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;
52 Cards in this Set
- Front
- Back
3 stages in epidemiological research
|
descriptive
analytic experimental |
|
a typical descriptive study
|
survey
|
|
2 typical analytic studies
|
cohort
case-control |
|
3 phenomena assessed in analytic epidemiology
|
host
agent environment |
|
4 typical host factors
|
genetics
behavior immunological state age |
|
numerator for period prevalence
|
total cases at beginning + all new cases
|
|
prevalence can be thought of as ___ times ___. an example of this is ___.
|
incidence
duration rise in HIV prevalence after tx became available |
|
incidence is useful for measuring ___ and assessing ___. it is used in ___ (2) studies
|
risk
disease etiology clinical trials cohort |
|
prevalence is useful for ___. it is used in ___ (2) studies.
|
planning health services
cross sectional case-control |
|
cumulative incidence is ___ divided by ___.
|
# of new cases during a time period
population at risk at the beginning |
|
incidence rate is ___ divided by ___.
|
number of new cases observed
total person-time of observation |
|
3 aims of descriptive epidemiology
|
describe natural history of disease
determine allocation of resource suggest hypotheses about disease causes |
|
direct age adjustment uses the ___ rates from a specfic population and applies them to the ___ population to obtain a ___ rate.
|
age-specific
standard age-adjusted |
|
2 kinds of studies
|
descriptive
analytical |
|
2 kinds of analytical studies
|
experimental
observational |
|
3 prospective studies
|
clinical trial
cohort historical prospective |
|
___ is the most common way to estimate risk
|
cumulative incidence
|
|
T/F: at the beginning of a cohort study, some of the cohort has the disease
|
false
|
|
the risk measure in a cohort study or ___ is ___. it is the ratio of ___ to ___.
|
randomized control trial
relative risk incidence with exposure incidence without exposure |
|
etiology of disease is assessed using risk ratio/difference
|
ratio
|
|
T/F: cohort studies are suitable for rare diseases
|
false
|
|
5 disadvantages of cohort study
|
expensive
large sample needed takes long time not suitable for rare diseases bias due to high dropout rate (loss to follow-up) |
|
single blinding means
|
patient doesn't know whether they're in tx or control
|
|
double blinding means
|
patient and person treating them don't know whether pt is in tx or control
|
|
triple blinding means
|
pt, person treating them, and data analyst don't know
|
|
number needed to treat (NNT) in terms of absolute risk reduction
|
NNT = 1/(ARR)
|
|
in case-control studies, ___ is used to approximate relative risk
|
odds ratio
|
|
formula for odds ratio
|
OR = (cases exposed*controls not exposed)/(cases not exposed*controls exposed)
|
|
in case-control study you should use ___ to measure cases
|
incidence
|
|
little improvement in precision is achieved once control/case ratio is above ___
|
4
|
|
4 disadvantages of case-control study
|
prone to recall bias
prone to selection bias in controls only an approximation of risk not suitable for rare exposures |
|
in cross-sectional study risk is measured by ___ (2)
|
prevalence odds ratio
prevalence ratio |
|
4 disadvantages of cross-sectional study
|
no temporal relationship
biased towards long-duration cases no risk ratio not suitable for rare diseases or exposures |
|
T/F: systematic error does not decrease with increased sample size
|
true
|
|
3 sources of systematic error
|
selection bias
information bias confounding |
|
3 a priori criteria for confounders
|
risk factor for disease
associated with exposure in source population NOT an intervening variable in the causal pathway |
|
3 ways to control confounding in experiment design
|
restriction of study
randomization matching |
|
3 ways to control confounding in data analysis
|
restriction of analysis
stratification multivariable regression |
|
validity is aka
|
accuracy
|
|
reliability is aka ___ (2)
|
repeatability
precision |
|
___ is a function of systematic errors
|
validity
|
|
___ is a function of random errors
|
reliability
|
|
___ errors are reduced with increasing n
|
random
|
|
T/F: bias only affects external validity
|
false
|
|
1st 5 of Hill's criteria for causality
|
strength of association
consistency of association specificity: single effect of exposure temporality biological gradient (dose-response) |
|
last 4 of Hill's criteria for causality
|
biological plausibility
coherence with literature experiment analogy |
|
___ determines the gain from performing a test
|
difference between pre-test and post-test probabilities
|
|
pre-test probability is estimated by ___
|
prevalence
|
|
post-test probability is estimated by ___ if test is positive, ___ if test is negative
|
PPV
1-NPV |
|
solution to lead-time bias in evaluating screening programs is ___
solution to length-time time bias in evaluating screening programs is ___ |
use mortality instead of survival time
use RCT |
|
screening test should have high ___
|
sensitivity
specificity PPV |
|
recall bias is an example of ___ bias
|
measurement/infomation
|