Final - Wk 9 (observational studies) Flashcards
True/False:
– P-values can tell you if the mean difference b/w groups in a study is statistically significant and if there is bias present????
false
Which type of study would be the best to assess associations in a rare disease?
case control
Observational vs Experimental:
– Which study design category do investigators have control of variables?
experimental
Observational vs Experimental:
– Which study design category do investigators not have control of variables?
observational
Observational vs Experimental:
– Which study design category is best to use if there are many unknown factors?
observational
Observational vs Experimental:
– Which study design category is best for investigating effect of single factors?
experimental
Observational vs Experimental:
– Which study design category is often the preferred design for studies of adverse events.
observational
Descriptive vs Interferential:
– Which type of analysis is useful mainly for prevalence and incidence - to characterize the population?
descriptive
Descriptive vs Interferential:
– Which type of analysis is useful for any studies looking at cause and effect or associations among risk factors (like OR and RR)?
interferential
Descriptive vs Interferential:
– Which type of analysis cannot test a hypothesis?
descriptive
Descriptive vs Interferential:
– Which type of analysis is used in testing hypotheses?
interferential
Cohort studies start with ____ and follow group over time to determine ____.
- exposure/risk factor
- - outcome
What are the 2 outcome measures of cohort studies?
relative risk or risk ratio
How do you interpret a relative risk ratio of:
– PR > 1
increased risk of outcome
How do you interpret a relative risk ratio of:
– PR = 1
No risk of outcome
How do you interpret a relative risk ratio of:
– PR < 1
Reduced risk of outcome
What is the correct interpretation of a RR of 1.36?
RRR = [ 1 - RR (1.36) ] x 100
– so RRR = risk of disease is increased by 36%
What is the interpretation of a RR of 0.80?
RRR = [1 - 0.8] = 0.2 x 100 = 20%
– Risk of the outcome in the exposed group was reduced by 20% (or occurred 20% less) relative to the unexposed group
What is the interpretation of a RR of 3.3?
RRR = [1 - 3.3] = 2.3 x 100 230%
– Outcome was 230% more likely to occur in the exposed group
What type of cohort study is this:
– Cohort with/without risk factor of interest followed forward in time to record outcome?
Prospective
What type of cohort study is this:
– Cohort is followed backward in time to examine risk factors present years in the past?
Retrospective
What are 3 advantages of cohort studies?
- show the natural course of a disease or condition
- show how Tx work in “real world” settings
- less possibility for bias
What are 2 limitations of cohort studies?
- expensive
- - cannot prove cause and effect as well as experimental designs like RCT
Which is able to suggest cause and effect than case-control or cross-sectional studies:
– Case-control OR cross-sectional study OR cohort study
cohort study
Case-control studies look at the difference between group ____ the disease/condition and one group ____.
- with
- - without
Case-control studies are useful for what type of conditions?
rare conditions
Can case-control studies use what types of statistics?
odds ratio
How many times more likely the odds of finding an exposure in someone w/ disease is compared to finding the exposure in someone w/o the disease.
odds ratio
What does an Odds Ratio > 1 indicate?
OR > 1 = increased frequency of exposure
What does an Odds Ratio = 1 indicate?
OR = 1 no change in frequency of exposure
What does an Odds Ratio < 1 indicate?
OR < 1 = decreased frequency of exposure
What is the purpose of cross-sectional studies?
look at prevalence
What is a limitation of cross-sectional studies?
can’t tell whether exposure preceded disease
What are 2 big problems of cross-sectional studies?
recall bias and self-reported bias
If you want to evaluate effectiveness, you must have a ____ group.
comparison group
In order to determine if cause precedes effect, measurements must be made when?
at least 2 points in time