biostats

Question 1

absolute risk reduction

Answer 1

ARR = event rate in control group - event rate in tx group

Question 2

appropriate calculation for case-control study

Answer 2

odds ratio

Question 3

appropriate calculation for cohort study

Answer 3

relative risk

Question 4

odds ratio calculation

Answer 4

ad/bc

Question 5

relative risk calculation

Answer 5

a/(a+b)/ c/(c+d)

Question 6

attributable risk calculation

Answer 6

a/(a+b) - c/(c+d)

Question 7

number needed to treat calculation

Answer 7

1/ARR

Question 8

number needed to harm calculation

Answer 8

1/AR

Question 9

confounding vs. effect modification

Answer 9

when confounding is controlled for it usually reveals no significant difference between groups unlike effect modification (outcome modified by presence of another factor)

Question 10

Hawthorne effect

Answer 10

study subjects change their behavior if they become aware that they are being studied

Question 11

Berkson’s bias

Answer 11

selection bias created by choosing hospitalized patients as control group