You are on page 1of 18

The Third factor-

Effect Modification
An Example of Stratified Analysis
 Salmonella after wedding dinner
 Disease seems to be associated with both chicken and rice
 But many had both chicken and rice

Exposure Cases Controls Odds ratio 95% ci


Rice 37 / 50 21 / 50 3,9 (1,7 - 9,2)
Chicken 40 / 50 20 / 50 6,0 (2,8 - 12,7)
Cake 32 / 50 27 / 50 1,5 (0,7 - 3,4)
Juice 16 / 50 20 / 50 0,7 (0,3 - 1,6
Confounding
Is rice a confounder for the chicken 
salmonellosis association?
Stratify: Make one 2x2 table for rice-eaters
and one for non-rice-eaters (e.g. in Episheet)

Chicken Salmonellosis

Rice
No confounding
Because:
OR for chicken alone = ORmh for
chicken ”controlled for rice”

Exposure Cases Controls Odds ratio 95% ci


Rice-eaters: Chicken 36 / 37 18 / 21 6,0 (0,6 - 62)
Non-rice-eaters: Chicken 4 / 13 2 / 29 6,0 (0,9 - 38)
Chicken "controlled for rice" 40 / 50 20 / 50 6,0 (1,4 - 26)
Confounding
Is chicken a confounder for the rice 
salmonellosis association?
Stratify: Make one 2x2 table for chicken-eaters
and one for non-chicken-eaters (e.g. in
Episheet)
Rice Salmonellosis

Chicken
Confounding
Because:
OR for rice alone = ORmh for rice
”controlled for chicken”

Exposure Cases Controls Odds ratio 95% ci


Chicken-eaters: rice 36 / 40 18 / 20 1,0 (0,17 - 1,0)
Non-chicken-eaters: rice 1 / 10 3 / 20 1,0 (0,09 - 11)
Rice "controlled for chicken" 37 / 50 21 / 50 1,0 (0,24 - 4,2)

Not 3,9
Conclusion
 Chicken is associated with salmonellosis
 Rice is not associated with salmonellosis
 confounding by chicken because many chicken-
eaters also had rice
 rice only appeared to be associated with
salmonellosis
 Stratification was needed to find confounding
 Compare crude OR to adjusted OR (ORmh)
 If > 10-20% difference  confounding!
Exercises

 Exercise 1
Confounding: exercise
Hypothetical case control study

Lung cancer No lung


cancer
Drinker 50 50

Non-drinker 50 150

100 200

Q> Calculate the crude OR

OR  50 150  3.0


50  50
Confounding: example
Smokers Non-smokers

Lung No Lung No
cancer lung cancer lung
cancer cancer
Drinke Drinke
45 15 5 35
r r
Non- 10 Non- 140
30 20
drinker drinker
75 25 25 175
Q> Calculate the OR for smokers Q> Calculate the OR for non-smokers

OR s  4510 1.0 ORn s  5140 1.0


1530 3520
Confounding: example
Interpretation
Smoking

Drinking X Lung cancer

• Drinking is not associated with lung caner


• Smoking is a confounder
Exercises

 Exercise 2
Effect modification: example
Lung cancer No lung
cancer
Drinker 50 50

Non-drinker 50 150

100 200

Q> Calculate the crude OR

OR  50 150  3.0


50  50
Effect modification: example
Smokers Non-smokers

Lung No Lung No
cancer lung cancer lung
cancer cancer
Drinke Drinke
45 15 5 35
r r
Non- 10 Non- 140
15 35
drinker drinker
60 25 40 175
Q> Calculate the OR for smokers Q> Calculate the OR for non-smokers

OR s  4510  2.0 ORn s  5140  0.57


1515 3535
Effect modification: example
Interpretation
Smoking

Drinking Lung cancer

• Drinking is associated with lung cancer


• Smoking is an effect modifier
Exercises

 Exercise 3
Exercises

 Exercise 3
The following table shows the relative risk (RR) of cigarette smoking and developing
oral cancer, grouped by whether having alcohol beverage drinking habit or not.

Condition Having alcohol Not having Combined both having


(A-D) habit group alcohol habit and not having alcohol
group habit groups
A 4.0 2.0 1.0
B 4.0 4.0 1.0
C 4.0 2.0 2.8
D 4.0 4.0 4.0

a. Which condition in the table 1 that alcohol beverage drinking is most likely to be a
confounder but not a modifier? B
b. Which condition in the table 1 that alcohol beverage drinking is most likely to be a
confounder and also a modifier? A
c. Which condition in the table 1 that alcohol beverage drinking is mostly likely to be
a modifier but not a confounder? C
d. Which condition in the table 1 that alcohol beverage drinking is mostly likely
neither to be a modifier nor a confounder? D

You might also like