Study designs: analytical Flashcards
Cross sectional study: describe
take exposure and outcome data AT THE SAME TIME
relationship can be seen between them
but can’t determine direction
do cross sectional studies show prevalence
yes
do cross sectional studies show incidence?
no - just a snapshot
2 additional advantages of cross sectional studies
quick/cheap/easy
generate hypotheses
are cross sectional studies good for rare diseases, why so
no
as they’re rare and unlikely to come up
example of cross sectional study for teenage alcohol consumption vs depression?
teens were surveyed about whether they consume alcohol and whether they’ve been diagnosed with depression.
describe case control studies
info collected at same time but
case/control data is specifically looked at first. (often matched)
then exposure prevalence is determined within cases and controls.
what statistic does case control studies look at
odds ratio
can you determine incidence with case control
no
is case control good for rare diseases? why so
yes - as you can select them from the start
additional advantage and disadvantage of case control
helps determine whether an exposure is a risk factor/potential cause
recall bias (when determining exposure) selection bias (if making mathces)
case control example of teenage alcohol consumption vs depression?
teens with and without depression are identified and matched based on other characteristics. Prevalence of alcohol consumption is determined in cases and controls.
cohort study: common ground?
starts with exposure (or not), then looks at whether cases/controls come about
what statistic do cohort studies look at
relative risk
describe prospective cohort study
disease free cohort
look at exposure in present time
then follow over time to see if people develop disease
if you increase time period, what’s a pro and con?
more info can be collected
but higher chance of loss to follow up
are PCS good for rare risk factors? why so
yes as you can pick them as a starting point
are PCS good for rare diseases? why so?
no. as they’re unlikely to develop
are PCS good for common diseases? why so
yes. as they’re likely to develop
why is prospective cohort study best evidence for causation out of analytical studies
as you can follow disease progression in real time
two other disadvantages of prospective cohort studies
confounding
selection bias
prospective cohort study example of teenage alcoholism and depression
start with population of teens without depression. Whether individuals consume alcohol or not are noted. Individuals in each group are tracked to see if they develop depression
describe retrospective cohort study
look at past exposure. then follow up in future to see if disease develops