Design and Analysis of Quality of Life Studies in Clinical Trials, Second Edition (Chapman & Hall CRC Interdisciplinary Statistics)

Cover Page......Page 1
Title: DESIGN and ANALYSIS of QUALITY of LIFE STUDIES in CLINICAL TRIALS, Second Edition......Page 5
ISBN 9781420061178......Page 6
Contents......Page 8
Datasets for the Examples......Page 18
Scope of the Book......Page 19
Self-Study and Discussion Groups......Page 20
Acknowledgments......Page 21
1.1 Health-Related Quality of Life (HRQoL)......Page 22
1.2.1 Health Status Measures......Page 24
1.2.2 Patient Preference Measures......Page 25
1.2.4 Generic versus Disease-Specific Instruments......Page 26
1.2.6 Response Format......Page 27
1.2.7 Period of Recall......Page 28
1.3 Study 1: Adjuvant Breast Cancer Trial......Page 29
1.3.2 Quality of Life Measure and Scoring......Page 30
1.3.3 Timing of HRQoL Assessments......Page 32
1.4 Study 2: Migraine Prevention Trial......Page 33
1.4.1 Patient Selection and Treatment......Page 34
1.4.2 Quality of Life Measure and Scoring......Page 35
1.4.4 Questionnaire Completion/Missing Data......Page 36
1.5.1 Treatment......Page 37
1.5.2 Quality of Life Measure and Scoring......Page 38
1.5.3 Timing of Assessments......Page 39
1.6 Study 4: Renal Cell Carcinoma Trial......Page 40
1.6.2 Quality of Life Measures and Scoring......Page 41
1.6.3 Timing of HRQoL Assessments......Page 42
1.6.4 Questionnaire Completion/Missing Data......Page 43
1.7.1 Patient Selection and Treatment......Page 44
1.7.2 Patient Reported Outcomes......Page 45
1.8.2 Timing and Frequency of Assessments......Page 46
1.9 Summary......Page 48
2.1.1 Purpose of the Protocol......Page 50
2.3 Research Objectives and Goals......Page 52
2.3.2 Pragmatic versus Explanatory Inference......Page 54
2.4 Selection of Subjects......Page 55
2.5.1 Event- or Condition-Driven Designs......Page 56
2.5.4 Timing of the Follow-Up HRQoL Assessments......Page 57
2.5.5 Frequency of Evaluations......Page 58
2.5.7 Assessment after Discontinuation of Therapy......Page 59
2.6 Selection of Measurement Instrument(s)......Page 60
2.6.1 Trial Objectives......Page 61
2.6.2 Validity and Reliability......Page 62
2.6.3 Appropriateness......Page 63
2.7.1 Order and Place of Administration......Page 64
2.7.3 Data Collection and Key Personnel......Page 65
2.7.4 Avoiding Missing Data......Page 66
Education......Page 67
Explicit Procedures for Follow-Up......Page 68
2.8.2 Scoring Multi-Item Scales......Page 69
2.8.3 Item nonresponse......Page 70
2.9 Summary......Page 72
3.1.1 Repeated Measures Models......Page 74
3.1.3 Selection between Models......Page 75
Renal Cell Carcinoma Trial (Study 4)......Page 76
3.2.1 Advantages of the General Linear Model (GLM)......Page 77
Likelihood Ratio Tests......Page 78
Cell Mean Model......Page 79
Center Point Model......Page 81
3.3.4 Covariates......Page 82
Cell Mean Model......Page 83
Reference Row Model......Page 84
3.3.6 Modeling the “Mean” Structure in SPSS......Page 85
Center Point Model with Common Baseline......Page 86
Cell Mean Model......Page 87
3.4 Building Repeated Measures Models: The Covariance Structure......Page 88
3.4.1 Unstructured Covariance......Page 89
3.4.2 Structured Covariance......Page 90
3.4.3 Building the Covariance Structure in SAS......Page 91
3.4.5 Building the Covariance Structure in R......Page 93
3.5 Estimation and Hypothesis Testing......Page 95
Cell Mean Model......Page 96
Center Point with a Common Baseline......Page 97
Cell Mean Model......Page 98
3.5.3 Estimation and Hypothesis Testing in R......Page 99
Center Point with a Common Baseline......Page 100
3.6 Summary......Page 102
4.1 Introduction......Page 104
4.2.1 Polynomial Models......Page 105
4.2.2 Piecewise Linear Regression......Page 106
4.2.3 Modeling the “Mean” Structure in SAS......Page 108
4.2.5 Modeling the “Mean” Structure in R......Page 109
4.3 Building Growth Curve Models: The Covariance Structure......Page 110
4.3.1 Variance of Random Effects (G)......Page 111
4.3.2 Variance of Residual Errors (Ri)......Page 112
4.3.3 Building the Covariance Structure in SAS......Page 114
4.3.4 Building the Covariance Structure in SPSS......Page 115
4.4 Model Reduction......Page 116
Area under the Curve......Page 117
Contrasts at Specific Points in Time......Page 118
Rates of Change......Page 119
4.5.2 Estimation and Testing in SPSS......Page 120
4.5.3 Estimation and Testing in R......Page 121
4.6 An Alternative Covariance Structure......Page 122
Structure 2......Page 123
4.6.3 Estimation and Testing......Page 124
4.7 Summary......Page 125
5.1 Introduction......Page 126
Continuous Covariates......Page 128
Departures from Linear Associations......Page 131
5.2.2 Change from Baseline......Page 133
5.2.3 Centering Covariates......Page 134
5.3 Mediation......Page 136
5.3.1 Mediation with Treatment as the Predictor......Page 137
5.3.2 Mediation with Time as the Predictor......Page 139
5.4.1 Mediation in Mixed Effects Models......Page 141
5.4.2 Non-Linear Relationships......Page 142
5.5 Summary......Page 144
6.1 Introduction......Page 146
6.1.3 How Much Data Can Be Missing?......Page 147
6.1.4 Prevention......Page 148
Lung Cancer Trial (Study 3)......Page 149
6.3.1 The Concept......Page 151
6.3.2 Notation......Page 152
6.4.1 The Concept......Page 153
Lung Cancer Trial (Study 3)......Page 154
6.4.4 Analytic Methods......Page 155
Formal Tests......Page 156
Analytic Explorations......Page 159
6.6.1 The Concept......Page 160
6.6.2 Identifying Dependence on Unobserved Data…......Page 161
Lung Cancer Trial (Study 3)......Page 162
6.7 Example for Trial with Variation in Timing of Assessments......Page 164
6.8 Example with Different Patterns across Treatment Arms......Page 166
6.9 Summary......Page 167
7.2 Missing Completely at Random......Page 170
Hypothetical Example......Page 171
7.2.2 Repeated Univariate Analyses......Page 172
Adding Baseline as a Covariate......Page 173
Change from Baseline......Page 174
Hypothetical Example......Page 175
7.3.1 Maximum Likelihood Estimation (MLE)......Page 176
7.3.2 Empirical Bayes Estimates......Page 177
7.3.5 Baseline Assessment as a Covariate......Page 178
7.3.6 Adding Other Baseline Covariates......Page 179
Exclusion of Observations......Page 180
7.4.1 Selection Models......Page 181
7.4.3 Final Comments......Page 182
7.5 Summary......Page 183
8.1 Introduction to Imputation......Page 184
8.1.2 Imputation in Multivariate Longitudinal Studies......Page 185
8.2 Missing Items in a Multi-Item Questionnaire......Page 186
8.3.1 Mean Value Substitution......Page 188
8.3.2 Explicit Regression Models......Page 189
Identification of the Imputation Model......Page 190
Simple Univariate Regression......Page 191
Conditional Predicted Values......Page 192
8.4.1 Last Value Carried Forward (LVCF)......Page 193
8.4.3 Arbitrary High or Low Value......Page 195
8.4.4 Hot Deck and Other Sampling Procedures......Page 196
8.6 Underestimation of Variance......Page 197
8.7.1 Sensitivity Analysis......Page 199
8.8 Summary......Page 200
9.2 Overview of Multiple Imputation......Page 202
Step 3: Analysis of M Datasets......Page 203
9.3.1 Identification of the Imputation Model......Page 204
9.3.2 Computation of Imputed Values......Page 205
9.3.3 Practical Considerations......Page 206
9.3.5 Extensions to Multiple HRQoL Measures......Page 207
9.3.7 Lung Cancer Trial (Study 3)......Page 208
Implementation in SAS......Page 210
Implementation in SPSS......Page 212
9.4.2 Predictive Mean Matching......Page 213
Implementation in SPSS......Page 214
9.5.1 Practical Considerations......Page 215
9.6 Multivariate Procedures for Non-Monotone Missing Data......Page 217
9.6.2 Implementation in SPSS......Page 218
9.6.3 Implementation in R......Page 220
Estimates......Page 221
Other Useful Information......Page 222
9.7.3 Analysis of M Datasets in SAS......Page 223
9.7.4 Analysis of M Datasets in SPSS......Page 224
9.7.5 Analysis of M Datasets in R......Page 225
9.8.1 Sensitivity Analyses......Page 226
9.8.4 Implications for Design......Page 228
9.9 Summary......Page 229
10.1.1 General Approach......Page 230
Lung Cancer Trial (Study 3)......Page 231
10.2.1 Specifying Patterns......Page 234
10.3 Restrictions for Growth Curve Models......Page 235
10.3.1 Linear Trajectories over Time......Page 236
10.3.2 Estimation of the Parameters......Page 238
Lung Cancer Trial......Page 239
Renal Cell Carcinoma Trial......Page 240
Average of Stratum Specific Estimates......Page 241
10.3.5 Implementation in R......Page 244
Average of Stratum Specific Estimates......Page 245
Centered Indicator Variables......Page 246
10.4.1 Bivariate Data (Two Repeated Measures)......Page 247
Complete-Case Missing Variable (CCMV) Restriction......Page 248
Brown’s Protective† Restrictions......Page 249
Large-Sample Inferences for Mu2......Page 250
Lung Cancer Trial......Page 251
Lung Cancer Trial......Page 252
Available Case Missing Value (ACMV) Restriction......Page 253
Comparison of CCMV, ACMV and NCMV Estimates......Page 255
10.5 Standard Errors for Mixture Models......Page 256
10.5.1 Delta Method......Page 257
10.5.2 Bootstrap Methods......Page 258
10.6 Summary......Page 259
Random Effects......Page 260
11.2 Conditional Linear Model......Page 262
11.2.1 Assumptions......Page 263
11.2.3 Lung Cancer Trial (Study 3)......Page 264
Implementation in SPSS......Page 268
Random-Coefficient versus Pattern Mixture Models......Page 269
11.3 Varying Coefficient Models......Page 270
11.3.1 Assumptions......Page 271
11.3.2 Application......Page 272
11.3.3 General Application......Page 273
11.4 Joint Models with Shared Parameters......Page 274
11.4.3 Alternative Parameterizations......Page 276
Choice of…......Page 277
Initial Estimates......Page 278
Maximization of Likelihood......Page 280
Results......Page 282
11.4.6 Implementation in R......Page 283
11.4.7 Multiple Causes of Dropout......Page 284
11.4.8 Other Model Extensions......Page 286
11.5 Summary......Page 287
12.1 Introduction......Page 288
12.2 Outcome Selection Model for Monotone Dropout......Page 289
Dropout as a Function of Time and Treatment (CRD)......Page 291
Dropout as a Function of Previously Observed Outcomes (RD)......Page 292
Dropout as a Function of the Current Measure of the Outcome (ID)......Page 294
12.3 Summary......Page 295
13.1 Introduction......Page 296
Relative Importance of Endpoints......Page 297
13.2 General Strategies for Multiple Endpoints......Page 298
13.2.1 Limiting the Number of Confirmatory Tests......Page 299
13.2.3 Multiple Testing Procedures......Page 300
13.3.1 Univariate versus Multivariate Test Statistics......Page 301
13.3.3 Error Rates......Page 302
13.4.1 Global Tests Based on Multivariate Test Statistics......Page 303
13.4.2 Equally Weighted Univariate Statistics......Page 304
13.4.3 Importance Weighting/ Spending alpha......Page 305
Step-Up Procedure......Page 306
13.5.3 Implementation in SPSS......Page 307
13.6 Closed Testing and Gatekeeper Procedures......Page 308
13.6.1 Closed Testing Based on a Bonferroni Correction......Page 309
Design 1: Single Primary Endpoint......Page 310
Design 3: Joint Co-Primary Endpoints......Page 311
13.6.3 Shortcuts for Closed Testing Procedures......Page 312
13.6.4 A Closed Testing Procedure Based on a Multivariate Test......Page 314
13.7 Summary......Page 315
14.1.1 Composite Endpoints versus Summary Measures......Page 316
Increased Power......Page 318
14.2 Choosing a Composite Endpoint or Summary Measure......Page 319
14.3 Summarizing across HRQoL Domains or Subscales......Page 320
14.3.2 Factor Analytic Weights......Page 322
Composite Endpoints......Page 324
14.4 Summary Measures across Time......Page 326
14.4.2 Simple Linear Functions......Page 327
Repeated Measures Models......Page 328
Mixed-Effects Growth Curve Models......Page 329
Repeated Measures......Page 330
14.4.3.1 Mixed-Effect Growth Curve Models......Page 331
14.5 Composite Endpoints across Time......Page 332
14.5.1 Notation......Page 333
14.5.2 Missing Data......Page 334
Lung Cancer Trial......Page 335
14.5.4 Area Under the Curve (AUC)......Page 336
Impact of Missing Data......Page 338
Impact of Missing Data......Page 339
14.5.6.1 Stratified Analysis of Composite Endpoints......Page 340
14.6 Summary......Page 341
15.2 QALYs......Page 344
15.2.1 Estimation of QALYi......Page 345
Limited Follow-Up......Page 346
Timing of Assessments......Page 347
15.2.2 Analysis of QALYi......Page 348
15.3 Q-TWiST......Page 349
15.3.1 Kaplan-Meier Estimates of Time in Health States......Page 350
15.3.2 Case 1: Known Estimates of UTOX and UREL......Page 352
15.3.3 Case 2: Two Unknown Weights That Are Equal across Treatments......Page 353
15.3.4 Proportional Hazards Estimates of Time in Health States......Page 355
15.4 Summary......Page 356
16.1 Introduction......Page 358
16.2.2 Primary versus Secondary Endpoints......Page 359
16.2.4 Comparison Procedures......Page 360
16.2.5 Who Is Included?......Page 361
16.2.7 Missing Data......Page 362
16.3 Sample Size and Power......Page 363
Familiar Special Cases......Page 365
Complete Designs Inflated for Dropout......Page 366
Incomplete Designs......Page 367
16.3.3 Longitudinal Data and Growth Curve Model......Page 371
Intermittent Missing Data Patterns and Time-Varying Covariates......Page 373
Multivariate Tests......Page 374
Restricted Maximum Likelihood Estimation......Page 375
16.5 Summary......Page 377
Loading Datasets......Page 380
Centering Variables......Page 381
Bootstrap Procedures for Longitudinal Data......Page 382
Loading Datasets......Page 384
Setting up Models for the “Means”......Page 385
Functions......Page 386
Bootstrap Procedures for Longitudinal Data......Page 388
Reverse Coding and Scoring Questionnaires......Page 390
Centering Variables......Page 391
Bootstrap Procedure for Longitudinal Data......Page 392
Putting it together......Page 394
Appendix C: Cubic Smoothing Splines......Page 378
References......Page 398
B......Page 416
C......Page 417
H......Page 418
L......Page 419
M......Page 420
N......Page 421
R......Page 422
S......Page 423
T......Page 424
Z......Page 425
Back Page......Page 426