WT - Biostats Flashcards
Studies with controls
clinical trial
case control
prospective/retrospective cohort
Studies w/ patients as their own controls
crossover study
Studies without control
cross-sectional
case series
study typically determining odds ratio
case control
-compares group with disease to a group w/o disease
study typically determining relative risk
cohort
-compares group with a given exposure to a group w/o such exposure
study typically determining disease prevalence
cross sectional
study comparing effect of a series of 2 or more tx on a participant
crossover study
study examining natural hx
case series
cohort studies are organized by
- selecting group who do NOT have disease of interest
- det their exposure status
- following to assess for development of disease
Odds ratio
ad/bc
measure of strength of association between exposure and outcome
Relative risk
(a/(a+b))/(c/(c+d))
probability of outcome occurring in exposed group vs unexposed group
relative risk reduction
1-RR
1-(a/(a+b))/(c/(c+d))
attributable risk
(a/a+b)-(c/c+d)
absolute risk reduction
(c/c+d)-(a/a+b)
NNT
1/ARR
lower #, better tx
NNH
1/AR
higher #, safer exposure
odds ratio/relative risk
measure strength of assoc btwn 2 categorical variables
exp
- categorical risk (smoke)
- categorical outcome (MI)
ANOVA
compares mean of quantitative variables of _/> 3 indep groups
exp
- serum ferritin levels in 3 dif age groups
- FEV1 in light/med/heavy smokers
results of _________ will be equiv when comparing dif btwn means of 2 groups
(2 tests)
t-test and ANOVA
regression analysis
describes effect of 1 or more independent variables (exposures, risk fx) which may be QUANT or QUAL, on one QUANT dependent variable (binary outcome)
exp
-egg and meat consumption affecting [serum ferritin]
chi square
compares 2 categorical variables (large sample size)
exp
FEV1 and smoking status
fisher exact test
compares 2 categorical variables (small sample size)
exp
FEV1 and smoking status
2 sample T-test
compares the dif between the numerical means of 2 groups
normal distribution
16 - 68 - 16
- 5 - 95 - 2.5
- 15 - 99.7 - 0.15
odds/risk ratio including __ –> exposure/risk is NOT significant outcome
1
if CI includes ___ or if ____, then no statistical diff between groups
zero
or if CI overlap between 2 groups
