Measures of effect

Question 1

What is absolute risk?

Answer 1

• The incidence of disease in a population
• Indicates magnitude of risk, but doesn’t take into consideration the risk of disease in an unexposed population
• We are often interested in finding out whether there is disease associated with a specific exposure

Question 2

What are approaches to measures of excess risk?

Answer 2

• Differences in risks
  
  • =(risk in exposed) – (risk in unexposed)- Attributable
• Ratio of risks
• Relative Risk
  
  • =(risk in exposed)/(risk in unexposed)

Relative risk estimates the magnitude of an association between exposure and disease, based on the incidence of disease in the exposed group relative to the unexposed group. … Attributable risk, or risk difference, is the absolute difference in incidence between an exposed and unexposed group.

Question 3

What is the formula for differences in risk?

Answer 3

=(risk in exposed) – (risk in unexposed)- Attributable

Question 4

What is relative risk? RR

Answer 4

• Measures the strength of association between a factor/characteristic/exposure and disease
• Larger the relative risk, stronger the association
• RR = 1.0 –> Risk of disease does not differ

between exposed and unexposed

RR > 1.0 –> Risk of disease greater in

exposed than unexposed

RR < 1.0 –> Risk of disease less in exposed

than unexposed (exposure is protective)

Question 5

What is the formula for odds?

Answer 5

(Probability event will occur) /(Probability event will not occur)

Question 6

How are odds used in cohort studies?

Answer 6

• Odds of disease occurring in exposed

a/(a+b)⁼ a

b/(a+b) ⁼ b

• Odds of disease occurring in unexposed

c/(c+d) ⁼c

d/(c+d) ⁼ d

Odds Ratio (OR) = odds that an exposed person develops disease/odds that a non-exposed person develops the disease

Question 7

Analyzing case-control data

Answer 7

• We don’t know the incidence in exposed or non-exposed because we selected diseased (cases) and non-diseased (controls) subjects
• We cannot calculate relative risk (RR) directly, but can calculate the odds ratio (i.e., exposure OR)
• Under certain circumstances, can obtain a good estimate of the RR from a case-control study using the OR

Question 8

How is the odds ratio calculated in case-control study?

Answer 8

OR=Cross product ratios in cohort and case-control studies

Question 9

When is the odds ratio (relative odds) obtained in a case-control study a good approximation of the relative risk in the population? (3 conditions)

Answer 9

1. When the cases studies are representative, with regards to history of exposure, of all people with the disease in the population from with the cases were drawn.
2. When the controls studies are representative, with regard to history of exposure, of all people without the disease in the population from which the cases were drawn.
3. When the disease being studies does not occur frequently.

(a/a+b)/(c/c+d) = (a/b)/(c/d)

Question 10

When is odds ratio and good proxy for relative risk with infrequesquent diseases/outcomes

Answer 10

• relative risk and odds ratio are similar.
• Works best when disease is rare. In this example, there are 300 people who have the disease which isn’t a lot in the population.

Question 11

What is the odds ratio in a unmatched case-control study?

Answer 11

(ad)/(bc)

Question 12

What is matching?

Answer 12

The process of selecting controls so that they are similar to cases with regard to specific characteristics such as age, race, and sex

Question 13

What are concordant pairs

Answer 13

1. Pairs in which both the cases and controls are exposed
2. Pairs in which neither the cases nor the controls are exposed

Question 14

What are discordant pairs?

Answer 14

1. Pairs in which the case is exposed but the control is not exposed
2. Pairs in which the case is not exposed but the control is exposed