Comparison: Measures of Association

Question 1

What are measures of association?

Answer 1

A quantity that expresses the direction and strength of association between two variables

Like:
Ratio measures (relative risk, prevalence ratio, odds ratio)

Difference measures (risk difference, prevalence different)

In epi, we’re mainly interested in exposure status and outcome status

Question 2

When interpreting a measure of association, what is a comparison group?

Answer 2

exposed (index group) vs. unexposed (reference group)

Question 3

When interpreting a measure of association, what is direction?

Answer 3

The direction (pos and neg) of the measure of association provides information on the nature of the relationship between the exposure and outcome

Question 4

When interpreting a measure of association, what is magnitude?

Answer 4

The magnitude of the measure of association provides information about the strength of the relationship between the exposure and outcome.

Question 5

How do you calculate ratio measures of association?

Answer 5

(amount in exposed)/(amount in unexposed)

Question 6

How do you calculate differences measures of association?

Answer 6

(amount in exposed) - (amount in unexposed)

Question 7

How do you interpret ratio measures of association?

Answer 7

Question 8

How do you calculate cumulative incidence (risk) in exposed in a longitudinal study (cohort or randomized trial)?

Answer 8

Question 9

How do you calculate cumulative incidence (risk) in unexposed in a longitudinal study (cohort or randomized trial)?

Answer 9

Question 10

What does relative risk/risk ratio (RR) mean?

Answer 10

RR compares the risk of a health event in one group with the risk in another group. (exposed vs. unexposed)

Question 11

How do you calculate relative risk (RR)?

Answer 11

Question 12

How do you interpret relative risk?

Answer 12

EX: Cohort study of the association between radon exposure and incident lung disease yields a relative risk (RR) of 3.50.

Question 13

How do you calculate the prevalence of outcome in exposed and unexposed groups in a cross-sectional study?

Answer 13

Question 14

How do you prevalence ratio in a cross-sectional study?

Answer 14

Question 15

How do you interpret prevalence ratio (PR)?

Answer 15

EX: In a cross-sectional study, the PR was 0.8 for the association between physical activity (exposure) and coronary heart disease (CHD) (outcome).

Question 16

Can you calculate risk (incidence) of outcome and relative risk, prevalence of outcome, and outcome ration in a case-control study and why?

Answer 16

Since we don’t know how many people were “at risk” for becoming a case (i.e. we don’t know the denominator), we cannot calculate risk (incidence) of outcome and relative risk.

Since we ourselves selected cases and controls (i.e. we dictated the proportion of cases), we cannot calculate prevalence of outcome and prevalence ratio.

Question 17

What kind of calculation can you do with case-control study data? And how do you calculate it?

Answer 17

Odds of exposure = probability (p) of being exposed divided by the probability of being unexposed (1-p) = p/(1-p)

Question 18

How do calculate odds ratio (OR) in case-control studies?

Answer 18

Question 19

How do you interpret odds ratio (OR) from a case-control study?

Answer 19

Ex: The OR was 2.3 for the association between cigarette smoking and emphysema a case-control study.

Question 20

How do you interpret the value of a difference measure of association?

Answer 20

Question 21

How do you calculate risk difference (cumulative incidence difference) in longitudinal studies? (difference measure of association)

Answer 21

Question 22

How do you calculate prevalence difference in cross-sectional studies? (difference measure of association)

Answer 22

Question 23

True or False: Odds difference is not used as a measure of association.

Answer 23

True

Question 24

What is attributable risk in those exposed?

Answer 24

Assuming a casual relationship between exposure and disease… How much of a disease in those exposed can be attributed to the exposure? What is the amount of risk of a disease that can be prevented in the exposed group if we are able to eliminate the exposure in question?

Question 25

Why is attributable risk important to public health?

Answer 25

Public health policy makers may ask… Would elimination of the exposure lead to a substantial reduction in related morbidity and mortality in exposed individuals?

Question 26

How do you calculate attributable risk of disease in the exposed group?

Answer 26

Question 27

Calculate the attributable risk of CHD in smokers. How do you interpret the value?

Answer 27

Question 28

How do you calculate the proportion attributable risk of disease in the exposed group?

Answer 28

Question 29

What is the significance of proportion attributable risk in the total population?

Answer 29

Useful for determining priorities for public health action

Public health policy makers may ask… Would elimination of the exposure lead ti a substantial reduction in related morbidity and mortality in the total population (which includes both exposed and non-exposed individuals)?

Question 30

How do you calculate proportion attributable risk of disease in the total population?

Answer 30

Question 31

What are the conditions under which you can calculate attributable risk of disease for the total population?

Answer 31

1) the incidence among exposed
2) the incidence among unexposed
3) the proportion of the total population that is exposed