Flashcards in Case Control Studies Deck (39):
How are case-control studies done?
Identify individuals with a disease (cases)
Identify `similar’ individuals without the disease (controls)
Determine previous exposure
Relate information on exposure to disease
Explain case-control studies simply.
You want cases that are representative of all people with the disease. Where do you source cases from? (2)
Could be incident cases from disease registry
Could be hospital based recruitment but this may give a biased sample (as they may be more ill anyway)
You want the controls to be as similar to the cases as possible. Where do you source them from?
Same population as cases
Why do we do 'matching' of cases and controls?
Why is this difficult?
To know about potential confounders e.g age/gender
It is difficult to recruit controls
What is always matched on? (2)
What is undermatching?
This is where the cases and controls aren’t similar enough.
This gives the impression that the exposure is related to the disease when it may not be.
What is overmatching?
This is where cases and controls may be too similar.
This would give impression that the exposure is not related to the disease when it may be.
What is the main type of bias that case control studies are susceptible to?
What is recall bias?
Case patients may remember more than control patients (because they have the disease so are more likely to remember/link certain exposures when trying to figure out why they have the disease)
What is meant by reverse causality?
Has disease caused changes in recent exposures
How might selection of cases in case control studies be biased?
If they aren't representative of all people with the disease
How might selection of controls in case control studies be biased?
If they aren't representative of all people without the disease
If they aren't similar to the cases (apart from the fact they don’t have the disease)
What is a nested case-control study?
This is where you take serum samples for example, and you store them. You receive the death certificates of these people as they die. After 20 years you look at the people who died from the disease you are looking at, e.g. prostate cancer. You chose the controls/a select group that didn’t die from prostate cancer. You then measure the serum samples for the potential marker of the disease, e.g. PSA.
What are the advantages of a nested case-control study? (3)
What is the disadvantage of nested case-control studies?
Need cohort study with stored serum samples
Which type of study is suitable for a rare disease?
Which type of study is suitable for a rare exposure?
Which is quicker to run – cohort or case control?
Is recall bias a problem in cohort or case control studies?
Can you use RR from a case-control study? What is used instead?
No, as we do not know risk of the disease (as you have started with cases with the disease).
When is the odds ratio a good estimate of the relative risk?
If the disease is RARE
How do you calculate odds ratios?
Odds Ratios = odds exposure in cases/odds exposure in controls
How do you calculate absolute excess risk? How is this done for case control studies?
Risk in Exposed – Risk in Unexposed
You use other sources to get the population risk.
How is attributable proportion worked out?
What is the formula?
Incidence in population attributable to exposure/
Incidence in population
p (RR - 1)/1 + p (RR - 1)
How do you estimate ‘p’ for the attributable proportion in case control studies?
Divide the number of controls who were exposed by the total number of controls
i.e. those that put the babies on their side/the control (the babies who did not die)
What do cross-sectional studies measure? How?
Existing disease and current exposure
They do a sample at one point in time without knowledge disease or exposure
What are the advantages of a cross-sectional study? (3)
Can look at exposures that won’t change e.g gender
Gives measures of prevalence and exposure rates
Quickest and cheapest way
What are the disadvantages of a cross-sectional study? (2)
No use for rare exposures or rare diseases
Not useful for assessing causality
Which type of study is more likely to be biased?
Which type is study is least likely to be biased?
Is a cohort or case-control study more likely to be biased?
Order the studies in terms of evidence/causality from strongest to weakest.
Three of the study types are likely to have confounding, but you can adjust. For which study is this unlikely?
What are the possible reasons for associations between an exposure and a disease? (6)
What types of bias are possible reasons for associations between an exposure and a disease? (2)
What is the Bradford Hill Criteria for Causation?
A way of determining if results are due to chance or there is causality.
What is meant by the strength of association?
How big the RR is