Case Control Studies

Question 1

What are the root meanings of “epidemiology”?

Answer 1

Epidemiology = study of distribution + determinants of health states in populations

Question 2

What is incidence?

Answer 2

Number of new cases of disease in a population at risk over a given time. Measures risk of developing disease.

Question 3

How is incidence calculated?

Answer 3

New cases ÷ population at risk.

Question 4

Why must population be “at risk”?

Answer 4

People already with disease cannot become new cases.

Question 5

What is prevalence?

Answer 5

Total cases (new + existing) at a point in time. Measures burden of disease.

Question 6

What is mortality rate?

Answer 6

(# deaths from a disease ÷ total population). Measures fatality in a population.

Question 7

What is morbidity?

Answer 7

Illness/disease presence (not death).
A person may have multiple morbidities at once.

Question 8

What is an exposure?

Answer 8

A potential cause (independent variable).

Question 9

What is an outcome?

Answer 9

Health condition potentially influenced by exposure (dependent variable).

Question 10

How do cross-sectional, case-control, and cohort differ?

Answer 10

• Cross-sectional: exposure & outcome measured simultaneously
• Case-control: start with outcome → look backward at exposure
• Cohort: start with exposure → follow forward to outcome

Question 11

Which study design CANNOT determine temporality?

Answer 11

Case-control (exposure measured after disease occurs).

Question 12

Define “cases.”

Answer 12

Individuals with the disease/outcome of interest.

Question 13

Define “controls.”

Answer 13

Individuals without the disease/outcome.
Must represent the population from which cases arose.

Question 14

Ideal situations for case-control studies?

Answer 14

• Rare outcomes
• Outcomes with long latency (e.g., cancers)
• When exposure data must be collected retrospectively
• When disease already occurred (efficient)

Question 15

What is required when selecting cases?

Answer 15

• Must truly have the outcome
• Must use a defined method of diagnosis
• Should avoid bias by using consistent criteria for all cases

Question 16

Key inclusion rule for controls?

Answer 16

Must NOT have the outcome.

Question 17

Why might controls have another condition?

Answer 17

To equalize healthcare system interaction (reduces confounding).

Question 18

Formal definition of odds?

Answer 18

Probability event occurs ÷ probability it does not.
OR: (# with event ÷ # without event).

Question 19

Why is odds different from risk?

Answer 19

• Odds compares event : non-event
• Risk compares event : total population

Question 20

OR > 1 means?

Answer 20

Exposure more common in cases → positive association.

Question 22

OR = 1 means?

Answer 22

No association. Null hypothesis.

Question 23

When is an OR significant?

Answer 23

When 95% CI does NOT include 1.0.

Question 24

Why are CIs narrower with larger sample sizes?

Answer 24

More precise estimate; less variability.

Question 25

What are the advantages of case-control studies

Answer 25

Efficient for rare diseases * Efficient for long latency periods * Requires fewer subjects than cohort studies * Quick + inexpensive * Can examine multiple exposures for one disease * No need to wait years for outcomes to develop

Question 26

Limitations of case-control studies

Answer 26

- no temporality - Exposure timing uncertainty - Heavy reliance on memory - Recall bias - Misclassification - Cannot calculate incidence or prevalence