Bradford Hill

Question 1

Temporality

Answer 1

Essential to establish a causal relation

Longitudinal CS –> measure E and O at the same time

Cause needs to come before the outcome

Question 2

Strength of association

Answer 2

The stronger the association, the more likely to be causal in the absence of known bias

Question 3

Reversability

Answer 3

Remove/absence of exposure = impact on outcome

Lowering exposure = change health outcome

Question 4

Biological gradient
(dose response)

Answer 4

More exposure = more health outcome

“The more they smoke, the more likely, they will develop lung cancer”

Question 5

Biological plausibility of association

Answer 5

Does the association make sense biologically?

Example: chemicals in tobacco that are known to promote cancers (carcinogens)

Question 6

Specificity of association

Answer 6

A cause leads to a single effect –> many disease share causes

An effect has a single cause –> disease have multiple causes

(this is the weakest criteria in the Bradford Hill model)

Question 7

Consistency of association

Answer 7

Replication of the findings by different investigations at different times and places with different methods

“There a multiple other studies with similar resutls”