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;
54 Cards in this Set
- Front
- Back
rate
|
number of cases/population/per unit time
|
|
measures of moribity:
|
a.incidence rate
b.attack rate c.secondary rate d.prevelance e.incidence d.point prevlance e.peroid prevelance |
|
Incidence rate
|
# of new cases/population (per time)
*10^5 |
|
Attack rate
|
# of new cases during epidemic/population at the start of the epidemic
*10^2 |
|
Secondary attack rate
|
# of new cases among contacts of known cases/size of at risk population
*10^2 *# ill/#contacts |
|
Prevlance
|
most effective in chronic illnes
|
|
Incidence
|
short duration
|
|
point prevalance
|
# current cases/estimated population
*10^5 |
|
peroid prevelance
|
# current cases in time interval/est pop
*10^5 |
|
Confidence Interval
|
%95 measns that you are 95% condifent all of the time
*so 95% of the time, you'll fall in the range *5% of the time, you'll be o/side the range |
|
Risk ratio
|
risk for group of interest: risk for comparison group:
1. coleslw ex-at coleslw 9 sick, 1=not sick attack rate9/10=90% 2. didn't eat coleslwa 1=sick 9=not sick attack rate=1/10=10% *so risk ratio is .9/.1=9 so ppl who ate the coleslwa re 9x as liely to get sick as the group that didn't eat the slaw |
|
Odds ratio
|
estimates the RR-is good when disease is rate
*(a*d)/(b*c) |
|
Measures of morality
|
a. crude death rat
b. specific death rate c. proptionate moraltiy d. case fatality rate |
|
a. crude death rate
|
#deaths/population
10^3 or 5 |
|
b. cause-specific death rate
|
# deaths d/t cause//population
*10^3 |
|
c. porportionate moratlity
|
#deaths d/t cause//total deaths
*10^2 |
|
d. case fatality rate
|
# fatalities d/t case/# cases
|
|
Case definition
|
used yb epidemologist s to define who has the disease*always look for this in case deinition in study
1. used for surveillance 2. used for studies-either observational or experimental 3. decides who they're going to include |
|
what's included in descriptive epidemiolgy
|
who
what where when |
|
What's included anaytical studies
|
1. experimental study includes the study design*** (see more notecards)
|
|
Study design: includes lots of things--anme the first 2
|
1. who are the subjects-establish entry criteria
2. need for a comparision gorup-new drug vs. old drug or placelbo |
|
What are the next three things a study design involves:
|
3. random assignment of treatment (not always possible)--observational comparison are rarely definitive
2. compulsive c+-treat grps identical 3. crossover deigsn-subj serves as their own contorl*bad if they drop out |
|
What the 6th part of the study disgn?
|
Intention to treat..
a. same outcome as income b. favors null hypthesis where there's no relationship bw exposure and outcome |
|
The 7th part of study design is outcome and includes what 4 things?
|
1. it must be measurable-death, lab results
2. must be clincaly relevant to health 3. How sure are we that there is a diff bw pops? 4. P<0.05-less than 5% chance that the difference is due to coincidence--statisically different |
|
Describe descriptive vs. analytical epidemology
well describe descriptive first |
descriptive-asks: who, waht when, where, when, where
1. there's no comparison group 2. measures morbidity and morality 3.use descriptive epidemology to suggest hypothesis, NOT TO TEST THEM!! |
|
Analytical
|
Asks: how, why
1. there is a comparison group 2. measures association..OR, RR 3. 2 types: observational(where OBSERVES and measures risk factor in population)and experimental |
|
What are the three types of observational studies:
|
a. cohort
b. case-controlled study c. cross sectional study |
|
Desribe cohort study
|
start w/dz free pop; good w/common dz
1. follow a pop over time and look at risk factors ass w/devleopment of dz 2. has limiations: need larg N, expensive, time consuming; lose subjects d/t floow up 3. advangtes: it decreases bias |
|
Describe case-control study
|
start w/dz pop; good w/rare dz; look at risk factors in diseased vs. non-diseased populations
1. limitations-accuracy, did dz cause rf or vice versa? 2. advantages-quick,cheap, more efficient for rare dz-use OR to estimate RR |
|
Describe cross-sectional study
|
not used often, good if money is low;
1. describes pop at time of study 2. exposure and dz are look at toghether 3. limiations 4. CANT MAKE CAUSAL INFERENCE-this is bc you don't know which came first-the risk factor or dz-- 5. advangtes are cheap and quick |
|
What is the other type of analytical epide study that's not observational?
|
experimental-where epi/scientist controls the exposure in the pop
|
|
Reiteration: RR
|
How big is the difference b/n the 2 groups?
RR=16 95% CI (8-667); then DZ is 16x more likey in exposed vs. unexposed *strong ass bc there's an increase RR and CI is far from 1 |
|
What does it mean if RR=3 with 95% (.2-.8)
|
There's a reduction in disease in 70% of ppl who were exposed; exposure was protective
*strong associ |
|
confidence intervals
|
how big is the zone of uncertainity around the RR
1. CI gives margin of error-gives the interval that your results will fall 95% of the time 2. if CI includes 1-there is no association (.7-1.9) |
|
What is casuation?
|
an approach to scientific inference that's easier to test in a population versus an inviduval
|
|
Causation: inductivisim
|
we make the same observation repeatedly then we make generalizations from it-
1. obseravation leads to generalization 2. rooster causes the sun to come up |
|
Casuation: refutationism
|
kill the rooster and the sun comes up--i hae refuted the hypthesis
|
|
Causation: consensus:
|
scienfitic community determines validity of theories
|
|
Causation: how does science advance?
|
by process of conjecture and refutation--therize then disprove
|
|
Hill Casual Criteria
|
developed by Bardford Hill: all scienfic work is incomplete
|
|
strength of association
|
a.strong assoc are more likey to be causal
b. weak assoc are more liley to be d/t bias |
|
Consisntency
|
lack of consistency doesn't rule out causation
|
|
specificty
|
single cause leads to single effect
|
|
specificity
|
single cause leads to a single effect
ex.vinyl chorlide-leads to liver cancer |
|
temporality
|
cause precedes effect
|
|
biologic gradient
|
more experiments-more disease
1. monotonic dose-response curve |
|
plausability
|
is there a logical biological explanation
|
|
coherence
|
cause and effect hypthesis gives a choherent explanation for evidence
|
|
experimental evidence
|
hypothesis is consistent w/results
|
|
analogy
|
the similarity bw this hypthesis and accpeted hypthesis
|
|
Surveillance:
|
ongoing collection of health data: involves three types
1. passive 2.active 3. enhanced |
|
passive ****
|
most used; health dept waits for cases to be rpsoted by docs/hospitals
|
|
active
|
health dept contacts docs by person/phone
|
|
enhanced passive
|
health dept mails/faxes sheets to docs requesting data
|