1
Q
2x2 tablet
A
- Exposure on left: treatment or placebo
- outcome present/outcome absent on right
2
Q
Control event rate CER
A
C/C+D
3
Q
Exposed event rate EER
A
A/A+B
4
Q
odds of event in treatment group
A
A/B
5
Q
Odds of event in placebo group
A
C/D
6
Q
Odds ratio
A
A/B divided by C/D
7
Q
Absolute benefit increase or ARR
A
-EER-CER
A/A+B)-(C/C+D
8
Q
Relative Risk
A
- EER/CER
- (A/A+B)/ (C/C+D)
9
Q
Relative benefit risk increase (relative risk reduction)
A
(EER-CER)/CER
ABI/CER
RR-1
10
Q
Number needed to treat
A
1/ARR (proportion)
or 100 divided ARR (%)
11
Q
Number needed to harm
A
1/ARR or 100/ARR
12
Q
Applying to local population
A
- Need to adapt the control event rate for the population
- this is called a PEER (patient expected event rate)
- given in exams
13
Q
Population ABI
A
RBIxPEER
-you can then calculate the new NNT as 1/new ABI
14
Q
F fraction
A
-susceptability of a local population to a particular outcome
NNT (new)=NNT (old)/F value
