Epidemiology - Key Concepts Flashcards
What is needed to calculate mortality rate ?
MEANINGFUL STATISTICS
A denominator population
A time frame
State some denominators
Health board
City
Hospital
Disease register
Recruited to a study
What must the denominator correspond to ?
The numerator
State the types of epidemiological study designs
2 major groups:
Observational Studies
Experimental Studies
What can observational studies be subdivided into ?
Studies that look at:
- Populations (as a whole)
- Individuals
Describe Observational Population studies
Descriptive study
“Ecological” population case series
Describe the subtypes of Observational Individual studies
Descriptive
- Looks at case studies
- Cross sectional studies
Analytic
- Cohort study
- Case-control study
State the types of Experimental study designs
Quasi-experimental studies
Randomised Clinical Trials
Describe ecological studies
The unit of study is a population.
Descriptive, retrospective (look several years back), observational (observe over a period of time)
What are ecological studies useful for ?
Useful to study signs and symptoms, look at characteristics of cases for casual hypothesis.
Important to create disease definitions, foundation for other studies.
What are case series ?
A series, often consecutive, of cases with the same disease.
Example of a case series
5 cases of Pneumocystis pneumonia and an unexpectedly high incidence of Kaposi’s sarcoma amongst young, previously healthy men in 1981 led to discovery of HIV.
Describe cross sectional study design
Sample a population
Estimate the proportion of:
- different exposures
- different signs/symptoms
- different outcomes
Use data to:
- describe prevalence/burden
- explore associations
Describe case control study deigns
Select cases with an outcome
Select controls without an outcome
Explore EXPOSURES in cases and controls.
Compare exposures in cases and controls.
Identify association.
Describe cohort study deigns
Select people without an outcome
Classify according to an exposure (subjected to an exposure)
Follow-up
- Prospective
- Retrospective
Compare RISK of disease in exposed and unexposed
Describe quasi-experimental study design
Non-Random allocation
- Intervention
- Control not required though commonly used
Researcher not in control of treatments, depends on existing groups.
Establish cause- and effect- relationship between dependent and independent variables.
Describe randomised control trial study design
Random allocation
- Intervention
- Control/comparator
Compare RISK of outcome in intervention and control groups
Objective of randomised control trials
Important in terms of describing the treatment effect.
Effect of treatment vs control
Objective of cohort study design
Good for defining:
- Cause
- Prognosis
- Incidence
of disease
Objective of quasi-experimental study design
Good for defining cause-effect relationship
Objective of case-control study design
Good for defining cause.
Is exposure the cause of outcome ?
Objective of cross-sectional study design
Good for determining prevalence of a condition/ disease.
Time-frame of RCT’s
Look towards the future
Time frame of cohort study designs
Can be:
- Prospective : future
- Retrospective : past
Time frame of quasi-experimental study designs
Look towards the future
Time frame of case control study designs
Look in the past
Time frame of cross-sectional study designs
Look in the past
X-axis
Independent variable
(exposure)
Y-axis
Dependent variable
(outcome)
Function of scatter plot
Used to test association of exposures and outcomes.
Robin Hood index
Inequality
- Higher inequality
- Higher mortality
Why is age adjusted in a scatterplot ?
Age is a confounder as it varies between states and affects mortality rates.
Function of standardisation of age
Streamlines the inequality associated mortality measurement across states.
Linear positive association
Higher the prevalence of something associated with an increase in an outcome.
e.g. Increase in inequality is associated with an increase in mortality across states.
Inverse (negative) correlation
Higher the exposure, the lower the outcome.
Crude mortality
Number of people with an outcome / the population of the area
Multiply by 1,000 (to give per 1,000)
Limitations of ‘crude’ rates
Of limited value when comparing 2 populations with different structures (i.e. confounding variables)
2 populations with the same crude rates for a particular outcome (e.g. death) will have different overall rates if the distribution of a confounder within the populations are different (e.g. age).
What is important to do when comparing mortality rates between different populations ?
It is crucial to standardise
Standardisation
A set of techniques, based on weighted averaging, used to remove as much as possible the effects of differences in age or other confounding variables in comparing 2 or more populations.
State the 2 types of standardisation
- Direct standardisation
- Indirect standardisation
Direct standardisation
Population of interest
(known: age-specific mortality rates)
Standard population
Compare the age-adjusted mortality rates
Standard Population
One in which all the confounders have been taken care of.
Indirect standardisation
Reference population
(unknown: age-specific mortality rates)
Population of interest
Compare the standardised mortality ratios of the population of interest
Compare the adjusted morality rates
Standardised mortality ratio
Number of observed deaths / Number of expected deaths
MULTIPLIED BY 100
What is the standardised mortality ratio ?
EXAM Q - MEMORISE
A weighted average of the age-specific mortality rates.
Where the weights are the proportions of persons in the corresponding age groups of a standard population.
How can you deal with confounding ?
Study Design
Data Analysis (standardisation) - free of bias
Confounding
EXAM Q - MEMORISE DEFINITION
The distortion of a measure of the effect on an exposure of an outcome due to the association of the exposure with other factors that influence the occurrence of the outcome.
Quick Definition of confounding
Unmeasured variable, which influences both the supposed cause and supposed effect.
Bias
EXAM Q - MEMORISE DEFINITION
An error in the conception and design of a study - or in the collection, analysis, interpretation, reporting, publication or review of data.
Leading to results or conclusions that are systematically (as opposed to randomly) different from the truth.
What can bias lead to ?
Wrong conclusions about:
- Disease causation
- Treatment effectiveness
Errors that arise from bias
Systematic error in:
- What data are collected
How data are:
- collected
- analysed
- interpreted
- reported
Describe the hierarchy of evidence
(TOP to BOTTOM)
Systematic reviews
RCT’s
Cohort studies
Case control studies
Case series and case reports
Editorials and expert opinions
Bradford Hill criteria for causality
EXAM Q
- Strength
- Consistency
- Specificity
- Temporality
- Biological gradient
- Plausibility
- Coherence
- Experiment
- Analogy
Consistency
A casual link is more likely if the association is observed in different studies and different sub-groups.
Specificity
A casual link is more likely when a disease is associated with one specific factor.
The more specific an association between a factor and an effect is, the bigger the probability of a casual relationship.
Temporality
A casual link is more likely if exposure to the putative cause has been shown to precede the outcome.
Biological gradient
A casual link is more likely if different levels of exposure to the putative factor lead to different risk of acquiring the outcome.
Plausibility
A casual link is more likely of a biologically plausible mechanism is likely or demonstrated.
But, knowledge of the mechanism is limited by current knowledge.
Coherence
A casual link is more likely if the observed association conforms with current knowledge.
Analogy
A casual link is more likely if an analogy exists with other diseases, species or settings.
Experiment
A casual link is very likely is removal or prevention of the putative factor leads to a reduced or non-existent risk of acquiring the outcome.
