Biostatistics

Question 1

Define a case-control study.

Answer 1

Two very similar groups are studied, one with the presence and one with the absence of an outcome.
Risk factors in both groups are studied and associations between variables can be formed but never proven (retrospective).

Question 2

Define a cohort study.

Answer 2

Prospective study–two very similar groups with a difference in an independent variable. Followed over time, but can be costly.

Question 3

Define randomized-control study.

Answer 3

Two similar groups, randomly assigned to the control group or intervention group. Participants are blinded to their assignment.

Question 4

Define Type 1 error.

Answer 4

Type 1 error– rejecting the null hypothesis when in actually true

Question 5

Define Type 2 error.

Answer 5

Type 2 error– failing to reject the null hypothesis when it is truly false.

Question 6

What is sensitivity?

Answer 6

Probability of obtaining a positive screen among subjects who truly have the disease/outcome of interest.

SNOUT- if a high sensitivity test has a negative test result, it definitely rules it out

Question 7

What is specificity?

Answer 7

Probability of obtaining a negative screen among subjects who truly do not have the disease/outcome of interest.

SPIN- if you have a high specificity test has a positive result, it rules it in

Question 8

Formula for number needed to treat?

Answer 8

1/ARR; ARR stands for absolute risk reduction

ARR= P(placebo) - P(treatment); P represents the proportion of negative outcomes each group.