Case Control Studies Flashcards
Describe how a case control study works
Case group (people who have the disease) - measure the exposure in the exposed group and unexposed group
Control group (people without the disease) - measured exposure and unexposure
What do case control studies do
Identify individuals with a disease (cases)
Identify `similar’ individuals without the disease (controls)
Determine previous exposure
Relate information on exposure to disease
What are the types of observational studies
- Cohort
- Case control studies
- Cross sectional studies
What are the types of interventional studies
- Randomised controlled studies
How do you source cases in a case control study
Source of cases
- Want representative of all people with the disease of interest
- Could be incident cases from disease registry
- Could be hospital based recruitment but this may give a biased sample
How do you source controls in a case controls study
Source of controls
- same population as cases
- If hospital based ensure the reason for being in hospital not also related to exposure of interest
- Usually more than one control per case
Why do we use matching
Why
- Know about potential confounders eg age/gender
- Not interested in examining the association of these confounders
Give an example of how we can use matching
How
If using GP register then can select by age and sex
Once matching variable is selected it is….
Once matching variable is selected it is not possible to analyse it as a risk factor later
What are the biases in a case control study
Recall bias
- Cases may remember more than controls
Reverse causality
- Has disease caused changes in recent exposures
Selection of cases
- Are they representative of all people with the disease
Selection of controls
- Are they representative of all people without the disease
- Are they similar to the cases
What happens in undermatching
- Cases and controls are not similar enough
- E.g. if they don’t match on age cases may be older and therefore more likely to smoke than controls – this would give the impression that smoking is related to the disease when it may not be
What happens in over matching
- Cases and controls may be too similar
- E.g. if chose siblings may not differ in exposure of interest for example parental smoking
- This would give the impression that parental (passive) smoking is not related to the disease when it may be
What is the difference between a case control study and a cohort study
Cohort study
- in a cohort study you identify your cohort and measure to see if they are exposed or not exposed over time and then see if they get the disease or not
Case control study
- this is when you identify those with the disease and measure to see if they were exposed or unexposed
- and identify those without the disease and measure to see if they were exposed or unexposed
What is a nested case control study
- Case control study nested within a cohort study
How does a nested case control study work
As an example
- identify a cohort and store the blood sample at the beginning
- then follow up and see who gets the disease and who doesn’t not
- everyone with the disease is called a case so you take the blood sample out of the freezer and look at the biomarkers
- everyone who does nto get the disease are controls
- for every case you might do 3 or 5 controls
What is a nested case control study useful for
This is useful for biomarker studies where biomarker is expensive to measure
What are the advantages and disadvantages of a nested case control study
Advantages
– Cheap, quick and easy
– Exposure before disease
– Can retrieve stored samples to look at new biomarkers
Disadvantages
- Need cohort study with stored serum samples
What type of study out of case control or cohort study would you use to study a risk factor for a rare disease
- Use a case control study
- - if a disease is extremely rare a cohort study may have be impractically large to get enough people with the disease
What type of study out of case control or cohort study would you use to study a rare exposure associated with the disease
- Cohort study – if an exposure is rare there may be too few people exposed in a case control study to be able to draw conclusions
What type of study out of case control or cohort study would you use if you need to investigate a factor for a disease quickly
case control study
What type of study is recall bias a problem in
case control studies and not cohort studies as they are remembering wether or not they can into contact with the exposure
Why can you not use relative risk in a case control study
Can’t use relative risk as do not know risk of the disease (as you have started with cases with the disease)
What do you you use instead of relative risk in a case control study
odds ratio
- Use odds ratio = odds exposure in cases/ odds exposure in controls
How do you work out the odds ratio
odds ratio = odds exposure in cases/ odds exposure in controls
if the disease is rare….
If the disease is RARE the odds ratio is a good estimate of the relative risk
How do you work out the absolute excess risk
- Risk in exposed – risk in unexposed
What is the problem with using absolute risk in a case control study
- you dont know the risk in a case control study
Describe an example of how to work out absolute excess risk in a case control study
- Babies on there sides are twice as likely to die as babies on their back
From ONS: Population risk of SIDs = 3.4 per 10,000 per year
Therefore estimate of risk in babies put on their sides
= 2 x 3.4 per 10,000 per year
= 6.8 per 10,000 per year
Therefore estimate of absolute excess risk in babies put on their sides =
6.8 per 10,000 per year – 3.4 per 10,000 per year
= 3.4 per 10,000 per year
What is attributable proportion
The same information allows you to calculate the proportion of disease in the exposed group that can be attributed to the exposure
describe the formula for attributable proportion
Incidence in population attributable to exposure/
Incidence in population
= p(relative risk - 1)/ 1+p(relative risk - 1)
P = proportion exposed in population
- work this out by using the control variable over the total number of controls
How does a cross sectional study work
- Measure existing disease and current exposure
- Sample at 1 point in time without knowledge of disease or exposure
What are the advantages and disadvantages of a cross sectional study
• Advantages
– Can look at exposures that won’t change eg gender
– Gives measures of prevalence and exposure rates
• Disadvantages
– No use for rare exposures or rare diseases
– Not useful for assessing causality
Does exposure to blue asbestos cause lung cancer - What study should you do
Cohort
Is prevalence of back pain more common in men or women ?
Cross sectional
What is the order of the risk of bias in studies
Likely - Cross section - case control - cohort - clinical studies Unlikely
What is the order of the risk of confounding in studies
likely - Cross section - case control - cohort - clinical studies unlikely
What is the order of strength of proof of causality in studies
likely - Cross section - case control - cohort - clinical studies unlikely
What are the reasons for associations between an exposure and disease
- Bias
- Reverse causality
- Confounding
- Incorrect analysis
- Chance
- Casual
A case control study of patient’s with Chron’s disease showed that they were much more likely to have reported ever eating Corn flakes.
what is the reason for association between the exposure and disease
Recall bias
Those with Chron’s disease more likely to remember everything they ate. Most people have eaten corn flakes at some time.
Non-randomised studies have shown an increased risk of death in preventive (multivessel) percutaneous coronary intervention versus infarct artery only PCI in patients with STEMI
what is the reason for association between the exposure and disease
Selection bias
Patients who had preventive PCI more likely to be sicker
Randomised controlled trials have shown the benefit of preventive PCI
A case control study of stomach cancer patients found that they had a higher antacid usage in the 6 months prior to diagnosis than did the controls.
what is the reason for association between the exposure and disease
Reverse Causality
People with undiagnosed stomach cancer tend to have the same symptoms for which antacids are taken
In a cohort study heavy drinkers were found to be more likely to develop lung cancer. Does drinking cause lung cancer ?
what is the reason for association between the exposure and disease
Confounding
Heavy smokers are more likely to be heavy drinkers.
Smoking causes lung cancer
Randomised placebo controlled trial of candesartan in patients with chronic heart failure found it increased risk; RR = 1.6 (95% CI 1.1-2.2)
what is the reason for association between the exposure and disease
Chance
This trial was the odd one out
What is the criteria for causality
Bradford Hill Criteria for Causation
Name the parts that make up the bradford hill criteria for causation
- Strength of association – how extreme is the relative risk – relative risk of 5 or 0.2 or lower or 0.2 or lower is a strong relative risk
- Dose response - does increasing the dose increase the risk
- Time sequence – exposure preceded the outcome
- Consistency of findings – similar studies on different populations
- Biological plausibility – e.g. tar is known to be carcinogeni
- Coherence of the evidence – cross-sectional, cohort and case-control studies all find the same increased risk
- Reversibility – the relative risk for ex cigarette smokers is between that of current smokers and non-smokers
Could the association be due to
Could the association be due to ?
Chance
Bias
Incorrect analysis
Confounding
Reverse causality
CausalityHow do you work out if the association is casual
Bradford Hill Criteria for Causation
