Measures of Association: RR and OR
What are the two types of matched-pairs in a case-control study? Describe them.
1. Concordant pairs: pairs that had the same exposure experience 2. Discordant pairs: pairs that had different exposure experience
Measuring association allows for what? (3)
1. Quantifies the relationship between exposure and the outcome (disease, injury, etc.) 2. Compare those exposed and unexposed (cohort) 3. Compare those with a specific outcome and those without that specific outcome (case-control)
What is relative risk (RR)?
A measure of association (between exposure and outcome) that tells you how strong an association is between an exposure and the development of a disease -Probability of an event -Risk in exposed/risk in nonexposed
What is odds ratio (OR)?
A measure of association that is a good estimate of the RR (not as good, but gives us an estimate) Can be obtained in a case-control and cohort study Can be used INSTEAD of the relative risk also known as relative odds
If OR is less than 1, then
Exposure is NEGATIVELY related to outcome
If OR is greater than 1, then
Exposure is POSITIVELY related to outcome
(T/F) Odds ratio formula is different for a case-control and cohort study.
FALSE Odds ratio formula is the same for BOTH cohort and case-control studies
What must you have to calculate relative risk?
INCIDENCE Incidence of the disease in the exposed (Ie) and the incidence of the disease in the non-exposed (Io)
Relative Risk (RR) =
Ie/Io = incidence in exposed/incidence in non-exposed [a/(a+b)]/[c/(c+d)] Ends with a whole number/decimal (not a percentage) 10^n cancels out
what did the Framingham Study find? (4)
Increased cholesterol level = increased risk of CVHD Increased age = increase risk of CVHD Younger women = increased risk of CVHD Men had an increased risk of CVHD when they were 30-49 than women due to hormones (provides a rationale for hormone replacement therapy for heart disease - but this can increase the risk of cancer)
What are the two types of studies in a case-control study?
Matched and unmatched
Which is better in a case-control study, matched or unmatched? Why?
Matched, because matching cases with control = you mare them alike on certain factors = increases power, and decreases underlying factors Unmatched can inflate the risk ==> makes it less accurate
What are the three steps in calculating OR?
Multiple across the 2x2 table (2 steps) Divide
Can relative risk be calculated in case-control studies?
NO, you must use odds ratio (an estimate of RR) Can NOT calculate incidence = can NOT calculate RR
If RR = 1, then
Risk in exposed = Risk in non-exposed No association
If RR > 1, then
Risk in exposed greater than risk in nonexposed Positive association, possibly causal
If RR <1, then
Risk in exposed is less than risk in nonexposed Negative association, possible protective Ex: sunscreen and skin cancer, exercise and heart disease
(T/F) The interpretation of odds ratio is the same as relative risk.
TRUE
What was the Framingham study?
The Framingham Heart Study was a prospective cohort study where 5,000 people were studied over more than a 20-year period. (RR used) The goal was to study the causes and risks of heart disease. There were a lot of questions about family history, diet, exercise, employment, measuring blood pressure, listening to the heart, and some blood tests
What is the ultimate goal in a case-control study?
To find exposure
Can relative risk be calculated directly in a cohort study?
YES, because we have INCIDENCE
How do you calculate the following in a cohort study? What is the probability that the disease will develop in an exposed person?
a/a+b
In a cohort study, how do we calculate incidence in exposed individuals (Ie)?
a/a+b
How do you calculate the following in a cohort study? What are the odds that the disease will develop in an exposed person?
a/b (both of them are in the exposed categories of the 2x2 table)
How do you calculate odds ratio in a case-control study?
ad/bc
How do you calculate odds ratio in a cohort study?
ad/bc
How do you calculate odds ratio in an unmatched case-control study?
ad/bc
How do you calculate odds ratio in matched case control study?
b/c -only uses discordant pairs and ignores concordant pairs
How do you calculate the following in a cohort study? What is the probability that the disease will develop in a non-exposed person?
c/c+d
In a cohort study, how do we calculate incidence in non-exposed individuals (Io)?
c/c+d
How do you calculate the following in a cohort study? What are the odds that the disease will develop in a non-exposed person?
c/d
If OR is 1, then
the exposure is NOT related to outcome
What is absolute risk?
the incidence of disease in a population -indicates the magnitude of the risk in a group of people with a certain exposure