Cohort

Question 1

Cohort
General info

Answer 1

• Incident or new cases
• Individual level
• Free of disease on baseline

1. Start with EXPOSED/NON-EXPOSED
2. Follow cohort longitudinally to see who develops DISEASE

Question 2

Cohort
Advantages

Answer 2

1. Rare exposures (We can select them)
2. Multiple outcomes can be studied
3. Temporal relationship is clearer (EXPOSURE before DISEASE)
4. Direct estimate of risk of developing the disease

Question 3

Cohort
Disadvantages

Answer 3

1. Not practical for rare DISEASES (that is C&C)
2. Prone to bias
   -Attrition: Loss to follow-up, long studies.
   -Selection: Non representative samples. No response, no participation.
   -Information: Missclasification of status, Recall of exposure, Interviewer collecting data
   -Confounding: Other variables
   -Survivorship: Only who survive skew results
3. Changes can be missed, in exposure or medical practice (Temporal)
4. Cohort effect: Variations over time influence outcome
5. Larger, longer and more expensive than C&C

Question 4

Cohort
Formula and interpretation

Answer 4

• RR (a/a+b)÷(c/c+d)
• AR (a/a+b) - (c/c+d)
• %AR (RR-1)/RR or (CIE-CIUE)/CIE

Other
* Incidence rate ratio (IRR) (a/T1) ÷ (c/T2)
* Incidence rate difference (a/T1) - (c/T2)
* Odds ratio (OR) (a/b)÷(c/d)

(a/T1) Incidence rate (IR) in exposed (a+b)
(c/T2) Incidence rate (IR) in unexposed (c+d)

Question 5

Cohort
Relative risk (RR)

Answer 5

It looks at the actual risk of getting the disease if you are exposed

Risk ratio (a/a+b)÷(c/c+d)

• (a/a+b) risk in exposed (CI)
• (c/c+d) risk in unexposed (CI)

The risk of DEPRESSION among DRINKERS is 1.5 times the risk of DEPRESSION among NON-DRINKERS

Similar to Cross Sectional PRR.

Interpretation of ACTUAL RISK:
* PRR = <1 Exposure associated with low risk (prevalence) of disease in exposed vs unexposed (Protective effect)
* PRR = 1 Exposure no effect (didnt prevent/harm)
* PRR = >1 Exposure associated with high risk (prevalence) of disease in exposed vs unexposed (Harmful effect)

Question 6

Cohort
Attributable risk (AR or PAR)

Answer 6

Attributable risk (AR) (a/a+b) - (c/c+d)
* (a/a+b) risk in exposed (CI)
* (c/c+d) risk in unexposed (CI)

The excess risk of DEPRESSION associated with HEAVY ALCOHOL USE is 0.015, assuming causality.

Similar to RR but is a substraction instead of division

Question 7

Cohort
Percent attributable risk (%AR)

Answer 7

Percent AR (%AR)
* (RR-1)/RR or
* (CIE-CIUE)/CIE

If causality established, 40 is the percentage of the total risk of DEPRESSION among HEAVY ALCOHOL USERS that it’s attributable to ALCOHOL consumption.
———————————-OR—————————-
40% of the cases of DEPRESSION with HEAVY ALCOHOL USE can be attributed to the fact of the HEAVY ALCOHOL USE

With person time years (incidence rate): (IR E- IR UE)/ IR E

Question 8

Cohort
Odds Ratio (OR)

Answer 8

Describes if exposure is associated with odds of disease vs unexposed

Odds Ratio (OR) ` (a/b)÷(c/d)`

• (a/b) odds of disease in exposed
• (c/d) odds of disease in unexposed

The odds of DEPRESSION among DRINKERS are 1.5 TIMES the odds of depression AMONG NON-DRINKERS

Similar to Cross sectional (POR EXPOSURE). Cases&Controls (OR DISEASE)

Interpretation of RISK ASSOCIATION:
* POR = <1 Exposure has asscn with low odds of disease vs unexposed
* POR = 1 Exposure not associated with odds of disease
* POR = >1 Exposure has asscn with high odds of disease vs unexposed

Question 9

Cohort
Prospective vs Retrospective

Answer 9

Prospective (Longitudinal)
* From today, moving forward
* The best observational study bc exposure is assesed before disease occurs
* Long periods of time

Retrospective
* Collect data from the past (like C&C)
* Ensuring exposure preceded disease