Randomised Clinical Trials

David Machin, Leicester Cancer Research Group, University of Leicester, UK; and Medical Statistics Group, School of Health and Related Research, University of Sheffield, UK.Peter M. Fayers, Institute of Applied Health, University of Aberdeen, UK.Bee Choo Tai, Saw Swee Hock School of Public Health, National University of Singapore, and National University Health System, Singapore; and Yong Loo Lin School of Medicine, National University of Singapore and National University Health System, Singapore.

• Preface xiiiPart I Basic Considerations 11 Introduction 31.1 Introduction 31.2 Some completed trials 41.3 Choice of design 131.4 Practical constraints 181.5 Influencing clinical practice 201.6 History 201.7 How do trials arise? 221.8 Ethical considerations 241.9 Regulatory requirements 241.10 Focus 251.11 Further reading 252 Design Features 272.1 Introduction 272.2 The research question 292.3 Patient selection 302.4 The consent process 322.5 Choice of interventions 332.6 Choice of design 352.7 Assigning the interventions 372.8 Making the assessments 382.9 Analysis and reporting 382.10 Technical details 422.11 Guidelines 432.12 Further reading 443 The Trial Protocol 453.1 Introduction 453.2 Abstract 473.3 Background 493.4 Research objectives 493.5 Design 523.6 Intervention details 533.7 Eligibility 563.8 Randomisation 583.9 Assessment and data collection 613.10 Statistical considerations 633.11 Ethical issues 663.12 Organisational structure 693.13 Publication policy 693.14 Trial forms 703.15 Appendices 713.16 Regulatory requirements 723.17 Guidelines 743.18 Protocols 744 Measurement and Data Capture 774.1 Introduction 774.2 Types of measures 784.3 Measures and endpoints 804.4 Making the observations 914.5 Baseline measures 924.6 Data recording 934.7 Technical notes 1014.8 Guidelines 1015 Randomisation 1035.1 Introduction 1035.2 Rationale 1045.3 Mechanics 1045.4 Application 1135.5 Carrying out randomisation 1155.6 Documentation 1195.7 Unacceptable methods 1205.8 Guidelines 1206 Trial Initiation 1216.1 Introduction 1216.2 Trial organisation 1226.3 Data collection and processing 1306.4 Internal data monitoring 1326.5 Ethical and regulatory requirements 1336.6 Launching the trial 1346.7 Trial registries 1346.8 Guidelines 1357 Trial Conduct and Completion 1377.1 Introduction 1377.2 Regular feedback 1377.3 Publicity 1417.4 Protocol modifications 1427.5 Preparing the publication(s) 1427.6 The next trial? 1457.7 Protocol 1468 Basics for Analysis 1478.1 Introduction 1478.2 The standard Normal distribution 1488.3 Confidence intervals 1498.4 Statistical tests 1508.5 Examples of analysis 1528.6 Regression methods 1698.7 Other issues 1798.8 Practice 1828.9 Technical details 1839 Trial Size 1859.1 Introduction 1859.2 Significance level and power 1869.3 The fundamental equation 1889.4 Specific situations 1909.5 Practical considerations 1989.6 Further topics 2039.7 Guideline 2069.8 Software 20610 Data and Safety Monitoring 20910.1 Introduction 20910.2 The DSMB 21110.3 Early reviews 21410.4 Interim reviews 21910.5 Protocols 22811 Reporting 22911.1 Introduction 22911.2 Publication 23011.3 Responsibilities 23311.4 Background 23511.5 Methods 23611.6 Findings 24811.7 When things go wrong 25811.8 Conclusions 25911.9 Guidelines 260Part II Adaptions of the Basic Design 26312 More Than Two Interventions 26512.1 Introduction 26512.2 Unstructured comparisons 26612.3 Comparisons with placebo (or standard) 27012.4 Dose-response designs 27512.5 Factorial trials 28012.6 Complex structure comparisons 28913 Paired and Matched Designs 29313.1 Matched-pair trials 29313.2 Cross-over trials 30513.3 Split-mouth designs 31113.4 Guidelines 31714 Repeated Measures Design 31914.1 Introduction 31914.2 Simplified analysis 32214.3 Regression models 32914.4 Auto-correlation 33114.5 Accounting for auto-correlation 33414.6 The design effect (DE) 33814.7 Trial size 34414.8 Practicalities 34714.9 Reporting 35014.10 Matched organs receiving the same intervention 35415 Non-Inferiority and Equivalence Trials 35715.1 Introduction 35715.2 Non-inferiority 35815.3 Analysis 36115.4 Trial size 36615.5 Equivalence 37015.6 Reporting 37315.7 Practical Issues 37315.8 Guidelines 37316 Cluster Designs 37516.1 Design features 37516.2 Procedures 37616.3 Regression models 37916.4 Intra-class correlation 38016.5 Trial size 38116.6 Analysis 38616.7 Practicalities 38816.8 Reporting 38816.9 Further reading 38917 Stepped Wedge Designs 39117.1 Introduction 39117.2 Notation 39217.3 Basic structure 39617.4 Randomisation 39817.5 Cross-sectional design 39817.6 Closed cohort design 40817.7 Practicalities 413Part III Further Topics 41518 Genomic Targets 41718.1 Introduction 41718.2 Predictive markers 41818.3 Enrichment design 42018.4 Biomarker-Stratified Designs 42218.5 Adaptive threshold designs 43119 Feasibility and Pilot Studies 43519.1 Introduction 43519.2 Feasibility studies 43619.3 External-pilot studies 43719.4 Considerations across external-pilot and main trial 44419.5 Internal-pilot studies 44519.6 Other preliminary studies 44719.7 Reporting 44920 Further Topics 45120.1 Introduction 45120.2 Adaptive approaches 45220.3 Large simple trials 46120.4 Bayesian methods 46320.5 Interim analyses 46720.6 Zelen randomised consent designs 47220.7 Systematic overviews 476Statistical Tables 483Glossary 493References 503Index 523