Studies, epidemiology Flashcards
What does Relative Risk = 1 mean
No association between exposure and disease
What does Relative Risk < 1 mean
Greater risk in exposed vs non-exposed
may have causal relationship
What does Relative Risk > 1 mean
Lower risk in exposed vs non-exposed
may have protective relationship
Describe a cross sectional study
- pros
- cons
Choose people and simultaneously measure their EXPOSURE and health outcomes (DISEASE)
PROS
-obtain disease prevalence
CONS
-cannot determine causation
Describe a cohort study
- pros
- cons
Studies EXPOSURE
PROS
- can test causation
- can test associated with time
CONS
-expensive
Describe a case control study
- pros
- cons
Studies DISEASE
PROS
-can test causation
CONS
- selection bias
- recall bias
Describe a randomised control study
- pros
- cons
Studies EFFECTS OF INTERVENSIONS
PROS
-best for valid results
CONS
-sample size may be too large to study small differences
Describe recall bias
Participants differ in the wall they remember data on outcomes
Describe selection bias
The way the participants are selected is not representative of the entire population
Describe information bias
Difference in way exposure or outcome is measured between groups
Bradford hill criteria for causation
-presence of these factors means that relationship is likely causal
- Temporality
- Plausibility
- Consistency/ replicability
- Strength
- Dose-response relationship
- Specificity
- Reversibility on cessation of exposure
- Coherence
- Not due to alternatiive explanations
What is temporality
Exposure to factor must happen before disease develops
What is considered consistent/ replicable
Able to replicate the association in different studies with different samples and different investigators
What is dose-response relationship
People exposed to increasingly higher levels have increasingly higher risks of disease
What is specificity in the context of Bradford hill criteria
1 to 1 relationship between single exposure and single disease
