Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
71 Cards in this Set
- Front
- Back
case control study design
|
compare group of people w/ disease to group w/o disease
|
|
case control study measures?
|
odds ratio
|
|
cohort study design
|
compare group w/ given exposure or risk factor to group w/o
|
|
cohort study measures?
|
relative risk
|
|
cross sectional study design
|
collect data from grp of people to assess frequency of disease
|
|
cross sectional study measures?
|
disease prevalence
|
|
twin concordance study design
|
compare frequency twins develop disease
|
|
twin concordance study measures?
|
heritability
|
|
adoption study design
|
compare siblings raised by biologic vs adoptive parents
|
|
adoption study measures?
|
heritability and influence of environment
|
|
Phase I clinical trial
|
small # pts (healthy); assess safety, toxicity, pharmokinetics
|
|
Phase II clinical trial
|
small # pts w/ disease; assess efficacy, dosing, adverse effects
|
|
Phase III clinical trial
|
large # pts randomly assigned to drug or placebo; compare new treatment to standard of care
|
|
Phase IV clinical trial
|
post marketing surveillance; detect rare or longterm adverse effects
|
|
meta-analysis
|
pools data from several studies to come to overall conclusion; greater power
|
|
Sensitivity definition
|
ability to detect disease when present; rule OUT, for screening
|
|
Sensitivity formula
|
TP/ TP + FN
|
|
Specificity definition
|
ability to test to indicate no disease when disease is not present; rule IN, for confirmation
|
|
specificity formula
|
TN / TN + FP
|
|
PPV definition
|
probably that person has disease given a (+) test result
|
|
PPV formula
|
TP/ TP + FP
|
|
NPV definition
|
probably that person is disease free given a (-) test result
|
|
NPV formula
|
TN/ FN + TN
|
|
point prevalence
|
total cases @ given time / total population
|
|
incidence
|
new cases over period / total population at risk
|
|
prevalence is approx equal to?
|
incidence x disease duration
|
|
prevalence
|
incidence for acute disease
|
|
odds ratio
|
odds of disease in exposed group / odds of disease in unexposed group
|
|
odds ratio approximates what? Under what conditions?
|
approximates RR if incidence is not too high
|
|
relative risk
|
relative probablilty of getting a disease in exposed grp compared to unexposed grp
|
|
relative risk formula
|
% disease in exposed grp / % disease in exposed grp
|
|
attributable risk
|
difference in risk b/t exposed + unexposed; OR proportion of disease occurrences attibutable to the exposure
|
|
absolute risk reduction (ARR)
|
reduction in risk assoc w/ treatment as compabe dto placebo
|
|
NNT
|
1/ ARR
|
|
NNH
|
1/AR
|
|
Precision
|
consistency + reproducibility of a test (reliability); absence of random error
|
|
what reduces precision?
|
random error
|
|
accuracy
|
trueness of test (validity)
|
|
what reduces accuracy?
|
systematic error
|
|
Selection bias
|
nonrandom assignment to a study group
|
|
Recall bias
|
knowledge of presence of disorder alters recall by subjects
|
|
sampling bias
|
subjects are not representative to general population
|
|
late-look bias
|
information gathered at inappropriate time
|
|
procedure bias
|
subjects in different grps not treated the same
|
|
confounding bias
|
effect of 1 factor distorts or confuses effect of the other
|
|
how to control confounding?
|
use matching (in case-control studies)
|
|
lead time bias
|
early detection confused w/ increased survival
|
|
pygmalion effect
|
when researcher's belief in efficacy of a treatment changes outcome of treatment
|
|
Hawthorne effect
|
when group being studied changes its behavior ownign to knowledge of being studied
|
|
Positive skew
|
mean > median > mode; tail on right
|
|
Negative skew
|
mean < median < mode; tail on left
|
|
which is least affected by outliers? Mean, median, or mode
|
mode
|
|
Null hypothesis
|
no difference
|
|
Alternative hypothesis
|
there is a difference
|
|
Type I error
|
stating there IS an effect/difference when there really isn't; = p
|
|
Type II error
|
stating there IS NOT an effect/difference when there really is; = β
|
|
which error is considered "false positive"? Which is "false negative?"
|
false (+) = type I error; false (-) = type II error
|
|
Power, definition
|
probability of rejecting null hypothesis when it is in fact false, or likelihood of finding a difference if one in fact exists
|
|
Power depends on?
|
1) total # of end points 2) difference in compliance b/t grps 3) size of expect effect
|
|
how to increase power?
|
bigger sample size
|
|
power formula
|
1-β
|
|
what is n?
|
sample size
|
|
what is σ?
|
standard deviation; = sqr(variance)
|
|
what is SEM?
|
standard error of the mean; = σ/sqr(n); how much variation in mean from sample to sample
|
|
what does it mean if the CI b/t 2 grps overlaps?
|
groups are not significantly different
|
|
confidence interval
|
range of values in which a specific probability of the means of repeated samples is expect to fall
|
|
t-test
|
checks difference b/t means of 2 groups
|
|
ANOVA
|
checks difference b/t means of 3 or more groups
|
|
chi squared test
|
checks difference b/t 2 or more % or proportions of categorical outcomes
|
|
correlation coefficient r
|
b/t -1, 1; closer to 1 = stronger correlation b/t 2 variables
|
|
coefficient of determination
|
r^2
|