Applied meta-analysis with R / Ding-Geng Chen, Karl E. Peace.

Format:

Book

Author/Creator:

Chen, Ding-Geng, author.

Peace, Karl E., 1941- author.

Contributor:

EBSCOhost.

Harry E. Humphreys Book Fund.

Series:

Chapman & Hall/CRC biostatistics series

Language:

English

Subjects (All):

Meta-analysis.

R (Computer program language).

Physical Description:

1 online resource (xxiv, 321 pages) : illustrations.

Place of Publication:

Boca Raton : CRC Press, 2013.

System Details:

text file

Summary:

In biostatistical research and courses, practitioners and students often lack a thorough understanding of how to apply statistical methods to synthesize biomedical and clinical trial data. Filling this knowledge gap, Applied Meta-Analysis with R shows how to implement statistical meta-analysis methods to real data using R. Drawing on their extensive research and teaching experiences, the authors provide detailed, step-by-step explanations of the implementation of meta-analysis methods using R. Each chapter gives examples of real studies compiled from the literature. After presenting the data and necessary background for understanding the applications, various methods for analyzing meta-data are introduced. The authors then develop analysis code using the appropriate R packages and functions. This systematic approach helps you thoroughly understand the analysis methods and R implementation, enabling you to use R and the methods to analyze your own meta-data. Features: Represents one of the first books on how to use R for analyzing meta-data, Provides up-to-date meta-analysis methods and models and illustrates their application to biomedical research, Describes a variety of real clinical trials with the associated clinical data, Offers easy access to computational methods using R functionality and packages, such as "meta," "rmeta," and "metafor", Gives step-by-step presentations of the code development and results, Requires no prior experience with R Book jacket.

Contents:

1 Introduction to R 1

1.1 What is R? 1

1.2 Steps for Installing R and Updating R Packages 3

1.2.1 First Step: Install R Base System 3

1.2.2 Second Step: Installing and Updating R Packages 4

1.2.3 Steps to Get Help and Documentation 5

1.3 Database Management and Data Manipulations 7

1.3.1 RMySQL to Microsoft Excel 7

1.3.2 Other Methods to Read Data into R 8

1.3.3 R Package foreign 9

1.4 A Simple Simulation on Multi-Center Studies 10

1.4.1 Data Simulation 10

1.4.1.1 R Functions 10

1.4.1.2 Data Generation and Manipulation 11

1.4.1.3 Basic R Graphics 16

1.4.2 Data Analysis 21

1.4.2.1 Data Analysis from Each Center 21

1.4.2.2 Data Analysis with Pooled Data from five-Centers 23

1.4.2.3 A Brief Introduction to Meta-Analysis 24

1.5 Summary and Recommendations for Further Reading 28

2 Research Protocol for Meta-Analyses 31

2.1 Introduction 31

2.2 Defining the Research Objective 32

2.3 Criteria for Identifying Studies to Include in the Meta-Analysis 33

2.3.1 Clarifying the Disease under Study 33

2.3.2 The Effectiveness Measure or Outcome 34

2.3.3 The Type of Control Group 34

2.3.4 Study Characteristics 35

2.3.5 Type of Patient 35

2.3.6 Length of Study 35

2.4 Searching for and Collecting the Studies 36

2.5 Data Abstraction and Extraction 38

2.6 Meta-Analysis Methods 40

2.7 Results 40

2.8 Summary and Discussion 41

3 Fixed-Effects and Random-Effects in Meta-Analysis 43

3.1 Two Datasets from Clinical Studies 44

3.1.1 Data for Cochrane Collaboration Logo: Binary Data 44

3.1.2 Clinical Studies on Amlodipine: Continuous Data 45

3.2 Fixed-Effects and Random-Effects Models in Meta-Analysis 45

3.2.1 Hypotheses and Effect Size 46

3.2.2 Fixed-Effects Meta-Analysis Model: The Weighted-Average 47

3.2.2.1 Fixed-Effects Model 47

3.2.2.2 The Weighting Schemes 48

3.2.3 Random-Effects Meta-Analysis Model: DerSimonian-Laird 50

3.2.3.1 Random-Effects Model 50

3.2.3.2 Derivation of DerSimonian-Laird Estimator of τ² 52

3.2.4 Publication Bias 54

3.3 Data Analysis in R 55

3.3.1 Meta-Analysis for Cochrane Collaboration Logo 55

3.3.1.1 Fitting the Fixed-Effects Model 55

3.3.1.2 Fitting the Random-Effects Model 58

3.3.2 Analysis of Amlodipine Trial Data 60

3.3.2.1 Load the Library and Data 60

3.3.2.2 Fit the Fixed-Effects Model 61

3.3.2.3 Fit the Random-Effects Model 65

3.4 Which Model Should We Use? Fixed-Effects or Random-Effects? 66

3.4.1 Fixed-Effects 67

3.4.2 Random-Effects 67

3.4.3 Performing Both Fixed-Effects and Random-Effects Meta-Analysis 68

3.5 Summary and Conclusions 68

4 Meta-Analysis with Binary Data 71

4.1 Data from Real Life Studies 71

4.1.1 Statin Clinical Trials 71

4.1.2 Five Studies on Lamotrigine for Treatment of Bipolar Depression 74

4.2 Meta-Analysis Methods 77

4.2.1 Analysis with Risk-Ratio 77

4.2.1.1 Definition 77

4.2.1.2 Statistical Significance 80

4.2.1.3 The Risk-Ratio Meta-Analysis: Step-by-Step 82

4.2.1.4 Risk-Ratio Meta-Analysis: R package meta 86

4.2.2 Analysis with Risk-Difference 87

4.2.2.1 Definition 87

4.2.2.2 Implementation in R Package meta 88

4.2.3 Meta-Analysis with Odds Ratio 90

4.2.3.1 Data Structure 90

4.2.3.2 Odds Ratio: Woolf's Method 90

4.2.3.3 R Implementation with R Package meta 94

4.2.4 Meta-Analysis using Mantel-Haenszel Method 94

4.2.4.1 Details of the Mantel-Haenszel Method 94

4.2.4.2 Step-by-Step R Implementation 97

4.2.4.3 Meta-Analysis Using R Library Meta 99

4.2.5 Peto's Meta-Analysis Method 100

4.2.5.1 Peto's Odds Ratio 100

4.2.5.2 Step by Step Implementation in R 102

4.2.5.3 R Implementation in meta 103

4.3 Meta-Analysis of Lamotrigine Studies 106

4.3.1 Risk-Ratio 106

4.3.2 Risk-Difference 112

4.3.3 Odds Ratio 117

4.4 Discussion 122

5 Meta-Analysis for Continuous Data 125

5.1 Two Published Datasets 125

5.1.1 Impact of Intervention 125

5.1.2 Tubeless vs Standard Percutaneous Nephrolithotomy 126

5.2 Methods for Continuous Data 128

5.2.1 Estimate the Mean Difference Δ 128

5.2.2 Estimate the Standardized Mean Difference δ 129

5.2.3 Step-by-Step Implementation in R 131

5.2.3.1 Load the Data Into R 132

5.2.3.2 Meta-Analysis using R Library meta 132

5.2.3.3 Step-by-Step Calculations in R 134

5.3 Meta-Analysis of Tubeless vs Standard Percutaneous Nephrolithotomy 141

5.3.1 Comparison of Operation Duration 141

5.3.2 Comparison of Length of Hospital Stay 145

5.3.3 Comparison of Postoperative Analgesic Requirement 148

5.3.4 Comparison of Postoperative Haematocrit Change 149

5.3.5 Conclusions and Discussion 150

5.4 Discussion 151

6 Heterogeneity in Meta-Analysis 153

6.1 Heterogeneity Quantity Q and the Test of heterogeneity in R meta 153

6.2 The Quantifying Heterogeneity in R meta 155

6.2.1 The τ² Index 155

6.2.2 The H Index 156

6.2.3 The I² Index 156

6.3 Step-By-Step Implementations in R 157

6.3.1 Cochrane Collaboration Logo Data 157

6.3.1.1 Illustration Using R Library meta 158

6.3.1.2 Implementation in R: Step-by-Step 159

6.3.2 Tubeless vs Standard PCNL Data 166

6.3.2.1 Implementation in R Library meta 166

6.3.2.2 Implementation in R: Step-by-Step 167

6.4 Discussion 175

7 Meta-Regression 177

7.1 Data 178

7.1.1 Bacillus Calmette-Guerin Vaccine Data 178

7.1.2 Ischaemic Heart Disease 180

7.1.3 Attention-Deficit/Hyperactivity Disorder for Children and Adolescents 180

7.2 Meta-Regression 182

7.2.1 The Methods 182

7.2.2 Example: BCG Data Analysis 184

7.2.2.1 Random-Effects Meta-Analysis 184

7.2.2.2 Meta-Regression Analysis 188

7.2.3 Meta-Regression vs Weighted Regression 191

7.3 Data Analysis Using R 194

7.3.1 IHD Data Analysis 194

7.3.1.1 Random-Effects Meta-Analysis 194

7.3.1.2 Meta-Regression Analysis 196

7.3.1.3 Comparison of Different Fitting Methods 199

7.3.2 ADHD Data Analysis 204

7.3.2.1 Data and Variables 204

7.3.2.2 Meta-Analysis 205

7.3.2.3 Meta-Regression Analysis 206

7.3.2.4 Summary 212

7.4 Discussion 212

8 Individual-Patient Level Data Analysis versus Meta-Analysis 213

8.1 Introduction 213

8.2 Treatment Comparison for Changes in HAMD 216

8.2.1 IPD Analysis 216

8.2.1.1 IPD Analysis by Each Study 216

8.2.1.2 IPD Analysis with Pooled Data 221

8.2.1.3 IPD Analysis Incorporating Covariates 222

8.2.1.4 Summary of IPD Analysis 224

8.2.2 Meta-Analysis 224

8.3 Treatment Comparison for Changes in MADRS 233

8.3.1 IPD Analysis 233

8.3.2 Meta-Analysis 239

8.4 Summary 246

8.5 Simulation Study on Continuous Outcomes 247

8.5.1 Simulation Data Generator 247

8.5.2 Simulation Data Estimator 251

8.5.3 Simulation 254

8.6 Discussion 259

9 Meta-Analysis for Rare Events 261

9.1 The Rosiglitazone Meta-Analysis 262

9.2 Step-by-Step Data Analysis in R 263

9.2.1 Load the Data 263

9.2.2 Data Analysis for Myocardial Infarction (MI) 264

9.2.3 Data Analysis for Cardiovascular Death (Death) 274

9.3 Discussion 281

10 Other R Packages for Meta-Analysis 283

10.1 Combining p-Values in Meta-Analysis 284

10.2 R Packages for Meta-Analysis of Correlation Coefficients 285

10.2.1 Introduction 286

10.2.2 Example 286

10.2.3 Discussion 291

10.3 Multivariate Meta-Analysis 292

10.3.1 The Model and the Package of mvmeta 292

10.3.2 Examples 293

10.3.3 Summary 298

10.4 Discussion 299.

Notes:

Description based on print version record.

Includes bibliographical references (pages 305-313) and index.

Electronic reproduction. Ipswich, MA Available via World Wide Web.

Local Notes:

Acquired for the Penn Libraries with assistance from the Harry E. Humphreys Book Fund.

ISBN:

1466506008

9781466506008

Publisher Number:

99954967787

Access Restriction:

Restricted for use by site license.