Lecture 29: Association and Confounding Flashcards
What is an association?
Does the exposure increase or decrease the likelihood of the outcome?
Outcomes can have multiple _______
causes
Define a cause
an event, condition or characteristic [or a combination of these factors] that plays an essential role in producing an occurrence of the disease
As per the causal pie model, many different exposures are sufficient to cause the
outcome
In the causal pie model, we call the whole pie a ______ _______ of the outcome
sufficient cause (SC)
In the causal pie model, each exposure is a ______ of the sufficient cause and therefore we call each of the exposures a
component
component cause
Give an example of a component of lung cancer?
smoking is one component of a sufficient cause for lung cancer
How can we prevent disease by eliminating exposures?
If we eliminate exposures, then the sufficient causes that the exposures are part of can therefore not causes the disease
What is a necessary cause?
A component cause which is necessary for the disease to occur and so must be part of every sufficient cause
What makes a valid association?
If the outcome is unlikely to be due to chance, bias or confounding
The presence of a valid statistical _______ in no way implies _______
association
causality
How can we determine causality?
We need to consider a series of guidelines and then make a judgement based on the totality of the evidence
What is the acronym to remember the 7 guidelines to causality?
BEST CDS
What are the 7 guidelines to causality?
Biological plausibility Experimental evidence Specificity Temporal sequence Consistency Dose-responsive relationship Strength of association
What is meant by biological plausibility?
Is there a plausible biological mechanism for the association?
What is meant by experimental evidence?
Is there evidence from human RCTs or animal experiments
What is meant by specificity?
is the exposure specifically associated with a particular outcome but not others?
What is meant by temporal sequencing?
Do we know whether it was the outcome or exposure that came first?
What is meant by consistency?
are the findings consistent with findings from other studies
What is meant by dose-response relationship?
Does the risk of the outcome change with increasing or decreasing amounts of exposure?
When do does-response relationships not apply?
is the relationship is not linear or if you are looking at binary (smoking or not smoking) instead of levels of exposure (number of cigarettes)
What is meant by the strength of the association?
The stronger the association, the less likely it is to be due to confounding or bias
