Case-control Studeis Flashcards
Table for IRR
Case | control
Exposed A. | B
Non-exposed C. | D
The rare disease assumption –> IRR ~ AD/CB
Odd ratio
OR
OR is an approximation of the IRR
Under the rare disease assumption IRR ~ AD/CB = the OR
95% CI
OD = ad/CB
Error factor = e^2x(square root of) 1/a + 1/b + 1/c + 1/d
95% CI: (OR/ef) to (OR x ef)
It is worth increasing the number of controls up to a point, typically up to 4 to 6 times as many controls as there are cases
Nested case-control studies
Advantages over conventional case-control study - incidence rates can be calculated, population for sampling of controls is already defined
Advantages over conventional cohort studies - can collect more detailed information from a minority of participants
Key issues for case-control studies
selection bias
- cases should be representative of all cases
- controls should be representative of the population
information bias
- systematic error of wrongly categorised as exposed
confounding
- minimised by matching using important con founders
- adjusted for by analysis with logistic regression
Comparison of Cohort and Case-control studies
COHORT STUDIES
- based on exposure status
- time-consuming and expensive
- not good for rare outcomes or diseases
- study a range of outcomes or diseases for each exposure
- limits bias but prone to losses to follow up
- can establish that the exposure precedes the outcome/disease
- directly measure incidence
CASE-CONTROL
- based on outcome status
- quicker and cheaper
- not good for rare exposure
- range of exposures for a single outcome of disease
- prone to selection and information bias
USE NESTED CASE-CONTROL STUDIES
Conducting a case-control study
analysis always looks backwards in time
Identify a group of cases
Identify a group of non-cases (controls)
Ascertain pervious exposure status of everyone
Compare level of exposure in cases and controls
