Epidemiology and Biostatistics: An Introduction to Clinical Research

282 Pages·2019·4.494 MB·English

by  Noel S. Weiss, MD PhD

Epidemiology and
Biostatistics
An Introduction to Clinical
Research
Bryan Kestenbaum
Second Edition
123
Epidemiology and Biostatistics
Bryan Kestenbaum
Epidemiology and
Biostatistics
An Introduction to Clinical Research
Second Edition
Editors
Noel S. Weiss, MD PhD
Department of Epidemiology
University of Washington
Seattle, WA
USA
Abigail Shoben, PhD
Biostatistics, College of Public Health
The Ohio State University
Columbus, OH
USA
Bryan Kestenbaum, MD, MS
Division of Nephrology
Department of Medicine
University of Washington
Seattle, WA
USA
ISBN 978-3-319-96642-7 ISBN 978-3-319-96644-1 (eBook)
https://doi.org/10.1007/978-3-319-96644-1
Library of Congress Control Number: 2018955511
© Springer Nature Switzerland AG 2009, 2019
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the
editors give a warranty, express or implied, with respect to the material contained herein or for any errors
or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims
in published maps and institutional affiliations.
This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface
This textbook was originally created from a disparate collection of materials used to
teach Epidemiology and Biostatistics to second-year medical students at the
University of Washington. These materials included handouts, practice problems,
guides to reading journal articles, quizzes, notes from student help sessions, and
student emails. The primary goal of these materials, and now this book, is to recre-
ate the perspective of learning Epidemiology and Biostatistics for the first time.
With critical editing assistance from Epidemiology faculty, graduate students in
Epidemiology and Biostatistics, and the students themselves, I have tried to pre-
serve the innate logic and connectedness of clinical research methods and demon-
strate their application.
The textbook is intended to provide students with the tools necessary to form
their own informed conclusions from research studies. More than ever, a clear
understanding of the fundamental aspects of Epidemiology and Biostatistics is
needed to successfully navigate the increasingly complex methods used by modern
research studies. The appetite for studies of human health has grown at a rapid pace;
yet, the interpretation of findings obtained from these studies has not always keep
pace.
In this second edition, I have tried to further clarify concepts that are most diffi-
cult for students and provide a more logical structure to the material. Many new
examples have been added throughout the textbook to reenforce these concepts.
This book could not have been created without the dedicated help of the editors,
the teaching assistants, and most importantly the students, who asked the important
questions. I would especially like to thank my family who patiently allowed me so
much time to write.
Seattle, WA, USA Bryan Kestenbaum, MD, MS
v
Contents
Part I Epidemiology
1 Causal Relationships in Health and Disease . . . . . . . . . . . . . . . . . . . . . . 3
1.1 Inferring Causation from Epidemiologic Studies . . . . . . . . . . . . . . . 6
1.2 Factors Favoring an Inference of Causation . . . . . . . . . . . . . . . . . . . 7
1.2.1 Evidence Arising from Randomized Studies . . . . . . . . . . . . . 7
1.2.2 Strength of Association . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.2.3 Temporal Relationship . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.2.4 Exposure-Varying Association . . . . . . . . . . . . . . . . . . . . . . . . 9
1.2.5 Biological Plausibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2 Basic Measures of Disease Frequency . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.1 Prevalence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.1.1 Definition of Prevalence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.1.2 Applications of Prevalence Data . . . . . . . . . . . . . . . . . . . . . . 12
2.1.3 Limitation of Prevalence Measures . . . . . . . . . . . . . . . . . . . . 13
2.2 Incidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.2.1 Definitions of Incidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.2.2 Applications of Incidence Data . . . . . . . . . . . . . . . . . . . . . . . 16
2.3 Relationship Between Prevalence and Incidence . . . . . . . . . . . . . . . . 17
2.4 Stratification of Disease Frequencies by Person, Place,
and Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.4.1 Measures of Disease Frequency Stratified
by Characteristics of Person. . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.4.2 Measures of Disease Frequency Stratified
by Characteristics of Place . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.4.3 Measures of Disease Frequency Stratified
by Characteristics of Time . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
vii
viii Contents
3 General Considerations in Epidemiologic Research . . . . . . . . . . . . . . . 21
3.1 Interventional Versus Observational Study Designs . . . . . . . . . . . . . 22
3.1.1 Interventional Studies Can Isolate the Causal
Impact of Specific Treatments . . . . . . . . . . . . . . . . . . . . . . . . 22
3.1.2 Interventional Studies Are Limited to Evaluation
of Specific Treatments and Diseases . . . . . . . . . . . . . . . . . . . 24
3.1.3 The Results of Interventional Studies May
Have Limited Applicability . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.2 Study Population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.2.1 Source Population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.2.2 Exclusion Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3.2.3 Where to Find Information About the Study Population
in a Research Article . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.3 Exposure and Outcome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.3.1 Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.3.2 Measuring the Study Data . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.3.3 Where to Find Information About the Exposure
and Outcome in a Research Article . . . . . . . . . . . . . . . . . . . . 32
3.4 Internal and External Validity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.5 Summary of Common Research Study Designs . . . . . . . . . . . . . . . . 33
Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
4 Case Reports and Case Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
5 Cross-Sectional Studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
6 Cohort Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
6.1 Cohort Study Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
6.1.1 Exclusion for Prevalent Disease . . . . . . . . . . . . . . . . . . . . . . 46
6.1.2 Creation of the Cohorts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
6.1.3 Determination of the Outcome. . . . . . . . . . . . . . . . . . . . . . . . 48
6.2 Quality of the Exposure Measurements . . . . . . . . . . . . . . . . . . . . . . . 49
6.2.1 Are the Measurements Accurate? . . . . . . . . . . . . . . . . . . . . . 49
6.2.2 Are the Measurements Precise? . . . . . . . . . . . . . . . . . . . . . . . 50
6.2.3 Are the Measurements Applied Impartially
to the Study Population? . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
6.2.4 Are the Measurements Performed at the Right Time? . . . . . . 50
6.2.5 Retrospective Versus Prospective Data Collection . . . . . . . . 51
6.3 Pharmacoepidemiology Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
6.4 Analysis of Cohort Study Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
6.4.1 Calculation of Disease Incidences Among the Cohorts . . . . . 53
6.4.2 Comparison of Disease Incidences Among the Cohorts . . . . 54
6.5 Advantages of Cohort Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Contents ix
6.5.1 Ability to Discern Temporal Relationships Between
Exposure and Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
6.5.2 Ability to Study Multiple Outcomes . . . . . . . . . . . . . . . . . . . 59
6.6 Limitations of Cohort Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
6.6.1 Confounding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
6.6.2 Inefficient Design for Rare Diseases and Those
with a Long Latency Period . . . . . . . . . . . . . . . . . . . . . . . . . . 60
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
7 Case-Control Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
7.1 Case-Control Study Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
7.2 Selection of Cases and Controls . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
7.2.1 Select Case Individuals Using a Specific Definition
of the Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
7.2.2 Select Case Individuals Close to the Time of Initial
Disease Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
7.2.3 Select Control Individuals from the Same Underlying
Population as the Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
7.2.4 Select Control Individuals Who Have the Same
Opportunity to Be Counted as a Case . . . . . . . . . . . . . . . . . . 67
7.2.5 Nested Case-Control Studies . . . . . . . . . . . . . . . . . . . . . . . . . 67
7.2.6 Matching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
7.2.7 Number of Controls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
7.3 Analysis of Case Control Study Data . . . . . . . . . . . . . . . . . . . . . . . . 69
7.3.1 Concept of the Odds Ratio . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
7.3.2 Practical Calculation of the Odds Ratio . . . . . . . . . . . . . . . . . 72
7.3.3 Odds Ratios and Relative Risk . . . . . . . . . . . . . . . . . . . . . . . . 72
7.4 Advantages of Case Control Studies . . . . . . . . . . . . . . . . . . . . . . . . . 74
7.4.1 Ideal for Studying Rare Diseases and Those
with a Long Latency Period . . . . . . . . . . . . . . . . . . . . . . . . . . 74
7.4.2 Efficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
7.4.3 Evaluation of Multiple Exposures . . . . . . . . . . . . . . . . . . . . . 76
7.5 Limitations of Case Control Studies . . . . . . . . . . . . . . . . . . . . . . . . . 76
7.5.1 Confounding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
7.5.2 Requires Ascertainment of Previous Exposures
in Retrospect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
7.5.3 Inability to Directly Determine the Incidence
of Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
8 Randomized Trials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
8.1 Rationale for Randomized Trials . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
8.2 General Design of Randomized Trials . . . . . . . . . . . . . . . . . . . . . . . . 82
8.3 Trial Populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
8.3.1 Definition of the Target Condition . . . . . . . . . . . . . . . . . . . . . 83
x Contents
8.3.2 Exclusion of People Suspected to Have Difficulty
Adhering to the Study Treatments . . . . . . . . . . . . . . . . . . . . . 83
8.3.3 Exclusion of People Who Have
Comorbid Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
8.3.4 Exclusion of People Who Are Already Receiving
the Study Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
8.3.5 Exclusion for Safety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
8.3.6 Broadly Inclusive Healthcare Settings Promote
Applicability of Trial Results . . . . . . . . . . . . . . . . . . . . . . . . . 86
8.4 Interventions and Control Procedures . . . . . . . . . . . . . . . . . . . . . . . . 86
8.4.1 Intervention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
8.4.2 Control Procedures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
8.5 Outcomes of Trials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
8.5.1 Capture Important Benefits and Harms
of the Intervention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
8.5.2 Types of Outcomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
8.5.3 Mortality Outcomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
8.5.4 Surrogate Outcomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
8.5.5 Premature Termination of Trials . . . . . . . . . . . . . . . . . . . . . . 90
8.6 Procedures to Promote Internal Validity of Trials . . . . . . . . . . . . . . . 91
8.6.1 Randomization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
8.6.2 Blinding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
8.6.3 Concealment of Treatment Allocation . . . . . . . . . . . . . . . . . . 93
8.6.4 Efficacy and Effectiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
8.6.5 Trial Reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
8.7 Specific Trial Designs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
8.7.1 Factorial Trials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
8.7.2 Crossover Trials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
8.7.3 Phases of Drug Development . . . . . . . . . . . . . . . . . . . . . . . . . 98
8.8 Analysis of Clinical Trial Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
8.8.1 Measures of Effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
8.8.2 Intention-To-Treat Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 103
8.8.3 Subgroup Analyses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
8.9 Limitations of Randomized Trials . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
8.9.1 Limited External Validity (Applicability)
of the Trial Population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
8.9.2 Limited External Validity (Applicability)
of the Trial Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
8.9.3 Narrow Study Question . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
8.9.4 Randomized Design Accounts Only
for Confounding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
8.9.5 Negative Trials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
Contents xi
9 Misclassification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
9.1 Definition of Miscalculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
9.2 Non-differential Miscalculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
9.2.1 Non-differential Misclassification of the Exposure . . . . . . . . 115
9.2.2 Non-differential Misclassification of the Outcome . . . . . . . . 118
9.2.3 Summary of Non-differential Misclassification . . . . . . . . . . . 120
9.3 Differential Misclassification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
9.3.1 Differential Misclassification of the Exposure. . . . . . . . . . . . 120
9.3.2 Differential Misclassification of the Outcome . . . . . . . . . . . . 122
9.3.3 Summary of Differential Misclassification . . . . . . . . . . . . . . 122
9.4 Assessment of Misclassification in Research Articles . . . . . . . . . . . . 123
Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
10 Confounding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
10.1 Confounding Obscure Understanding of Causal Relationships . . . . 125
10.2 Evaluation of a Potential Confounding Characteristic . . . . . . . . . . . 127
10.2.1 Confounding Characteristic Associated
with the Exposure 127
10.2.2 Confounding Characteristic Associated
with the Outcome 129
10.2.3 C onfounding Characteristic Not on the Causal
Pathway of Association 129
10.2.4 Evaluation of Confounding Characteristics
in Research Articles 131
10.3 Residual Confounding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
10.4 Confounding by Indication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
10.5 Methods to Control for Confounding . . . . . . . . . . . . . . . . . . . . . . . 134
10.5.1 Method of Restriction 135
10.5.2 Method of Stratification plus Adjustment 136
10.5.3 Method of Matching 139
10.5.4 Method of Regression 141
10.6 Interpreting Results After Adjustment for Confounding . . . . . . . . . 141
Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
11 Effect Modification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
11.1 Concept of Effect Modification . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
11.2 Evaluation of Effect Modification . . . . . . . . . . . . . . . . . . . . . . . . . . 146
11.2.1 C riteria for Assessing the Presence of Effect
Modification 147
11.2.2 S tatistical Considerations for Evaluating Effect
Modification 148
11.3 Effect Modification and Confounding Are Distinct
Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
11.4 Effect Modification on the Relative and Absolute Scales . . . . . . . . 154
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155