Term
|
Definition
Positive Skew - tail to the right
Mean > median |
|
|
Term
|
Definition
Negative Skew - tail to the left
Mean < median |
|
|
Term
|
Definition
| √((x1 - xmean)2.../(n-1)) |
|
|
Term
|
Definition
|
|
Term
|
Definition
total cases of dx at time t /
total # of at risk individuals in population |
|
|
Term
| Cumulative Incidence (CI) |
|
Definition
# of New cases of dx over a time period/
Population at risk over time period |
|
|
Term
|
Definition
# of new cases of dx over a time period/
Sum of each persons’ time at risk (person-time) over time period |
|
|
Term
|
Definition
CI = IR * time
only if IR is constant over time period |
|
|
Term
|
Definition
# deaths due to disease/
# cases of disease |
|
|
Term
|
Definition
# of deaths over time period/
total person-time in time period |
|
|
Term
|
Definition
# deaths from disease in a population
total # of deaths from all reasons |
|
|
Term
|
Definition
# deaths from disease in a population
total # of deaths from all reasons |
|
|
Term
|
Definition
| doesn't account for age or population |
|
|
Term
|
Definition
| crude rate * standard population = expected # of events for each population |
|
|
Term
|
Definition
Risk Diff = Cumulative incidence (exposed) – cumulative incidence (non-exposed)
Rate diff = Incidence rate (exposed) – incidence rate (non-exposed) |
|
|
Term
|
Definition
Cumulative incidence (exposed)/
Cumulative incidence (non-exposed) |
|
|
Term
|
Definition
Incidence rate (exposed)/
Incidence rate (non-exposed) |
|
|
Term
|
Definition
Odds of being exposed (among diseased)/
Odds of being exposed (non-diseased) |
|
|
Term
|
Definition
| IR(exposed) - IR(non-exposed) |
|
|
Term
|
Definition
IR(exposed) - IR (non-exposed)/
IR(exposed) (x100) |
|
|
Term
|
Definition
| 1/absolute risk reduction |
|
|
Term
| Population attributable risk |
|
Definition
| IR(pop) - IR(non-exposed) |
|
|
Term
| Population attributable risk % |
|
Definition
IR(pop) - IR(non-exposed)/
IR(pop) (x100) |
|
|
Term
Case Report Study/series
pros & cons |
|
Definition
Pros: cheap, can ID rare / new diseases - early warning
Cons: no comparison, small n, could be exception |
|
|
Term
Ecological studies
pros & cons |
|
Definition
Pros: inexpensive, data already available, can ID potential risk factors / generate hypothesis
Cons: ecological fallacy, can't test hypotheses |
|
|
Term
|
Definition
Pros: just at one time point so relatively cheap & quick
Cons: prevalence bias, selection bias, hard to study if there are rare diseases/exposures |
|
|
Term
|
Definition
Pros: rare exposures can be followed, can look at multiple outcomes, less selection bias, can find incidence rates, can look at temporality
Cons: expensive, long f/u, hard to study rare outcomes, loss to f/u |
|
|
Term
|
Definition
Pros: cheapest, can study rare dxs, OR estimates RR
Cons: can't get relative risk, prevalence bias, |
|
|
Term
|
Definition
Pros - can determine causality, reduces bias, multiple outcomes, measure incidence rates
Cons - most expensive, placebo effect, loss to f/u, non-compliance |
|
|
Term
|
Definition
true+/all+ diagnoses
a/a+c
test's ability to correctly ID individuals who have the disease |
|
|
Term
|
Definition
true-/all- diagnoses
D/B+D
tells you test's ability to correctly ID individuals who truly don't have dx |
|
|
Term
| Predictive Value Positive |
|
Definition
A/A+B
true+ / all+ tests
If test positive, what is chance that have disease |
|
|
Term
| Predictive Value Negative |
|
Definition
true-/all- tests
D/C+D
If I test negative, how sure are you I don't have disease |
|
|
Term
|
Definition
| prevalence of untreated dx in population |
|
|
Term
| Posttest probability after positive test |
|
Definition
| A/A+B (same as predicted value positive) |
|
|
Term
| Posttest probability negative |
|
Definition
C/C+D or 1-PVN
What is the chance that I have the disease after a negative test |
|
|
