Please enable JavaScript.
Coggle requires JavaScript to display documents.
Causation: The Logic of Epi (Guidelines for determining causation…
Causation:
The Logic of Epi
"Causality" defined
The potential for changing
an outcome by prevention
Most clinical, epi & PH research
Relating of causes to the effects
they produce;
ideas about the nature of the relations
of cause & effect
Integration of biological, clinical,
epi & social evidence
Epi is probablistic
Association does not imply causation
Basis of scientific inquiry:
search for alternative explanations
rather than try to find "absolute proof"
that one expl must be correct
Stat assoc implies that an exposure
(risk factor) is related to change in the
probability
of the outcome (disease, death)
Multicausality
Complex Diseases
have complex, multiple
causes
BOTH
necessary & sufficient
a single factor,
by itself,
can cause the disease
AND the disease is not present
when the factor is absent
eg Asbestosis &
exposure to asbestos
Necessary but
not sufficient-
AND
Each factor
alone
does not cause the disease,
But it occurs when
all risk factors are present
eg TB:
need bacillus
AND
susceptibility
Sufficient-
with it, disease
Sufficient but
not necessary-
OR
eg leukaemia
caused by radiation
OR
benzene
independent pathway
to disease
One factor can produce a disease
but so can other factors
& it is not necessary
that both occur together
Necessary-
without it, no disease
Neither sufficient,
nor necessary
but the disease might not occur
if the factor is present
AND
the disease can occur
when the factor is absent
A factor by itself
may be related
to risk of disease
eg heart disease
The hypothetico-deductive method
From this a specific
hypothesis
is
developed which can be
tested
The hypothesis is then rigorously
tested
using experiments &
by collecting observations
A
research question
is developed
based on established
theories
proposing an explanation for,
or cause of, phenomena
The observations are then used
to judge whether the hypothesis
has been
falsified
Guidelines for determining causation
Biological Plausibility
Consideration of alternative explanations
Dose-response relationship
Experimentation
(esp cessation of exposure)
Replication of the findings
Consistency with other knowledge
Strength of association
Specificity of association
Temporal relationship