolphe Quetelet (1796–1874) 162.8      George Biddell Airy (1801-1892)2.9 Francis Galton (1822–1911) 162.10 Karl Pearson (1857–1936) 172.11 ‘Student’ (1876–1937) 172.12 R.A. Fisher (1890–1962) 172.13 Modern mathematical statistics 182.14 Medical statistics 192.15 Statistics in clinical trials today 202.16 The current debate 212.17 A living science 212.18 Further reading 23References 233 Design and Interpretation of Clinical Trials as Seen by a Statistician 273.1 Prefatory warning 273.2 Introduction 273.3 Defining effects 283.4 Practical problems in using the counterfactual argument 283.5 Regression to the mean 293.6 Control in clinical trials 333.7 Randomization 343.8 Blinding 363.9 Using concomitant observations 373.10 Measuring treatment effects 383.11 Data generation models 393.12 In conclusion 413.13 Further reading 41References 414 Probability, Bayes, P-values, Tests of Hypotheses and Confidence Intervals 434.1 Introduction 434.2 An example 444.3 Odds and sods 444.4 The Bayesian solution to the example 454.5 Why don’t we regularly use the Bayesian approach in clinical trials? 464.6 A frequentist approach 474.7 Hypothesis testing in controlled clinical trials 484.8 Significance tests and P-values 494.9 Confidence intervals and limits and credible intervals 504.10 Some Bayesian criticism of the frequentist approach 514.11 Decision theory 514.12 Conclusion 524.13 Further reading 52References 535 The Work of the Pharmaceutical Statistician 555.1 Prefatory remarks 555.2 Introduction 565.3 In the beginning 575.4 The trial protocol 575.5 The statistician’s role in planning the protocol 585.6 Sample size determination 595.7 Other important design issues 605.8 Randomization 605.9 Data collection preview 615.10 Performing the trial 615.11 Data analysis preview 615.12 Analysis and reporting 625.13 Other activities 635.14 Statistical research 635.15 Further reading 64References 65Part 2 Statistical Issues: Debatable and Controversial Topics in Drug Development 676 Allocating Treatments to Patients in Clinical Trials 696.1 Background 696.2 Issues 71References 876.A Technical appendix 887 Baselines and Covariate Information 957.1 Background 957.2 Issues 98References 1087.A Technical appendix 1098 The Measurement of Treatment Effects 1138.1 Background 1138.2 Issues 114References 1298.A Technical appendix 1309 Demographic Subgroups: Representation and Analysis 1339.1 Background 1339.2 Issues 134References 1449.A Technical appendix 14510 Multiplicity 14910.1 Background 14910.2 Issues 150References 16110.A Technical appendix 16211 Intention to Treat, Missing Data and Related Matters 16511.1 Background 16511.2 Issues 167References 17811.A Technical appendix 18012 One-sided and Two-sided Tests and other Issues to Do with Significance and P-values 18312.1 Background 18312.2 Issues 184References 19213 Determining the Sample Size 19513.1 Background 19513.2 Issues 198References 21114 Multicentre Trials 21314.1 Background 21314.2 Issues 213References 23014.A Technical appendix 23115 Active Control Equivalence Studies 23515.1 Background 23515.2 Issues 237References 24715.A Technical appendix 24916 Meta-Analysis 25116.1 Background 25116.2 Issues 253References 26816.A Technical appendix 27017 Cross-over Trials 27317.1 Background 27317.2 Issues 275References 28418 n-of-1 Trials 28718.1 Background 28718.2 Issues 289References 29319 Sequential Trials 29519.1 Background 29519.2 Issues 302References 31320 Dose-finding 31720.1 Background 31720.2 Issues 319References 33421 Concerning Pharmacokinetics and Pharmacodynamics 33721.1 Background 33721.2 Issues 343References 35822 Bioequivalence Studies 36122.1 Background 36122.2 Issues 362References 37923 Safety Data, Harms, Drug Monitoring and Pharmaco-epidemiology 38323.1 Background 38323.2 Issues 388References 40324 Pharmaco-economics and Portfolio Management 40524.1 Background 40524.2 Issues 407References 42925 Concerning Pharmacogenetics, Pharmacogenomics and Related Matters 43325.1 Background 43325.2 Issues 437References 45025.A Technical appendix 451Glossary 453Index 483