Case controlled studies Flashcards
What is a case controlled study and how is it carried out?
This is a study looking at past exposure
Identify individuals with a disease i.e. a case
Identify similar individuals without the disease i.e. the controls
Determine both they’re previous exposure
Relate information on their exposure to the prevalence of the disease
What is meant by ‘undermatching’ and ‘overmatching’ and what is the issue with these?
Undermatching - if the cases and the controls selected are not similar enough
Overmatching - if the cases and the controls are too similar e.g. siblings may not differ from parental smoking
This can alter the results
What are the different types of bias that can occur from case-control studies?
Recall bias - the cases will often remember more about past exposure/past history than the controls
Reverse causality - was the disease responsible for the recent exposures e.g. effect of antacids on stomach cancer incidence - the people may have been taking the antacids due to the symptoms of the stomach cancer which had already developed
Selection of cases - they may not be representative of all the people with the disease
Selection of controls - they may not be representative of all the people with the disease - ensure that they are similar enough (but not too much) to the cases
What is the fundamental difference between a cohort study and a case control study?
Cohort study - start with cases and look at the initial exposure and then look forwards to assess the prevalence of the disease
Case control - start with the cases effected by the disease and the controls unaffected by the disease and then look back at past exposure and risk factors
What are nested case-control studies?
This is where you identify a cohort and study them throughout the years e.g. events of death in prostate cancer
In later years, those who died of prostate cancer are the cases and those that did not die become the controls
Can then look back at the initial serum samples of the cases and the controls and compare these
What are the advantages of nested case-control studies?
What are the disadvantages
They are cheap, quick and easy
They work via exposure prior to the disease
Need a cohort study with stored serum samples
Why can you not use relative risk to analyse data from case-control studies?
Because you do not know the risk of the disease - you have started with cases of the disease
How can you analyse data from case control studies?
Use the ‘odds ratio’ instead of the relative risk:
Odds exposure in cases/
Odds exposure in controls
Can also use the absolute excess risk
What is the significance of the odds ratio if the disease is rare?
If the disease is rare, then the odds ratio is a good estimate of the relative risk
i.e. rare disease: odds ratio is almost identical to the relative risk
What is a cross-sectional study?
This is like a snapshot study - measure the existing disease and the current exposure
Take a sample at one point in time without any knowledge of the disease or exposure - no follow up and no questions about past history
What are the advantages of cross sectional studies?
What are the disadvantages?
Can look at exposures that won’t change e.g. gender
Provides measures of prevalence and exposure rates
Not useful for measurement of rare exposures or rare diseases
Not useful for assessing causality
Other than the actual cause of the disease, what might the the cause of the disease be as a result of a trial?
Chance Bias Incorrect analysis Confounding Reverse causality Causality
What is assessed in the Bradford Hill criteria?
Strength of association Dose response/biological gradient Time sequence/temporality Consistency of findings Biological plausability Coherence of evidence Reversibility
