Miscellaneous Flashcards

1
Q

How do you calculate the odds ratio (OR)?

Example dataset:
(a) Ill people who eat ice-cream (IIC): 13
(b) Ill people who do not eat ice-cream (InIC): 17
(c) Well people who eat ice-cream (WIC): 32
(d) Well people who do not eat ice-cream (WnIC): 23

A
  1. Divide conditions of group 1
  2. Divide conditions of group 2
  3. Divide the results of the divisions of each group
  4. IIC / InIC = x
    a / c = x
  5. WIC / WnIC = y
    c / d = y
  6. x / y = z , OR = z
  7. 13 / 17 = 0.76
  8. 32 / 23 = 1.39
  9. 0.76 / 1.39 = 0.54 (0.55 rounded up) , OR = 0.54
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What does an odds ratio of exactly 1 mean?

A

Property A does not affect the odds of property B

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

What does an odds ratio of more than 1 mean?

A

There is a higher odds of property B happening with exposure to property A

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What does an odds ratio less than 1 mean?

A

Lower odds

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Risk ratio can also be called relative risk. TRUE or FALSE?

A

TRUE. Risk ratio and relative risk are the same thing.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

How do you calculate relative risk (RR)?

A

Divide the category in question by the total number of the group for each category, then divide the answers.

For example:
100 obese people either have had a stroke (20) or have not had a stroke (80).
500 healthy people either have had a stroke (50) or have not had a stroke (500).

The risk of having a stroke if you are obese is 0.2 (20 / 100), compared to 0.1 (50 / 500) if you are healthy (not obese).
0.2 / 0.1 = 2
RR = 2

The RR of having a stroke if you are obese compared to if you are not obese is 2.
Obese people have twice the risk of having a stroke than healthy people.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Odds ratios (ORs) are important in case-controlled studies. TRUE or FALSE?

A

TRUE As in case-controlled studies it is not statistically appropriate to calculate RR.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Is it statistically appropriate to calculate RR (relative risk) in case-controlled studies?

A

No, it isn’t. The appropriate caculation is OR (odds ratio)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly