Cohort Study Design

Question 1

Historical / Retrospective Cohort Study

Answer 1

• exposed in past
  (Classify initial exposure from history)
• follow for outcome now + future

Question 2

Case-Control

Answer 2

• find cases (now)

- check back (past) for exposure

Question 3

How to: Cohort Study

Answer 3

1. Recruit outcome-free individuals
2. Exposed vs unexposed
3. Follow up each group
4. Analyse separate
5. Compare exposed groups

Question 4

Incidence Rate Ratio

Answer 4

• Compare 2 rate ratios
• relative risk

IRR = IR(exposed) / IR(unexposed)

No units

Question 5

Advantages Cohort

Answer 5

• study expo & personal characteristics not routinely collected
• obtain more detailed info (outcomes & expo
• collect additional data (pot. Confounding factors)
• can identify cause + effect
• easier to adjust for

Question 6

Concurrent / Prospective Cohort

Answer 6

• recruit outcome-free individuals

Classify according to exposure status

Question 7

Cohort Disatvantages

Answer 7

• long (ethical dilemma) + large scale (expensive)
• expensive investigations (to define outcome / exposure)
• survivor bias (due to risk of dropping out, ill more likely to drop)
• not good for rare outcomes
• difficulty w/ confounding

Question 8

Incidence Rate

Answer 8

• number of cases in given pop over a period time

IR(expo) = d(expo) / p-y(expo)
Ir(unexpo) = d(unexpo) / p-y(unexpo

• related to prevalence (P=I x D)