Statistical and Epidemiological Concepts Flashcards
Define ‘absolute measure’.
Difference between exposure
groups.
Define ‘continuous outcome’ related to absolute measure.
Difference in mean levels of the outcome
(e.g. BMI, BP, HbA1c, cognition, lung function).
Define ‘binary outcome’ related to absolute measure.
Difference in occurrence % (or risk) of the outcome (e.g. death, hypertension, diabetes, dementia).
Define ‘relative measure’.
Ratio between exposure groups.
Define ‘continuous outcome’ related to relative measure.
Percentage change of the outcome.
Define ‘binary outcome’ related to relative measure.
Relative risk (RR) or odds ratio (OR) of
occurrence of the outcome.
How do you calculate Odds?
Number of cases divides by number of non-cases.
How do you calculate Odds Ratio (OR)?
Odds 1/odds 2
How do you calculate Risk?
Number of cases divided by total number of cases and non-cases.
How do you calculate Risk Ratio (RR)?
P1/P2
What is a confounding factor?
Factors that partially or completely account for the exposure-outcome
association (mask the underlying true association).
List features of confounding factors.
- Related to both outcome and exposure
- Often occur before the exposure
- Exists when the exposure-outcome association is altered after adjusting (or controlling) for these factors.
What is a mediator?
It is on the pathway between the exposure and the outcome.
List features of mediation.
- Occurs after exposure and before outcome
- Exposure-outcome association reduces (to close to zero in case of total mediation) after adjusting (or controlling) for the mediator.
What is an interaction (effect modifier)?
- A factor that changes the strength
of exposure-outcome association.
