Forewords
Preface
Acknowledgements
1 INTRODUCTION AND CAUSAL CONCEPTS
1.1 Introduction
1.2 A brief history of multiple causation concepts
1.3 A brief history of scientific inference
1.4 Key components of epidemiologic research
1.5 Seeking causes
1.6 Models of causation
1.7 Counterfactual concepts of causation for a single exposure
1.8 Experimental versus observational evidence of causation
1.9 Constructing a causal diagram
1.10 Causal criteria
2 SAMPLING
2.1 Introduction
2.2 Non-probability sampling
2.3 Probability sampling
2.4 Simple random sample
2.5 Systematic random sample
2.6 Stratified random sample
2.7 Cluster sampling
2.8 Multistage sampling
2.9 Targeted (risk-based) sampling
2.10 Analysis of survey data
2.11 Sample-size determination
2.12 Sampling to detect disease
3 QUESTIONNAIRE DESIGN
3.1 Introduction
3.2 Designing the question
3.3 Open question
3.4 Closed question
3.5 Wording the question
3.6 Structure of questionnaires
3.7 Pre-testing questionnaires
3.8 Validation
3.9 Response rate
3.10 Data-coding and editing
4 MEASURES OF DISEASE FREQUENCY
4.1 Introduction
4.2 Counts, proportions, odds, and rates
4.3 Incidence
4.4 Calculating risk
4.5 Calculating incidence rates
4.6 Relationship between risk and rate
4.7 Prevalence
4.8 Mortality statistics
4.9 Other measures of disease frequency
4.10 Standard errors and confidence intervals
4.11 Standardisation of risks and rates
5 SCREENING AND DIAGNOSTIC TESTS
5.1 Introduction
5.2 Attributes of the test per se
5.3 The ability of a test to detect disease or health
5.4 Predictive values
5.5 Interpreting test results that are measured on a continuous scale
5.6 Using multiple tests
5.7 Evaluation of diagnostic tests
5.8 Evaluation when there is no gold standard
5.9 Other considerations in test evaluation
5.10 Sample size requirements
5.11 Group-level testing
5.12 Use of pooled samples
6 MEASURES OF ASSOCIATION
6.1 Introduction
6.2 Measures of association
6.3 Measures of effect
6.4 Study design and measures of association
6.5 Hypothesis testing and confidence intervals
6.6 Multivariable estimation of measures of association
7 INTRODUCTION TO OBSERVATIONAL STUDIES
7.1 Introduction
7.2 A unified approach to study design
7.3 Descriptive studies
7.4 Observational studies
7.5 Cross-sectional studies
7.6 Estimating incidence from one or more cross-sectional studies
7.7 Inferential limitations of cross-sectional studies
7.8 Repeated cross-sectional versus cohort studies
7.9 Reporting of observational studies
8 COHORT STUDIES
8.1 Introduction
8.2 Selecting the study group
8.3 The exposure
8.4 Disease as exposure
8.5 Ensuring exposed and non-exposed groups are comparable
8.6 Follow-up period
8.7 Measuring the outcome
8.8 Analysis
8.9 Reporting of cohort studies
9 CASE–CONTROL STUDIES
9.1 Introduction
9.2 The study base
9.3 The case series
9.4 Principles of control selection
9.5 Selecting controls and data layout in risk-based designs
9.6 Sampling controls and data layout in rate-based designs
9.7 Other sources of controls
9.8 The number of controls per case
9.9 The number of control groups
9.10 Exposure and covariate assessment
9.11 Keeping the cases and controls comparable
9.12 Analysis of case–control data
9.13 Reporting guidelines for case–control data
10 HYBRID STUDY DESIGNS
10.1 Introduction
10.2 Case-crossover studies
10.3 Case–case studies
10.4 Case-case–control studies
10.5 Case–series studies
10.6 Case–cohort studies
10.7 Case-only studies
10.8 Two-stage sampling designs
11 CONTROLLED STUDIES
11.1 Introduction
11.2 Background, objectives, and summary trial design
11.3 Participants: the study group
11.4 Specifying the intervention
11.5 Measuring the outcome
11.6 Sample size
11.7 Allocation of study subjects
11.8 Follow-up/compliance
11.9 Statistical methods and analysis
11.10 Conclusions
11.11 Clinical trial designs for prophylaxis of communicable organisms
11.12 Reporting of clinical trials
12 VALIDITY IN OBSERVATIONAL STUDIES
12.1 Introduction
12.2 Selection bias
12.3 Examples of selection bias
12.4 Reducing selection bias
12.5 Information bias
12.6 Bias from misclassification
12.7 Validation studies to correct misclassification
12.8 Measurement error
12.9 Errors in surrogate measures of exposure
12.10 The impact of information bias on sample size
13 CONFOUNDING: DETECTION AND CONTROL
13.1 Introduction
13.2 Control of confounding prior to data analysis
13.3 Matching on confounders
13.4 Detection of confounding
13.5 Analytic control of confounding
13.6 Multivariable modelling to control confounding
13.7 Other approaches to control confounding and estimate causal effects
13.8 Propensity scores for controlling confounding
13.9 External adjustment and sensitivity analysis for unmeasured confounders
13.10 Understanding causal relationships
13.11 Summary of effects of extraneous variables
14 LINEAR REGRESSION
14.1 Introduction
14.2 Regression analysis
14.3 Hypothesis testing and effect estimation
14.4 Nature of the X-variables
14.5 Detecting highly correlated (collinear) variables
14.6 Detecting and modelling interaction
14.7 Causal interpretation of a multivariable linear model
14.8 Evaluating the least squares model
14.9 Evaluating the major assumptions
14.10 Assessment of individual observations
14.11 Time-series data
15 MODEL-BUILDING STRATEGIES
15.1 Introduction
15.2 Steps in building a model
15.3 Building a causal model
15.4 Reducing the number of predictors
15.5 The problem of missing values
15.6 Effects of continuous predictors
15.7 Identifying interaction terms of interest
15.8 Building the model
15.9 Evaluate the reliability of the model
15.10 Presenting the results
16 LOGISTIC REGRESSION
16.1 Introduction
16.2 The logistic model
16.3 Odds and odds ratios
16.4 Fitting a logistic regression model
16.5 Assumptions in logistic regression
16.6 Likelihood ratio statistics
16.7 Wald tests
16.8 Interpretation of coefficients
16.9 Assessing interaction and confounding
16.10 Model-building
16.11 Generalised linear models
16.12 Evaluating logistic regression models
16.13 Sample size considerations
16.14 Exact logistic regression
16.15 Conditional logistic regression for matched studies
17 MODELLING ORDINAL AND MULTINOMIAL DATA
17.1 Introduction
17.2 Overview of models
17.3 Multinomial logistic regression
17.4 Modelling ordinal data
17.5 Proportional odds model (constrained cumulative logit model)
17.6 Adjacent-category model
17.7 Continuation-ratio model
18 MODELLING COUNT AND RATE DATA
18.1 Introduction
18.2 The Poisson distribution
18.3 Poisson regression model
18.4 Interpretation of coefficients
18.5 Evaluating Poisson regression models
18.6 Negative binomial regression
18.7 Problems with zero counts
19 MODELLING SURVIVAL DATA
19.1 Introduction
19.2 Non-parametric analyses
19.3 Actuarial life tables
19.4 Kaplan–Meier estimate of survivor function
19.5 Nelson–Aalen estimate of cumulative hazard
19.6 Statistical inference in non-parametric analyses
19.7 Survivor, failure, and hazard functions
19.8 Semi-parametric analyses
19.9 Parametric models
19.10 Accelerated failure time models
19.11 Frailty models and clustering
19.12 Multiple outcome event data
19.13 Discrete-time survival analysis
19.14 Samples sizes for survival analyses
20 INTRODUCTION TO CLUSTERED DATA
20.1 Introduction
20.2 Clustering arising from the data structure
20.3 Effects of clustering
20.4 Simulation studies on the impact of clustering
20.5 Introduction to methods for dealing with clustering
21 MIXED MODELS FOR CONTINUOUS DATA
21.1 Introduction
21.2 Linear mixed model
21.3 Random slopes
21.4 Contextual effects
21.5 Statistical analysis of linear mixed models
22 MIXED MODELS FOR DISCRETE DATA
22.1 Introduction
22.2 Logistic regression with random effects
22.3 Poisson regression with random effects
22.4 Generalised linear mixed model
22.5 Statistical analysis of GLMMs
22.6 Summary remarks on analysis of discrete clustered data
23 REPEATED MEASURES DATA
23.1 Introduction to repeated measures data
23.2 Univariate and multivariate approaches to repeated measures data
23.3 Linear mixed models with correlation structure
23.4 Mixed models for discrete repeated measures data
23.5 Generalised estimating equations
24 INTRODUCTION TO BAYESIAN ANALYSIS
24.1 Introduction
24.2 Bayesian analysis
24.3 Markov chain Monte Carlo estimation
24.4 Statistical analysis based on MCMC estimation
24.5 Extensions of Bayesian and MCMC modelling
25 ANALYSIS OF SPATIAL DATA: INTRODUCTION AND VISUALISATION
25.1 Introduction
25.2 Spatial data
25.3 Spatial data analysis
25.4 Additional topics
26 ANALYSIS OF SPATIAL DATA
26.1 Introduction
26.2 Issues specific to statistical analysis of spatial data
26.3 Exploratory spatial analysis
26.4 Global spatial clustering
26.5 Localised spatial cluster detection
26.6 Space-time association
26.7 Modelling
27 CONCEPTS OF INFECTIOUS DISEASE EPIDEMIOLOGY
27.1 Introduction
27.2 Infection vs disease
27.3 Transmission
27.4 Mathematical modelling of infectious disease transmission
27.5 Methods of control of infectious disease
27.6 Estimating R0 and other parameters
27.7 Developing more complex models
27.8 Using models
27.9 Summary
28 SYSTEMATIC REVIEWS AND META-ANALYSIS
28.1 Introduction
28.2 Narrative reviews
28.3 Systematic reviews
28.4 Meta-analysis—introduction
28.5 Fixed- and random-effects models
28.6 Presentation of results
28.7 Heterogeneity
28.8 Publication bias
28.9 Influential studies
28.10 Outcome scales and data issues
28.11 Meta-analysis of observational studies
28.12 Meta-analysis of diagnostic tests
28.13 Use of meta-analysis
29 ECOLOGICAL AND GROUP-LEVEL STUDIES
29.1 Introduction
29.2 Rationale for group level studies
29.3 Types of ecologic variable
29.4 Issues related to modelling approaches in ecologic studies
29.5 The linear model in the context of ecologic studies
29.6 Issues related to inferences
29.7 Sources of ecologic bias
29.8 Analysis of ecologic data
29.9 Non-ecologic group-level studies
30 A STRUCTURED APPROACH TO DATA ANALYSIS
30.1 Introduction
30.2 Data-collection sheets
30.3 Data coding
30.4 Data entry
30.5 Keeping track of files
30.6 Keeping track of variables
30.7 Program mode versus interactive processing
30.8 Data-editing
30.9 Data verification
30.10 Data processing—outcome variables
30.11 Data processing—predictor variables
30.12 Data processing—multilevel data
30.13 Unconditional associations
30.14 Keeping track of your analyses
31 DESCRIPTION OF DATASETS
