What email address or phone number would you like to use to sign in to Docs.com?
If you already have an account that you use with Office or other Microsoft services, enter it here.
Or sign in with:
Signing in allows you to download and like content, and it provides the authors analytical data about your interactions with their content.
Embed code for: Sufficient Component Cause
Select a size
Incorporating notions of necessity and sufficiency into causal inference.
Sufficient Component Cause
Incorporating notions of necessity and sufficiency into causal inference
Patrick F. McArdle, Ph.D.
Causal inference in epidemiology is better viewed as an exercise in measurement of an
effect rather than as a criterion-guided process for deciding whether an effect is present
Rothman, Greenland: 2005
In what sense does flipping a switch cause a light to go on?
In the counterfactual sense, the switch needs to be in the on position for their to be light.
And if the switch is instead in the off position there is no light.
Therefore we conclude the switch is a cause of light in the room.
This pre-supposes the presence of other factors, such as available power, in tact wiring and an operating bulb.
Many causes, like the light switch, are only components in a complex causal mechanism.
An event is rarely sufficient in and of itself to cause an outcome.
Particularly in the epidemiological study of complex disease states.
Rothman’s Definition of a Cause
“We can define a cause of a specific disease
event as an antecedent event, condition,
or characteristic that was necessary for the
occurrence of the disease at the moment it
occurred, given that other conditions are fixed.”
- Modern Epidemiology
Mackie – American Philosophical Quarterly - 1965
An insufficient but necessary part of a condition which is itself unnecessary but sufficient for the result.
Rothman – American Journal of Epidemiology - 1976
Sufficient component cause definition
Wright – California Law Review - 1985
A necessary element of a sufficient set
Sufficient Causal Mechanism
A set of minimal conditions and events that inevitably produce disease;
“Minimal” implies that all of the conditions or events are necessary to that occurrence.
Together, the minimal conditions are sufficient to produce the disease.
Note the use of a deterministic framework. Given the set of conditions, disease will occur.
(1) A given event can be caused by more than one causal mechanism
(2) Every causal mechanism involves the joint action of a multitude of component causes.
If we eliminate a component cause, we can prevent some disease, even if we do not know all the other components in the causal mechanism.
If we eliminate a component cause, the disease will still exist if there are other causal mechanisms which do not include that component.
How do we normally estimate the strength of a cause in epidemiology?
The change in disease frequency when a factor is introduced or removed from the population.
Where “change” is usually estimated by either the difference or ratio of disease frequency or trait values.
Strength of a Cause
The “strength” of cause G, or the amount of cases caused by G, is directly proportional to the prevalence of factor H.
The make-up of each causal mechanism is determined by “nature” but the relative presence of each mechanism is driven by the prevalence of each component.
Therefore the “strength” of a causal component can change with time, place etc.
What fraction of disease is attributable to component A? component G?
What does that tell us about causality?
60% of breast cancer is due to smoking.
If that is true, what percentage of breast cancer could be due to genes?
Assume I perform a study of breast cancer in the Amish, a population where smoking among women is very rare. What percentage of breast cancer would you expect to be attributable to genes?
Relative risk of a disease due to some factor is different in the absence or presences of another factor.
Rothman suggests that all components acting within the same causal mechanism interact with one another to produce disease.
Practical Issues with SCC
Each piece of the pie needs to specify a causal contrast.
At what level does one describe each causal component?
Each causal component is likely the result of other, previous causes.
Where does one stop?
This is the sense of causation we will be mostly speaking of in this course. An exercise in measurement, rather than a criterion guided process as suggested by Bradford-Hill.
Minimality: The removal of any component cause will prevent the disease from that mechanism.
What are some interpretations of that statement?
What happens if we presume smoking is at least partially due to inherited factors?
What is a definition of interaction?ense, the switch needs to be in the on position for their to be light.