Case-Control Studies

Question 1

Outline the principles of a case control study

Answer 1

• Identify group of diseased people (cases) and a group of disease-free people (control)
• Look backwards in time to compare the EXPOSURE STATUS of the two groups

Question 2

What are the advantages of retrospective studies?

Answer 2

Uses historical data so is CHEAPER and FASTER than a prospective cohort study

Question 3

What is the main difference between a ‘case control’ and a ‘cohort’ study?

Answer 3

• Case control identifies outcomes and compares disease exposures
• Cohort identifies exposures and compares disease outcomes

Question 4

What are the disadvantages of a cohort study compared to a case control?

Answer 4

• Good for rare outcomes but not rare diseases

- Would need large number of person-years to accumulate rare disease events before data has any use

Question 5

Explain the analysis of case control studies

Answer 5

• Two samples: one from diseased population and one from non-diseased population
• Compare exposure history between two samples using an ODDS RATIO

Question 6

How would you calculate the odds ratio of being exposed in CASES?

Answer 6

• Exposed cases / unexposed cases

- a / c

Question 7

How would you calculate the odds ratio of being exposed in CONTROLS?

Answer 7

• Exposed controls / unexposed controls

- b / d

Question 8

How would you calculate an ODDS RATIO for a case control study?

Answer 8

• ad / bc
• a = exposed cases
• b = exposed controls
• c = unexposed cases
• d = unexposed controls

Question 9

Explain how you could estimate the range in which the true underlying odds ratio (OR) lies

Answer 9

• Calculate error factor using all 4 values in odds ratio equation
• Calculate the 95% CI using (OR / e.f.) and (OR x e.f.)

Question 10

How could you increase the precision of the odds ratio?

Answer 10

• Increase number of controls in study (b and d)
• Reduces error factor
• Narrower 95% CI

Question 11

What is the difference between a BINARY and a DOSE response?

Answer 11

• BINARY response uses 2 distinct groups (exposed and unexposed)
• DOSE response compares a range of exposures associated with a disease outcome (smoking and lung cancer - can compare the effects of smoking different amounts of cigarettes a day and how this affects the disease outcome)

Question 12

Explain how you would compare odds ratios in a dose response

Answer 12

• Calculate a separate odds ratio for each distinct category in the dose response
• Calculate subsequent error factors and 95% CIs for each odds ratio
• Compare the statistical significance of each

Question 13

Which study would you use to identify rare diseases within a population?

Answer 13

• CASE CONTROL
• Suitable for rare diseases but not rare outcomes
• Cohorts are suitable for rare outcomes but not rare diseases

Question 14

Give 3 advantages of case control studies over cohort studies

Answer 14

• Can look at many different potential exposures using a dose response (useful when there are several candidates for the causal exposure)
• Cheaper
• Quicker (don’t have to follow up as you are using data from the past)

Question 15

What is one major disadvantage of case control compared to cohort studies?

Answer 15

• Difficult to be sure that exposure precedes the disease outcome with case control
• In cohort studies it is clear that the exposure identified precedes the disease outcome
• Cohorts are LESS PRONE TO BIAS such as recall bias