This book covers classic epidemiological designs that use a reference/control group, including case-control, case-cohort, nested case-control and variations of these designs, such as stratified and two-stage designs. It presents a unified view of these sampling designs as representations of an underlying cohort or target population of interest. This enables various extended designs to be introduced and analysed with a similar approach: extreme sampling on the outcome (extreme case-control design) or on the exposure (exposure-enriched, exposure-density, countermatched), designs that re-use prior controls and augmentation sampling designs. Further extensions exploit aggregate data for efficient cluster sampling, accommodate time-varying exposures and combine matched and unmatched controls. Self-controlled designs, including case-crossover, self-controlled case series and exposure-crossover, are also presented. The test-negative design for vaccine studies and the use of negative controls for bias assessment are introduced and discussed.
This book is intended for graduate students in biostatistics, epidemiology and related disciplines, or for health researchers and data analysts interested in extending their knowledge of study design and data analysis skills.
This book
- Bridges the gap between epidemiology and the more mathematically oriented biostatistics books.
- Assembles the wealth of epidemiological knowledge about observational study designs that is scattered over several decades of scientific publications.
- Illustrates the performance of methods in real research applications.
- Provides guidelines for implementation in standard software packages (Stata, R).
- Includes numerous exercises, covering simple mathematical proofs, consideration of proposed or published designs, and practical data analysis.
Author(s): Marie Reilly
Series: Chapman & Hall/CRC Biostatistics Series
Publisher: CRC Press/Chapman & Hall
Year: 2023
Language: English
Pages: 471
City: Boca Raton
Cover
Half Title
Series Page
Title Page
Copyright Page
Table of Contents
Preface
List of Figures
List of Tables
List of Abbreviations
1. Classic Epidemiological Designs
1.1. Review of Measures of Disease Occurrence and Risk
1.1.1. Prevalence
1.1.2. Incidence
1.1.3. Relative measures of disease occurrence: risks and ratios
1.2. Study Population, Study Base
1.2.1. Primary and secondary study base
1.3. Sampling Designs
1.3.1. Cross-sectional study (survey)
1.3.2. Cohort study
1.3.3. Case-control study
1.3.4. Comparison of cohort and case-control design
1.4. Sources of Bias
1.4.1. Sampling bias
1.4.2. Response bias
1.4.3. Measurement bias (information bias)
1.4.4. Time-related bias
1.4.5. Confounding bias
1.5. Which Design?
1.6. Electronic Data Resources
1.7. Exercises
2. From Tables to Logistic Regression Models
2.1. Estimating RR or OR from 2-by-2 Tables
2.2. Sampling Distribution of a RR or OR
2.3. Stratification and Confounding
2.3.1. Interaction (effect modification)
2.3.2. Confounding of a risk estimate
2.3.3. Mantel-Haenszel odds ratio
2.4. Association, Homogeneity and Trend
2.4.1. Chi-squared test of association
2.4.2. Test of association in paired data: McNemar’s Test
2.4.3. Test of homogeneity
2.4.4. Effect modification of an exposure-disease relationship
2.4.5. Dose-response, test for trend
2.5. Logistic Regression
2.5.1. Adjusted OR from logistic regression
2.5.2. Logistic regression model with interaction term
2.5.3. Modelling a linear effect of a continuous variable
2.5.4. Multivariable logistic regression
2.5.5. From prospective to retrospective models
2.5.6. Matched data
2.6. Individually Matched Data
2.6.1. OR from paired data
2.6.2. Conditional logistic regression
2.7. Exercises
3. Extensions to Classic Epidemiological Studies
3.1. Missing/Incomplete Data
3.1.1. Intentionally missing data
3.2. Two-stage Studies
3.2.1. Statistical explanation
3.2.2. Two-stage illustration: Framingham data
3.2.3. Computation of sampling fractions
3.2.4. Two-stage survey of H. Pylori in school-children
3.2.5. Unintentional two-stage design
3.2.6. Summary of two-stage studies
3.3. Secondary Analysis of Case-control Data
3.3.1. What standard analysis is valid/invalid, and when?
3.3.2. Two-stage approach to reusing case-control data
3.4. Reusing Controls from Case-control Data
3.5. Exercises
4. Including Time: Cox Regression and Related
4.1. Inclusive, Exclusive, Concurrent Sampling
4.2. Time-to-event Data
4.2.1. Hazard and survival
4.2.2. Proportional hazards
4.3. Cox Regression
4.3.1. Adjusted hazard ratio
4.3.2. Stratified Cox regression
4.4. Nested Case-control Sampling
4.4.1. Illustration: Cox and conditional logistic regression
4.5. Case-cohort Sampling
4.5.1. Approaches to case-cohort analysis
4.5.2. Illustration: nested case-control and case-cohort designs
4.6. Comparison of Risk Sets
4.7. Comparison of Nested Case-control and Case-cohort
4.8. Exercises
5. Estimates Available from Standard Designs
5.1. Measures of Exposure Impact
5.1.1. Number needed to be exposed, NNE
5.1.2. Adjusted NNE
5.1.3. Attributable risks and impact numbers
5.1.4. Confidence intervals for measures of impact
5.2. Estimating RR from Logistic Regression
5.2.1. Doubling the cases in cohort or cross-sectional data
5.2.2. Mantel-Haenszel OR after doubling the cases
5.2.3. Adjusted RR from logistic regression
5.2.4. Estimating RR from case-control ddata
5.3. Risk of Transient Effects Using a ‘Quasi-Cohort’
5.4. Modelling Complex Exposure measurements
5.4.1. Estimating several aspects of the same exposure
5.4.2. Recoding the different measures of exposure
5.4.3. Coding interactions
5.4.4. Illustration of analysis of complex exposure
5.5. Exercises
6. Estimates from Matched and Nested Designs
6.1. Matched Designs
6.1.1. Matched case-control studies
6.2. Ignoring or Breaking the Matching
6.2.1. Ignoring the matching in cohort studies
6.2.2. Unconditional analysis of matched cohort data
6.2.3. Unconditional analysis of matched case-control data
6.2.4. Ignoring the matching in case-control analysis
6.3. Breaking the Time Matching
6.3.1. Kaplan-Meier type weights
6.3.2. Data necessary for reweighting
6.3.3. Illustration of weighted risk sets
6.4. Weighted Cox Likelihood
6.5. Illustration of Weighted Analysis of Nested Case-control Data
6.5.1. Estimation of HR from nested case-control data
6.5.2. Estimation of absolute risk from case-control data
6.6. Advantages of Breaking the Matching
6.6.1. Illustration of breaking the (over)matching
6.6.2. Further uses of reweighted case-control data
6.7. Exercises
7. Reusing Case-Control Data
7.1. Using Classic Case-control Data for new Outcomes
7.1.1. Explanatory variable as outcome
7.1.2. Reusing controls for a new outcome
7.2. Reusing Nested Case-control Data
7.2.1. Illustration in a realistic cohort
7.2.2. New outcome in restricted follow-up time
7.2.3. Application to study of breast cancer
7.2.4. Supplementing controls
7.2.5. Combining two nested case-control studies
7.3. Value of Reused Data
7.4. Analysis of Subgroups from Nested Case-control Data
7.4.1. Subgroups defined by outcome
7.5. Conclusion
7.6. Exercises
8. More Complex Designs
8.1. Case-cohort Design as a Two-stage Study
8.1.1. Stratified case-cohort
8.1.2. Post-stratification
8.2. Optimal Two-stage Designs for Binary Outcome
8.2.1. Optimal sampling
8.3. Efficient Sampling for a Time-to-event Outcome
8.3.1. Optimal selection to improve efficiency
8.4. Exposure-related Sampling
8.4.1. Counter-matching
8.4.2. Exposure enriched case-control study
8.5. Extreme Case-Control Design
8.5.1. Illustration
8.5.2. Data application
8.5.3. Power of ECC vs. NCC
8.5.4. Variations of extreme sampling
8.6. Exercises
9. More Complex Data Structures
9.1. Clustered Data
9.1.1. Two-stage design using aggregate cluster data
9.1.2. Efficient adjustment for cluster interventions
9.1.3. Case-control sampling within clusters
9.2. Two-stage Augmentation Sampling
9.3. Time-dependent Exposure
9.3.1. Exposure density sampling
9.3.2. Nested case-control sampling
9.3.3. Detailed history of exposure in case-control studies
9.4. Time-varying Associations
9.4.1. Time-varying associations and case-control designs
9.5. Combining Matched and Unmatched Case-control Data
9.5.1. Joint likelihood of matched and unmatched data
9.5.2. Missing indicator method
9.5.3. Cases with matched and unmatched controls
9.6. Exercises
10. Other Controlled Epidemiological Studies
10.1. Self-controlled designs
10.1.1. Case-crossover design
10.1.2. Extensions to the case-crossover design
10.1.3. Self-controlled case series
10.1.4. Exposure-crossover design
10.2. Test-negative Design
10.2.1. Bias in test-negative designs
10.2.2. Cluster-randomised test-negative design
10.3. Negative Controls
10.3.1. Confounding bias
10.3.2. Selection bias
10.3.3. Measurement error bias
10.3.4. Negative self-control
10.4. Active Comparators
10.4.1. Self-controlled active comparator
10.5. Summary
Bibliography
Index
