Veterinary Epidemiologic Research is a comprehensive text covering the key principles and methods used in veterinary epidemiologic research. It is written primarily for researchers and graduate students in veterinary epidemiology, but the material is equally applicable to those in related disciplines (human epidemiology, public health etc).

The first 13 chapters are devoted to issues related to the design and execution of observational studies and controlled trials.

Chapters 14 through 24 cover the statistical (multivariable) methods commonly used in the analysis of epidemiologic studies, including extensive coverage of mixed (random effects) models.

The book concludes with chapters dealing with spatial data, infectious disease epidemiology, meta-analysis, and ecologic studies. Extensive use is made of worked examples to demonstrate the principles being covered. All datasets referred to in the book are described in the book (Chapter 31) and on this website. Listings of program files (primarily Stata -do- files) used in all examples are provided on this website.

TABLE OF CONTENTS

1. INTRODUCTION AND CAUSAL CONCEPTS
1.1
Introduction
2
1.2
A brief history of multiple causation concepts
2
1.3
A brief history of scientific inference
5
1.4
Key components of epidemiologic research
8
1.5
Seeking causes
9
1.6
Models of causation
10
1.7
Counterfactual concepts of causation for a single exposure
17
1.8
Experimental versus observational evidence of causation
20
1.9
Constructing a causal diagram
21
1.10
Causal criteria
23

2. SAMPLING
2.1
Introduction
34
2.2
Non-probability sampling
34
2.3
Probability sampling
37
2.4
Simple random sample
37
2.5
Systematic random sample
38
2.6
Stratified random sample
38
2.7
Cluster sampling
38
2.8
Multistage sampling
40
2.9
Targeted (risk-based) sampling
41
2.10
Analysis of survey data
42
2.11
Sample-size determination
46
2.12
Sampling to detect disease
53

3. QUESTIONNAIRE DESIGN
3.1
Introduction
58
3.2
Designing the question
60
3.3
Open question
61
3.4
Closed question
61
3.5
Wording the question
64
3.6
Structure of questionnaires
65
3.7
Pre-testing questionnaires
66
3.8
Validation
67
3.9
Response Rate
67
3.10
Data-coding and editing
68

4. MEASURES OF DISEASE FREQUENCY
4.1
Introduction
74
4.2
Count, proportion, odds and rate
74
4.3
Incidence
75
4.4
Calculating risk
76
4.5
Calculating incidence rates
77
4.6
Relationship between risk and rate
79
4.7
Prevalence
80
4.8
Mortality statistics
81
4.9
Other measures of disease frequency
81
4.10
Standard errors and confidence intervals
83
4.11
Standardisation of risks and rates
85

5. SCREENING AND DIAGNOSTIC TESTS
5.1
Introduction
92
5.2
Attributes of the test per se
92
5.3
The ability of a test to detect disease or health
100
5.4
Predictive values
103
5.5
Interpreting test results that are measured on a continuous scale
105
5.6
Using multiple tests
111
5.7
Evaluation of diagnostic tests
114
5.8
Evaluation when there is no gold standard
116
5.9
Other considerations in test evaluation
122
5.10
Sample size requirements
123
5.11
Herd-level testing
123
5.12
Use of pooled samples
127

6. MEASURES OF ASSOCIATION
6.1
Introduction
136
6.2
Measures of association
137
6.3
Measures of effect
139
6.4
Study design and measures of association
143
6.5
Hypothesis testing and confidence intervals
143
6.6
Multivariable estimation of measures of association
148

7. INTRODUCTION TO OBSERVATIONAL STUDIES
7.1
Introduction
152
7.2
A unified approach to study design
154
7.3
Descriptive studies
156
7.4
Observational studies
157
7.5
Cross-sectional studies
158
7.6
Repeated cross-sectional versus cohort studies
162

8. COHORT STUDIES
8.1
Introduction
168
8.2
Study group
169
8.3
The exposure
171
8.4
Ensuring exposed and non-exposed groups are comparable
174
8.5
Follow-up period
175
8.6
Measuring the outcome
175
8.7
Analysis
176
8.8
Reporting of cohort studies
177

9. CASE-CONTROL STUDIES
9.1
Introduction
182
9.2
The study base
182
9.3
The case series
183
9.4
Principles of control selection
184
9.5
Selecting controls in risk-based designs
185
9.6
Selecting controls in rate-based designs
187
9.7
Other sources of controls
190
9.8
The number of controls per case
193
9.9
The number of control groups
193
9.10
Exposure and covariate assessment
193
9.11
Keeping the cases and controls comparable
193
9.12
Analysis of case-control data
194
9.13
Reporting guidelines for case-control studies
195

10. HYBRID STUDY DESIGNS
10.1
Introduction
200
10.2
Case-crossover studies
200
10.3
Case-case studies
203
10.4
Case-series studies
204
10.5
Case-cohort studies
206
10.6
Case-only studies
208
10.7
Two-stage sampling designs
209

11. CONTROLLED STUDIES
11.1
Introduction
214
11.2
Stating the objectives
215
11.3
The study group
216
11.4
Allocation of study subjects
221
11.5
Specifying the intervention
225
11.6
Masking (blinding)
225
11.7
Follow-up/compliance
226
11.8
Measuring the outcome
227
11.9
Analysis
227
11.10
Clinical trial designs for prophylaxis of communicable organisms
230
11.11
Ethical considerations
233
11.12
Reporting of clinical trials
235

12. VALIDITY IN OBSERVATIONAL STUDIES
12.1
Introduction
244
11.2
Examples of selection bias
244
12.3
Examples of selection bias
249
12.4
Reducing selection bias
254
12.5
Information bias
255
12.6
Bias from misclassification
257
12.7
Validation studies to correct misclassification
263
12.8
Measurement error
264
12.9
Errors in surrogate measures of exposure
265
12.10
The impact of information bias on sample size
266

13. CONFOUNDING: DETECTION AND CONTROL
13.1
Introduction
272
13.2
Control of confounding prior to data analysis
275
13.3
Matching on confounders
276
13.4
Matching using propensity scores
281
13.5
Detection of confounding
283
13.6
Analytic Control of Confounding
288
13.7
Other approaches to control confounding and estimate causal effects
295
13.8
Multivariable modelling to control confounding
301
13.9
Instrumental variables to control confounding
302
13.10
External adjustment and sensitivity analysis for unmeasured confounders
304
13.11
Understanding causal relationships
306
13.12
Summary of effects of extraneous variables
315

14. LINEAR REGRESSION
14.1
Introduction
324
14.2
Regression analysis
324
14.3
Hypothesis testing and effect estimation
326
14.4
Nature of the X-variables
333
14.5
Detecting highly correlated (collinear) variables
338
14.6
Detecting and modelling interaction
340
14.7
Causal interpretation of a multivariable linear model
341
14.8
Evaluating the least squares model
344
14.9
Evaluating the major assumptions
349
14.10
Assessment of individual observations
356
14.11
Time-series data
360

15. MODEL-BUILDING STRATEGIES
15.1
Introduction
366
15.2
Steps in building a model
367
15.3
Building a causal model
367
15.4
Reducing the number of predictors
368
15.5
The problem of missing values
374
15.6
Effects of continuous predictors
375
15.7
Identifying interaction terms of interest
381
15.8
Building the model
383
15.9
Evaluate the reliability of the model
388
15.10
Presenting the results
390

16. LOGISTIC REGRESSION
16.1
Introduction
396
16.2
The logistic model
396
16.3
Odds and odds ratios
397
16.4
Fitting a logistic regression model
398
16.5
Assumptions in logistic regression
399
16.6
Likelihood ratio statistics
400
16.7
Wald tests
401
16.8
Interpretation of coefficients
402
16.9
Assessing interaction and confounding
405
16.10
Model-building
408
16.11
Generalised linear models
408
16.12
Evaluating logistic regression models
410
16.13
Sample size considerations
421
16.14
Exact logistic regression
421
16.15
Conditional logistic regression for matched studies
422

17. MODELLING ORDINAL AND MULTINOMIAL DATA
17.1
Introduction
428
17.2
Overview of models
429
17.3
Multinomial logistic regression
431
17.4
Modelling ordinal data
436
17.5
Proportional odds model (constrained cumulative logit model)
437
17.6
Adjacent-category model
441
17.7
Continuation-ratio model
443

18. MODELLING COUNT AND RATE DATA
18.1
Introduction
446
18.2
The Poisson distribution
447
18.3
Poisson regression model
448
18.4
Interpretation of coefficients
449
18.5
Evaluating Poisson regression models
451
18.6
Negative binomial regression
454
18.7
Problems with zero counts
461

19. MODELLING SURVIVAL DATA
19.1
Introduction
468
19.2
Non-parametric analyses
473
19.3
Actuarial life tables
473
19.4
Kaplan-Meier estimate of survivor function
475
19.5
Nelson-Aalen estimate of cumulative hazard
478
19.6
Statistical inference in non-parametric analyses
479
19.7
Survivor, failure and hazard functions
480
19.8
Semi-parametric analyses
485
19.9
Parametric models
503
19.10
Accelerated failure time models
507
19.11
Frailty models and clustering
510
19.12
Multiple outcome event data
517
19.13
Discrete-time survival analysis
518
19.14
Sample sizes for survival analyses
522

20. INTRODUCTION TO CLUSTERED DATA
20.1
Introduction
530
20.2
Clustering arising from the data structure
530
20.3
Effects of clustering
536
20.4
Simulation studies on the impact of clustering
540
20.5
Introduction to methods for dealing with clustering
542

21. MIXED MODELS FOR CONTINUOUS DATA
21.1
Introduction
554
21.2
Linear mixed model
555
21.3
Random slopes
560
21.4
Contextual effects
564
21.5
Statistical analysis of linear mixed models
565

22. MIXED MODELS FOR DISCRETE DATA
22.1
Introduction
580
22.2
Logistic regression with random effects
580
22.3
Poisson regression with random effects
584
22.4
Generalised linear mixed model
587
22.5
Statistical analysis of GLMMs
593
22.6
Summary remarks on analysis of discrete clustered data
603

23. REPEATED MEASURES DATA
23.1
Introduction to repeated measures data
608
23.2
Univariate and multivariate approaches to repeated measures data
611
23.3
Linear mixed models with correlation structure
616
23.4
Mixed models for discrete repeated measures data
624
23.5
Generalised estimating equations
627

24. INTRODUCTION TO BAYESIAN ANALYSIS
24.1
Introduction
638
24.2
Bayesian analysis
638
24.3
Markov chain Monte Carlo (MCMC) estimation
642
24.4
Statistical analysis based on MCMC estimation
647
24.5
Extensions of Bayesian and MCMC Modelling
651

25. ANALYSIS OF SPATIAL DATA: INTRODUCTION AND VISUALISATION
25.1
Introduction
664
25.2
Spatial data
664
25.3
Spatial data analysis
667
25.4
Additional topics
673

26. ANALYSIS OF SPATIAL DATA
26.1
Introduction
680
26.2
Issues specific to statistical analysis of spatial data
680
26.3
Exploratory spatial analysis
682
26.4
Global spatial clustering
690
26.5
Localised spatial cluster detection
697
26.6
Space-time association
700
26.7
Modelling
704

27. CONCEPTS OF INFECTIOUS DISEASE EPIDEMIOLOGY
27.1
Introduction
716
27.2
Infection vs disease
718
27.3
Transmission
719
27.4
Mathematical modelling of infectious disease transmission
721
27.5
Estimating R0 and other infectious disease parameters
725

28. SYSTEMATIC REVIEWS AND META-ANALYSIS
28.1
Introduction
740
28.2
Narrative reviews
740
28.3
Systematic Reviews
741
28.4
Meta-analysis – Introduction
745
28.5
Fixed- and random-effects models
746
28.6
Presentation of results
749
28.7
Heterogeneity
750
28.8
Publication bias
758
28.9
Influential studies
760
28.10
Outcome scales and data issues
760
28.11
Meta-analysis of observational studies
764
28.12
Meta-analysis of diagnostic tests
766
28.13
Use of meta-analysis
766

29. ECOLOGICAL AND GROUP-LEVEL STUDIES
29.1
Introduction
774
29.2
Rationale for group level studies
774
29.3
Types of ecologic variable
775
29.4
Issues related to modelling approaches in ecologic studies
776
29.5
Issues related to inferences
778
29.6
Sources of ecologic bias
778
29.7
Non-ecologic group-level studies
782

30. A STRUCTURED APPROACH TO DATA ANALYSIS
30.1
Introduction
790
30.2
Data-collection sheets
790
30.3
Data coding
791
30.4
Data entry
791
30.5
Keeping track of files
792
30.6
Keeping track of variables
792
30.7
Program mode versus interactive processing
793
30.8
Data-editing
794
30.9
Data verification
795
30.10
Data processing—outcome variable(s)
795
30.11
Data processing—predictor variables
796
30.12
Data processing—multilevel data
796
30.13
Unconditional associations
796
30.14
Keeping track of your analyses
797