Study design Flashcards
observational vs experimental study
researcher observes exposure vs assigns it
Descriptive studies measure…
what, association, not causality
Case report/series
Detailed report on a patient/group and their cases. Simple description without comparison group.
Case report/series are good for?
cons?
generating hypothesis
maybe biased selection
Cross sectional studies are?
a survey/measure of one point in time
Cross sectional studies pros/cons
pros: can look at specific population of interest
cons: potential bias if people choose not to participate
Ecological studies study what?
Groups, not individuals
trends over time
Ecological studies pros/cons
pros: good for hypothesis generating, uses available data (no gathering)
cons: correlation does not equal causation (cant link risk factors to disease)
Observational analytical studies study what?
compares groups but researcher has no control over assignment
Cross-sectional analytical studies measure?
prevalence
strength of cross-sectional analytical studies? Cons?
Can look at specific population prevalence
Cons: not for rare/short lasting diseases, cannot determine temporal relationship between risk factors and disease onset
What is a case-control study how done? What data obtained?
ID subjects by outcome at the beginning of study, control group from same exposure but no outcome. Data about exposure to a risk factor. Is retrospective–outcome already known, and is observational–no intervention.
Case control vs cohort study?
IN cohort, take exposed and non-exposed group and follow until disease occurs, compare incidence. Prospective (from present to future). Can also be retro- and bi-directional.
In Case-control. Take diseased and non-diseased and compare odds of exposure/prevalence between groups. Retrospective (from outcome back)
Case-control good for?
Bad for?
good for rare and long diseases
subject to bias, incidence can’t be calculated bc you are selecting groups, temporal relationship between exposure and disease not clear.
Cohort studies pros
cons
Can establish timing, assess causality
Can standardize selection
Can calculate rates/incidence
Good for rare exposures
Potential unknown confounders
exposure status and follow-up success may change