7. Case-control Studies Flashcards
What are the benefits of conducting a case control study over a cohort study?
Conventional cohort studies take a long time whereas case control studies can be completed fairly quickly, cohort studies are very expensive but case control studies are cheaper, cohort studies aren’t good for studying rare outcomes but case controls are.
How are cohort studies analysed? (Give the equation).
Using incidence rate ratios and the rare disease assumption so IRR= (exposed with disease x unexposed without disease)/ (unexposed with disease x exposed without disease).
How are case-control studies analysed? (Give the equation).
Using the odds ratio, OR = (exposed cases x unexposed controls)/ (unexposed cases x exposed controls).
How is the 95% confidence interval calculated?
Odds ratio/ error factor to odds ratio x error factor.
How many controls should be used per case and why?
4-6, as the number of controls increases, the odds ratio precision increases but only up to this number of controls per case.
What is the data collection method used in a conventional case control study?
Retrospective collection of data.
What is the data collection method used in a nested case control study?
Collection of data from the evolving outcome and exposure database of a concurrent or prospective cohort study.
What are the advantages of a nested case control study over the following?
a. A conventional case control study
b. A conventional cohort study
a. Incidence rates can be calculated and the population for sampling controls is already defined.
b. More detailed information can be collected for a minority of patients.
What are three key issues for case control studies?
Selection bias, information bias and confounding.
How is selection bias a problem in case control studies?
It can underestimate the odds ratio. If cases and controls are picked from the same population, e.g. the same hospital ward, they’re likely to have had the same exposures so the link between exposure and outcome is likely going to be underestimated.
How is selection bias of case control studies dealt with?
The controls have to be representative of the whole population.
How is information bias a problem in case control studies?
There could be randomly inaccurate measurements or systematic misclassifciations like recall bias, assessor bias etc.
What tends to be the effect of information bias on case control studies?
Tends to underestimate the odds ratio.
How can confounding be dealt with in case control studies?
By matching using important cofounders, also adjusted for by analysing with logistic regression.
How is a case control study conducted?
A group of cases are identified, a group of non-cases/ controls are then identified too. The previous exposure status of everyone is ascertained and the levels of exposure in cases and controls is compared.
