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;
42 Cards in this Set
- Front
- Back
def. an event, condition, or characteristic that increasese the likelihood of the disease of advances its occurrence
|
cause
|
|
define necessary
|
cause of outcome MUST be present for the outcome to occur
|
|
defin sufficient
|
cause if present ASSURES an outcome
|
|
define component cause
|
event or condition that when combined with other agents or conditions will produce disease
|
|
is "does it have to be there?" necessary or sufficient?
|
necessary
|
|
necessary? sufficient?
mycobacterium TB --> TB |
yes no
|
|
necessary? sufficient?
guillotine --> death? |
no yes
|
|
necessary? sufficient?
smoking --> lung cancer |
no no
|
|
high chol --> CHD
|
no no
|
|
cigarettes + poor detox + poor diet = _______ cause
|
component
|
|
why do we need inference?
|
hard to really PROVE something
might need action we resort to inference for public health decisions |
|
Explain the levels of evidence
I II1 II2 II3 III |
I RCT
II1 randomization II2 cohort or case II3 series and uncontrolled trials III opinions |
|
What are the eight basic Hill guidelines for judging f something is causal
|
temporal relationship
strength of association dose response replication of findings biologic plasibiltity consistency with other knowledge cessation of exposure consideration of alternative |
|
what studies can establish a temporal relationship
|
clinical trials
cohort studies nested case control studies time series |
|
The ______ the RR the higher the likelihood that the association is causal
|
RR
|
|
T or F. The further the measure is from one the stronger the assoIciation
|
true!
|
|
is RR=.3 or RR=.8 stronger?
|
.3 (it's further from one)
|
|
How can you measure strength
|
RR
ARR Attr risk population attributable risk |
|
what is levin's fromla
|
pop AR = Pe (RR-1)/pe (RR-1) +1
|
|
T or F. Lack of dose response rules out causality
|
FALSE
|
|
define population attributable risk
|
proprotion of risk in general population that would be removed if exposure were eliminated
|
|
Pop AR =
|
(Ipop - I0)/Ipop
|
|
what are the main limitations of meta analysis
|
quantitative results give false sense of certainty
unrecognized heterogeneity publication bias |
|
define validity
|
ability of the test to distinguish who has the disease and who does not
|
|
define reliability
|
reproducibility of the test
|
|
does validity lack bias or random error? how about reliability?
|
validity lacks bias
reliability lacks random error |
|
what are two basic ways to measure validity
|
sensitivity and specificity
|
|
define sensitivity
|
true +/all disease
|
|
how do you setup table to find sensitivity and specificity
|
test gold standard
calculate going down |
|
what does PPV tell you
|
the proportion of the positive test results that are truly positive
|
|
For a given level of sensitivity/specificity, the PPV and NPV depend heavily on ______ of the condition
|
prevalnence
|
|
what does the ROC durve look like
|
sensitivity vs (1-spec)
the upper left the better the test |
|
LR =
|
sensitivity/(1-spec)
|
|
likelihood ratio combines info from ____ and _____ of a test
|
sensitivity and specificity
|
|
Post test odds = _____ x pretest odds
|
LR+
|
|
LR - =
|
(1-sense)/(spec)
|
|
T or F. The likelihood ratio is calculated as the ratio of probability of positive result in people with the disease over the probability of a positive result in people without disease
|
true
|
|
deifne likelihood ratio
|
how many times more likely a positive test result is to be found indiseased individuals compared to non diseased individuals
|
|
what is a disadvantage of cost benefit analysis
|
hard to put money value on life
|
|
what are the purporses of screening
|
treatment
counseling research forensic exclusion |
|
name some ethical elements for screening
|
test should have known sensitivity specificity and predictive value in a defined populations.
If treatment is goal, well designed peer reviewed studies shoudl establish benefits, risks, and cost. labeling should be presumed to have risks and no inherent benfits |
|
what are the two most common purposes of screening
|
treatment and counseling
|