Epidemiology and Medical Statistics (eBook)

Front cover 1
Epidemiology and Medical Statistics 4
Copyright page 5
Table of contents 6
Preface 14
Contributors 16
Ch.1. Statistical Methods and Challenges in Epidemiology and Biomedical Research 20
1. Introduction 20
2. Characterizing the study cohort 22
3. Observational study methods and challenges 25
4. Randomized controlled trials 31
5. Intermediate, surrogate, and auxiliary outcomes 36
6. Multiple testing issues and high-dimensional biomarkers 37
7. Further discussion and the Women’s Health Initiative example 40
References 41
Ch.2. Statistical Inference for Causal Effects, With Emphasis on Applications in Epidemiology and Medical Statistics 47
1. Causal inference primitives 47
2. The assignment mechanism 55
3. Assignment-based modes of causal inference 60
4. Posterior predictive causal inference 66
5. Complications 74
References 77
Ch.3. Epidemiologic Study Designs 83
1.Introduction 83
2.Experimental studies 84
3. Nonexperimental studies 92
4.Cohort studies 92
5. Case-control studies 103
6. Variants of the case-control design 116
7. Conclusion 123
References 123
Ch.4. Statistical Methods for Assessing Biomarkers and Analyzing Biomarker Data 128
1. Introduction 128
2. Statistical methods for assessing biomarkers 129
3. Statistical methods for analyzing biomarker data 145
4. Concluding remarks 162
References 163
Ch.5. Linear and Non-Linear Regression Methods in Epidemiology and Biostatistics 167
1. Introduction 167
2. Linear models 170
3. Non-linear models 186
4. Special topics 195
References 201
Ch.6. Logistic Regression 206
1. Introduction 206
2. Estimation of a simple logistic regression model 207
3. Two measures of model fit 210
4. Multiple logistic regression 211
5. Testing for interaction 213
6. Testing goodness of fit: Two measures for lack of fit 214
7. Exact logistic regression 215
8. Ordinal logistic regression 220
9. Multinomial logistic regression 223
10. Probit regression 225
11. Logistic regression in case-control studies 226
References 228
Ch.7. Count Response Regression Models 229
1. Introduction 229
2. The Poisson regression model 231
3. Heterogeneity and overdispersion 243
4. Important extensions of the models for counts 249
5. Software 266
6. Summary and conclusions 269
References 270
Ch.8. Mixed Models 272
1. Introduction 272
2. Estimation for the linear mixed model 278
3. Inference for the mixed model 280
4. Selecting the best mixed model 283
5. Diagnostics for the mixed model 287
6. Outliers 289
7. Missing data 289
8. Power and sample size 291
9. Generalized linear mixed models 292
10. Nonlinear mixed models 293
11. Mixed models for survival data 294
12. Software 295
13. Conclusions 295
References 296
Ch.9. Survival Analysis 300
1. Introduction 300
2. Univariate analysis 301
3. Hypothesis testing 307
4. Regression models 314
5. Regression models for competing risks 329
References 336
Ch.10. A Review of Statistical Analyses for Competing Risks 340
1. Introduction 340
2. Approaches to the statistical analysis of competing risks 343
3. Example 346
4. Conclusion 358
References 359
Ch.11. Cluster Analysis 361
1. Introduction 361
2. Proximity measures 363
3. Hierarchical clustering 369
4. Partitioning 374
5. Ordination (scaling) 377
6. How many clusters? 380
7. Applications in medicine 383
8. Conclusion 383
References 384
Ch.12. Factor Analysis and Related Methods 386
1. Introduction 386
2. Exploratory factor analysis (EFA) 387
3. Principle components analysis (PCA) 394
4. Confirmatory factor analysis (CFA) 394
5. FA with non-normal continuous variables 398
6. FA with categorical variables 399
7. Sample size in FA 401
8. Examples of EFA and CFA 402
9. Additional resources 408
Appendix A: PRELIS and LISREL code for the CFA example with continuous MVs 410
Appendix B: Mplus code for CFA example with categorical MVs 410
References 410
Ch.13. Structural Equation Modeling 414
1. Models and identification 414
2. Estimation and evaluation 418
3. Extensions of SEM 429
4. Some practical issues 434
References 437
Ch.14. Statistical Modeling in Biomedical Research: Longitudinal Data Analysis 448
1. Introduction 448
2. Analysis of longitudinal data 450
3. Design issues of a longitudinal study 475
References 479
Ch.15. Design and Analysis of Cross-Over Trials 483
1. Introduction 483
2. The two-period two-treatment cross-over trial 486
3. Higher-order designs 495
4. Analysis with non-normal data 501
5. Other application areas 504
6. Computer software 507
References 508
Ch.16. Sequential and Group Sequential Designs in Clinical Trials: Guidelines for Practitioners 510
1. Introduction 511
2. Historical background of sequential procedures 512
3. Group sequential procedures for randomized trials 513
4. Steps for GSD design and analysis 526
5. Discussion 527
References 528
Ch.17. Early Phase Clinical Trials: Phases I and II 532
1. Introduction 532
2. Phase I designs 533
3. Phase II designs 545
4. Summary 558
References 560
Ch.18. Definitive Phase III and Phase IV Clinical Trials 565
1. Introduction 565
2. Questions 567
3. Randomization 569
4. Recruitment 570
5. Adherence/sample size/power 571
6. Data analysis 573
7. Data quality and control/data management 577
8. Data monitoring 577
9. Phase IV trials 582
10. Dissemination – trial reporting and beyond 583
11. Conclusions 584
References 584
Ch.19. Incomplete Data in Epidemiology and Medical Statistics 588
1. Introduction 588
2. Missing-data mechanisms and ignorability 590
3. Simple approaches to handling missing data 592
4. Single imputation 593
5. Multiple imputation 597
6. Direct analysis using model-based procedures 600
7. Examples 602
8. Literature review for epidemiology and medical studies 605
9. Summary and discussion 606
Appendix A 607
Appendix B 611
References 617
Ch.20. Meta-Analysis 621
1. Introduction 621
2. History 622
3. The Cochran–Mantel–Haenszel test 623
4. Glass’s proposal for meta-analysis 625
5. Random effects models 626
6. The forest plot 628
7. Publication bias 629
8. The Cochrane Collaboration 633
References 633
Ch.21. The Multiple Comparison Issue in Health Care Research 635
1. Introduction 635
2. Concerns for significance testing 636
3. Appropriate use of significance testing 637
4. Definition of multiple comparisons 638
5. Rational for multiple comparisons 639
6. Multiple comparisons and analysis triage 640
7. Significance testing and multiple comparisons 642
8. Familywise error rate 644
9. The Bonferroni inequality 645
10. Alternative approaches 648
11. Dependent testing 650
12. Multiple comparisons and combined endpoints 654
13. Multiple comparisons and subgroup analyses 660
14. Data dredging 670
References 670
Ch.22. Power: Establishing the Optimum Sample Size 675
1. Introduction 675
2. Illustrating power 677
3. Comparing simulation and software approaches to power 682
4. Using power to decrease sample size 691
5. Discussion 696
References 696
Ch.23. Statistical Learning in Medical Data Analysis 698
1. Introduction 698
2. Risk factor estimation: penalized likelihood estimates 700
3. Risk factor estimation: likelihood basis pursuit and the LASSO 709
4. Classification: support vector machines and related estimates 712
5. Dissimilarity data and kernel estimates 719
6. Tuning methods 723
7. Regularization, empirical Bayes, Gaussian processes priors, and reproducing kernels 726
References 727
Ch.24. Evidence Based Medicine and Medical Decision Making 731
1. The definition and history of evidence based medicine 731
2. Sources and levels of evidence 734
3. The five stage process of EBM 736
4. The hierarchy of evidence: study design and minimizing bias 737
5. Assessing the significance or impact of study results: Statistical significance and confidence intervals 740
6. Meta-analysis and systematic reviews 741
7. The value of clinical information and assessing the usefulness of a diagnostic test 741
8. Expected values decision making and the threshold approach to diagnostic testing 745
9. Summary 746
10. Basic principles 746
References 747
Ch.25. Estimation of Marginal Regression Models with Multiple Source Predictors 749
1. Introduction 749
2. Review of the generalized estimating equations approach 751
3. Maximum likelihood estimation 754
4. Simulations 756
5. Efficiency calculations 759
6. Illustration 760
7. Conclusion 762
References 764
Ch.26. Difference Equations with Public Health Applications 766
1. Introduction 766
2. Generating functions 767
3. Second-order nonhomogeneous equations and generating functions 769
4. Example in rhythm disturbances 771
5. Follow-up losses in clinical trials 777
6. Applications in epidemiology 784
References 792
Ch.27. The Bayesian Approach to Experimental Data Analysis 794
Preamble: and if you were a Bayesian without knowing it? 794
1. Introduction 795
2. Frequentist and Bayesian inference 797
3. An illustrative example 802
4. Other examples of inferences about proportions 814
5. Concluding remarks and some further topics 822
References 827
Subject Index 832
Handbook of Statistics Contents of Previous Volumes 842