Essential Epidemiology : an Introduction for Students and Health Professionals.

Cover
Half-title
Title page
Copyright information
Table of contents
Foreword
Preface
	Symbols
	Acknowledgements
List of Contributors
1 Epidemiology is . . .
	A case of food poisoning
	Subdisciplines of epidemiology
	On epidemics
	An historical epidemic
	The beginnings
	What does epidemiology offer?
		Description of health status of populations
		Causation
		Evaluation of interventions
		Natural history and prognosis
	What do epidemiologists do?
		Descriptive studies: person, place and time
			By ‘person’
			By ‘place’
			By ‘time’
		Analytic studies
		Intervention studies
	A natural experiment
	Conclusions
	References
	Recommended for further reading
2 How long is a piece of string? Measuring disease frequency
	What are we measuring?
	The concepts: prevalence and incidence
		Prevalence
		Incidence
		Population at risk
		The relationship between incidence and prevalence
	Measuring disease occurrence in practice: epidemiological studies
		Incidence rates versus incidence proportion
		Example
			Prevalence
			Incidence proportion (cumulative incidence)
			Incidence rate
	Measuring disease occurrence in practice: using routine data
		Crude incidence and mortality rates
		Age-specific incidence and mortality rates
		Standardised incidence and mortality rates
			A note about standard populations
		Measuring risk using routine statistics
	Other measures commonly used in public health
		Standardised incidence and mortality ratios
		The proportional (or proportionate) mortality ratio (PMR)
		The case-fatality ratio (CFR)
		Survival rate and relative survival rate
	Global health indicators
		Mortality indicators
		Life expectancy
		Disability-free life expectancy
		Years of life lost (YLL)
		Quality-adjusted life years (QALYs)
		Health-adjusted life expectancy (HALE)
		Disability-adjusted life years (DALYs)
	Summary
	Questions
	References
3 Who, what, where and when? Descriptive epidemiology
	Case reports and case series
	Vital statistics and mortality data
		Census data
		Civil registration systems
			National death registers
			Verbal autopsy
		Health and demographic surveillance systems
		Challenges in using mortality Data
	Morbidity data
		Disease registries
		Health records
		Prevalence surveys
			Demographic and health surveys
	Creative use of existing data
		Migrant studies
		Ecological or correlation studies
		E-data
	Confidentiality
	Summary
	Questions
	References
	Recommended for Further Reading
4 Healthy research: study designs for public health
	The ideal study
	Intervention studies or trials
		Randomised controlled trials (RCTs)
			Crossover trials
			n-of-1 trials
			Cluster randomised controlled trials
			Community trials
		Non-randomised designs
	Observational studies
		Cohort studies
			Historical cohort studies
			Record linkage
			Prognostic or survival studies
			Case-cohort studies
			Nested case-control studies
		Case-control studies
			Case-crossover studies
		Cross-sectional studies
		Ecological studies
	A word about ethics
	Summary
	Questions
	References
	Recommended for Further Reading
5 Why? Linking exposure and disease
	Looking for associations
	Ratio measures (relative risk)
		Rate ratios
		Risk ratios
		Prevalence ratios
		A note about relative risks
		Standardised incidence and mortality ratios
	Difference measures (attributable risk)
		Rate differences
		Risk differences
		Attributable fractions (AFs)
			Interpretation of the attributable risk
		Population attributable risks (PARs)
		Population attributable fractions (PAFs)
			Interpretation of the population attributable risk
		A word of caution regarding attributable risks
	Relative risk versus attributable risk: an example
	Case-control studies
		Relative risk in case-control studies
			Interpreting odds ratios
			Odds ratios in cross-sectional studies
		Attributable risk in case-control studies
	Looking for associations when the measures are continuous
	Summary
	Questions
	References
	Recommended for Further Reading
6 Heads or tails: the role of chance
	Random sampling error
	Statistical significance: could an apparent association have arisen by chance?
	Confidence intervals
		The relationship between p-values and confidence intervals
	Power: could we have missed a true association?
	Interpreting p-values and confidence intervals
	Statistical versus clinical significance
	Summary
	Questions
	References
	Recommended for further reading
7 All that glitters is not gold: the problem of error
	Sources of error in epidemiological studies
	Selection bias
		Some specific sources of selection bias
			Volunteers
			Low response rates
			Loss to follow-up
			Ascertainment or detection bias
			The healthy-worker effect
		Control of selection bias
		Assessing the likely effects of selection bias on the results of a study
			External comparisons
			Sensitivity analysis
			Quantitative bias analysis
	Measurement or information error
		Random error
		Systematic error
		The effects of measurement error
			Non-differential misclassification
			Differential misclassification
		Sources of measurement error
			Recall bias
			Interviewer or observer bias
		Control of measurement error
			Definitions
			Choice of instrument
			Quality control
		Assessment of measurement error
			Assessing accuracy
			Assessing precision
		Assessing the likely effects of measurement error on the results of a study
	Summary
	Questions
	References
	Recommended for further reading
8 Muddied waters: the challenge of confounding
	An example of confounding: is alcohol a risk factor for lung cancer?
	Characteristics of a confounder
	The effects of confounding
		How can we tell if an association is confounded?
		When will a possible confounder actually be a confounder in practice?
	Control of confounding
		Control of confounding through study design
			Randomisation
			Restriction
			Matching
			Does increasing the size of a study help?
		Control of confounding in data analysis
			Stratification
			Multivariable modelling
			Residual confounding
	Confounding: the bottom line
	Questions
	References
	Recommended further reading
9 Reading between the lines: reading and writing epidemiological papers
	The research question and study design
	Internal validity
		The study sample: selection bias
			Example 1: case-control studies of blood transfusion and Creutzfeldt-Jakob disease
			Example 2: a case-control study of oesophageal cancer and smoking in Australia
		Measuring disease and exposure: measurement bias
			Example 3: a case-control study of body mass index (BMI) and asthma in Mexico
		Confounding
			Example 4: a cross-sectional study of risk factors for depression in the UK
			Example 5: a cohort study of statin use and atrial fibrillation in the USA
		Interpreting results from RCTs
			Example 6: the Women’s Health Initiative (WHI) trial of menopausal hormone therapy
		Chance
		Overall internal validity
	So what? Are the results important?
	Generalisability (external validity)
	Descriptive studies
	Writing papers
	Summary: one swallow doesn’t make a summer
	Questions
	References
	Recommended for further reading
10 Who sank the boat? Association and causation
	What do we mean by a cause?
		Some definitions
	Association versus causation
	Evaluating causation
		Temporality
		Strength of association
		Consistency
		Dose-response relationships
		Biological plausibility
		Specificity
		Pulling it all together
	An example: does H. pylori cause stomach cancer?
	Conclusion
	Questions
	References
	Recommended for Further Reading
11 Assembling the building blocks: reviews and their uses
	What is a systematic review?
	Identifying the literature
		Publication and related biases
		Study inclusion and exclusion
	Appraising the literature
	Summarising the data
		Graphical display of results
		Assessing heterogeneity
		Meta-analysis
			Pooled analysis
			A word of caution
	Drawing conclusions
	Assessing the quality of a systematic review
	Making judgements in practice
		The US Preventive Services Task Force (USPSTF)
		The International Agency for Research on Cancer (IARC): monographs programme
		The World Cancer Research Fund and American Institute of Cancer Research
	The end result
	Conclusion
	Questions
	References
	Recommended for further reading
12 Surveillance: collecting health-related data for epidemiological intelligence and public health action
	The scope of surveillance
	Why conduct surveillance?
	Surveillance essentials
		Defining a case for surveillance purposes
		Collection of surveillance data
		Analysis of surveillance data
		Evaluation of surveillance systems
	Types of surveillance
		Indicator-based surveillance
		Event-based surveillance
			Digital surveillance - a new era for event-based surveillance
			Mass gathering surveillance
		Sentinel surveillance - the health status of sentinels
		Other forms of surveillance
	Summary
	Questions
	References
	Recommended for further reading
13 Outbreaks, epidemics and clusters
	Outbreaks, epidemics and clusters
	Epidemiology of infectious diseases
		A causal model
			The infectious agent
			The host
			Transmission
			The environment
	Non-infectious clusters and outbreaks
	Outbreak management and investigation
		Management of outbreaks
		Investigating outbreaks
			The identification phase
			The hypothesis-generation and testing phase
			The confirmation phase
	Evidence for causation
	Summary
	Questions
	References
	Recommended for Further Reading
14 Prevention: better than cure?
	Disease prevention in public health
	The scope for preventive medicine
		Population versus individual risk
	Strategies for prevention
		The high-risk strategy
		The mass strategy
	The population attributable fraction as a guide to prevention
		Attributable and avoidable disease
	Prevention in practice
	Evaluation of preventive interventions in practice
	A final (cautionary) word
	Questions
	References
	Recommended for Further Reading
15 Early detection: what benefits at what cost?
	Why screen?
		The disease process
		Screening versus case-finding
	The requirements of a screening programme
		The disease
		The screening test
			Test quality: sensitivity and specificity
			Test performance in practice: positive and negative predictive values
			An example - testing blood donors for HIV infection
			Parallels with clinical diagnostic tests
			The trade-off between sensitivity and specificity
		The screening programme
			Facilities required
			Treatment
			Cost
	Evaluation of a screening programme
		Health outcomes to be considered
		Potential sources of bias in the evaluation of a screening programme
			Volunteer bias
			Lead-time bias
			Length bias
		Design of a study to evaluate a screening programme
			Randomised studies
			Non-randomised studies
		The negative consequences of a screening programme
	Summary
	Questions
	References
	Recommended for Further Reading
16 Epidemiology and the public’s health
	Translating epidemiological research into practice
	Challenges
	Synthesis and integration
	Limiting error
	Improving measurement
	A final word
	References
	Recommended for further reading
Answers to questions
	Chapter 2
	Chapter 3
	Chapter 4
	Chapter 5
	Chapter 6
	Chapter 7
	Chapter 8
	Chapter 10
	Chapter 11
	Chapter 12
	Chapter 13
	Chapter 14
	Chapter 15
Appendix 1 Direct standardisation
	An example: standardising the IHD mortality rate for males in Germany to the world standard population
Appendix 2 Standard populations
	References
Appendix 3 Calculating risk and lifetime risk from routine data
	The ‘quick and dirty’ method
	The proper method
Appendix 4 Indirect standardisation
	An example: calculating the SMR for IHD in males in Brazil compared with Germany
Appendix 5 Calculating life expectancy from a life table
Appendix 6 Why the odds ratio approximates the relative risk for a rare disease
Appendix 7 Formulae for calculating confidence intervals for common epidemiological measures
Appendix 8 The Mantel-Haenszel method for calculating pooled odds ratios
	Meta-analysis
	Reference
Glossary
Index