Lecture 20: Case Control Studies Flashcards
What do case control studies do?
Ascertain outcome status, then find the exposures
What is the difference between a cohort study and a case control study?
Cohort studies ascertain exposure status and then find the outcome whereas case-control studies ascertain the outcome status and then find the exposures
Case control studies are what types of studies?
observational and analytic
Case control studies address the issues associated with what other type of study?
Cohort
Case control studies address the issues associated with another type of study. What are the issues?
- cohort studies can be inefficient with rare/slow to develop outcomes
- cohort studies can be inefficient with transient/acute exposures
What are the steps to carrying out a case control study?
- identify source population
- identify participants with a known outcome status (cases)
- sample people without the outcome from the same population (controls)
- measure exposure prior to the outcome in cases and controls
- compare odds of exposure to calculate the measure of association
What measure of association can be calculated from a case-control study?
odds ratio
How do you calculate the odds ratio?
odds of exposure (cases) divided by the odds of exposure (controls)
If the odds of exposure (cases) is greater than the odds of exposure (controls) then the outcome is more likely to occur in the _____ and therefore the exposure is a potential _______ factor
cases
risk
If the odds of exposure (controls) is greater than the odds of exposure (cases) then the outcome is more likely to occur in the _____ and therefore the exposure is a potential _______ factor
controls
protective
What can not be calculated from a case control study?
incidence
relative risk
risk difference
Why can we not calculate prevalence in a case control study?
because we have selected the number of people in the study with and without the outcome
Why can we not calculate incidence in a case-control study?
Because the incidence in the controls would be zero and because we have selected the number of people in the study with and without the outcome
In the box, what letter are used to calculate the odds of exposure in cases?
a/c
In the box, what letter are used to calculate the odds of exposure in controls?
b/d
What is the null value for odds ratio?
1
What is the template for the interpretation of odds ratio?
we can just use the RR interpretation:
[exposed group] were [value] times as likely to [outcome] compared to [comparison group]
Why can we use the relative risk interpretation for the odds ratio interpretation?
because when disease is are, OR approximates the RR
If controls don’t have the outcome, how can we carry out step 4: measure exposure prior to the outcome in cases and controls?
by using index dates
What are index dates?
We pretend that the control had the event on the same date as the case (or close to it)
Case-control studies can only look at one _______
outcome
When selecting the cases, do we look at prevalent cases or incident cases and why?
we look at incident cases because we only want to consider the development of an exposure and not the duration
What two things must the control group be/do in a case-control study?
- they must be capable of becoming a case
- they need to represent the exposure distribution in a source population
What is the purpose of the control group?
to estimate the prevalence of exposure in a population from which the cases come from
What are the strengths of a case-control study?
- Good for rare outcomes and transient exposures
- Assess multiple exposures
- Temporal sequencing
- Quick and inexpensive
What are the limitations of a case-control study?
- Only studies one outcome
- Hard to get an appropriate control group
- Prone to information bias
