Observational and Epidemiological Studies

Question 1

What is the difference between observational and experimental studies?

Answer 1

Observational: investigator does not influence what the research subjects are exposed to– only tracks their natural course/progression
Experimental: investigator controls exposures that may influence the outcome of interest

Question 2

What are the 4 components of a true experimental study?

Answer 2

1. Manipulation
2. Control
3. Random assignment
4. Random selection

Question 3

What are 3 characteristics of observational studies?

Answer 3

Start with exposure, follow to outcome
Less rigid than controlled studies
Determine association, not cause-and-effect

Question 4

What is the exposure event rate (EER)

Answer 4

The proportion of subject in the exposure group experiencing the event
A/(A+B)

Question 5

What is the control event rate?

Answer 5

The proportion of subjects in the control group experiencing the event
C/(C+D)

Question 6

What is relative risk?

Answer 6

The risk of developing a disease in people exposed to a specific variable compared to those not exposed to that variable.
=EER/CER

Question 7

What does a relative risk of:
=1
<1
>1
mean?

Answer 7

=1: no association
<1: negative association
>1: positive association

Question 8

Are prospective or retrospective cohort studies better?

Answer 8

Both equally good, but prospective collection may introduce additional bias

Question 9

How can we calculate relative risk reduction? How can we interpret it?

Answer 9

(CER-EER)/CER

Exposure is associated with an x% relative reduction of the outcome compared to the control

Question 10

How can we calculate absolute risk reduction? How can we interpret it?

Answer 10

=CER-EER

Exposure is associated with an absolute reduction of the outcome by x%

Question 11

How can we calculate number needed to treat?

Answer 11

1/AAR

Question 12

What are the guidelines for determining causation?

Answer 12

Temporal relation
Biological plausibility
Consistency
Strength of the association
Dose-response relationship
Study design
Judging the evidence

Question 13

How can we calculate odds ratio?

Answer 13

ad/bc

Question 14

How does odds ratio compare to relative risk?

Answer 14

OR is always an overestimation of RR when RR>1

Question 15

What are the 4 types of bias?

Answer 15

Selection bias
Information bias
Data analysis bias
Survival bias