Absolute Risk
- incidence of disease in population.
o can indicate magnitude of risk in group with certain exposure
§ does not take into account risk of disease in non-exposed individuals.
o implicit comparison
How Do We Determine Whether a Certain Disease is Associated with a Certain Exposure
- whether there is excess risk of disease in person who has been exposed
o ratio of the risks (or incidence rates)
§ disease risk in exposed / disease risk in non-exposed
o difference in the risks
§ disease risk in exposed – disease risk in non exposed
Relative Risk
- RR = risk in exposed / risk in non exposed
- probability of event (developing disease) occurring in exposed people compred to probability of event in non exposed people, as ratio
- interpreting relative risk
- calculating RR is cohort studies
o RR can be calculated directly
§ RR = incidence in exposed / incidence in nonexposed
Odds Ratio
- In order to calculate a RR à must have incidence rates
- in case control study, incidence not available
- another measure of risk à odds ratio
defining odds ratio in cohort and case control studies
- odds of event – ratio of number of ways event can occur to number of ways event cannot occur
- odds = P / (1-p)
- what are the odds that disease will develop in an exposed person
|
| Develop Disease | Do not develop disaese |
| Exposed | a | b |
| Not Exposed | c | d |
OR = odds that exposed person develops disease / odds that non-exposed person develops disease
= ( a /b) / ( c/d)
= ad / bc
Odds Ratio in Case Control Study
- cannot calculate RR
o no info re incidence
- ask – what are the odds that a case was exposed
|
| Cases (with disease) | Controls (No Disaese) |
| History of exposure | a | b |
| No History of exposure | c | d |
odds of case having been exposed a/c
odds of control having been exposed b/d
OR = (a/c) / (b/d) = ad / bc
OR – cross products ratio
OR à ratio of the product of the two cells that support the hypothesis of association
(diseased people who weret exposed and non diseased people who were not exposed)
to the the product of the two cells that negate the hypothesis
(non-diseased people who were exposed and diseased people who were not exposed)
When Is the Odds Ratio a Good Estimate of RR
When following 3 conditions met
- cases studied are representative , with regard to history of exposure, of all people with the disease in the population from which cases are drawn.
- controls …..
- disease being studied does not occur frequently.
Examples of calculating odds ratio in case control studies
- calculating OR in unmatched case control study
- calculating OR in matched pairs case control study
No comments:
Post a Comment