Professional Documents
Culture Documents
Ukuran Asosiasi
Ukuran Asosiasi
Exposure Disease
Ukuran-ukuran asosiasi
• Ukuran rasio (perbandingan relatif)
– rasio dua frekuensi penyakit
membandingkan kelompok terpajan dengan
kelompok tidak terpajan
• Ukuran perbedaan efek (perbandingan
absolut)
– perbedaan antara ukuran frekuensi penyakit
suatu kelompok terpajan dan kelompok yang
tidak terpajan
Arithmetic
Disease
Yes No Total
Yes a b a+b
Exposure
c d c+d
No
Yes No Total
E (exposed) a b a+b
NE (unexposed) c d c+d
a Incidence in
FIRST, Exposed a b a+b a+b = exposed
SELECT
c Incidence in
Not =
c d c+d c+d non-
Exposed
exposed
a
Relative Risk ( RR ) = incidence in exposed a+b
incidence in non-exposed = c
c+d
Example : A Prospective Study of 3,000 Smokers and 5,000 Non-
smokers to Investigate Smoking and Coronary Heart Disease
(CHD)
THEN FOLLOW UP TO
SEE HOW MANY
Do Not Incidence
Develop
Develop Total per 1,000
CHD per year
CHD
FIRST, Healthy Smokers 84 2,916 3,000 28.0
SELECT
Healthy Non-
87 4,913 5,000 17.4
Smokers
Ukuran-ukuran asosiasi
• Ukuran rasio
– Rasio odds (Odds ratio = OR)
• Nama lain: Odds relative; rasio kros-produk
• rasio dua odds yang digunakan dalam studi
kasus-kontrol untuk mengestimasi rasio rate atau
rasio risiko
Ukuran-ukuran asosiasi
• Ukuran rasio
– Rasio odds (Odds ratio = OR)
• odds untuk satu kelompok dibagi dengan odds
untuk kelompok yang lain
• Bisa diinterpretasikan sama seperti risiko relatif,
dg syarat tertentu (misal: incidence rendah
rare disease assumption)
Ukuran-ukuran asosiasi
• Odds suatu kejadian
– rasio probabilitas bahwa kejadian terjadi
terhadap probabilitas kejadian tidak terjadi
P
Odds suatu peristiwa
1 P
CASES CONTROLS
( with ( without
disease) disease)
THEN,
MEASURE Were
a b
Exposed
PAST
EXPOSURE Were Not
c d
Exposed
Totals a+c b+d
Proportion a b
exposed a+c b+d
Figure 11-5 A, Odds ratio (OR) in a cohort study. B, Odds ratio (OR) in a case-control study.
a
b ad a c ad
= bc b : d = bc
c
d
In case-control studies, the odds ratio is the ratio of the odds of the cases having been
exposed to the odds of the controls having been exposed, i.e:
a
c ad a b ad
= bc c : = bc
b d
d
Tabel 1. Tabulasi silang pemajan dan status sakit,
insidens sakit dan Probabilitas odds sakit
pada studi kohort.
Status sakit
a x d 650 x 350
Odds Ratio 4,8
bxc 950 x 50
Perokok mempunyai risiko menjadi kasus 4,8 kali dari yang bukan
perokok.
Interpretasinya: odds perokok menjadi kasus 4,8 kali lebih besar
dari odds bukan perokok
Prevalence Odds Ratio (POR) = Cross Product
Ratio bila data didasarkan pada kasus-kasus
prevalens
650 x 350
Prevalence Odds Ratio 4,8
950 x 50
650 / 1600 0,40625
Prevalence ( proportion ) Ratio 3,25
50 / 400 0,125
exposure;
3.The disease being studied is
Odds Ratio & Risk Ratio
The odds ratio will provide a good estimate of the
risk ratio when:
OR = (a / c) / (b / d)
a / (a +b ) a / b ad
OR = (ad) / (bc) RR = ------------ = ------ =-- = OR
c / (c +d) c / d bc
Figure 11-6 Example: The odds ratio is a good estimate
of the relative risk when a disease is infrequent.
Downloaded from: StudentConsult (on 8 October 2009 11:44 AM)
© 2005 Elsevier
Figure 11-7 Example: The odds ratio is not a
good estimate of the relative risk when a
disease is not infrequent.
Downloaded from: StudentConsult (on 8 October 2009 11:44 AM)
© 2005 Elsevier
In a prospective study/cohort study,
the Relative Risk can be calculated
directly
20 / 1000 0,02
Incidence proportion (risk ) Ratio 2,00
10 / 1000 0,01
Risk
Risk for disease Risk of disease Risk for disease
comparison
is lower in the are equal for is higher in the
between
exposed than in exposed and exposed than in
exposed and
the unexposed unexposed the unexposed
unexposed
Exposure
reduces disease Exposure
Exposure as a Particular
risk increases
risk factor for exposure is not a
disease risk
the disease? (Protective risk factor
(Risk factor)
factor)
M easu re of
A ssociation
Chapter 8
Measures of Association