Case-control studies Flashcards
Concept of case control studies
Find cases
Find controls
Look at past exposures - RETROSPECTIVE
Compare exposure levels
What can we measure with case control study?
Odds ratio
Problem with odds ratio
It’s not a rate ratio - doesn’t show increased risk
error factor calculation
e to the power of 2x square root of 1/a + 1/b +1/c + 1/b
Calculating odds ratio
AD/BC
Why case control study vs cohort?
Cohort takes long time
Cohort is expensive
Cohort is not good for studying rare diseases
Confidence interval =
Odds ratio / error factor (lower value)
Odds ratio x error factor (upper value)
What does confidence interval tell us?
If null value (or 1 if ratio) is within interval then there is no statistical significance of data
BUT this does not show less risk (NOT A RATE RATIO)
What is the precision of odds ratio affected by?
Number of healthy people and cases
Its worth increasing controls up to a point (4-6x cases)
Case control vs cohort
Case control looks backwards (retrospective)
Cohort looks forwards (prospective)
Different types of cohort study?
Concurrent - prospective waiting for events to happen
Historical cohort - retrospective collection of data (eg from pre existing records)
Different types of case control study
Conventional - retrospective collection of data (from recall)
Nested case control study - collection of data from concurrent cohort study
What is a nested case control study?
Case control study is nested within a cohort study
Advantages of nested case control study over normal case control
AIncidence rates can be calculated
population for sampling of controls is defined
Advantages of nested case control over cohort study
Can collect more detailed information for minority of participants (eg blood samples or additional info)