09.04.18 Clinical Epidemiology

Question 1

degree to which the results obtained from a study sample reflect the truth

Answer 1

Internal validity

Question 2

results that would be obtained if everyone in the source population participated in the study, and all the measures were completely accurate

Answer 2

Truth

Question 3

population from which the sample was drawn

Answer 3

Source population

Question 4

generalizability; degree to which the results of the study are applicable to other (external) populations that are not being sampled

Answer 4

External validity

Question 5

observed findings are farther from the null value that the truth is from the null value (stronger association)

Answer 5

Away from the null

Question 6

observed finding are closer to the null value than the truth is from the null value (weaker association)

Answer 6

Bias toward the null

Question 7

distortion in the results due to the effects of a third, extraneous factor

Answer 7

Confounding

Question 8

distortion in the results due to a difference between subjects who participated and those who did not

Answer 8

Selection bias

Question 9

distortion in results due to error in measurement of the exposure, outcome or both

Answer 9

Measurement bias

Question 10

errors in measurement of the exposure that are independent of the outcome, or vice versa (alcohol, bias towards null)

Answer 10

Nondifferential

Question 11

errors in measurement of the exposure depend on the disease, or vice versa (bias away from null)

Answer 11

Differential

Question 12

• Enroll non-diseased
  
  • Collect exposure data and outcome data and use this to determine risk factors
• Can measure causality because exposure comes prior to outcome

Answer 12

Cohort (prospective study)

Question 13

• Instead of starting at baseline and following over years look backwards in time to follow them until incidence of disease
  
  • Use data already collected and follow forward from there
  • Retrospectively identify exposures, outcomes, and determine risk factors
• Not as strong as prospective

Answer 13

Cohort (retrospective study)

Question 14

• Start with individuals of outcome of interest
  
  • Rare, so not practical to study as a cohort: find people with outcome and compare to those who did not by matching
  • Look for exposures that happened prior to outcome
  • Sometimes these exposures have been measured sometimes not, so individuals need to recall or we need to reconstruct through records
    ○ Can have bias because data collected backwards
  • Measuring RR of exposure to development of outcome
  • Use odds ratio
  • Risk or incidence can’t be measured directly
  • Select cases and then we match them to others
• Arbitrary

Answer 14

Case control study

Question 15

• Randomize to control and experimental

- Avoids systematic error

Answer 15

Randomized control trial