Lecture 3, part 1

Question 1

What is the purpose of comparison?

Answer 1

Summarize relationship btwn exposure and dz by comparing at least 2 measures of dz frequency in different pops
It’s the essence of epidemiology
The rate of dz in an exposed group says nothing about whether exposure is a RF for (or cause of) the dz
-Can only be evaluated by comparing dz occurrence in an exposed grp to another grp (unexposed or referenced grp)

Question 2

What are measures of effect?

Answer 2

Difference in dz occurrence btwn 2 grps of ppl who differ with respect to causal characteristics
An effect is the endpoint of a causal mechanism
Measures what would happen to one pop under 2 possible conditions (counterfactual)

Question 3

What are measures of association

Answer 3

Preferred term when characteristic under study is not causal (or of unknown causality)
Theoretically equivalent to the effect measure
Compares occurrence in 2 different pops

Question 4

What are relative effects?

Answer 4

Ratio of risks or probability

Question 5

What do relative effects do?

Answer 5

Calculate the ratio of 2 measures of dz frequency
-Measure in exposed grp
-Measure in unexposed grp

Question 6

Examples of relative effects

Answer 6

Risk ratios
Odds ratios

Question 7

What are absolute effects?

Answer 7

Differences in risks

Question 8

What do absolute effects do?

Answer 8

Calculate the difference btwn 2 measures of dz frequency
-Measure in exposed grp
-Measure in unexposed comparison grp

Question 9

Examples of absolute effects

Answer 9

Risk differences
Attributable proportions

Question 10

Purpose of risk ratio

Answer 10

Gives info on the relative effect of the exposure on the dz (or outcome)
-Estimate of how many times higher (or lower) the risk of dz is among the exposed (index group) as compared to the unexposed (comparison group)
-Commonly used in etiologic research (attempt to ID causes of dz)

Question 11

Formula for risk ratio

Answer 11

Risk in exposed group/risk in unexposed group

Question 12

What is the implication when RR = 1.0?

Answer 12

Risk in exposed is same as risk in non-exposed
No association between exposure and dz

Question 13

What is the implication when RR > 1?

Answer 13

Risk in exposed > risk in non-exposed
Positive association; causal?

Question 14

What is the implication when RR < 1?

Answer 14

Risk in exposed < risk in non-exposed
Negative association; protective

Question 15

To what group can risk ratios and odds ratios be compared if there are more than 2 different groups?

Answer 15

The reference group

Question 16

What is the default value for the reference group in RRs or ORs?

Answer 16

1.00

Question 17

Describe reference group selection

Answer 17

It’s arbitrary, often based on study question or clinical relevance
Usually the group with the lowest risk of the dz/outcome

Question 18

Definition of odds of an event

Answer 18

The ratio of the probability that the event will occur to the probability that the event will not occur

Question 19

Definition of odds of an exposure

Answer 19

The ratio of the probability that the exposure was present to the probability that the exposure was absent

Question 20

What do odds ratios do?

Answer 20

Quantifies the relationship between an exposure and a health outcome
-Used when do not know the total pop in the margin of a two-by-two table, or when do not know the rate or risk of dz

Question 21

When does an odds ratio approximate the risk ratio?

Answer 21

When the health outcome is uncommon

Question 22

Odds ratio formula

Answer 22

ad/bc

Question 23

What is the implication of OR = 1.0?

Answer 23

Exposure is not related to dz
No association; independent

Question 24

What is the implication of OR > 1.0?

Answer 24

Exposure is positively related to dz
Positive association

Question 25

What is the implication of OR < 1.0?

Answer 25

Exposure is negatively related to dz
Negative association