Lecture 4 (Cohort Studies) Flashcards
Cohort study
Grp exposed/not exposed are followed and compared w/respect to incidence rate of disease
Other names for cohort study
longitudinal study
follow up studies
pro/retrospective studies
Pros of cohort studies
- Gold standard among obsv studies (exposure precedes onset of disease, necessity of causality)
- No recall bias
- Not assuming measurement apply at any other point in time
Cons of cohort studies
- Large # of participants
- Many yrs of follow up ($)
- Losses to follow up
- Obsv study: hard to determine causality
Cohort study: types of comparison grps
- Internal: cohort includes exposed/unexposed
- External: entire cohort exposed, need external grp for comparison
Types of cohorts
- Defined by exposure (tests hyps, study rare exposures)
- Defined by factor unrelated to exposure (research platform, sample of pop based on willingness)
Relative Risk
-For analysis of cohort study
RR=1 : equal risk
RR > 1: risk exposed > risk non-exposed
RR < 1: risk exposed < risk non-exposed (protective effect)
*But association =/= causation
I (exp) / I (non-exp)
Attributable Risk
-The magnitude of disease incidence attributable to a specific exposure
Background exposure
-Incidence in non-exposed grp, occurs regardless of exposure
AR increases w/
- Increasing RR
- Increasing pop prevalence of exposure
Prospective cohort study
- Concurrent
- Assemble cohort/measure exposures at present, follow into future
- Pros: designed specifically for study (ask whatever)
- Cons: timely and $
Retrospective cohort study
- Uses historical data f/existing records to go back in time and assemble a cohort
- Pros: less $, less time
- Cons: reliance on records for other purposes –> inferior quality data
Indirect age-adjustment comparison measures
Standardized mortality ratio (SMR)
= (observed deaths/expected deaths) x 100
Standardized incidence ratio (SIR)
= (observed cases/expected cases) x 100
*Both relative risks
