Methods and Applications of Statistics in Clinical Trials, Volume 1 – Concepts, Principles, Trials, and Designs

Contributors xxiii
<p>Preface xxix</p>
<p>1 Absolute Risk Reduction 1</p>
<p>1.1 Introduction 1</p>
<p>1.2 Preliminary Issues 1</p>
<p>1.3 Point and Interval Estimates for a Single Proportion 2</p>
<p>1.4 An Unpaired Difference of Proportions 5</p>
<p>1.5 Number Needed to Treat 8</p>
<p>1.6 A Paired Difference of Proportions 10</p>
<p>References 11</p>
<p>Further Reading 12</p>
<p>2 Accelerated Approval 14</p>
<p>2.1 Introduction 14</p>
<p>2.2 Accelerated Development Versus Expanded Access in the U.S.A 14</p>
<p>2.3 Sorting the Terminology Which FDA Initiatives Do What? 15</p>
<p>2.4 Accelerated Approval Regulations: 21 C.F.R. 314.500, 314.520, 601.40 16</p>
<p>2.5 Stages of Drug Development and FDA Initiatives 16</p>
<p>2.6 Accelerated Approval Regulations: 21 CFR 314.500, 314.520, 601.40&nbsp;17</p>
<p>2.7 Accelerated Approval with Surrogate Endpoints 18</p>
<p>2.8 Accelerated Approval with Restricted Distribution 20</p>
<p>2.9 Phase IV Studies/Post Marketing Surveillance 20</p>
<p>2.10 Benefit Analysis for Accelerated Approvals Versus Other Illnesses&nbsp;21</p>
<p>2.11 Problems, Solutions, and Economic Incentives 22</p>
<p>2.12 Future Directions 24</p>
<p>References 25</p>
<p>Further Reading 26</p>
<p>3 AIDS Clinical Trials Group (ACTG) 27</p>
<p>3.1 Introduction 27</p>
<p>3.2 A Brief Primer on HIV/AIDS 27</p>
<p>3.3 ACTG Overview 28</p>
<p>3.4 ACTG Scientific Activities 29</p>
<p>3.5 Development of Potent Antiretroviral Therapy (ART) 29</p>
<p>3.6 Expert Systems and Infrastructure 36</p>
<p>References 37</p>
<p>4 Algorithm–Based Designs 40</p>
<p>4.1 Phase I Dose–Finding Studies 40</p>
<p>4.2 Accelerated Designs 43</p>
<p>4.3 Model–Based Approach in the Estimation of MTD 46</p>
<p>4.4 Exploring Algorithm–Based Designs With Prespecified Targeted Toxicity Levels 48</p>
<p>References 51</p>
<p>5 Alpha–Spending Function 53</p>
<p>5.1 Introduction 53</p>
<p>5.2 Alpha Spending Function Motivation 54</p>
<p>5.3 The Alpha Spending Function 56</p>
<p>5.4 Application of the Alpha Spending Function 57</p>
<p>5.5 Confidence Intervals and Estimation 59</p>
<p>5.6 Trial Design 59</p>
<p>5.7 Conclusions 61</p>
<p>References 61</p>
<p>Further Reading 63</p>
<p>6 Application of New Designs in Phase I Trials 65</p>
<p>6.1 Introduction 65</p>
<p>6.2 Objectives of a Phase I Trial 65</p>
<p>6.3 Standard Designs and Their Shortcomings 66</p>
<p>6.4 Some Novel Designs 67</p>
<p>6.5 Discussion 72</p>
<p>References 72</p>
<p>Further Reading 73</p>
<p>7 ASCOT Trial 74</p>
<p>7.1 Introduction 74</p>
<p>7.2 Objectives 74</p>
<p>7.3 Study Design 74</p>
<p>7.4 Results 75</p>
<p>7.5 Discussion and Conclusions 77</p>
<p>References 78</p>
<p>8 Benefit/Risk Assessment in Prevention Trials 80</p>
<p>8.1 Introduction 80</p>
<p>8.2 Types of B/RAs Performed in Prevention Trials 81</p>
<p>8.3 Alternative Structures of the Benefit/Risk Algorithm used in Prevention Trials 82</p>
<p>8.4 Methodological and Practical Issues with B/RA in Prevention Trials 84</p>
<p>References 87</p>
<p>9 Biased Coin Randomization 90</p>
<p>9.1 Randomization Strategies for Overall Treatment Balance 90</p>
<p>9.2 The Biased Coin Randomization Procedure 91</p>
<p>9.3 Properties 92</p>
<p>9.4 Extensions to the Biased Coin Randomization 92</p>
<p>9.5 Adaptive Biased Coin Randomization 94</p>
<p>9.6 Urn Models 99</p>
<p>9.7 Treatment Balance for Covariates 102</p>
<p>9.8 Application of Biased Coin Designs to Response–Adaptive Randomization 103</p>
<p>References 104</p>
<p>10 Biological Assay, Overview 106</p>
<p>10.1 Introduction 106</p>
<p>10.2 Direct Dilution Assays 108</p>
<p>10.3 Indirect Dilution Assays 109</p>
<p>10.4 Indirect Quantal Assays 113</p>
<p>10.5 Stochastic Approximation in Bioassay 116</p>
<p>10.6 Radioimmunoassay 117</p>
<p>10.7 Dosimetry and Bioassay 118</p>
<p>10.8 Semiparametrics in Bioassays 119</p>
<p>10.9 Nonparametrics in Bioassays 119</p>
<p>10.10 Bioavailability and Bioequivalence Models 120</p>
<p>10.11 Pharmacogenomics in Modern Bioassays 121</p>
<p>10.12 Complexities in Bioassay Modeling and Analysis 122</p>
<p>References 122</p>
<p>Further Reading 124</p>
<p>11 Block Randomization 125</p>
<p>11.1 Introduction 125</p>
<p>11.2 Simple Randomization 125</p>
<p>11.3 Restricted Randomization Through the Use of Blocks 126</p>
<p>11.4 Schemes Using a Single Block for the Whole Trial 130</p>
<p>11.5 Use of Unequal and Variable Block Sizes 131</p>
<p>11.6 Inference and Analysis Following Blocked Randomization 134</p>
<p>11.7 Miscellaneous Topics Related to Blocked Randomization 135</p>
<p>References 136</p>
<p>Further Reading 138</p>
<p>12 Censored Data 139</p>
<p>12.1 Introduction 139</p>
<p>12.2 Independent Censoring 140</p>
<p>12.3 Likelihoods: Noninformative Censoring 143</p>
<p>12.4 Other Kinds of Incomplete Observation 143</p>
<p>References 141</p>
<p>13 Clinical Data Coordination 146</p>
<p>13.1 Introduction 146</p>
<p>13.2 Study Initiation 147</p>
<p>13.3 Study Conduct 151</p>
<p>13.4 Study Closure 158</p>
<p>13.5 Summary 161</p>
<p>References 162</p>
<p>14 Clinical Data Management 164</p>
<p>14.1 Introduction 164</p>
<p>14.2 How Has Clinical Data Management Evolved? 165</p>
<p>14.3 Electronic Data Capture 166</p>
<p>14.4 Regulatory Involvement with Clinical Data Management 167</p>
<p>14.5 Professional Societies 167</p>
<p>14.6 Look to the Future 168</p>
<p>14.7 Conclusion 169</p>
<p>References 169</p>
<p>15 Clinical Significance 170</p>
<p>15.1 Introduction 170</p>
<p>15.2 Historical Background 170</p>
<p>15.3 Article Outline 171</p>
<p>15.4 Design and Methodology 171</p>
<p>15.5 Examples 181</p>
<p>15.6 Recent Developments 181</p>
<p>15.7 Concluding Remarks 185</p>
<p>References 185</p>
<p>16 Clinical Trial Misconduct 191</p>
<p>16.1 The Scope of this Article 191</p>
<p>16.2 Why Does Research Misconduct Matter? 191</p>
<p>16.3 Early Cases 192</p>
<p>16.4 Definition 193</p>
<p>16.5 Intent 194</p>
<p>16.6 What Scientific Misconduct was Not 194</p>
<p>16.7 The Process 194</p>
<p>16.8 The Past Decade 195</p>
<p>16.9 Lessons from the U.S. Experience 196</p>
<p>16.10 Outside the United States 197</p>
<p>16.11 Scientific Misconduct During Clinical Trials 198</p>
<p>16.12 Audit 198</p>
<p>16.13 Causes 199</p>
<p>16.14 Prevalence 200</p>
<p>16.15 Peer Review and Misconduct 200</p>
<p>16.16 Retractions 201</p>
<p>16.17 Prevention 201</p>
<p>References 202</p>
<p>17 Clinical Trials, Early Cancer and Heart Disease 205</p>
<p>17.1 Introduction 205</p>
<p>17.2 Developments in Clinical Trials at the National Cancer Institute (NCI) 205</p>
<p>17.3 Developments in Clinical Trials at the National Heart, Lung, and Blood Institute (NHLBI) 209</p>
<p>References 213</p>
<p>18 Cluster Randomization 216</p>
<p>18.1 Introduction 216</p>
<p>18.2 Examples of Cluster Randomization Trials 217</p>
<p>18.3 Principles of Experimental Design 218</p>
<p>18.4 Experimental and Quasi–Experimental Designs 219</p>
<p>18.5 The Effect of Failing to Replicate 220</p>
<p>18.6 Sample Size Estimation 221</p>
<p>18.7 Cluster Level Analyses 222</p>
<p>18.8 Individual Level Analyses 223</p>
<p>18.9 Incorporating Repeated Assessments 225</p>
<p>18.10 Study Reporting 226</p>
<p>18.11 Meta–Analysis 227</p>
<p>References 228</p>
<p>19 Coherence in Phase I Clinical Trials 230</p>
<p>19.1 Introduction 230</p>
<p>19.2 Coherence: Definitions and Organization 230</p>
<p>19.3 Coherent Designs 232</p>
<p>19.4 Compatible Initial Design 233</p>
<p>19.5 Group Coherence 234</p>
<p>19.6 Real–Time Coherence 235</p>
<p>19.7 Discussion 238</p>
<p>References 238</p>
<p>20 Compliance and Survival Analysis 240</p>
<p>20.1 Compliance: Cause and Effect 240</p>
<p>20.2 All–or–Nothing Compliance 241</p>
<p>20.3 More General Exposure Patterns 242</p>
<p>20.4 Other Structural Modeling Options 242</p>
<p>References 244</p>
<p>21 Composite Endpoints in Clinical Trials 246</p>
<p>21.1 Introduction 246</p>
<p>21.2 The Rationale for Composite Endpoints 246</p>
<p>21.3 Formulation of Composite Endpoints 247</p>
<p>21.4 Examples 248</p>
<p>21.5 Interpreting Composite Endpoints 250</p>
<p>21.6 Conclusions 251</p>
<p>References 251</p>
<p>22 Confounding 252</p>
<p>22.1 Introduction 252</p>
<p>22.2 Confounding as a Bias in Effect Estimation 252</p>
<p>22.3 Confounding and Noncollapsibility 258</p>
<p>22.4 Confounding in Experimental Design 260</p>
<p>References 261</p>
<p>23 Control Groups 263</p>
<p>23.1 Introduction 263</p>
<p>23.2 History 263</p>
<p>23.3 Ethics 264</p>
<p>23.4 Types of Control Groups: Historical Controls 266</p>
<p>23.5 Types of Control Groups: Randomized Controls 268</p>
<p>23.6 Conclusion 271</p>
<p>References 271</p>
<p>24 Coronary Drug Project 273</p>
<p>24.1 Introduction 273</p>
<p>24.2 Objectives 273</p>
<p>24.3 Study Design and Methods 273</p>
<p>24.4 Results 275</p>
<p>24.5 Conclusions and Lessons Learned 281</p>
<p>References 282</p>
<p>Further Reading 284</p>
<p>25 Covariates 285</p>
<p>25.1 Universal Character of Covariates 285</p>
<p>25.2 Use of Covariates in Clinical Trials 286</p>
<p>25.3 Continuous Covariates: Categorization or Functional Form? 293</p>
<p>25.4 Reporting and Summary Assessment of Prognostic Markers 295</p>
<p>References 296</p>
<p>26 Crossover Design 300</p>
<p>26.1 Introduction 300</p>
<p>26.2 The Two–Period, Two–Treatment Design 301</p>
<p>26.3 Higher Order Designs 304</p>
<p>26.4 Model–Based Analyses 307</p>
<p>References 308</p>
<p>27 Crossover Trials 310</p>
<p>27.1 Introduction 310</p>
<p>27.2 2 x 2 Crossover Trial 312</p>
<p>27.3 Higher–Order Designs for Two Treatments 312</p>
<p>27.4 Designs for Three or More Treatments 312</p>
<p>27.5 Analysis of Continuous Data 314</p>
<p>27.6 Analysis of Discrete Data 315</p>
<p>27.7 Concluding Remarks 317</p>
<p>References 317</p>
<p>28 Diagnostic Studies 320</p>
<p>28.1 Introduction 320</p>
<p>28.2 Diagnostic Studies 320</p>
<p>28.3 Reliability 324</p>
<p>28.4 Validity 331</p>
<p>References 338</p>
<p>Further Reading 339</p>
<p>29 DNA Bank 340</p>
<p>29.1 Definition and Objectives of DNA Biobanks 340</p>
<p>29.2 Types of DNA Biobanks 343</p>
<p>29.3 Types of Samples Stored 344</p>
<p>29.4 Quality Assurance and Quality Control in DNA Biobanks 345</p>
<p>29.5 Ethical Issues 346</p>
<p>29.6 Current Biobank Initiatives 348</p>
<p>29.7 Conclusions 350</p>
<p>References 350</p>
<p>30 Up–and–Down and Escalation Designs 353</p>
<p>30.1 Introduction 353</p>
<p>30.2 Up–and–Down Designs 353</p>
<p>30.3 Escalation Designs 357</p>
<p>30.4 Comparing U&amp;D, Escalation and Model–Based Designs 359</p>
<p>References 359</p>
<p>Further Reading 361</p>
<p>31 Dose Ranging Crossover Designs 362</p>
<p>31.1 Introduction 362</p>
<p>31.2 Titration Designs and Extension Studies 369</p>
<p>31.3 Randomized Designs 373</p>
<p>31.4 Discussion and Conclusion 376</p>
<p>References 379</p>
<p>Further Reading 382</p>
<p>32 Flexible Designs 383</p>
<p>32.1 Introduction 383</p>
<p>32.2 The General Framework 384</p>
<p>32.3 Conditional Power and Sample Size Reassessment 387</p>
<p>32.4 Extending the Flexibility to the Choice of the Number of Stages&nbsp;392</p>
<p>32.5 Selection of the Test Statistic 393</p>
<p>32.6 More General Adaptations and Multiple Hypotheses Testing 393</p>
<p>32.7 An Example 395</p>
<p>32.8 Conclusion 395</p>
<p>References 396</p>
<p>33 Gene Therapy 399</p>
<p>33.1 Introduction 399</p>
<p>33.2 Requirements for Successful Therapeutic Intervention 399</p>
<p>33.3 Pre–Clinical Research 402</p>
<p>33.4 Translational Challenges of Gene Therapy Trials 404</p>
<p>33.5 Clinical Trials &middot; 407</p>
<p>33.6 Lessons Learned 408</p>
<p>33.7 The Way Forward 411</p>
<p>References 411</p>
<p>Further Reading 422</p>
<p>34 Global Assessment Variables 423</p>
<p>34.1 Introduction 423</p>
<p>34.2 Scientific Questions for Multiple Outcomes 423</p>
<p>34.3 General Comments on the GST 424</p>
<p>34.4 Recoding Outcome Measures 424</p>
<p>34.5 Types of Global Statistical Tests (GSTs) 425</p>
<p>34.6 Other Considerations 428</p>
<p>34.7 Other Methods 430</p>
<p>34.8 Examples of the Application of GST 434</p>
<p>34.9 Conclusions 435</p>
<p>References 435</p>
<p>35 Good Clinical Practice (GCP) 438</p>
<p>35.1 Introduction 438</p>
<p>35.2 Human Rights and Protections 438</p>
<p>35.3 Informed Consent 439</p>
<p>35.4 Investigational Protocol 439</p>
<p>35.5 Investigator′s Brochure 440</p>
<p>35.6 Investigational New Drug Application 440</p>
<p>35.7 Production of the Investigational Drug 440</p>
<p>35.8 Clinical Testing 441</p>
<p>35.9 Sponsors 442</p>
<p>35.10 Contract Research Organization 444</p>
<p>35.11 Monitors 444</p>
<p>35.12 Investigators 444</p>
<p>35.13 Documentation 444</p>
<p>35.14 Clinical Holds 445</p>
<p>35.15 Inspections/Audits 446</p>
<p>References 446</p>
<p>Further Reading 446</p>
<p>36 Group–Randomized Trials 448</p>
<p>36.1 Introduction 448</p>
<p>36.2 Group–Randomized Trials in Context 449</p>
<p>36.3 The Development of Group–Randomized Trials in Public Health&nbsp; 450</p>
<p>36.4 The Range of GRTs in Public Health 451</p>
<p>36.5 Current Design and Analytic Practices in GRTs in Public Health&nbsp; 452</p>
<p>36.6 The Future of Group–Randomized Trials 453</p>
<p>36.7 Planning a New Group–Randomized Trial 456</p>
<p>References 462</p>
<p>37 Group Sequential Designs 467</p>
<p>37.1 Introduction 467</p>
<p>37.2 Classical Designs 469</p>
<p>37.3 The &aacute;–Spending Function Approach 474</p>
<p>37.4 Point Estimates and Confidence Intervals 477</p>
<p>37.5 Supplements 478</p>
<p>References 479</p>
<p>38 Hazard Ratio 483</p>
<p>38.1 Introduction 483</p>
<p>38.2 Definitions 483</p>
<p>38.3 Illustration of Hazard Rate, Hazard Ratio and Risk Ratio 484</p>
<p>38.4 Example on the Use and Usefulness of Hazard Ratios 486</p>
<p>38.5 Ad–hoc Estimator of the Hazard Ratio 486</p>
<p>38.6 Confidence Interval of the Ad–hoc Estimator 487</p>
<p>38.7 Ad–hoc Estimator Stratified for the Covariate Renal Function 491</p>
<p>38.8 Properties of the Ad–hoc Estimator 493</p>
<p>38.9 Class of Generalized Rank Estimators of the Hazard Ratio 493</p>
<p>38.10 Estimation of the Hazard Ratio with Cox′s Proportional Hazards Model 494</p>
<p>38.11 Discussion 497</p>
<p>Further Reading 499</p>
<p>References 499</p>
<p>39 Large Simple Trials 500</p>
<p>39.1 Large, Simple Trials 500</p>
<p>39.2 Small but Clinically Important Objective 500</p>
<p>39.3 Eligibility 502</p>
<p>39.4 Randomized Assignment 502</p>
<p>39.5 Outcome Measures 504</p>
<p>39.6 Conclusions 506</p>
<p>References 506</p>
<p>Further Reading 508</p>
<p>40 Longitudinal Data 510</p>
<p>40.1 Definition 510</p>
<p>40.2 Longitudinal Data from Clinical Trials 510</p>
<p>40.3 Advantages 512</p>
<p>40.4 Challenges 512</p>
<p>40.5 Analysis of Longitudinal Data 513</p>
<p>References 514</p>
<p>Further Reading 514</p>
<p>41 Maximum Duration and Information Trials 515</p>
<p>41.1 Introduction 515</p>
<p>41.2 Two Paradigms: Duration versus Information 516</p>
<p>41.3 Sequential Studies: Maximum Duration versus Information Trials&nbsp;516</p>
<p>41.4 An Example of a Maximum Information Trial 519</p>
<p>References 521</p>
<p>42 Missing Data 522</p>
<p>42.1 Introduction 522</p>
<p>42.2 Methods in Common Use 524</p>
<p>42.3 An Alternative Approach to Incomplete Data 525</p>
<p>42.4 Illustration: Orthodontic Growth Data 527</p>
<p>42.5 Inverse Probability Weighting 531</p>
<p>42.6 Multiple Imputation 531</p>
<p>42.7 Sensitivity Analysis 532</p>
<p>42.8 Conclusion 533</p>
<p>References 533</p>
<p>43 Mother to Child Human Immunodeficiency Virus Transmission Trials 536</p>
<p>43.1 Introduction 536</p>
<p>43.2 The Pediatric Aids Clinical Trials Group 076 Trial 538</p>
<p>43.3 Results 538</p>
<p>43.4 The European Mode of Delivery Trial 540</p>
<p>43.5 The HIV Network for Prevention Trials 012 Trial 541</p>
<p>43.6 The Mashi Trial 544</p>
<p>References 545</p>
<p>Further Reading 549</p>
<p>44 Multiple Testing in Clinical Trials 550</p>
<p>44.1 Introduction 550</p>
<p>44.2 Concepts of Error Rates 551</p>
<p>44.3 Union–Intersection Testing 552</p>
<p>44.4 Closed Testing 553</p>
<p>44.5 Partition Testing 555</p>
<p>References 556</p>
<p>Further Reading 557</p>
<p>45 Multicenter Trials 558</p>
<p>45.1 Definitions 558</p>
<p>45.2 History 560</p>
<p>45.3 Examples 561</p>
<p>45.4 Organizational and Operational Features 563</p>
<p>45.5 Strengths 564</p>
<p>45.6 Counts 565</p>
<p>Readings 569</p>
<p>References 569</p>
<p>46 Multiple Endpoints 570</p>
<p>46.1 Introduction 570</p>
<p>46.2 Multiple Testing Methods 571</p>
<p>46.3 Multivariate Global Tests 573</p>
<p>46.4 Conclusions 574</p>
<p>References 575</p>
<p>47 Multiple Risk Factor Intervention Trial 577</p>
<p>47.1 Introduction 577</p>
<p>47.2 Trial Design 577</p>
<p>47.3 Trial Screening and Execution 579</p>
<p>47.4 Findings at the End of Intervention 580</p>
<p>47.5 Long–Term Follow–Up 581</p>
<p>47.6 Epidemiologie Findings from Long–Term Follow–up of 361,662 MRFIT Screenees 582</p>
<p>47.7 Conclusions 583</p>
<p>References 583</p>
<p>Further Reading 586</p>
<p>48 N–of–1 Randomized Trials 587</p>
<p>48.1 Introduction 587</p>
<p>48.2 Goal of N–of–1 Studies 587</p>
<p>48.3 Requirements 588</p>
<p>48.4 Design Choices and Details for N–of–1 Studies 589</p>
<p>48.5 Statistical Issues 592</p>
<p>48.6 Other Issues 593</p>
<p>48.7 Conclusions 596</p>
<p>References 596</p>
<p>49 Noninferiority Trial 598</p>
<p>49.1 Introduction 598</p>
<p>49.2 Essential Elements of Noninferiority Trial Design 598</p>
<p>49.3 Objectives of Noninferiority Trials 600</p>
<p>49.4 Measure of Treatment Effect 600</p>
<p>49.5 Noninferiority Margin 601</p>
<p>49.6 Statistical Testing for Noninferiority 603</p>
<p>49.7 Medication Nonadherence and Misclassificat ion/Measurement Error&nbsp;604</p>
<p>49.8 Testing Superiority and Noninferiority 605</p>
<p>49.9 Conclusion 605</p>
<p>References 605</p>
<p>50 Nonrandomized Trials 609</p>
<p>50.1 Introduction 609</p>
<p>50.2 Randomized vs. Nonrandomized Clinical Trials 609</p>
<p>50.3 Control Groups in Nonrandomized Trials 611</p>
<p>50.4 Statistical Methods in Design and Analyses 613</p>
<p>50.5 Conclusion and Discussion 616</p>
<p>References 617</p>
<p>51 Open–Labeled Trials 619</p>
<p>51.1 Introduction 619</p>
<p>51.2 The Importance of Blinding 619</p>
<p>51.3 Reasons Why Trials Might Have to be Open–Label 622</p>
<p>51.4 When Open–Label Trials Might be Desirable 623</p>
<p>51.5 Concluding Comments 623</p>
<p>References 623</p>
<p>Further Reading 624</p>
<p>52 Optimizing Schedule of Administration in Phase I Clinical Trials 625</p>
<p>52.1 Introduction 625</p>
<p>52.2 Motivating Example 627</p>
<p>52.3 Design Issues 627</p>
<p>52.4 Trial Conduct 631</p>
<p>52.5 Extensions and Related Research 632</p>
<p>References 632</p>
<p>53 Partially Balanced Designs 635</p>
<p>53.1 Introduction 635</p>
<p>53.2 Association Schemes 635</p>
<p>53.3 Partially Balanced Incomplete Block Designs 641</p>
<p>53.4 Generalizations of PBIBDs and Related Ideas 648</p>
<p>References 655</p>
<p>54 Phase I/II Clinical Trials 658</p>
<p>54.1 Introduction 658</p>
<p>54.2 Traditional Approach 659</p>
<p>54.3 Recent Developments 660</p>
<p>54.4 Illustrations 663</p>
<p>References 665</p>
<p>55 Phase II/III Trials 667</p>
<p>55.1 Introduction 667</p>
<p>55.2 Description and Legal Basis 668</p>
<p>55.3 Better Dose–Response Studies with Phase 2/3 Designs 672</p>
<p>55.4 Principles of Phase 2/3 Designs 673</p>
<p>55.5 Inferential Difficulties 676</p>
<p>55.6 Summary 678</p>
<p>References 679</p>
<p>Further Reading 680</p>
<p>56 Phase I Trials 682</p>
<p>56.1 Introduction 682</p>
<p>56.2 Phase I in Healthy Volunteers 683</p>
<p>56.3 Phase I in Cancer Patients 684</p>
<p>56.4 Perspectives in the Future of Cancer Phase I Trials 687</p>
<p>56.5 Discussion 688</p>
<p>References 688</p>
<p>57 Phase II Trials 692</p>
<p>57.1 Introduction 692</p>
<p>57.2 Proof–of–Concept (Phase Ha) Trials 693</p>
<p>57.3 Dose–Ranging (Phase lib) Trials 695</p>
<p>57.4 Efficacy Endpoints 697</p>
<p>57.5 Oncology Phase II Trials 697</p>
<p>References 697</p>
<p>Further Reading 699</p>
<p>58 Phase III Trials 700</p>
<p>58.1 Introduction 700</p>
<p>58.2 Research Methodology in Phase III 700</p>
<p>58.3 Type of Design 706</p>
<p>58.4 Discussion 708</p>
<p>References 709</p>
<p>59 Phase IV Trials 711</p>
<p>59.1 Introduction 711</p>
<p>59.2 Definitions and Context 711</p>
<p>59.3 Different Purposes for Phase IV Trials 712</p>
<p>59.4 Essential and Desirable Features of Phase IV Trials 715</p>
<p>59.5 Examples of Phase IV Studies 715</p>
<p>59.6 Conclusion 717</p>
<p>References 717</p>
<p>Further Reading 718</p>
<p>60 Phase I Trials in Oncology 719</p>
<p>60.1 Introduction 719</p>
<p>60.2 Dose–Limiting Toxicity 719</p>
<p>60.3 Starting Dose 720</p>
<p>60.4 Dose Level Selection 720</p>
<p>60.5 Study Design and General Considerations 720</p>
<p>60.6 Traditional, Standard, or 3 + 3 Design 721</p>
<p>60.7 Continual Reassessment Method and Other Designs that Target the MTD722</p>
<p>60.8 Start–Up Rule 722</p>
<p>60.9 Phase I Trials with Long Follow–Up 722</p>
<p>60.10 Phase I Trials with Multiple Agents 723</p>
<p>60.11 Phase I Trials with the MTD Defined using Toxicity Grades 723</p>
<p>References 723</p>
<p>Further Reading 724</p>
<p>61 Placebos 725</p>
<p>61.1 History of Placebo 725</p>
<p>61.2 Definitions 725</p>
<p>61.3 Magnitude of the Placebo Effect 726</p>
<p>61.4 Influences on the Placebo Effect 727</p>
<p>61.5 Ethics of Employing Placebo in Research 728</p>
<p>61.6 Guidelines for the Use of Placebos in Research 729</p>
<p>61.7 Innovations to Improve Research Involving Placebo 731</p>
<p>61.8 Summary 732</p>
<p>References 732</p>
<p>62 Planning a Group–Randomized Trial 736</p>
<p>62.1 Introduction 736</p>
<p>62.2 The Research Question 736</p>
<p>62.3 The Research Team 737</p>
<p>62.4 The Research Design 737</p>
<p>62.5 Potential Design Problems and Methods to Avoid Them 738</p>
<p>62.6 Potential Analytic Problems and Methods to Avoid Them 739</p>
<p>62.7 Variables of Interest and Their Measures 739</p>
<p>62.8 The Intervention 740</p>
<p>62.9 Power 742</p>
<p>62.10 Summary 742</p>
<p>References 743</p>
<p>63 Postmenopausal Estrogen/Progestin Interventions Trial (PEPI) 744</p>
<p>63.1 Introduction 744</p>
<p>63.2 Design and Objectives 744</p>
<p>63.3 Study Design 746</p>
<p>63.4 Outcomes 747</p>
<p>63.5 Results 749</p>
<p>63.6 Conclusions 753</p>
<p>References 754</p>
<p>Further Reading 756</p>
<p>64 Preference Trials 759</p>
<p>64.1 Introduction 759</p>
<p>64.2 Potential Effects of Preference 759</p>
<p>64.3 The Patient Preference Design 761</p>
<p>64.4 Advantages and Disadvantages of the Patient Preference Design 761</p>
<p>64.5 Alternative Designs 764</p>
<p>64.6 Discussion 767</p>
<p>References 768</p>
<p>Further Reading 769</p>
<p>65 Prevention Trials 770</p>
<p>65.1 Introduction 770</p>
<p>65.2 Role Among Possible Research Strategies 771</p>
<p>65.3 Prevention Trial Planning and Design 773</p>
<p>65.4 Conduct, Monitoring, and Analysis 775</p>
<p>References 776</p>
<p>66 Primary Efficacy Endpoint 779</p>
<p>66.1 Defining the Primary Endpoint 779</p>
<p>66.2 Fairness of Endpoints 780</p>
<p>66.3 Specificity of the Primary Endpoint 782</p>
<p>66.4 Composite Primary Endpoints 782</p>
<p>66.5 Missing Primary Endpoint Data 784</p>
<p>66.6 Censored Primary Endpoints 784</p>
<p>66.7 Surrogate Primary Endpoints 785</p>
<p>66.8 Multiple Primary Endpoints 786</p>
<p>66.9 Secondary Endpoints 786</p>
<p>References 786</p>
<p>Further Reading 788</p>
<p>67 Prognostic Variables in Clinical Trials 789</p>
<p>67.1 Introduction 789</p>
<p>67.2 A General Theory of Prognostic Variables 791</p>
<p>67.3 Valid Covariates and Recognizable Subsets 792</p>
<p>67.4 Stratified Randomization and Analysis 793</p>
<p>67.5 Statistical Importance of Prognostic Factors 795</p>
<p>References 797</p>
<p>68 Randomization Procedures 799</p>
<p>68.1 Basics 799</p>
<p>68.2 General Classes of Randomization: Complete Versus Imbalance–Restricted Procedures 800</p>
<p>68.3 Procedures for Imbalance–Restricted Randomization 801</p>
<p>68.4 Randomization–Based Analysis and the Validation Transformation&nbsp;809</p>
<p>68.5 Conclusions 810</p>
<p>References 810</p>
<p>69 Randomization Schedule 813</p>
<p>69.1 Introduction 813</p>
<p>69.2 Preparing the Schedule 814</p>
<p>69.3 Schedules for Open–Label Trials 817</p>
<p>69.4 Schedules to Mitigate Loss of Balance in Treatment Assignments Because of Incomplete Blocks 818</p>
<p>69.5 Issues Related to the use of Randomization Schedule 822</p>
<p>69.6 Summary 824</p>
<p>References 825</p>
<p>Further Reading 826</p>
<p>70 Repeated Measurements 827</p>
<p>70.1 Introduction and Case Study 827</p>
<p>70.2 Linear Models for Gaussian Data 828</p>
<p>70.3 Models for Discrete Outcomes 831</p>
<p>70.4 Design Considerations 836</p>
<p>70.5 Concluding Remarks 837</p>
<p>References 838</p>
<p>71 Simple Randomization 841</p>
<p>71.1 Introduction 841</p>
<p>71.2 Concept of Randomization 841</p>
<p>71.3 Why is Randomization Needed? 842</p>
<p>71.4 Methods: Simple Randomization 842</p>
<p>71.5 Advantages and Disadvantages of Randomization 845</p>
<p>71.6 Other Randomization Methods 846</p>
<p>71.7 Stratified Randomization 846</p>
<p>References 849</p>
<p>Further Reading 849</p>
<p>72 Subgroups 850</p>
<p>72.1 Introduction 850</p>
<p>72.2 The General Problem 851</p>
<p>72.3 Definitions 851</p>
<p>72.4 Subgroup Effects and Interactions 852</p>
<p>72.5 Tests of Interactions and the Problem of Power 853</p>
<p>72.6 Subgroups and the Problem of Multiple Comparisons 856</p>
<p>72.7 Demographic Subgroups 858</p>
<p>72.8 Physiological Subgroups 861</p>
<p>72.9 Target Subgroups 861</p>
<p>72.10 Improper Subgroups 863</p>
<p>72.11 Summary 865</p>
<p>References 865</p>
<p>73 Superiority Trials 867</p>
<p>73.1 Introduction 867</p>
<p>73.2 Clinicians Ask One–Sided Questions, and Want Immediate Answers&nbsp;867</p>
<p>73.3 But Traditional Statistics Is Two–Sided 867</p>
<p>73.4 The Consequences of Two–Sided Answers to One–Sided Questions&nbsp;868</p>
<p>73.5 The Fallacy of the "Negative" Trial 868</p>
<p>73.6 The Solution Lies in Employing One–Sided Statistics 868</p>
<p>73.7 Examples of Employing One–Sided Statistics 868</p>
<p>73.8 One–Sided Statistical Analyses Need to be Specified Ahead of Time&nbsp;869</p>
<p>73.9 A Graphic Demonstration of Superiority and Noninferiority 869</p>
<p>73.10 How to Think about and Incorporate Minimally Important Differences 870</p>
<p>73.11 Incorporating Confidence Intervals for Treatment Effects 871</p>
<p>73.12 Why We Should Never Label an "Indeterminate" Trial Result as "Negative" or as Showing "No Effect" 871</p>
<p>73.13 How Does a Treatment Become "Established Effective Therapy"?&nbsp;872</p>
<p>73.14 Most Trials are Too Small to Declare a Treatment "Established Effective Therapy" 872</p>
<p>73.15 How Do We Achieve a Superiority Result? 872</p>
<p>73.16 Superiority and Noninferiority Trials when Established Effective Therapy Already Exists 872</p>
<p>73.17 Exceptions to the Rule that It Is Always Unethical to Substitute Placebos for Established Effective Therapy 873</p>
<p>73.18 When a Promising New Treatment Might be Added to Established Effective Therapy 873</p>
<p>73.19 Using Placebos in a Trial Should Not Mean the Absence of Treatment 874</p>
<p>73.20 Demonstrating Trials of Promising New Treatments Against (or in Addition to) Established Effective Therapy 874</p>
<p>73.21 Why We Almost Never Find, and Rarely Seek, True "Equivalence"&nbsp;874</p>
<p>73.22 The Graphical Demonstration of "Superiority" and "Noninferiority"&nbsp;876</p>
<p>73.23 Completing the Circle: Converting One–Sided Clinical Thinking into One–Sided Statistical Analysis 876</p>
<p>73.24 A Final Note on Superiority and Noninferiority Trials of "Me–Too" Drugs 877</p>
<p>References 877</p>
<p>Further Reading 877</p>
<p>74 Surrogate Endpoints 878</p>
<p>74.1 Introduction 878</p>
<p>74.2 Illustrations 879</p>
<p>74.3 Validation of Surrogates 880</p>
<p>74.4 Auxiliary Variables 883</p>
<p>74.5 Conclusions 884</p>
<p>References 885</p>
<p>75 TNT Trial 887</p>
<p>75.1 Introduction 887</p>
<p>75.2 Objectives 887</p>
<p>75.3 Study Design 887</p>
<p>75.4 Results 888</p>
<p>75.5 Conclusions 892</p>
<p>References 892</p>
<p>Further Reading 893</p>
<p>76 UGDP Trial 894</p>
<p>76.1 Introduction 894</p>
<p>76.2 Design and Chronology 895</p>
<p>76.3 Results 906</p>
<p>76.4 Conclusion and Discussion 909</p>
<p>References 914</p>
<p>77 Women′s Health Initiative Hormone Therapy Trials 918</p>
<p>77.1 Introduction 918</p>
<p>77.2 Objectives 918</p>
<p>77.3 Study Design 918</p>
<p>77.4 Results 919</p>
<p>77.5 Conclusions 927</p>
<p>References 928</p>
<p>78 Women′s Health Initiative Dietary Modification Trial 931</p>
<p>78.1 Rationale for Biomarker Calibration of Self–Report Measures of Diet&nbsp;931</p>
<p>78.2 Nutrient Biomarker Study Energy and Protein Calibration 932</p>
<p>78.3 Measurement Error Properties of 4DFR, 24HR, and FFQ 933</p>
<p>78.4 Calibration of Self–Report Measures of Physical Activity 933</p>
<p>78.5 Psychosocial Measures and Biomarker–Calibrated Intake 936</p>
<p>78.6 Calibrated Energy, Protein, Protein Density, and Cardiovascular Disease Incidence 937</p>
<p>78.7 Diabetes and Calibrated Consumption 938</p>
<p>78.8 Cancer and Calibrated Intake 940</p>
<p>78.9 Associations Between Protein Intake, Frailty, and Renal Function&nbsp;940</p>
<p>78.10 Summary and Future Directions 941</p>
<p>References 943</p>
<p>Index 945</p>