Session 6 Flashcards
What is a case-control study?
A study where a case is selected based on an outcome, and then a group is selected without the outcome and the exposures are compared.
We can then see whether there is an association between an exposure and an outcome.
What type of study is a case-control study?
An observational, analytical study.
What do the case and control refer to?
They both refer to having or not having an outcome, and not the exposure.
Case = outcome, event, effect.
Control = no outcome, no event, no effect.
What are the steps to forming a case-control study?
Cases are selected - those with the outcome.
Controls are selected - those without the outcome.
The controls are then matched to the cases - criteria must be reached to match them.
What is a nested case-control study?
Those that do not have an outcome are picked from the same population from those that have the outcome.
What is a non-nested case-control study?
The control is not picked from the same population of those with the outcome.
What are the strengths of a case-control study?
They are relatively inexpensive.
It is efficient when the outcome is a rare condition.
Efficient for when the conditions take a long time to develop as the outcome has already occurred.
It allows for studying multiple exposures for a particular outcome.
The sample size can be small - there can be larger matching ratios to try and remove any chance.
What are the weaknesses of a case-control study?
Cannot determine incidence and prevalence as the number of cases and controls are fixed - the outcome has already occurred and so no new cases will develop.
It is not suitable for rare exposures - cases are selected on their outcomes, not their exposures.
What is statistical power?
Refers to the likelihood of finding a statistically significant result.
What is statistical power influenced by?
It is influenced by sample size.
Increasing the control to case ratio will increase the statistical power.
What does matching cases do?
Matching reduces the risk of confounding.
We can match based on age, sex, and many factors to try and keep remove many potential exposures that we are not studying for.
What is the outcome of a case-control study?
An odds ratio.
What are the different biases of case-control studies?
Selection bias - matching the controls with the cases can make them unrepresentative of the population as a whole.
Recall bias - when looking at different exposures that people had, this could be of significance.
Misclassification bias - some exposures are difficult to define and so they may be put in incorrect categories.
