Preface
Tribute
List of Contributors
About the Companion Website
1 Systematic Reviews in Health Research: An Introduction
Matthias Egger, Julian P.T. Higgins, George Davey Smith
1.1 Systematic Review, Meta-Analysis, or Evidence Synthesis?
1.2 The Scope of Meta-Analysis
1.3 Historical Notes
1.4 Why do we Need Systematic Reviews? The Situation in the 1980s
1.5 Traditional Reviews
1.6 Limitations of a Single Study
1.7 A More Transparent and Thorough Appraisal
1.8 The Epidemiology of Results
1.9 What was the Evidence in 1981?
1.10 An Exercise in Mega-Silliness?
1.11 Conclusions
Part I: Principles and Prodcedures
2 Principles of Systematic Reviewing
Julian P.T. Higgins, George Davey Smith, Douglas G. Altman, and Matthias Egger
2.1 Developing a Review Protocol
2.2 Presenting, Combining, and Interpreting Results
2.3 Interpreting Findings
2.4 Conclusions
3 Identifying Randomized Controlled Trials
Julie Glanville and Carol Lefebvre
3.1 Searching CENTRAL to Identify Randomized Controlled Trials
3.2 Sources to Search in Addition to CENTRAL
3.3 Searching for Studies Other than Randomized Controlled Trials
3.4 Building Search Strategies
3.5 Conclusion
4 Assessing the Risk of Bias in Randomized Trials
Matthew J. Page, Douglas G. Altman, and Matthias Egger
4.1 Risk of Bias and Quality
4.2 The Evidence Base for Risk of Bias
4.3 Sources of Bias in Randomized Trials
4.4 Approaches to Assessing Risk of Bias in Randomized Trials
4.5 Incorporating Risk of Bias in Meta-Analysis
4.6 Conclusions
5 Investigating and Dealing with Publication Bias and Other Reporting Biases
Matthew J. Page, Jonathan A.C. Sterne, Julian P.T. Higgins, and Matthias Egger
5.1 The Evidence Base for Reporting Biases in Health Research
5.2 Approaches to Minimize Risk of Bias Due to Missing Results
5.3 Approaches to Assess Risk of Bias Due to Missing Results
5.4 Conclusions
6 Managing People and Data
Eliane Rohner, Julia Bohlius, Bruno R. da Costa, and Sven Trelle
6.1 The Team
6.2 The Data
6.3 Outlook:Automation and Data Sharing
7 Reporting and Appraisal of Systematic Reviews
Larissa Shamseer, Beverley Shea, Brian Hutton and David Moher
7.1 Consequences of Poor Reporting
7.2 Reporting Systematic Review Protocols
7.3 Reporting Systematic Reviews
7.4 Reporting Systematic Reviews Without Meta-Analyses
7.5 Other guidance for Reporting Systematic Reviews
7.6 Reporting Other Types of Systematic Reviews
7.7 Optimizing Reporting in Practice
7.8 Appraisal of Systematic Reviews
7.9 Conclusions
Part II: Meta Analysis
8 Effect Measures
Julian P.T. Higgins, Jonathan J. Deeks, and Douglas G. Altman
8.1 Individual Study Estimates of Intervention Effect: Binary Outcomes
8.2 Individual Study Estimates of Intervention Effect: Continuous Outcomes
8.3 Individual Study Estimates of Intervention Effect: Time-to-Event Outcomes
8.4 Individual Study Estimates of Intervention Effect: Rates
8.5 Individual Study Estimates of Intervention Effect: Ordinal Outcomes
8.6 Criteria for Selection of a Summary Statistic
8.7 Case Studies
8.8 Discussion
9 Combining Results Using Meta-Analysis
Jonathan J. Deeks, Richard D. Riley, and Julian P.T. Higgins
9.1 Meta-Analysis
9.2 Formulae for Deriving a Summary Estimate of the Intervention Effect in a New Trial by Combining Trial Results (Meta-Analysis)
9.3 Confidence Interval for Overall Effect
9.4 Test Statistic for Overall Effect
9.5 Prediction Interval for the Intervention
9.6 Meta-Analysis with Individual Participation Data
9.7 Additional Analyses
9.8 Some Practical Issues
9.9 Discussion
10 Exploring Heterogeneity
Julian P.T. Higgins and Tianjing Li
10.1 Clinical, Methodological, and Statistical Variability Across Students
10.2 Real and Spurious Heterogeneity
10.3 Subgroup Analysis: Dividing the Evidence into Subsets
10.4 Meta-Regression
10.5 Practical Problems in the Exploration of Heterogeneity
10.6 Closing Remarks
11 Dealing with Missing Outcome Data in Meta-Analysis
Ian R. White and Dimitris Mavridis
11.1 Analysis of a Single Study with Missing Data
11.2 Meta-Analysis with Missing Data
11.3 Method 1: Using Reasons for Missing Data and Simple Assumptions
11.4 Method 2: Quantifying Departures from MAR
11.5 Two Worked Examples
11.6 Recommendations
12 Individual Participant Data Meta-Analysis
Mark C. Simmonds and Lesley A. Stewart
12.1 Advantages and Challenges of Collecting Individual Participant Data
12.2 Performing a Systematic Review Using Individual Participant Data
12.3 Methods for Meta-Analysis with Individual Participant Analysis
12.4 Going Beyond Estimating the Summary Effect
12.5 Individual Participant Data Analysis of Observational Studies
12.6 Combining Individual Participant Data with Published Data
12.7 Reporting Findings
12.8 Conclusion
13 Network Meta-Analysis
Georgia Salanti and Julian P.T. Higgins
13.1 Indirect Comparison and Transitivity
13.2 Indirect and Direct Evidence
13.3 Network Plots of Interventions
13.4 Systematic Reviews Underlying Network Meta-Analysis
13.5 Synthesis of Data
13.6 Intransitvity and Inconsistency
13.7 Ranking Interventions
13.8 Conclusions
14 Dose—Response Meta Analysis
Nicola Orsini, Susanna C. Larsson, and Georgia Salanti
14.1 Example: Coffee Consumption and Mortality Risk
14.2 Estimating Dose—Response Association
14.3 A Linear Trend for a Single Study
14.4 A Quadratic Trend for a Single Study
14.5 A Restricted Cubic Spline Model for a Single Study
14.6 Synthesizing Dose—Response Association Across Studies
14.7 Testing Departure from a Linear Dose—Response Relationship
14.8 Extensions, Limitations,and Developments
14.9 Conclusions
Part III: Specific Study Designs
15 Systematic Reviews of Nonrandomized Studies of Interventions
Jelena Savović, Penny F. Whiting, and Olaf M. Dekkers
15.1 The Importance of Nonrandomized Studies in the Evaluation of Interventions
15.2 Defining the Research Question and Eligibility Criteria for the Review
15.3 Searching for Nonrandomized Studies of Interventions
15.4 Risk of Bias
15.5 Synthesizing Results
15.6 Conclusions
16 Systematic Reviews of Diagnostic Accuracy
Yemisi Takwoingi and Jonathan J. Deeks
16.1 Rationale for Undertaking Systematic Reviews of Studies of Test Accuracy
16.2 Features of Studies of Test Accuracy
16.3 Summary Measures of Diagnostic Accuracy
16.4 Measures of Diagnostic Accuracy
16.5 Systematic Reviews of Studies of Diagnostic Accuracy
16.6 Meta-Analysis of Studies of Diagnostic Accuracy
16.7 General Principles of Diagnostic Accuracy Meta-Analysis
16.8 Methods for Meta-Analysis of a Single Test
16.9 Quantifying and Investigating Heterogeneity
16.10 Comparisons of the Accuracy of Two or More Tests
16.11 Software Options and Model Fitting Issues
16.12 Interpretation and Reporting
16.13 Discussion
17 Systematic Reviews of Prognostic Factor Studies
Richard D. Riley, Karel G.M. Moons, Douglas G. Altman, Gary S. Collins, and Thomas P.A. Debray
17.1 Defining the Review Question
17.2 Searching and Selecting Eligible Studies
17.3 Data Extraction
17.4 Evaluating Applicability and Quality of Primary Studies
17.5 Meta-Analysis
17.6 Quantifying and Examining Heterogeneity
17.7 Examining Small-Study Effects
17.8 Reporting and Interpretation of Results
17.9 Meta-Analysis Using Individual Participant Data
17.10 Conclusions
18 Systematic Reviews of Prediction Models
Gary S. Collins, Karel G.M Moons, Thomas P.A. Debray, Douglas G. Altman, and Richard D. Riley
18.1 Framing the Review Question
18.2 Identifying Relevant Publications
18.3 Data Extraction
18.4 Assessing Methodological Quality
18.5 Meta-Analysis of Clincal Prediction Model Studies
18.6 Case Study: Meta-Analysis of EuroSCORE II
18.7 Discussion
19 Systematic Reviews of Epidemiological Studies of Etiology and Prevalence
Matthias Egger, Diana Buitrago-Garcia, and George Davey Smith
19.1 Why do we Need Systematic Reviews of Epidemiological Studies?
19.2 Meta-Analysis of Epidemiological Studies
19.3 Preparing the Systematic Review
19.4 Triangulation of Evidence
19.5 Conclusion
20 Meta-Analysis in Genetic Association Studies
Gibran Hemani
20.1 Study Designs for Detecting Genetic Associations
20.2 The Role of Meta-Analysis in Genome-Wide Association Studies
20.3 Future Prospects
Part IV: Cochrane and Guideline Development
21 Cochrane: Trusted Evidence. Informed Decisions. Better Health
Gerd Antes, David Tovey, and Nancy Owens
21.1 Background and History
21.2 Cochrane Groups
21.3 Cochrane's Product
21.4 Cochrane in the Twenty-First Century
21.5 Cochrane in Transition: Challenges and Opportunities
22 Using Systematic Reviews in Guideline Development: The GRADE Approach
Holger J. Schünemann
22.1 Introduction
22.2 The Certainy in The Evidence, Quality of the Evidence, or Strength of the Evidence
22.3 Developing Recommendations and Making Decisions
22.4 Outlook
Part V: Outlook
23 Innovations in Systematic Review Production
Julian Elliott and Tari Turner
23.1 Workflow Platforms
23.2 Semi-Automation
23.3 Crowdsourcing
23.4 Data Structures
23.5 Evidence Use
23.6 Living Systematic Reviews
23.7 Diverse Data
23.8 Data Analytics
23.9 Conclusions
24 Future for Systematic Reviews and Meta-Analysis
Shah Ebrahim and Mark D. Huffman
24.1 The Demand for Systematic Reviews
24.2 Increasing Demand is Good
24.3 The Suplpy Side of Systematic Review
24.4 New Frontiers for Systematic Reviews
24.5 Is the Current World of Systematic Reviews Sustainable?
24.6 Methods for Improving the Process of Creating and Updating Systematic Reviews
24.7 Multiple Interventions and Network Meta-Analysis
24.8 Improving Trial Registration, Reporting and Detecting Fraud
24.9 Prioritization of Reviews and Updates
24.10 Conclusion
Part VI: Software
25 Meta-Analysis in Stata
David J.Fisher, Marcel Zwahlen, Matthis Egger, and Julian P.T. Higgins
25.1 Getting Started
25.2 Commands to Perform a Standard Meta-Analysis
25.3 Cumulative and Influence Meta-Analysis
25.4 Funnel Plots and Tests for Funnel Plot Asymmetry
25.5 Meta-Regresson
25.6 Multivariate and Network Meta-Analysis
26 Meta-Analysis in R
Guido Schwarzer
26.1 Getting Started
26.2 Installing R Packages for Meta-Analysis
26.3 Loading Meta-Analysis Packages
26.4 Getting Help
26.5 Aspirin in Preventing Death after Myocardial Infarction (Example 1)
26.6 Beta-Blocker in Preventing Short-Term Mortality After Myocardial Infarction (Example 2)
26.7 Meta-Regression - Influence of Distance from the Equator on Tuberculosis Vaccine Effectiveness
26.8 Evaluation of Bias in Meta-Analysis - Tests for Small-Study Effects and Trim-and-Fill Method
26.9 Other Statistical Methods for Meta-Analysis in R Packages Meta and Metaphor
26.10 Overview of Other R Packages for Meta-Analysis
27 Comprehensive Meta-Analysis Software
Michael Borenstein
27.1 Motivating Example
27.2 Data Entry
27.3 Basic Analysis
27.4 High-Resolution Plot
27.5 Subgroup Analysis
27.6 Meta-Regression
27.7 Publication Bias
27.8 Additional Features in Comprehensive Meta-Analysis
27.9 Teaching Elements
27.10 Documentation
27.11 Availability
Index
)