Lecture 5A, part 3

Question 1

Strengths of case-control studies?

Answer 1

More efficient for rare dzs, or when little is known about the dz
Can evaluate multiple exposures or RFs for a single outcome
Require less time and money than cohort studies (good for conditions with long latent or incubation periods)
Useful for dynamic pops

Question 2

Limitations of case-control studies

Answer 2

May be difficult to select adequate controls
Often more subject to greater risk of bias (recall, selection, etc)

Question 3

What should be considered when choosing a study design?

Answer 3

What is the best study design from a scientific perspective?
What study design is ethical?
What study design is feasible to implement?

Question 4

Scientific perspective study considerations

Answer 4

Concerns about validity and efficiency
Consider natural hx of dz (dz progression, latency periods, etc)

Question 5

Ethical considerations in study design

Answer 5

Concepts of beneficence
Respect for persons
Justice
Consideration of vulnerable pops

Question 6

Considerations in study feasiblity

Answer 6

Consider logistics: money/funding, access to data, staff, time, dedication

Question 7

What should be known in order to say that a change in X causes a change in Y?

Answer 7

We must know that Y is after X is changed
We must know that Y would have been if X had not been changed (the “counterfactual”)

Question 8

Definition of counterfactual

Answer 8

A fact that did not happen but would have happened if something else would have been changed
-What if?

Question 9

What are counterfactuals also called?

Answer 9

Potential outcomes

Question 10

What must be measured when we want to measure the effect of a particular exposure or potential cause?

Answer 10

Observed amount of effect in a pop who are exposed to that cause
Imagine the amount of the effect which would have been observed, if the same pop would not have been exposed to that cause, all other conditions remaining identical
The difference of the two effect measures is the effect due to the cause we are interested in

Question 11

What do we estimate given the counterfactual effect is unobservable?

Answer 11

A proxy or surrogate amount of effect from a pop who are not exposed to the cause and otherwise comparable to study pop