Confounding: Guilty by Association?

Question 1

Define “Confounding”

Answer 1

Confounding refers to when a relationship between an exposure and an outcome is distorted by their shared relationship with something else

Question 2

What are two examples of Observational Studies?

Answer 2

Cohort Studies

Case Control Studies

Question 3

Are Cohort / Case control studies prospective or retrospective?

Answer 3

Cohort = Prospective

Case control = Retrospective

Question 4

What 4 things can researchers do to remove a confounding factor from the design and analysis of their study?

Answer 4

1. Restriction (exclusion)
2. Matching
3. Stratification
4. Multiple Variable Regression

Question 5

What are 2 pitfalls to using Restriction as a means to remove a confounding factor?

Answer 5

1. Less data available

2. Difficult if there are multiple confounding factors

Question 6

What is “Matching” as a means to eliminate Confounding in research?

Answer 6

Creating a comparison group matched on the possible confounder

Question 7

Using Matching as a means to eliminate Confounding typically is used in what study design?

Answer 7

Case-control studies