Concepts of Causality

Question 1

Explain what is meant by a cause-effect relationship in an epidemiological context

Answer 1

necessary = cause always precede disease
sufficient = cause alone can lead to disease

e.g. measles virus = measles

aim is to use knowledge gained to remove, avoid or protect against harmful factors

Question 2

How can associations be present in the absence of a true cause-effect relationship?

Answer 2

study design

confounding

bias

chance

common cause

reverse causality

Question 3

How do you distinguish causal from non-causal associations?

Answer 3

epidemiological reasoning = hypothesis, analytical study, observed associations, cause-effect relationship

hierarchy of evidence

Question 4

Explain how to evaluate the strength of evidence in favour of a cause-effect relationship

Answer 4

Bradford Hill Criteria for Causality =

A causal link is more likely if:

• strength of association = high OR, IRR
• specificity of association = outcome associated with specific factor
• consistency of association = observed in diff studies
• temporal sequence = exposure before outcome
• dose response = diff levels of exposure cause diff levels of risk
• reversibility = removals leads to reduced/no risk
• coherence of theory = association conforms with current knowledge
• biological plausibility = biological plausible mechanism is likely/demonstrated
• analogy = analogy exists with other diseases