Statistics

Question 1

What are advantages to case report/case series?

Answer 1

1) Cheap
2) Easy to do

Question 2

What are the disadvantages to a case report/case series?

Answer 2

1) Generates hypothesis.
2) Cannot test for association.
3) Prone to selection bias.

Question 3

What are the advantages to a cross-sectional study?

Answer 3

1) Cheap
2) Easy to do
3) Good at determining extent of disease.
4) Can determine prevalence of a disease ( total # cases / total population)

Question 4

What are the disadvantages to cross-sectional studies?

Answer 4

1) No temporal association.
2) Not good for rare disease.
3) Follow-up not possible.

Question 5

What are the advantages to a case control study?

Answer 5

1) Easier and cheaper vs RCT.
2) Need fewer patients vs RCT.
3) Good for rare disease or disease with long latency.

Question 6

What are the disadvantages to a case control study?

Answer 6

1) Locked into the disease -have to pick outcome at the onset.
2) Easy to pick wrong controls.
3) Recall bias because asking to look back in time at exposures.
4) Cannot be sure of temporal association (cause and effect)

Question 7

What statistic can be generated from a case control study?

Answer 7

OR = ad / bc
Cannot calculate RR as you generate # cases / controls.

Question 8

What are the advantages to a cohort study?

Answer 8

1) Can establish temporal relationship.
2) Can show association.
3) No recall bias.
4) Good for rare exposures.

Question 9

What are the disadvantages to a cohort study?

Answer 9

1) Long follow-up required.
2) Expensive.
3) Not good for rare disease.
4) Role of confounders.
5) Locked into the variables you select.

Question 10

What statistics can be generated from a cohort study?

Answer 10

1) Incident = # new cases in time / total population at risk
2) RR = EER / CER = (a/(a+b)) / (c/(c+d))

EER - experimental event rate

CER - control event rate

Question 11

What are the advantages to a RCT?

Answer 11

1) Can test hypothesis.
2) Can control for bias if sample size large enough.
3) Best evidence to prove causation.
4) Placebo arm can show natural history of disease.

Question 12

What are the disadvantages of RCT?

Answer 12

1) Expensive.
2) Difficult for rare diseases.
3) Difficult to select proper study population.
4) Length of follow-up required.

Question 13

How do you calculate relative risk reduction (RRR) in a RCT?

Answer 13

1) RRR = (CER - EER) / CER
2) RRR = 1 - RR

• A positive RRR means increased RR by X %.

Question 14

How do you calculate absolute risk reduction (ARR)?

Answer 14

ARR = Absolute value (CER - EER)

Question 15

How do you calculate NNT

Answer 15

NNT = 1/ARR