What qualities should a

disease have to make it

worthwhile to investigate?
Disease investigations should have
some public health significance

The disease is important in terms of

the number of individuals it affects
The disease is important in terms of
the types of populations it affects
The disease is important in terms of its
causal pathway or risk characteristics
 Research Questions/Hypotheses

Is there an association between

Exposure (E) & Disease (D)?
Hypothesis: Do persons with exposure
have higher levels of disease than
persons without exposure?
Is the association real, i.e. causal? Sever
 Big Picture
Look for links between exposure &
disease
to intervene and prevent disease

Look to identify what may cause disease

Basic definition of cause
exposure that leads to new cases of disease
remove exposure and most cases do not occur
 Big Picture
On a population basis
An increase in the level of a causal
factor will be accompanied by an
increase in the incidence of disease
(all other things being equal).
If the causal factor is eliminated or
reduced, the frequency of disease will
decline
 Infectious Disease Epidemiology
Investigations/studies are undertaken to
demonstrate a link [relationship or
association] between an agent (or a vector
or vehicle carrying the agent) and disease

Exposure Disease
[ Agent ]
[ Vector/Vehicle ]
 Injury Epidemiology
Studies are undertaken to demonstrate
a link [association] between an agent /
condition and an injury outcome

Exposure Disease
[ Agent Energy Transfer ]
[ Vehicle carrying the agent automobile ]
[ Condition Risk taking behaviour ]
 Chronic Disease Epidemiology
Studies are undertaken to demonstrate
a link [relationship or association]
between a condition/agent and disease

Exposure Disease
[ Condition e.g. gene, environment ]
 Issues to consider
Etiology (cause) of chronic disease is often
difficult to determine
Many exposures cause more than one
outcome
Outcomes may be due to a multiple
exposures or continual exposure over time
Causes may differ by individual
 Exposure
OR
Genetic Background
OR
Combination of Both

Association
? Causation ?

Disease or Other Outcome

Suppose we determine that an exposure is associated with disease.

How do we know if the observed association reflects a causal
relationship?
 First step in determining causation:
Understanding disease etiology
Experimental studies
in vitro systems
animal studies in controlled environments
Allows for
control of precise dose
control of environmental conditions
loss to follow up kept to a minimum
Problems with
extrapolating data to human populations
human diseases with no good animal models
Clinical pathologies
Second step in determining causation: Conducting
Studies in Human Populations

Heres where Epidemiology is important.

Epidemiology capitalizes on natural or

unplanned experiments. We take
advantage of groups who have been exposed
for non-study purposes.

All of the study designs are important here and

provide different evidence for or against a causal
hypothesis.
Figure 14-3 A frequent sequence of studies
in human populations.
Downloaded from: StudentConsult (on 11 September 2009 07:00 PM)
2005 Elsevier
 Causation and Association
Epidemiology does not determine the
cause of a disease in a given individual

Instead, it determines the relationship or

association between a given exposure
and frequency of disease in populations

We infer causation based upon the

association and several other factors
 Association vs. Causation
Association - an identifiable
relationship between an exposure and
disease
implies that exposure might cause disease
exposures associated with a difference in
disease risk are often called risk factors

Most often, we design interventions

based upon associations
 Association vs. Causation
Causation - implies that there is a
true mechanism that leads from
exposure to disease

Finding an association does not

make it causal
 ASOSIASI VS KAUSASI
Untuk memutuskan apakah
eksposur A menyebabkabn penyakit
B, kita pertama kali harus
menemukan apakah dua variabel itu
berasosiasi, misal apakah satu
ditemukan lebih umum pada adanya
yang lain.
 CAUSAL ASSOCIATION

ADALAH SUATU ASOSIASI ANTARA 2

KONDISI ATAU KARAKTERISTIK
DIMANA BILA TERJADI PERUBAHAN
(MIS: PERUBAHAN JUMLAH ATAU
PERUBAHAN KUALITAS) PADA 1
KONDISI, MAKA AKAN DIIKUTI OLEH
PERUBAHAN PADA KONDISI YANG
LAINNYA
 CAUSALITY OR CAUSAL
ASSOCIATION
BERHUBUNGAN DENGAN CAUSE EFFECT
RELATIONSHIPS UNTUK
MENENTUKAN SEBERAPA BERBEDA
SUATU KONDISI (MIS: FAKTOR
RISIKO) BERHUBUNGAN DENGAN
KONDISI LAINNYA (MIS: PENYAKIT)
MENENTUKAN PENYEBAB ATAU
ETIOLOGI
Two step process to carry out studies and
evaluate evidence

1. Determine if an association is present

- Ecologic studies: studies of group characteristics
- Cross-sectional studies: studies at one particular
time
- Case-control or cohort studies: studies of
individual characteristics.
2. If an association is demonstrated,
determine whether the observed
association is likely to be a causal one
using pre-determined criteria.
 General Models of Causation
Cause: event or condition that plays an
role in producing occurrence of a disease

How do we establish cause in situations

that involve multiple factors/conditions?

For example, there is the view that

most diseases are caused by the
interplay of genetic and
Environmental factors.
General Models of Causation
How do we establish cause?

Exposure Disease

Additional Factors
 Web of Causation

There is no single cause

Causes of disease are interacting
Illustrates the interconnectedness
of possible causes

RS Bhopal
 Understanding Causality
Types of Association
causal
noncausal
Types of Causal relationships
direct
indirect
Types of causal factors
sufficient
necessary
Jenis-jenis hubungan sebab-akibat
A.Tidak berhubungan secara statistik

A.Berhubungan secara statistik

kausal langsung
kausal tidak langsung
 TYPES OF ASSOCIATION
A. Not statistically
associated (independent)
B. Statistically associated
1. Noncasually (secondarily
associated)
2. Causially associated
a. Indirectly associated
b. Directly casual
 Hampir semua statistik berusaha menemukan
apakah dua variabel berhubungan, dan jika
demikian, seberapa kuat, dan apakah chance
(kebetulan) dapat menjelaskan asosiasi yang
diamati.

Statistik terutama dirancang untuk menilai

peranan chance dalam asosiasi itu.

Suatu nilai p hanya menceritakan kepada

kita seberapa mungkin asosiasi itu mempunyai
peningkatan secara kebetulan.

Oleh sebab itu, Analisis statistis sendiri tidak

dapat membangun bukti hubungan kausal.
Figure 14-5 Types of associations.
Downloaded from: StudentConsult (on 11 September 2009 07:00 PM)
2005 Elsevier
 The relationship between coffee
consumption and pancreatic cancer
In 1981, MacMahon et al. reported results from a
case-control study of cancer of the pancreas.

There was an apparent dose response

relationship between coffee consumption and
cancer of the pancreas, particularly in women.

Was the disease caused by coffee consumption

or by some factor closely related to coffee
consumption?

MacMahon B, et al. N Engl J Med 1981 304:630 - 33

 The relationship between coffee
consumption and pancreatic cancer
Smoking is closely associated with both
pancreatic cancer and coffee consumption.

There were many issues with control

selection and measurement of exposure
levels in cases and controls.

Subsequent studies were unable to reproduce

the result.

MacMahon B, et al. N Engl J Med 1981 304:630 - 33

Interpreting Associations - Causal and Non-Causal
Non-Causal (due to
Causal
confounding)
Coffee Consumption Coffee Consumption
Real
Association
Spurious
Association
Smoking

Real
Association

Pancreatic Cancer Pancreatic Cancer

 Two Types of Association:
Real and Spurious
A real association is present if
the probability of occurrence of an event or
the quantity of a variable
depends upon the occurrence of one or more
other events, characteristics or variables.

Spurious associations refer to non-causal

associations due to chance, bias, failure to
control for extraneous variables
(confounding), etc.
 Why is it important to distinguish
between causal and non-causal
associations?

Causal relationships are used to make public

health decisions and design interventions.

In our example, if smoking was indeed

causal, it would be irresponsible to target
coffee drinking as an intervention.

Very important to consider all confounders.

 Hubungan faktor dg penyakit

1. Hubungan statistik KAUSAL:

a. Langsung/DIRECT
* Dua arah
* Searah
b. Tak langsung/INDIRECT
2. Hubungan substantif
 HUBUNGAN KAUSAL

DIRECT vs INDIRECT

DIRECT (LANGSUNG):
FAKTOR OUTCOME

INDIRECT (TIDAK
LANGSUNG):
FAKTOR OUTCOME
Figure 14-12 Direct versus indirect
causes of disease.
Downloaded from: StudentConsult (on 11 September 2009 07:00 PM)
2005 Elsevier
 Types of Causal Relationships:
Direct vs Indirect

Direct Indirect

Factor Factor 1

Factor 2
Factor 3
Factor 4

Disease Disease
 Types of Causal Relationships:
Direct vs Indirect

Direct Indirect
F508 Polymorphism High cholesterol

Artery thickening

Hemostatic factors

Cystic Fibrosis Myocardial infarction

 Steps in causality
Need to answer 2 major questions

Is there actually an association?

If there is an association, is it
likely to be causal?
 Steps in causing causality
Is there actually an association?

Association actually exists and is

statistically meaningful
Association is not due to chance so
is statistically significant
The association occurs at
individual level and not on
ecological level (aggregate or
geographical unit)
 Steps in causing causality
If there is an association,
is it likely to be causal

Association is not due to bias

Selection bias
Information or measurement bias
Confounding bias

Confirmatory criteria for causality is satisfied

Based on specific qualities of association between
risk factor and disease
Pembuatan kesimpulan kausal

Penggunaan kriteria kausal

dalam pembuatan kesimpulan
dari data.
 Evaluasi Hubungan Kausal

Postulat Koch:
Hanya berlaku untuk penyakit
infeksi, tidak berlaku untuk
penyakit non-infeksi

Kriteria Bradford Hill:

Berlaku pada seluruh kondisis
 Postulat Koch-Henle berlaku
pada penyakit-penyakit infeksi,
tetapi tidak berlaku pada
penyakit non-infeksi

Pada penyakit non-infeksi: pada

media kultur tidak akan tumbuh
kuman
 Understanding Causality

Lets say you have determined:

there is a real association,
you believe it to be causal (ruled out
confounding),
figured out that it is a direct causal factor
sorted out the necessary vs. sufficient
factor issue
NOW have your proven CAUSALITY?
Nine guidelines for judging whether
an association is causal
Temporal relationship Consideration of
alternate explanations
Strength of association
Cessation of exposure
Dose response
relationship Specificity of the
association
Replication of the
findings Consistency with other
knowledge
Biologic plausibility
Bila melihat data dari studi epidemiologis, kita
sering menggunakan kriteria kasual untuk
membantu dalam pembobotan bukti.

Hal yang paling umum digunakan adalah

sebagai berikut,
dinyatakan pertama kali dari kerja Ahli
statistik dari Inggris Austin Bradford Hill, dan
kemudian dikembangkan lebih lanjut oleh
Surgeon General's Office, Amerika Serikat,
dalam laporannya tahun 1964 tentang merokok
dan kanker.
 16.3 Hills Framework
Hill, A. B. (1965). The environment and disease: association or causation?
Proceedings of the Royal Society of Medicine, 58, 295-300.

1. Strength
2. Consistency
3. Specificity
4. Temporality
5. Biological gradient
6. Plausibility
7. Coherence
8. Experimentation
9. Analogy
 Hills Postulates
1. Strength of Association the stronger the association, the less
likely the relationship is due to chance or a confounding variable

2. Consistency of the Observed Association has the

association been observed by different persons, in different places,
circumstances, and times? (similar to the replication of laboratory
experiments)

3. Specificity if an association is limited to specific persons, sites

and types of disease, and if there is no association between the
exposure and other modes of dying, then the relationship supports
causation

4. Temporality the exposure of interest must precede the outcome

by a period of time consistent with any proposed biologic mechanism

5. Biologic Gradient there is a gradient of risk associated with the

degree of exposure (dose-response relationship)
 Hills Postulates (cont)
6. Biologic Plausibility there is a known or postulated
mechanism by which the exposure might reasonably alter the
risk of developing the disease

7. Coherence the observed data should not conflict with known

facts about the natural history and biology of the disease

8. Experiment the strongest support for causation may be

obtained through controlled experiments (clinical trials,
intervention studies, animal experiments)

9. Analogy in some cases, it is fair to judge cause-effect

relationships by analogy With the effects of thalidomide and
rubella before us, it is fair to accept slighter but similar evidence
with another drug or another viral disease in pregnancy
 Criteria 1: Strength
Relative risk as a measure of
strength
Stronger RRs carry more weight
More difficult to explain away by
confounders
But, do not dismiss small RRs
some causal relations will by
nature have a weak association
 KRITERIA KAUSAL
Postulat Hill (1)
1. Kekuatan asosiasi
asosiasi yang lebih kuat,
kurang mungkin
berhubungan secara
kebetulan atau suatu variabel
perancu (confounding)
 KEKUATAN
Apakah asosiasi itu kuat?
Perokok berat berasosiasi
dengan duapuluh kali lipat lebih
tinggi tingkat kanker paru-paru,
dan dua kali lipat tingkat
penyakit jantung.
Asosiasi merokok dengan kanker
paru karena itu lebih kuat dari
pada asosiasi dengan penyakit
 Criterion 1: Strength
Stronger associations are less
easily explained by confounding
than weak associations
Ratio measures (like RR, SMR,
OR) are the best way to quantify
the strength of an association
Example: An RR of 10 is much
stronger evidence for causality
than an RR of 2
 Strength of association
Which odds ratio would you be more likely to infer
causation from?

OR#1: OR = 1.4 95% CI = (1.2 - 1.7)

OR#2: OR = 9.8 95% CI = (1.8 - 12.3)

OR#3: OR = 6.6 95% CI = (5.9 - 8.1)

 Criteria 2: Consistency

Similar findings
using diverse methods
in different populations
under a variety of circumstances
Consistency alone does not prove
causality
You can have consistently biased
 KRITERIA KAUSAL
Postulat Hill

2. Konsistensi Asosiasi yang diamati

Mempunyai asosiasi yang
diamati oleh orang yang berbeda,
tempat, persoalan dan waktu
yang berbeda? (mirip dengan
replikasi eksperimen
laboratorium)
 KONSISTENSI
Konsistensi dapat juga berarti :
Replikasi pasti, sebagai ilmu
laboratorium, atau
Replikasi dalam banyak persoalan
yang berbeda.
Dalam epidemiologi, replikasi pasti
adalah tidak mungkin (impossible)
 Criterion 2: Consistency
Consistency: studies using
diverse methods in different
populations under a variety of
circumstances lead to similar
conclusions
Example: Ecological, cohort,
and case-control done by
independent researchers
studying different populations
all showed a strong association
between smoking and lung
cancer.
Consistency (I)
 KRITERIA KAUSAL
Postulat Hill (2)
3. Spesifisitas
Jika suatu asosiasi terbatas
pada orang, tempat dan tipe
penyakit tertentu (spesifik),
dan jika tidak ada asosiasi
antara ekposur dan model lain
kematian, kemudian
hubungan itu mendukung
 Criteria 3: Specificity
Factor leads to a specific disease
Requires knowledge at cellular level
Converse is NOT true
Some casual relations are non-
specific
e.g., smoking causes multiple
diseases
 SPESIFISITAS

Kausalitas diperkuat jika eksposur

diasosiasikan dengan suatu penyakit
spesifik, dan bukan dengan
keseluruhan varitas penyakit-
penyakit
Contoh 1.
Asbestos sebab penyakit paru-paru spesifik,
asbestosis, dapat dibedakan dari berbagai
penyakit paru-paru lainnya.

Tetapi eksposure timbal pada tingkat rendah

dihubungkan dengan IQ (Intelligent Quotient)
yang lebih rendah daripada suatu sindrom otak
yang dapat dibedakan.
Jadi timbal (Pb = Plumbum = timah hitam)
lebih tidak tentu sebagai sebab karena
kemungkinan rancu dengan sebab-sebab yang
lain, ini bukan efek yang spesifik, IQ rendah
(misal SES = Social Economic Status).
 Criterion 3: Specificity
The factor is linked to a
specific causal mechanism
Example: Smoking is linked to
physical and chemical
carcinogenesis of epithelial
cells
Comment: Mechanisms are
difficult to establish when there
Aristotle
is a vacuum of knowledge (384 322 BCE)
 KRITERIA KAUSAL
Postulat Hill

4. Temporalitas/Urutan waktu
Eksposur yang menjadi
perhatian harus mendahului
outcome (penyakit) menurut
periode waktu yang konsisten
dengan berbagai usulan
mekanisme biologik
 URUTAN WAKTU

Ini kriteria yang sangat penting

secara sederhana menyatakan
bahwa orang harus mengetahui
pasti bahwa sebab mendahului
akibat dalam waktu. Kadang-
kadang ini sulit mengetahui ,
terutama dalam studi
kroseksional (penelitian survei).
Contoh 1.
Studi telah menemukan
hubungan terbalik antara
tekanan darah seseorang dengan
kadar kalsium serum. Tetapi
yang mana sebab dan yang mana
akibat?

Urutan waktu dapat juga

menjadi tidak tentu bila
 Criterion 4: Temporality
Exposure precedes disease by a reasonable
amount of time
Lead encephalopathy scenario 1:

Lead encephalopathy scenario 2:

Gerstman Chapter 16 69
 KRITERIA KAUSAL
Postulat Hill (3)

5. Gradien biologik
Ada suatu gradien risiko
berhubungan dengan derajat
eksposur (hubungan dosis-
respons)
 Hubungan Dosis-respons
Jika suatu gradien teratur risiko
penyakit ditemukan paralel
terhadap gradien eksposur
(misal: perokok ringan
mendapat kanker paru pada
tingkat menengah antara bukan
perokok dengan perokok berat)
kemungkinan hubungan kausal
diperjelas.
Contoh:

Untuk setiap peningkatan

jumlah rokok yang dihisap,
risiko kanker paru meningkat.
Cohort study: Tobacco smoking and lung cancer,
England & Wales, 1951

Source: Doll & Hill

 Biological Gradient
There is evidence of a dose-response relationship
Changes in exposure are related to a trend in
relative risk
 CRITERION 5: Biological Gradient
Increases in exposure dose dose-response in risk

Gerstman Chapter 16 75
 KRITERIA KAUSAL
Postulat Hill (3)

6. Plausibilitas biologik
Diketahui atau ada
mekanisme yang
dipostulasikan menurut
ekposur yang mungkin
beralasan setelah risiko
perkembangan penyakit
 Criteria 6: Plausibility

Plausible = makes sense in face

of known biological and and
other facts
But what of new previously
unexplained associations?
Where does new knowledge
come from?
 Criterion 6: Plausibility
Comment: All that is
Plausible mechanism plausible is not always true
in face of known
biological facts
Plausibility (defined):
appearing worthy of
belief

Gerstman Chapter 16 78
 Biologic Plausibility
does this make biologic sense?

is there a known mechanism by

which an exposure can lead to
disease?
Biologic Plausibility (cont.)
Example: daily consumption of
fatty foods increases risk of
heart disease
there is a biologic mechanism
fatty foods raises LDL cholesterol,
which increases the risk of heart
disease
 KRITERIA KAUSAL
Postulat Hill (4)

7. Koherens
Data yang diamati tidak
harus konflik dengan fakta
yang diketahui tentang
riwayat alamiah dan biologi
penyakit
 KOHERENS

Apakah asosiasi sesuai (cocok)

dengan pengetahuan biologis?
Seseorang harus mencari dukungan
pemeriksaan laboratorium, atau
dari aspek kondisi biologi yang lain.
Criteria 7: Coherence

Do the facts cohere

(i.e., to stick together)?
 Criterion 7: Coherence
All facts stick together to
form a coherent whole.
Example: Epidemiologic,
pharmacokinetic,
laboratory, clinical, and
biological data create a
cohesive picture about the
smoking and lung cancer.
Gerstman Chapter 16 84
 KRITERIA KAUSAL
Postulat Hill (4)
8. Eksperimen
Dukungan yang paling kuat
untuk mendukung penyebab
mungkin dapat diperoleh
melalui ekperimen yang
dikontrol (percobaan klinis,
studi intervensi, percobaan
hewan)
 Criteria 8: Experimentation
Experimental evidence should
support observational studies
Types of experiments
Epidemiologic (trials)
In vitro
Animal models
Natural experiments
Criterion 8: Experimentation
Experimental evidence
supports the
epidemiologic evidence
In vitro and in vivo
experiments
Experimentation is
often not possible in
humans
Animal models of
human disease

Gerstman Chapter 16 87
 KRITERIA KAUSAL
Postulat Hill (5)
9. Analogi
Pada beberapa kasus, adalah wajar
menilai hubungan sebab akibat
menurut analogi.
Dengan efek talidomid dan rubella
sebelum kita, adalah wajar bersikap
menerima tetapi pembuktian yang
mirip dengan obat atau virus yang
menyebabkan penyakit pada
CONTOH:

Adanya penanda (marker)

serologis infeksi Hepatitis B
dihubungkan dengan laju
peningkatan yang besar kanker
hati. Bahwa infeksi Hepatitis B
adalah sebab yang benar dari
kanker hati, juga ditunjang oleh
Sebaliknya, Reserpine (suatu
obat anti- hipertensif) dipikirkan
menjadi suatu sebab kanker
payudara berdasarkan atas studi
yang dilakukan awal tahun
1970an. Tetapi tidak ada
informasi biologis yang
menunjang, atau berbagai
mekanisme biologis yang dapat
dijelaskan secara benar.
 Criteria 9: Analogy
Similarities among things that are
otherwise different
e.g., before HIV was discovered,
epidemiologists noticed that AIDS
and Hepatitis B had analogous risk
groups
Evidence of similar transmission
Note: analogy is a weak form of
evidence
 Criterion 9: Analogy
Similarities among things
that are otherwise different
Weak form of evidence
Example: Before the HIV
was discovered,
epidemiologists noticed that
AIDS and Hepatitis B had
analogous risk groups,
suggesting similar types of
agents and transmission

Gerstman Chapter 16 92
 Criteria for Causation:
Smoking and Lung Cancer
Temporal relationship Smoking before Ca
Biologic plausibility Yes
Consistency > 36 studies
Alternatives

?
Cessation effects
Specificity of
Yes
association Point of attack
Strength of Association 25 x > 25+
cigarettes /day*
Dose-response Yes
*.Estimated that 80% of all Lung cancer due to Cigarette smoking
 Why was it relatively easy to determine that
smoking was a cause of lung cancer?

History of exposure to cigarettes can be

assessed with reasonable accuracy.
Cigarette smoking is common and present in
persons whose environment is otherwise
similar to that of nonsmokers.
Lung cancer incidence in smokers is much
greater than in nonsmokers.
Lung cancer is uncommon in nonsmokers.
 Why will it be relatively hard to determine if
community air pollution is a cause of lung
cancer?

Difficult to measure pertinent exposure

Long latent period of disease
Migration
Little variation in exposure among individuals
within a community
Lung cancer is common, even among persons
not exposed to pollution.
 NO DIRECT
ASSOCIATION ASSOCIATION

NO POSSIBILITY POSSIBILITY DIRECT CAUSE-

ASSOCIATED
FOR ASSOCIATION ASSOCIATED EFFECT

-Physically not -Remote -Cause-effect

associated -Causality
cause-effect a assurance
possible possibility exists
-Causally -Association
-Scientifically/ -Not affirmed
associated
medically not statistically but an indirect -Scientifically
associated at association and
probable an acceptable -Scientifically biomedically
connected proven
-Statistically level (50%- -Physically
60%) Statistically
not associated associated possible
-Secondarily but not -Statistically
association causative proven
association
Figure 10.4 Continuum of association
 Associations are observed
Causation is inferred

It is important to remember that these

criteria provide evidence for causal
relationships.

All of the evidence must be considered

and the criteria weighed against each
other to infer the causal relationship.
 Causal Inference: Realities
No single study is sufficient for causal
inference
Causal inference is not a simple process
consider weight of evidence
requires judgment and interpretation
No way to prove causal associations for
most chronic diseases and conditions
 Judging Causality

Weigh weaknesses Weigh quality

in data and other of science and
explanations results of causal
models
RS Bhopal
 Epidemiology Kept Simple

Chapter 16
From Association to Causation
(Causal Inference)
Gerstman Chapter 16 102
 References
1. Porta M. A dictionary of epidemiology. New York,
Oxford: Oxford University Press, 2008.
2. Susser MW. What is a cause and how do we know one ? A
grammar for pragmatic epidemiology. American
Journal of Epidemiology 1991; 133: 635- 648.
3. Rothman J, Greenland S. Modern epidemiology. third
edition. Lippincott - Raven Publishers, 2008.
4. Gordis L, Epidemiology .fourth edition .Saunders
5. Elsevier,2009