Causal Relationships Flashcards Preview

Epidemiology > Causal Relationships > Flashcards

Flashcards in Causal Relationships Deck (15)
Loading flashcards...
1

An event, condition, or characteristic w/o which the dz. would not have occurred.

Cause

2

What 3 factors do descriptive epidemiology aim to address?

  • Agent factors
  • Environmental influences
  • Host factors

(all interact)

3

List 3 Agent Factors

  • Infection Rate
  • Pathogenicity
  • Virulence

4

What are the 8 criteria for identifying cause-effect relationships?

  • Strenght of association
  • Temporal relationship
  • Dose-response relationship
  • Biological plausibility
  • Consistency
  • Elimination
  • Reversible associations
  • strength of study design

5

List 3 types of Associations

(association doesn't = cause)

  • Non-statistical (chance)
  • Statistical
  • Risk Factors

6

Casual factors can be:

(4)

  • necessary & sufficient
  • necessary but not sufficient
  • sufficient but not necessary
  • neither sufficient nor necessary (but contributes)

7

Reversible associations?

Removal of a factor results in a decr. or incr. frequency of dz.

8

List the epidemiologic study types in order of strongest to weakest for providing evidence that an association may be causal.

(5)

  1. Radonmized clinical trials
  2. Cohort & case-control
  3. Cross-sectional
  4. Case series
  5. Case report

9

List the 2 measures of strength of association & the tests the are used in

  • Relative Risk (RR): cohort

 

  • Odds Ratio (OR): case-control

10

List 3 factors that impair making causal inferences.

  • lengthy time interval btwn "cause" & dz.
  • mulitple "causes" leading to same dz.
  • causal factor requires other factors for dz. occurrence

11

What are 3 characteristics of a Cause?

  1. must precede the event
  2. can be (+) or (-) in the presence of exposure
    1. vaccines
  3. one or multiple casual factors

12

Causal factors are

  • Disease agents
  • Environmental influences
  • Host factors

(interactive)

13

Characteristics of Confidence Interval

  • includes P value
  • same pop. sampled, resulting intervals would bracket the TRUE estimate of the pop. ~95 or 99% of the time
  • depends on sample size (larger is better)

14

What does (+) statistical association possibly indicate?

Causality

15

What does (-) statistical association possibly indicate?

a protective factor