Models of causality II

Question 1

Strength of association

Answer 1

An association is the effect of E in the disease

Ex: E is in 10% of population. If the population with the rest of the contributing causes is 90% (A+B+C+D is 90%) = probability of developing the disease is 9% (10% of 90%)

Even though the prevalence of E is the =, the incidence of the disease depends also on the prevalence of the rest of the contributing factors (we need all of them to develop the disease)

Question 2

Interactions

Answer 2

If the probability of developing the disease only having B is 25% and the probability with only F is 25% as well, the probability of developing the disease when both are present is 100% (given that the rest of contributing factors are present)

Question 3

Attributable fraction (Excess or risk fraction)

Answer 3

% of disease that could be eliminated if we removed the factor

(proportion of cases aiributed to each cause. We must assure the interaction is causal)

Question 4

Induction period

Answer 4

Time between initiation of exposure and initiation of the disease (it is different in different exposures or diseases)

In the induction period, we must be exposed to all the other factors to be able to complete the causal mechanism

Question 5

Counterfactual model

Answer 5

Hypothetical model that tries to find what would have happened if the exposure changed

(what would have happened if the same person with the same headache did not take the pills, compare actual outcome with counterfactual outcome. How would this have affected the headache?)

Question 6

Falsifiability (Popper’s Theory)

Answer 6

We try to prove wrong the opposite of our hypothesis

Ex: instead of proving “the probability of having lung cancer his higher in smokers”, we try to show that “the probability of having lung cancer his lower in smokers than non smokers is wrong” (null hypothesis)