Lecture 21: Causality Flashcards
What is the definition of the public/population health model?
“To provide the maximum benefit for the largest number of people- at the same time reducing inequities in the distribution of health and wellbeing”
What is individual healthcare concerned with?
The treatment or restoration of health for the individual by a clinician (so it is reactive)
What is population health concerned with?
The health of groups of individuals, in the context of their environment (a comprehensive, proactive population approach to clinical practice)
Does epidemiology determine the cause of disease in an individual?
No
What does epidemiology determine?
The relationship or association between a given exposure and disease in a population
Can preventative action be taken before the cause is identified? What is an example?
Yes, for example, James Lind’s treatment of scurvy (1747)
What are two frameworks to determine causality?
- Bradford Hill Criteria “aids to thought” 1965
2. Rothman’s Causal Pies
What are the seven components of Bradford Hill’s Criteria?
- Temporality
- Strength of Association
- Consistency of Association
- Biological Gradient (dose-response)
- Biological Plausibility of Association
- Specifity of Association
- Reversibility
JUDGEMENT
What is the weakest component of Bradford Hill’s criteria?
- Specifity of association, because a single cause often leads to multiple effects
Define temporality
The cause comes before the disease (essential to establish a causal relation)
Eg smoking comes before lung cancer deaths
Define strength of association
The stronger the association, the more likely to be causal in the absence of known bias (selection, information and confounding)
Eg smoking/lung cancer risk ratio >10
Define consistency of association
Replication of the fi dings by different investigations, at different times, in different places, with different methods
Eg multiple studies showed similar results
Define biological gradient (dose exposure)
Incremental change in disease rates in conjunction with corresponding changes in exposure
Eg increased smoking per day showed higher risk of lung cancer
Define biological plausibility of association
Does the association make sense biologically?
Eg chemicals in tobacco (carcinogens) are known to promote cancer
Define specificity of association
A single cause leads to a single effect (often a single cause has multiple effects, however)
Eg smoking has multiple outcomes
- Define reversibility
The demonstration they under controlled conditions, changing the exposure causes a change in outcome
Eg reduced risk in lung cancer after quitting smoking
Does causal phenomena usually equal 1:1?
No, causal phenomena is usually complex and exposure-outcome relationships are not usually 1:1
What is the definition of a cause of disease?
An event, condition, characteristic (or combination of these factors) which play an essential role in producing the disease
What is the “sufficient” cause of a causal pie?
Is a factor/s that will inevitably produce the specific disease (ie causal mechanism or 1 pie)
What is a “component” cause of a causal pie?
A factor that contributes towards a disease causation but is not sufficient to cause a disease on its own
Eg poor sanitation in regards to TB disease (a slice of pie that’s not necessary)
What is a “necessary” cause of a causal pie?
A factor (or component cause) that must be present for a specific cause to occur
Eg the TB bug
Can you intervene at any number of points in the pie?
YES
Knowledge of the complete pathway is not w prerequisite for introducing preventative measures
Ie TB prevention occurred before its cause was discovered
What do we use association and several other factors for?
To infer causation and to intervene to prevent disease
Is a necessary cause still a component cause?
Yes, because it’s still part of the pie
What do epidemiologists look for links between?
Exposure and outcome
