18. Case/Control Studies

Question 1

standard set of summary questions to asses the quality of a case-control study

Answer 1

1. Was there a pre-specified hypothesis defining a relationship between an exposure and outcome?
2. Were the exposure and health outcome clearly and operationally defined?
3. Was the control group appropriate?
4. Was the measurement of exposure both in the cases and controls accurate and unbiased?
5. Was the measurement of the outcome both in the cases and controls accurate and unbiased?
6. Were the important confounding variables accounted for and controlled for in the statistical analysis?

Question 2

design of case-control study

Answer 2

• Select representative cases (with study outcome)
• Select comparable controls (without study outcome)
• Look historically in both groups for exposure
• Compare odds of exposure in cases and controls

Question 3

strengths of case control studies

Answer 3

• Good for outcomes with long latent periods
• Inexpensive
• Optimal for evaluation of rare diseases or rare outcomes
• Can examine multiple risk factors for a single disease or outcome

Question 4

weaknesses of case-control studies

Answer 4

• Particularly prone to biases and confounding
  
  • Statistical association but not true cause and effect
• Inefficient for evaluation of rare exposures

Question 5

bias

Answer 5

• error in design or conduct of a study
  
  • Any systematic error in the design, conduct, or analysis of a study that results in a mistaken estimate of an exposure’s effect on the risk of disease or risk of outcome
  • Systematic error results from flaws either in the method of selection of study participants (cases and controls) or in the procedures for gathering relevant exposure and/or outcome info
  • Validity depends upon methods by which subjects are selected for study and classified by the four categories (diseased, non-diseased, exposed, and unexposed)
  • If bias exists, the observed study results will be different from true results ( = tendency towards erroneous results)
  • Classification of bias
    
    • Selection bias

Present when individuals have different probabilities of being included in the study sample according to relevant study characteristics; most often the exposure and outcome of interest

        * medical surveillance bias
            * Eg OCPs and side effects
            * (someone on OCPs is more likely to see a doctor)
        * Choice of conrols
            * Eg hospital controls may not come from the base population of interest

Information bias
Results from a systematic tendency for individuals selected for inclusion in the study to be erroneously placed in different exposure/outcome categories, thus leading to misclassification
- recall bias

Question 6

information bias vs selection bias

Answer 6

• Selection bias

Present when individuals have different probabilities of being included in the study sample according to relevant study characteristics; most often the exposure and outcome of interest

* medical surveillance bias
    * Eg OCPs and side effects
    * (someone on OCPs is more likely to see a doctor)
* Choice of conrols
    * Eg hospital controls may not come from the base population of interest

Information bias
Results from a systematic tendency for individuals selected for inclusion in the study to be erroneously placed in different exposure/outcome categories, thus leading to misclassification
- recall bias

Question 7

confounders

Answer 7

Variables entangled with the study factor that masks the true relationship between the study factor and outcome

Question 8

Odds Ratio

Answer 8

OR = (a/c) / (b/d)

Outcome +

Outcome -

Total

Exposure +

a

b

Exposure -

c

d

Odds

a/c

b/d

Odds ratio of X means that the ratio of exposure among cases compared to controls is X.

If incidence of cases among exposed is rare (<5%), it means that OR = RR (those who are exposed are X timre more likely to develop case than non-exposed)