Observational Studies in Epidemiology

Question 1

What are the different kinds of observation study?

Answer 1

Question 2

How would you design a cohort study?

Answer 2

Question 3

What is the difference between prospective cohort study and retrospectice cohort study?

Answer 3

Prospective cohort study:

• defined population and risk factor of interest
• participants are classifed as exposed or unexposed

Retrospective cohort study:

• exposure data are collected from past records
• outcomes have already occured
• use historical records to classify individuals as exposed or unexposed to the risk factor of interest

Question 4

What is a confounder?

Answer 4

= a variable which is associated with the exposure of interest and indepentdently associated with the risk of developing the outcome

• it can lead to under or overestimation of the link between exposure and outcome

Question 5

How can you calculate risk ratio/relative risk (RR)?

Answer 5

• can be calculated from the results of a cohort study

RR=1 = no difference between groups

RR>1 = increased risk in exposed group compared to unexposed group

RR<1 = reduced risk in exposed group compared to unexposed group

Question 6

What are the advantages and disadvantages of cohort studies?

Answer 6

Question 7

How can you design a case-control study?

Answer 7

Question 8

How can you reduce confounders in case-control studies?

Answer 8

• take account at the design stage by matching
  
  • e.g. if age is a potential confounder of the association of interest, you may recruit one (or more) control of the same age for every case you recruit to the study –> Ensures equal distribution of the confounder across cases and controls

could introduce bias due to over-matching - occurs if the matching variable is on the casual pathway between exposure and disease

Question 9

How can you calculate the odds ratio?

Answer 9

• Cannot get risk ratio from case-control study so we use odds ratio

OR=1 no difference in the odds of exposure between diseased and control groups

OR>1 increased odds of exposure in the diseased group compared to the control group

OR<1 reduced odds of exposure in the diseased group compared to the control group

**useful for do 2x2 table of exposed yes or no and case or control

Question 10

What is the signifcance of having ‘1’ contained in the confidence interval?

Answer 10

Example:

95% CI = 0.68 to 2.01 tells us that we are 95% certain that the true odds ratio in the population from which this sample were taken is between 0.68 and 2.01

• 1 is contained within this interval and therefore we do not have enough evidence to reject the null hypothesis that the odds of women having an insufficient level were different between the cases and controls

Example 2:

95% CI = 1.10 to 2.50 tells us that we are 95% certain that the true odds ratio in the population from which this sample were taken is between 1.10 and 2.50

• 1 is not contained within this interval -> we conclude that there is evidence to suggest that the odds of women having a deficient level were higher amongst the cases compared to the controls

Question 11

What are the advantages and disadvantages of case-control studies?

Answer 11