Оглавление
Iain K. Crombie. Evidence in Medicine
Table of Contents
List of Tables
Guide
Pages
Evidence in Medicine. The Common Flaws, Why They Occur and How to Prevent Them
Preface
Aims of this Book
REFERENCES
CHAPTER 1 The Rationale for Treatment: A Brief History
THEORY AS JUSTIFICATION FOR TREATMENT
TESTING ON A SERIES OF PATIENTS
COMPARING GROUPS
COMPARING SIMILAR GROUPS
CASTING LOTS AND TREATMENT ALLOCATION
RANDOM NUMBERS FOR TREATMENT ALLOCATION
THE NEED FOR BLINDING
CONCLUSION
REFERENCES
CHAPTER 2 Sources of Bias in Randomised Controlled Trials
METHOD OF TREATMENT ALLOCATION
Generation of the Random Assignment
Importance of Concealed Allocation
Evidence that the Randomisation Process Is Subverted
Does Integrity of Allocation Concealment Matter?
PROBLEMS IN MEASURING THE OUTCOME
Switching Primary Outcomes
Blinding of Outcome Assessment
Reporting of Adverse Events
FOLLOW‐UP AND MISSING OUTCOMES
Extent of Loss to Follow‐up
Characteristics of Patients Lost
Bias from Loss to Follow‐up
MISSING OUTCOME DATA AND INTENTION TO TREAT
Methods of Imputation
Modified Intention to Treat
OTHER METHODOLOGICAL CONCERNS. Unregistered Trials and Bias
Small Studies
Low Power
Industry‐Funded Trials
CONCLUSIONS
REFERENCES
CHAPTER 3 Wasted and Unhelpful Trials
WASTED STUDIES. Uncompleted Trials
Unpublished Trials
Unnecessary Trials
NEGLECTED AREAS OF RESEARCH. Less Favoured Clinical Areas
Less Popular Types of Interventions
UNHELPFUL OUTCOME MEASURES
Surrogate Outcomes
Composite Outcomes
Relevance to Patients and Clinicians
COMET: Core Outcome Measures in Effectiveness Trials
LACK OF GENERALISABILITY
Possible Reasons for Exclusions
Poor Reporting of Patient Characteristics
The Average and the Individual
WEAK AND MISLEADING EVIDENCE
Reversals in Medicine
Lack of Reproducibility
CONCLUSION
REFERENCES
CHAPTER 4 Can the Analysis Bias the Findings?
THE P‐VALUE PROBLEM
Definition of the P‐value
The Frequency of Reported P‐values
Calculating and Reporting P‐values
QUESTIONABLE RESEARCH PRACTICES
HARKing (Hypothesising After the Results are Known)
Excluding Data
Multiple Comparisons
Subgroup Analyses
Selection of Covariates
ENSURING HIGH QUALITY ANALYSIS: THE STATISTICAL ANALYSIS PLAN
CONCLUSIONS
REFERENCES
CHAPTER 5 Systematic Reviews and Meta‐Analysis. INTRODUCTION
Meta‐Analysis
Aims of the Chapter
IDENTIFYING RELEVANT TRIALS
Unpublished Studies
Bias from Unpublished Studies
Adequacy of the Sources Searched
Problems with the Search Strategies
Screening for Relevant Trials
EXTRACTING TRIAL DATA
Incorrect Data
Multiple Outcomes
Outcome Reporting Bias
Adverse Events
THE QUALITY OF PRIMARY TRIALS
Taking Quality into Account
POOLING EFFECT SIZES ACROSS TRIALS
Heterogeneity
Exploring Heterogeneity
OTHER METHODOLOGICAL ISSUES. Missing Outcome Data
The Sparse Data Problem
Conflicting Findings
Wasted and Redundant Systematic Reviews
CONCLUSIONS
REFERENCES
CHAPTER 6. Fabrication, Falsification and Spin
FABRICATION
Impact of Fabrication
Identifying and Managing Data Fabrication
Reluctance to Report
FALSIFICATION
QUESTIONABLE RESEARCH PRACTICES
SPIN
Focussing on Misleading Significance
Drawing Misleading Conclusions
Misleading Abstracts
Words to Mislead
RETRACTIONS
DISCUSSION
REFERENCES
CHAPTER 7 Why Do Researchers Falsify Data or Manipulate Study Findings?
THE RESEARCH ENVIRONMENT
Impact Factors
Perverse Incentives
Gaming the System
RESEARCH OVERSIGHT
Peer Review
Predatory Journals and Conferences
CONFLICT OF INTEREST
Financial Conflicts
Non‐Financial Conflicts
INDIVIDUAL LEVEL EXPLANATIONS FOR RESEARCH MISCONDUCT
Data Fabrication and the Dark Triad
The Fraud Triangle, Data Falsification and Questionable Research Practices
HOW HONEST PEOPLE RATIONALISE MISCONDUCT
Societal Factors
Cognitive Biases
Blind Spot Bias
Confirmation Bias
Optimism Bias
Hindsight Bias
Moral Licensing
Motivation for Research
DISCUSSION
REFERENCES
CHAPTER 8 Developing a Strategy to Prevent Poor Quality and Misleading Research
RESEARCH ENVIRONMENT
Research Assessment
Research Training
RESEARCH TRANSPARENCY
Trial Registration
Quality of Reporting of Trial Methods
Data Sharing
RESEARCH OVERSIGHT
Strengthen Peer Review
Monitor Selective Outcome Reporting
RESEARCH INTEGRITY
Reduce the Use of Spin
Reporting Conflict of Interest
Integrity Training
ESSENTIAL ELEMENTS OF A TRANSFORMATIONAL STRATEGY
Commitment from Key Stakeholders
Monitoring the Research Process
Sanctions and Incentives
Availability of Expertise
Coordinated Approaches
Financial Support
The Cost of Inaction
Who Should Pay?
IMPLEMENTING A PROGRAMME FOR ACTION
REFERENCES
Appendix 1: Summary of the Key Findings on Poor Quality Research
PROBLEMS IN THE DESIGN, CONDUCT, ANALYSIS AND REPORTING OF STUDIES. Trials
Statistical Analysis
Systematic Reviews
FREQUENCY OF DATA FABRICATION AND FALSIFICATION
THE CAUSES OF POOR QUALITY AND MISLEADINGRESEARCH
THE FINDINGS IN PERSPECTIVE
REFERENCES
Appendix 2: Initiatives to Improve the Quality of Research
CHANGE THE RESEARCH ENVIRONMENT
Value Negative Findings
Change Research Assessment
Improve Training
Certificates of Research Competence
Trial Forge
PRECIS‐2
INCREASE RESEARCH TRANSPARENCY
Data Sharing
Open Science
The RIAT Initiative
QUALITY OF TRIAL METHODOLOGY. The CONSORT Statement for Trials
The SPIRIT 2013 Statement for Study Protocols
Reporting the Statistical Analysis: The SAMPL Guidelines
Reporting Harms
Reporting Trial Findings
TRIAL REGISTRATION
Enforcement of Trial Registration
REPORTING OF THE METHODS OF SYSTEMATIC REVIEWS
Registration of Protocols of Systematic Reviews
INCREASING ACCESS TO AND USE OF REPORTING GUIDELINES. EQUATOR and Reporting Guidelines
Interventions to Promote Guideline Use
IMPLEMENT VIGOROUS RESEARCH OVERSIGHT. Strengthen Peer Review
Alternative Forms of Peer Review
Reward Peer Reviewers
Audit Trial Publication
Monitor for Selective Outcome Reporting
Preprints and Pre‐Publication Review
Post‐Publication Review
PROMOTE RESEARCH INTEGRITY. National Level Initiatives
Training in Research Integrity
Monitor Data Quality
Promote Reporting of Fraud and Misconduct
Reduce the Use of Spin
Promote Honesty about Conflict of Interest
Research Institutions and Research Integrity
Criminalise Research Fraud
EXAMPLES OF COORDINATED INITIATIVES. The AllTrials Campaign
The QUEST Initiative
The Lancet REWARD Campaign
The UK Reproducibility Network
REFERENCES
Index. A
B
C
D
E
F
G
H
I
J
L
M
N
O
P
Q
R
S
U
V
W
WILEY END USER LICENSE AGREEMENT
