Chapter 14: From Association to Causation Flashcards
What is causality?
Causality is the relationship between two events or variables.
What study design does epidemiologist use?
They use natural experiments.
Why is natural experiments ethical?
Natural experiments are ethical because you are observing someone who presently has the disease how they are affected by the disease rather than having to affect people with disease just to study how it affects them.
What is real association?
Demonstrates that there is a causal association: association of exposure and the exposure that includes development of the disease.
What is spurious association?
Demonstrates that is noncausal association between exposure, due to confounding factors that contribute to the disease.
True or false are causal pathways direct and indirect?
True
What are direct causal pathways?
A factor that directly causes a disease without any intermediate steps or factors.
What are indirect pathways?
A factor that causes a disease but only through an intermediate step or steps.
If a relationship is causal, what are the four types of causal relationship that are possible?
- Necessary and sufficient
- Necessary but not sufficient
- Sufficient but not necessary
- Sufficient not necessary
What is necessary and sufficient?
The factor is both necessary and sufficient for producing the disease.
(without the factor present the disease will never develop/with the factor present the disease will always develop)
True or false, necessary and sufficient often happen
False, necessary and sufficient rarely happen because you can’t determine who will and will not get the disease.
What is necessary but not sufficient?
Factor is necessary but each factor alone is not sufficient to cause the disease. Multiple factors are required, usually in a very specific order. (need exposure before you get agent)
What is sufficient but not necessary?
One factor can produce the disease but so can other factor. These other factors can act alone to cause the disease. (the criterion of the sufficient is rarely met by a single factor)
If our statistics tell us there is a relationship between our exposure and outcome
We have to figure out if the association is causal.
What is neither necessary nor sufficient?
The most complex of all the models but it most accurately represents causal relationships.
Which type of causal relationships rarely occurs?
Necessary and sufficient.
Which type of causal relationships is the most complex?
Neither necessary not sufficient.
Temporal relationship
The outcome must occur after the exposure (exposure before disease)
Length of the interval between exposure and disease occurrence.
Strength of association
The statistical association between exposure and an outcome.
Dose-response relationshiop
The dose of an exposure increases, the risk of disease also increases.
Replication of the findings
If a relationship is causal, we should find it consistently in different studies and different populations.
Biologic plausibility
The association and relationship observed should be biologically possible. (Back up by biologic knowledge)
Consideration of alternate explanations
Investigators should examine potential confounding and external factors to figure out if the observed association could be mediated and explain away by any of those factors.
Cessation of exposure
If a factor is associated with an outcome, then disease incidence and prevalence should decline if we eliminate the factor.
