Reading...
Play button
Play button
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;
31 Cards in this Set
- Front
- Back
|
cohort
|
group of at-risk individuals followed over time for the development of the endpoint of interest
|
|
|
study base
|
person-time at risk contributed by the study population
|
|
|
cumulative incidence
|
risk=1- s(t), or events/number at risk
|
|
|
incidence rate
|
total number of events during T/total person-time at risk during T
|
|
|
prevalence
|
number of existing cases/size of population
|
|
|
odds
|
rate of disease/rate of no disease
|
|
|
when cumulative incidence is very low, odds ~
|
approximates cumulative incidence
|
|
|
Prevalence, at steady state
|
incidence * disease duration
|
|
|
R(t), where incidence is constant over time and I*t<0.1
|
~I*t
|
|
|
confidence intervals
|
difference measure +/- 1.96*SE(difference measure), or e^(ln relative measure +/- 1.96*se(ln relative measure)
|
|
|
risk difference or cumulative incidence difference
|
R1-R0
|
|
|
incidence rate difference
|
I1 - I0
|
|
|
cumulative incidence ratio
|
r1/r0
|
|
|
incidence rate ratio
|
I1/I0
|
|
|
relative excess risk
|
[r1(t) - r0(t)]/ro(t), or RR - 1
|
|
|
relative excess rate
|
(incidence1 - incidence0)/incidence0, or IR -1
|
|
|
attributable risk %
|
[r1(t) - r0(t)]/r1(t), or (RR-1)/RR
|
|
|
attributable rate %
|
(I1 -I0)/I1 = (IR -1)/IR
|
|
|
PAR %
|
Rtotal(t) - R0(t)/Rtotal(t), or [P*(RR-1)]/[P*(RR-1) +1]
|
|
|
components of error in a measured value
|
bias + random error
|
|
|
net sensitivity in parallel testing
|
probability of being in the a cell for either tests
|
|
|
kappa
|
observed agreement - agreement expected by chance alone/(100% - agreement expected by chance alone)
|
|
|
coefficient of variation, intraclass correlation
|
measures of reliability
|
|
|
cluster as a method of probability sampling from the source population
|
randomly sample from a set of geographic regions, then randomly sample individual from within the selected regions (often used in population health research in low resource settings or during disasters)
|
|
|
fixed cohort
|
cohort is closed on entry alone
|
|
|
strengths of prospective cohort study design
|
collect information in real time, an efficient design when exposure is rare, can study many exposures
|
|
|
limitation of prospective cohort design
|
often requires very long follow up and thus is labor intensive and expensive
|
|
|
active v passive follow-up in cohort studies
|
active: maintain contact with the participants throughout follow-up to ascertain changing exposure and outcome status v. no contact with participants use registries and other records to ascertain outcome
|
|
|
purpose of counterfactual
|
allows us to understand that to identify a causal association btwn exposure and disease, we need to select groups of individuals that differ only on exposure, and make a comparison of the risks or rates btwn two exposures states
|
|
|
optimal comparison group
|
the entire population of a country within strata of age and calendar time WHEN EXPOSURE IS RARE, estimate the incidence rate to reflect the at-risk experience of unexposued individuals
|
|
|
risk in a closed cohort:
|
cumulative incidence with denominator being all cases at risk at beginning
|
